The prehistory and early history of the Šumava Mountains (Czech Republic) as seen through anthropogenic pollen indicators and charcoal data

Abstract

In the lower forested mountain ranges of Europe, human impact on nature is usually confined to the written history of the Middle Ages. Our research in the Šumava mountains aims to specify the nature and intensity of human impact on vegetation, especially during agricultural prehistory. We use results from a multidisciplinary study of the unique La Tène archaeological site in the Sklářské Valley (elevation 802 m a.s.l.) and from a pollen and charcoal record 60 m away. With knowledge of this reference site we focus on the meaning of anthropogenic pollen indicators in 13 other pollen sites from central Šumava. From ca 3300 cal yr BP we detect an increase in NAP, Betula, Pinus and secondary anthropogenic indicators in pollen records – vegetation changes of anthropogenic origin. Charcoal records show a significant peak around 3200 cal yr BP. We found weak anthropogenic influence on the La Tène archaeological site in the Sklářské Valley, and much stronger anthropogenic pollen signals at other pollen sites dated to the Late Bronze and Iron Ages. Some of these sites are situated on trade routes which have been known since Medieval times but which most likely have much older origins. During prehistory, pollen data reveal no specific human activity such as pasturing or arable farming but reflect small-scale disturbances that supported growth of Betula and Pinus and an abundance of herbs. Such human impact could be connected primarily to activities along trade routes and to hunting, but other factors cannot be excluded.

Keywords

anthropogenic pollen indicators sedimentary charcoal Central Europe human impact fire Holocene

Introduction

The evolution of mountain landscapes in Europe has often been shaped by human activity. It has been repeatedly shown that mountains, especially the treeless alpine zone, have attracted people for various reasons since prehistory, with mining, logging, hunting and seasonal pastoralism being the most common forms of exploitation (Gilck and Poschlod, 2019; Houfková et al., 2019; Stopp, 2015). We now have extensive and detailed knowledge of the changes in land cover that have resulted from human activity, including the use of fire for expanding forest-alpine grassland ecotones, especially in the Alps (Dietre et al., 2014; Pini et al., 2017; Schwörer et al., 2015), Pyrenees (Rius et al., 2011) and Carpathians (Feurdean et al., 2016; Vincze et al., 2017; Wacnik et al., 2015). Much less is known about the mid-mountain ranges of Central Europe, including the largely treeless areas of the Black Forest, Harz, Krušné Mountains/Erzgebirge and the Šumava Mountains/Bavarian Forest. Prehistoric human impact was thought to have been less intense in these regions (Henkner et al., 2018; Robin et al., 2013; Rösch, 2000) or virtually absent until the High Middle Ages (Kozáková et al., 2015). The harsh climate, unfavourable soils, and lack of natural alpine grasslands suitable for grazing have all been offered as plausible explanations for the limited use by humans of these mountain areas (Dreslerová, 2015; Rösch et al., 2020). Archaeological research is difficult to carry out in heavily forested regions, and the resulting sparsity of archaeological evidence prevents an unequivocal assessment of the extent of human activity. However, in the Šumava Mountains/Bavarian Forest, the largest forested mid-mountain range in Central Europe, the occurrence of secondary anthropogenic indicators (Behre, 1981) and cereal pollen have been repeatedly reported in pollen records dating back to the Neolithic (Svobodová et al., 2002), thus indicating some kind of anthropogenic activity. Increased evidence of fire during the Iron Age further suggests a period of intensified human land use (Carter et al., 2018b). Recent advances in archaeological knowledge of the Šumava Mountains (hereafter referred to as Šumava) have revealed evidence of human presence during the Mesolithic (Eigner et al., 2017; Kapustka et al., 2019) and later prehistory (Dreslerová et al., 2020). These new findings have prompted a debate over the recognition of prehistoric human activity, the relationship between fire and humans, and the way people used the mountain landscape.

Tracing human impact on vegetation using the presence and abundance of anthropogenic pollen indicators is an established approach that has been widely applied in Europe (Deza-Araujo et al., 2020). Although the pollen of cultivated crops is considered a direct indication of human occupation (i.e. arable farming), its effectiveness is limited because of the large pollen loads produced by dense forest cover. Furthermore, long pollen sequences have often been analysed with a low resolution and without detailed radiocarbon dating, which has greatly limited the interpretative potential of pollen data (especially with regard to an assessment of human impact) and hampered the making of precise comparisons between pollen sites. Little attention has been paid to human impact before the High Medieval colonisation, although the occurrence of secondary anthropogenic indicators has been reported (Svobodová et al., 2002). The ability to detect anthropogenic impact in mountain environments relies mostly on the native plant species that relate to particular human activities (i.e. secondary pollen indicators), such as grazing or woodland clearing, and the frequent and intentional use of fire to promote optimal habitats. Although it is not possible to reliably distinguish between the natural and anthropogenic origins of a fire event, regional divergences in fire regimes in climatically and environmentally homogenous areas can be explained by human-related factors (Dietze et al., 2018). Accordingly, reconstructing biomass burning using sedimentary charcoal appears to be a promising tool for tracing anthropogenic activity, and one that complements the pollen signal.

In this paper, we use anthropogenic pollen indicators to compare the local pollen record from the Sklářské Valley pollen site with the 13 pollen sequences currently available for Šumava (Figure 1). To obtain an independent proxy for human presence, we also reconstruct local fire activity based on an analysis of macro-charcoal (>125 μm) in three peat bog sequences. Regional biomass burning has been estimated by compositing new and existing charcoal records. Our principal research questions are: (1) How strongly and specifically do pollen and charcoal signals indicate anthropogenic activity in the Šumava profiles compared to the reference site of the Sklářské Valley? (2) What can be deduced from the individual profiles about the nature of human activity in prehistory and early history? (3) How strongly was vegetation affected by human activity and fire disturbances during prehistory and early history?

Figure 1.

Šumava and the Bavarian Forest with Czech (dark green dots) and Bavarian (light green dots) pollen sites, charcoal sites (black dots), main trade routes: (a) Bohemian route, (b) smuggling branch of the Golden Trail, (c) three branches of the Golden Trail); archaeological evidence in the central part of the mountains (archaeological periods expressed by the same colours as in Figure 2); non-settlement/solitary find (cross), settlement (triangle). Pollen and charcoal sites: Březník A (1), Březník B (2), Hůrecká slať (3), Chalupská slať (4), Mrvý luh Chlum (5), Mrtvý luh (6), Malá niva (7), Rokytecká slať (8), Rybárenská slať (9), Stráženská slať (10), Zadní chalupy (11), Plešné jezero (12), Sklářské Valley (13), Stará Jímka (14), Šmauzy (15), Sonndorf (16), Dösingerried (17), Haidmühle (18), Finsterauer Filz (19), Ludwigsthal (20), Bachlern (21), Föhraufilz (22), Sonnen (23), Seefilz (24), Dreckiger Filz (25), Markfilz (26), Grosser Filz am Grossen Spitz-Berg (27), Kulz (28), Rachelsee (29), Kugelstattmoos (30). Background map: coniferous and mixed forest (green), urban area (orange), agro-forestry areas (grey).

The study area

Šumava/Bavarian Forest was formed by a tectonically elevated Paleogene plain and is one of the largest mid-mountain ranges in Europe. A large part of the original surface is well preserved in the central part of the region; its elevation today is over 1000 m a.s.l. More than 15% of the area is bog and it is one of the most important peatland regions in Central Europe (Soukupová, 1996). The timberline is reached at the highest summits of Plechý (1378 m a.s.l.) and Großer Arber (1456 m a.s.l.). The geological bedrock consists mostly of acidic granite and gneiss. Annual mean temperatures are about 5°C at the higher elevations. Average rates of precipitation are higher in the NW (1200–1300 mm) than in the SW (730–870 mm) (Culek, 1996). The lower mountain zone is covered mainly by mixed Fagus sylvatica–Abies alba forests. These areas has been affected by extensive logging and the subsequent planting of Picea abies. The upper mountain forest zone and margins of bogs are dominated by natural Picea abies stands (Chytrý, 2012). The vegetation development and fire history is well known thanks to a dense network of pollen sites (Carter et al., 2018a; Jankovská, 2006; Rüther and Nelle, 2006; Stalling, 1987; Svobodová et al., 2001, 2002; van der Knaap et al., 2020) and an increasing number of charcoal records (Bobek et al., 2019; Carter et al., 2018b).

The archaeological context

The earliest archaeological evidence from Šumava/Bavarian Forest is represented by chipped stone industry using cherts from the Ortenburg Jurassic in eastern Bavaria (Figure 1). Mesolithic hunter-gatherer camps (ca. 10500–7500 cal yr BP) have been found on the banks of the Roklanský brook in the vicinity of the pollen sites Rybárenská and Rokytecká slať (Katarína Čuláková, personal communication). It is likely that the camps were situated directly on a trading route along which the raw material was transported over the mountains into Bohemia (Eigner et al., 2017). In the late phase of the Neolithic Linear Pottery Culture, the Ortenburg and Arnhofen cherts (Figure 1) account for a substantial proportion of the chipped stone industry in Bohemia (Vondrovský et al., 2018). The high levels of use of the raw material suggests that people made frequent crossings of the mountains. The increased use of fire reported from the lacustrine sediments of Prášilské lake suggests a period of intensified human land use in the Early Iron Age (Carter et al., 2018b). This observation is supported by the recent discovery of a unique high-altitude archaeological La Tène site (802 m a.s.l.) located near the abandoned village of Frauenthal in the Sklářské Valley, dated to ca. 300–390 BC (Figure 1; Dreslerová et al., 2020). A person (or a small group of people) left traces of repeated activity over a small platform (ca. 20×20 m) situated above a distinct meander of the Křemelná river. The exact nature of this activity has been difficult to determine, but evidence gained from multi-disciplinary research suggests that the site might have been a camp on a trade route, a guardhouse, or perhaps a hermitage (Dreslerová et al., 2020). A peat bog located just 60 m away provided pollen and macro-charcoal records. The clear spatial relationship between the archaeological and palaeoecological records enabled a detailed assessment of the nature and intensity of prehistoric settlement in the mid-mountains. The beginning of the permanent colonisation of Šumava is associated with the establishment of the glass industry and logging in the High Middle Ages (Fanta et al., 2020; Zavřel and Anděra, 2003) although earlier activities were connected mainly with the access routes passing through the mountains in several directions, presumably since prehistory. Three branches of the ‘Golden Trail’ (Goldener Steig) and the north-trending Bohemian route were functioning at least from the eleventh century (Figure 1; Zavřel and Kubů, 2007a, 2007b). Prehistoric settlement in the foothills of Šumava/Bavarian Forest ceased at an altitude of around 600 m, above which arable farming seems to have been unprofitable (Figure 2).

Figure 2.

Archaeological evidence from Šumava, the Bavarian Forest and the surrounding foothills dated to particular archaeological periods: (a) Neolitic (pink), Eneolitic (violet). (b) Bronze Age (red). (c) Iron Age (blue). (d) Medieval Period (brown), Modern Age (yellow). Archaeological data kindly provided by Institute of Archaeology of the CAS, Prague and Bayerisches Landesamt für Denkmalpflege, München.

Materials and methods

Chronological control

The chronology of the whole dataset is based on 101 radiocarbon dates (for details, see Supplementary Table S1). A subset of 66 radiocarbon accelerated mass spectrometry (AMS) dates aimed to improve the chronological precision of selected profiles during the Bronze and Iron Ages. Additional plant material (mostly seeds, charcoal, Sphagnum stems, peat) was extracted from respective layers, pre-treated using acid-alkali-acid (AAA) and sealed-tube graphitisation in the Czech Radiocarbon Laboratory (CLR) in Prague; isotopic ratio ¹⁴C/¹²C was measured in the Hertelendi Laboratory of Environmental Studies (DebA), ATOMKI HAS in Debrecen, Hungary. Age-depth models were obtained from the Czech Quaternary Palynological Database (https://botany.natur.cuni.cz/palycz/), and its performance is summarised in Supplementary Table S1. Sedimentation rates in the reference site Sklářské Valley and Sklářské Valley 2, Šmauzy peat core were modelled using Bayesian age-depth modelling in the Bacon software (Blaauw and Christeny, 2011). Age-depth model for Zadní chalupy was built using Clam 2.2 (Blaauw, 2010). The IntCal13 calibration curve (Reimer et al., 2013) was used for converting ¹⁴C age to calendar years. All calibrated dates are reported in years Before Present (hereafter, cal yr BP). Summed probability distribution of radiocarbon dates from Sklářské Valley archaeological site (Dreslerová et al., 2020) was used to identify phases of human occupation. All the calculations on ¹⁴C dates were performed in Oxcal, version 4.3 (Ramsey, 2009). When referencing archaeological periods, we use the chrono-typological dating proposed by Jiráň and Venclová (2013), which is listed in Supplementary Table S2.

Pollen records

New data were gained from a peat bog situated in the Sklářské Valley (Figure 1). We used an Edelman corer to extract two parallel sedimentary cores: Sklářské Valley (two sections: S1 0–100 cm; S2 40–140 cm) and Sklářské Valley 2 (one section: 0–100 cm). Pollen sub-samples were taken from the Sklářské Valley core with increments of 1 cm. We also collected 13 pollen profiles from the Šumava region stored in the Czech Quaternary Palynological Database (Kuneš et al., 2009) or personal repository (Table 1). Four of these (Březník B, Šmauzy, Malá niva, Stará Jímka) were additionally sampled for both pollen analysis and radiocarbon dating. The pollen data from nine sites are yet to be published. All profiles are situated outside the zone of settlement but within the protected area of the Šumava National Park.

Table 1.

List of pollen and charcoal sites with their characteristics.

Site name	Location (WGS 84)	Elevation (m a.s.l.)	Area (ha)	Site description	Proxy	Number of ¹⁴C dates	Reference	Note
Březník A	48°57ʹ44.316ʺN; 13°29ʹ19.601ʺE	1250	10.7	Ombrothrophic raised bog	Pollen	6	Svobodová Svitavská, unpublished data	¹⁴C dates added
Březník B	48°57ʹ44.316ʺN; 13°29ʹ19.601ʺE	1150	10	Ombrothrophic raised bog	Pollen	3	Svobodová Svitavská, unpublished data	New layers and ¹⁴C dates added
Hůrecká slať	49°9ʹ7.992ʺN; 13°19ʹ39.18ʺE	870	62	Forested domed raised bog	Pollen	6	Svobodová et al. (2002)
Chalupská slať	49°0ʹ2.196ʺN; 13°39ʹ46.296ʺE	906	137	Transition mire	pollen	7	Svobodová Svitavská, unpublished data	¹⁴C dates added
Malá niva	48°54ʹ49.536ʺN; 13°48ʹ57.816ʺE	754	127	Forested raised bog	Pollen, charcoal	7	Bobek et al. (2019), Svobodová et al. (2002)	New layers and ¹⁴C dates added
Mrtvý luh	48°52ʹ0.480ʺN; 13°52ʹ58.512ʺE	735	352	Large transition valley bog	Pollen	4	Svobodová et al. (2001)
Mrtvý luh Chlum	48°52ʹ0.480ʺN; 13°52ʹ58.512ʺE	737	352	Large transition valley bog	Pollen	7	Svobodová Svitavská, unpublished data	¹⁴C dates added
Plešné jezero L2	48°46ʹ36.264ʺN; 13°51ʹ56.556ʺE	1105	7.5	Glacial lake	Pollen	3	Svobodová Svitavská, unpublished data	¹⁴C dates added
Rokytecká slať B	49°0ʹ55.080ʺN; 13°24ʹ43.920ʺE	1089	142	Large ombrotrophic raised bog	Pollen	7	Svobodová Svitavská, Jánský, unpublished data	¹⁴C dates added
Rybárenská slať	49°1ʹ52.644ʺN; 13°27ʹ42.804ʺE	1014	134	Large ombrotrophic raised bog	Pollen	7	Svobodová et al. (2002)	¹⁴C dates added
Sklářské Valley	49°8ʹ36.619ʺN; 13°23ʹ43.17ʺE	810	0.3	Small mire	Pollen	15	This study	Reference site
Sklářské Valley 2	49°8ʹ36.619ʺN; 13°23ʹ43.17ʺE	810	0.3	Small mire	charcoal	4	This study	Reference site
Stará Jímka	49°4ʹ8.000ʺN; 13°24ʹ10.000ʺE	1110	35	Transition valley bog, infilled lake	Pollen, charcoal	9	Bobek et al. (2019)
Stráženská slať	48°53ʹ55.932ʺN; 13°44ʹ32.136ʺE	812	114	Transition valley bog	Pollen	6	Svobodová et al. (2001)	¹⁴C dates added
Šmauzy	49°11ʹ33.987ʺN; 13°16ʹ15.515ʺE	1010	0.8	Spring mire	Pollen, charcoal	5	This study	Analysed recently
Zadní chalupy	49°7ʹ28.205ʺN; 13°23ʹ44.323ʺE	852	0.9	Forested raised bog	charcoal	5	This study	Analysed recently

Pollen samples were prepared according to Faegri and Iversen (1989) using 10% KOH, 10% HCl, concentrated HF and a mixture of acetic anhydride and sulphuric acid in a ratio of 1:9. In Sklářské Valley, Šmauzy and Stará Jímka, more than 700 grains were counted in most samples. Pollen sums in other profiles were usually 300–600 grains. Pollen types were defined according to the Northwest European Pollen Flora, Beug (2004) Punt (1976), Punt et al. (1988, 1995), Punt and Blackmore (1991), Punt and Clarke (1980, 1981, 1984), Punt and Hoen (2009), Reille (1992) and the reference collection of the Institute of Botany of the CAS. Non-pollen palynomorphs (NPP) were determined according to available publications (Pals et al., 1980; van Geel, 1978; van Geel et al., 1986, 1989) and follow the terminology reviewed by Miola (2012). The percentage pollen diagram was plotted using Tilia v. 1.7.16 (Grimm, 2004). Numerical zonation was performed in R (R Development Core Team, 2017) using constrained hierarchical clustering (CONISS), and statistically significant splits were determined with the broken-stick model (Juggins, 2017). The numerical approach is useful for depicting the major compositional changes in vegetation but is less suitable for highlighting low-intensity human impact. We therefore distinguished several additional zones.

Anthropogenic pollen indicators

We adopted the concept of anthropogenic indicators (AI) established by Behre (1981) to detect various human activities in the pollen record. Within this framework, primary anthropogenic indicators (PAI) are the pollen of cultivated plants and secondary anthropogenic indicators (SAI) are represented by the species associated with human activity (pasture, trampled areas, ruderal sites, etc.), but which may also occur in natural plant communities. We classified the pollen taxa in our dataset as follows: PAI – Secale cereale (hereafter Secale), Cerealia type; SAI – Plantago lanceolata (hereafter Plantago l), Plantago major/media (hereafter Plantago mm), Rumex acetosa type (hereafter Rumex), Artemisia, Chenopodiaceae, Urtica, Calluna vulgaris (hereafter Calluna). When deciphering anthropogenic signals, we focused on the following characteristics of a pollen record: (1) repeated occurrence of AI (several consecutive samples) and finds of more than one indicator in a single sample; (2) trends in non-arboreal pollen (NAP) and Poaceae sum reflecting the openness of the landscape (Broström et al., 1998); (3) increases in the pioneer trees Betula and Pinus, suggesting anthropogenic or natural disturbance of the forest canopy and spontaneous reforestation.

In densely forested areas, AI are generally rare because of the enormous pollen influx from trees. The possibility of recording AI depends on the pollen productivity of the dominant taxa and on the pollen sum counted (Odgaard, 2013). We neglected the different pollen production as all the tree dominants Fagus, Abies and Picea are present for the whole of agricultural prehistory in Šumava (Svobodová et al., 2002), and the effect is similar at all sites. However, to account for unequal pollen sums, we followed a re-sampling technique similar to Connor et al. (2019) and Seddon et al. (2015). We re-sampled individual pollen records to a target sum of 360 grains (quantile 25% of all pollen sums in the dataset) and a time window of 150 years (quantile 75% of all timespans between samples) (see Supplementary Figure S1). During the re-sampling procedure, we picked 360 random pollen grains from each sample 100 times and calculated the number of taxa for each sample. The assemblage with the number of taxa corresponding to the median was selected for further calculation and visualisation. The re-sampled dataset and samples with fewer than 360 grains were pooled within time windows and re-sampled again to 360 grains per time window. The results are visualised in the form of heat map (hereafter tetris plot) which depict each AI and individual site. In these plots, the number of sites per 150-year window in which a given AI is present and its percentage is displayed by a colour. In order to highlight the different stages of human impact across the whole of Šumava and to compare with our reference site Sklářské Valley, we visually delimited phases according to the following criteria: (1) changes in NAP curve representing the degree of landscape openness. NAP was a criterion for divisions at 5000 cal yr BP and 3300 cal yr BP; (2) chronological criteria including a division at the end of the Iron Age at 2000 cal yr BP and a division at 800 cal yr BP, which represents the beginning of Medieval colonisation in Šumava.

Charcoal records

Five macro-charcoal series (>125 μm) were used to reconstruct local and regional fire activity in Šumava (Table 1). These were Sklářské Valley 2, Zadní chalupy, Šmauzy, Malá niva and Stará Jímka (Bobek et al., 2019). Sediment cores were collected using an Edelman corer and sliced at 1 cm intervals. Sub-samples of 2–4 cm^-3 were disaggregated with the aid of a 10% NaCl solution and subsequently bleached by sodium hypochloride (NaOCl). Samples were wet-sieved through a 125 μm mesh, and non-charred organic material was manually removed from the coarse fraction under a stereomicroscope at 6.3–60× magnification to facilitate counting. The charcoal particles were quantified by visual inspection or using a custom-built image analysis system (http://www.microspock.cz/). Charcoal values (counts) were converted to accumulation rates (CHAR, pieces cm^–2 yr^–1) by dividing charcoal concentration (pieces cm^–3) by corresponding deposition time (yr cm^–1). Standardisation is required in order to account for analytical differences and the large variation in CHAR in the different records or in the characteristics of the sites (Power et al., 2010). Individual CHAR series were pre-binned using non-overlapping 100-year bins to counterbalance the impact of high-resolution records. The resulting sequences were rescaled using ‘minimax’, followed by a Box-Cox transformation to homogenise intra-record variability and, finally, z-score-transformed at the base period of –50 to 5000 cal yr BP. The zero z-score value corresponds to the mean charcoal influx of all transformed series at the scale of the whole Holocene period. To highlight trends in regional fire activity, a charcoal composite curve was constructed as proposed by Marlon et al. (2008). Transformed charcoal records were then smoothed using a locally weighted scatter plot smoother (LOWESS) with a window width of 1000 years. Confidence intervals (95%) were then calculated by 1000 bootstrap re-sampling with replacement of individual CHAR series. Statistical analysis was performed in the ‘paleofire’ R package (Blarquez et al., 2014).

Results

Sklářské Valley pollen site

The beginning of peat accumulation was dated to 3960 cal yr BP (Figure 3). Species composition of the forest from this time until colonisation during the Middle Ages does not undergo any significant changes. Pinus, Betula, Picea, Abies and Fagus are the dominant taxa around the pollen site. In the bog and along the stream entering and leaving it, Alnus was common, but Pinus, Betula and Picea were also present.

Figure 3.

Bayesian age-depth models for reference pollen profile Sklářské Valley (a) and parallel core Sklářské Valley 2 (b) The shaded area indicates the 95% probability interval of the model. The dashed line in red colour indicates the best-fit age-depth relationship.

In the oldest section (S2, 100–136 cm, 3400–3900 cal yr BP; Figure 4), Picea has the highest pollen ratio among terrestrial tree taxa. The proportion of Abies pollen gradually increases and the proportions of deciduous trees such as Corylus, Tilia, Ulmus and Fraxinus are slightly higher than the above in the section S1c. SAI were frequently detected, but mostly as individual (single pollen grain) finds. Only the pollen of Artemisia occurs almost continuously in this section and Plantago l in the oldest four samples, but in very low numbers. Trilete spores, Cyperaceae and Alnus are the dominant vegetation in the bog.

Figure 4.

Percentage pollen diagram from Sklářské Valley (silhouettes show 5× exaggeration curves). The period corresponding to occupation phase of the La Tène archaeological site (Dreslerová et al., 2020) is marked by a red frame. Solid zone lines are statistically significant.

Ratios of both Abies and Fagus remain relatively stable throughout S1c: Abies 15–20%, Fagus around 15%. The proportion of Picea gradually decreases from 20% to 15% or less. There are also several minima on the Picea curve, mostly corresponding to Pinus maxima (90 cm and 66–69 cm, 3100 cal yr BP and 2890–2930 cal yr BP) and thus most likely reflecting periodic fluctuations in pollen production. SAI were detected with only slightly higher frequency compared to S2. Apart from the oldest layers within S1c, Artemisia has a continual curve. Plantago l has one significant continual occurrence (82–90 cm, 3070–3140 cal yr BP; Figure 4). An increase in AI, Avena type and Rumex, is registered at 90–92 cm. There are peaks of Melampyrum and Calluna at the same level.

In S1b, the proportions of the pollen of Corylus, Tilia, Ulmus and Fraxinus are marginal, especially the latter three taxa. Carpinus gradually increases but Abies and Picea slowly decrease throughout this section. There is an abrupt increase in Pinus and Betula and an abrupt drop in Alnus and Cyperaceae in the oldest part of the section.

Layers corresponding to the nearby La Tène archaeological site are within the red frame (40–43 cm, 2000–2280 cal yr BP; Figure 4). In these layers, the arboreal pollen sum begins to decrease, particularly Betula, Picea, Abies and Alnus. Poaceae and trilete spores increase, Melampyrum remains high, and there are small peaks of Artemisia, Plantago l and Rumex and individual finds of Secale and Asteraceae-Cichorioidae.

In the remainder of S1b (between 40 and 30 cm, 2000-720 cal yr BP; Figure 4), there is an increase in Pinus and Carpinus, a sharp onset of Calluna, and individual occurrences of cereals (Secale, Triticum and Hordeum types). Almost every sample contained the non-pollen palynomorphs (NPPs) Gelasinospora, Glomus and Sporormiella.

For the uppermost section, S1a, the continual presence of cereals is typical and Secale is especially common. From SAI, Plantago l and Rumex have significant and continual curves. Other herbs such as Potentilla type, Rubiaceae, Ranunculus acris type and Trifolium repens type were also detected in substantial numbers in some samples. Abies and Fagus were present in minimal quantities.

Regional pollen data

General trends in the pollen percentages of Poaceae, Betula, Pinus and NAP are shown in Figure 5 (black curves). It is clear that in the oldest layers, an initial high proportion of NAP (–5%) in the Z1 pollen zone gradually decreases and a minimum is detected within Z2, between 5000 and 3300 cal yr BP. An increase in NAP is clear in Z3 and continues until the present day. The same can be observed with Poaceae and Betula, although the trends are less pronounced. Similar trends are seen with Pinus, except here the minimum covers a shorter period, between 5000 and 4200 cal yr BP. In Z5 NAP, Poaceae, Pinus and Betula have maxima; only Betula decreases in the youngest samples. A comparison with data from the pollen profile Sklářské Valley is shown in red. Here, the pollen record starts around 3300 cal yr BP. The NAP proportion is higher than in other pollen sites, especially Poaceae. Within Z3 the peaks in NAP and Poaceae are followed by two significant drops around 2600 and 2400 cal yr BP and a subsequent sharp increase between 2200 and 2000 cal yr BP. Drops in NAP and Poaceae are often co-incident with peaks of Betula and Pinus, which is most visible in Z3 between 2600 and 2000 cal yr BP. This could signify a clearance of the forest followed by rapid succession of pioneer trees.

Figure 5.

Comparison of summary pollen diagram (in black) from 13 profiles in Šumava (Š) with pollen record from reference site (R) Sklářské Valley (in red). Only selected taxa are shown including non-arboreal pollen (NAP). Horizontal black bars indicate the number of sites per 150-yrs window covered with pollen record. Note that length of the records is not equal.

Figure 6 show that most of the SAI taxa are detected in pollen records throughout the Holocene. Artemisia, Calluna, Plantago mm and Rumex occur most regularly (almost continually from Z1 to Z5, at least at one site). A general increase in the frequency of occurrences – more SAI taxa at more sites – is clear from 3000 cal yr BP. Another increase, especially in Artemisia, Plantago l, Plantago mm and Rumex, is visible from 2400 to 2000 cal yr BP. In section Z4 (2000-800 cal yr BP), Secale becomes continually present at least at one site. SAI taxa are frequent in section Z4. Significant human impact can be traced in a layer around 1200 cal yr BP where especially Cerealia type, Secale, Chenopodiaceae, Plantago l and Urtica are frequent. Within section Z5, which corresponds to the High Medieval and Modern period, maxima of Calluna, Plantago l, Rumex and Secale are detected.

Figure 6.

Tetris plot showing occurrences of anthropogenic pollen indicators in 13 records from Šumava compared to the reference site Sklářské Valley. Pollen percentages standardised per taxa corresponds to the intensity of the black colour. Each individual square represents one 150-yrs window at one site.

Figure 7 shows the abundance of AI at particular sites. In the oldest section, Z1, the continual or nearly continual presence of several SAI taxa were detected at Mrtvý luh (mostly Rumex, Plantago mm, Urtica) and Hůrecká slať (Artemisia, Calluna, Rumex; individual finds of other SAI). Three cases of single layers with increased occurrence of AI are significant in Z1: before 7200 cal yr BP, at Rybárenská slať (Cerealia type, maximum of Plantago mm, Urtica, Rumex, Artemisia, Calluna) and layers at 7200 cal yr BP at Chalupská slať (Secale, Rumex, Plantago mm, Chenopodiaceae, Artemisia, Calluna) and Stará Jímka (Artemisia, Calluna, Chenopodiaceae, Urtica).

Figure 7.

Frequency of anthropogenic pollen indicators at particular sites. Pollen percentages standardised per taxa corresponds to the intensity of the black colour (the darkest hue represents maximum pollen frequency). Each individual square represents one 150-years window at one site. Squares in yellow correspond to time windows with missing data.

In Z2 (5000–3300 cal yr BP; Figure 7), apart from the most common Artemisia and Calluna, the occurrence of AI is rarely continual; the exceptions are Mrtvý luh (Plantago mm) and Stará Jímka, with the most prominent human impact signal between 4300 and 4100 cal yr BP (Artemisia, Plantago l, Plantago mm, Rumex, Urtica). Layers with increased finds of AI are dated to 4200 cal yr BP at Malá niva (Rumex, Plantago l, Artemisia); to 4100 cal yr BP at Šmauzy (Cerealia type, Chenopodiaceae, Plantago l, Plantago mm) and Hůrecká slať (Secale, Plantago l, Rumex, Calluna); ca. 3600 cal yr BP at Březník B (Artemisia, Plantago l, Rumex); and ca. 3300 cal yr BP at Rybárenská slať (Secale, Rumex, Plantago mm, Plantago l).

A general increase in human impact pollen signals is evident in section Z3 (3300–2000 cal yr BP; Figure 7). Plantago l and Chenopodiaceae, both pollen taxa relatively rare before this point, start to be more frequent. The oldest layer with increased occurrence of AI is 3300 cal yr BP at Březník A (Secale, Plantago mm, Plantago l, Chenopodiaceae, Artemisia). Around 3000 cal yr BP, marked human impact is registered at Mrtvý luh (all SAI taxa except Urtica). Between ca. 3000 and 2600 cal yr BP a strong human impact pollen signal is registered at Plešné jezero (maximum of Urtica, Chenopodiaceae, Plantago l, Plantago mm, Rumex, Secale). Between 2400 and 1900 cal yr BP, Březník B registers continual or nearly continual records of all SAI except Chenopodiaceae and a single find of Cerealia type pollen, and the same is true of Malá niva between 2900 and 2000 cal yr BP (Rumex, Plantago l and Artemisia continually; other SAI individually). A less striking signal is registered at Stará Jímka between 3000 and 2300 cal yr BP (Artemisia, Calluna, Plantago l, Rumex, Urtica). Within Z3 there are also single layers with increased finds of SAI, such as 2900 cal yr BP at Rokytecká slať B. At Sklářské Valley, we register the near-continual presence of SAI (Artemisia, Calluna, Plantago l, Rumex) in the periods ca. 2900–2000 cal yr BP and ca. 2300–2000 cal yr BP); significant human impact is registered at six other sites (Březník B, Hůrecká slať, Stráženská slať [Cerealia type and Secale pollen], Mrtvý luh, Malá niva and Rybárenská slať).

Section Z4 (2000-800 cal yr BP; Figure 7) is characterised by the first continual occurrence of Secale. This is registered at Mrtvý luh Chlum between ca. 2000 and 1400 cal yr BP. In addition to Secale, there is also Cerealia type pollen, Rumex, Plantago l, Chenopodiaceae at a maximum, Artemisia and Calluna. Secale is further registered at Chalupská slať between ca. 2000 and 1100 cal yr BP together with all the SAI taxa except Urtica, and similarly but later at Mrtvý luh. At Hůrecká slať, Secale is first registered around 1400 cal yr BP, and later occurs together with Cerealia type pollen and all the SAI taxa. At Stará Jímka we see the start of continual Secale and Cerealia type from ca. 1100 cal yr BP. Secale and Cerealia type pollen were both found in one layer (ca. 1600 cal yr BP) at Rybárenská slať together with Rumex and Calluna. At Šmauzy, individual Secale occurrence (1200 cal yr BP) and the abundant and frequent presence of Plantago l is registered in the whole of Z4. A single find of Cerealia type pollen was made at Sklářské Valley (ca. 1100 cal yr BP). Here we can further trace human activity from a Plantago l record continuing from Z3 (2300–1800 cal yr BP) and from a marked increase in Calluna from ca. 1400 cal yr BP. A significant human impact signal, although without the pollen of cereals, is also registered at Březník B between ca. 1400 and 800 cal yr BP (maximum of Urtica, Rumex, Plantago l, Chenopodiaceae, Calluna, Artemisia).

Section Z5 covers a period from the supposed start of the colonisation of Šumava until the present day. Permanent human impact is registered at Sklářské Valley (Secale, Cerealia type, Rumex, Plantago l and strong signal of Calluna containing a maximum), Malá niva and Stará Jímka. Abundant Secale together with Cerealia type and SAI taxa is further registered at Šmauzy, Mrtvý luh and Hůrecká slať (here only the older part of Z5; Figure 7).

The beginning of Medieval arable farming in Šumava has been estimated using pollen curves of Cerealia type and Secale and compared to the establishment of the nearest village (Figure 8; Zavřel and Anděra, 2003).

Figure 8.

Pollen diagrams showing the original proportions of Secale cereale (a) and Cerealia type (b). Foundation of the nearest village (after Zavřel and Anděra, 2003) is marked by the the red arrow: Sklářské Valley: AD 1750 Frauenthall; Březník A and Březník B: AD 1820 Březník; Chalupská slať: AD 1760 Svinná Lada; Mrtvý luh Chlum: AD 1359 Volary; Stará jímka: AD 1750 Neubrunn; Rokytecká slať: AD 1614 Modrava; Rybárenská slať: AD 1614 Modrava; Šmauzy: before AD 1600 Gerlova huť; Malá niva: AD 1709 České Žleby; Mrtvý luh: AD 1359 Volary; Hůrecká slať: after AD 1617 Glassewald; Plešné jezero: no village within a 5 km radius; Stráženská slať: AD 1359 Strážný.

Fire record

The charcoal record from the catchment area of the Křemelná river (for individual site see Supplementary Figure S2) indicates increased fire activity during the Bronze Age, particularly at ca. 3800 cal yr BP (Supplementary Figure S2; Zadní chalupy, Šmauzy) and 3100 cal yr BP (Figure 9c and d; Sklářské Valley 2). During these two periods, we found pronounced peaks in charcoal influx and large charcoal particles (>2 mm), both indicating on-site fire occurrence. Between 3000 and 900 cal yr BP there is a generally low deposition of charcoal. In addition, a minor peak in charcoal accumulation was observed in the Sklářské Valley 2 profile (Figure 9d) during the La Tène period at ca. 2300 cal yr BP. This is followed by a rapid increase in charcoal accumulation between 900 and 650 cal yr BP. The regional CHAR composite curve (Figure 9f) indicate three phases of increased fire activity between 4000-3700 cal yr BP, 3300-3050 cal yr BP and 900-460 cal yr BP.

Figure 9.

Simplified pollen record from the Sklářské Valley peat bog showing (a) selected pollen types related to succesional stages of vegetation after fire disturbance. The charcoal influx (c) and concentration values (d) document periods of increased fire activity in the area adjacent to the sampling site. Black crosses (b) depict the presence of charcoal particles >2 mm which are indicative of the in-situ occurrence of fire. Some of the single charcoal particles were 14C dated (red triangle) to refine the temporal precision of fire occurrence. An occurrence of coprophilous fungi Sporormiella (green dots; local indicator of grazing herbivores or omnivores) is given in panel (e). Fungal spore indicator of fire and dead organic matter Gelasinospora is depicted as red dots. Regional fire activity (f) (median and 95% confidence interval) is based on a composite fire signal of five charcoal sequences from Šumava (see Supplementary Figure S2). Summed probability distribution of radiocarbon dates (g) from the archaeological site in Sklářské Valley (Dreslerová et al., 2020) show the phases of human presence. Yellow vertical bars show the visual correlation between increased regional fire activity and other proxies.

Discussion

Assessing the diagnostic capacity of anthropogenic indicators

Pollen records from Šumava suggest that human impact is largely reflected by the most common SAI such as Plantago l, Plantago mm, Rumex, Artemisia, Chenopodiaceae, Urtica and Calluna; there are also rare findings of Cerealia type and Secale pollen grains. The SAI represent anemogamous species with high pollen production, and pollen taxa that are considered regional and non-specific or less specific AI, especially when found in low numbers (Behre, 1981; Brun, 2011; Court-Picon et al., 2006). These plants are also sparsely distributed in montane forest communities, and every kind of disturbance, whether natural or anthropogenic, enhances their abundance in the vegetation. All the SAI taxa appear in pollen records from the beginning of the Holocene and clearly belong to the native vegetation (Figures 6 and 7). Artemisia, Calluna, Plantago mm and Rumex pollen are always common at least at one site and this reduces their indicative value. Artemisia can grow on disturbed or nutrient-enriched ground, such as eroded riverbanks, natural and anthropogenic forest openings, or places where nitrogen is concentrated. Calluna is typical of raised bogs and can be connected with human impact only with caution, and only in cases of a marked increase. Its pollen can indicate pastureland, grazed forest or shallow or poor acid soils (Behre, 1981; Gaillard et al., 1994). A stronger indicative value can be ascribed to pollen of Plantago l, Chenopodiaceae and Urtica. Plantago l and Chenopodiaceae are rare until 5000 cal yr BP and start to be abundant only from ca. 3300 cal yr BP. When found in low numbers, Plantago l pollen offers only a general indication of human activity (Behre and Kucan, 1986; Brun, 2011; Court-Picon et al., 2006); Pollen of Plantago mm has the disadvantage of representing two species with different ecologies. It can grow along paths as it is resistant to trampling, or in meadow-like habitats especially when they are grazed. Chenopodiaceae, which mostly include annual taxa, are connected with the early stages of succession after a disturbance and with nitrogen-rich sites. Pollen of Rumex includes a number of species, some of them typical of riverbank vegetation. As AI they are indicative with marked increases and in combination with other SAI. Pollen of Urtica is not generally common and usually occurs together with other SAI. It seems to have a particular connection with local nutrient enrichment.

The amount of NAP in the sample is a crucial criterion for the detection of representative numbers of AI and for increasing the chances of detecting rare pollen types. The proportion of NAP is closely related to the openness of the landscape. For comparison, Brun (2011) summarises the indicative value of various pollen taxa taken in present-day open landscape and works with a minimum NAP sum of 400. In the forested area of Šumava, the NAP value per sample ranging from 50 to 180 pollen grains (excluding Cyperaceae).

Evidence of human activity in a local pollen and charcoal record: The Sklářské Valley case study

Because of the low NAP sum, the 60 m distance between the pollen and archaeological sites (too far for specific indicators), and the characteristics and size of the archaeological site, the pollen record from Sklářské Valley yielded mostly non-specific AI. Our case study nevertheless shows that even basic SAI capture activity at an archaeological site, especially when more indicators occur together and in several consecutive samples, and if other changes in vegetation, such as peaks in light-demanding taxa, are detected in the same layer. A pollen signal contemporaneous with the La Tène site can be traced in ca. three samples (red frame in Figure 4). Small peaks of SAI (Plantago l, Rumex and Artemisia) represent minor human impact with no specific character. An increase in these pollen taxa suggests human presence connected with trampling, an increase in nutrients, or small-scale forest clearance. The reduction of the forest is indicated by a decrease in the arboreal pollen sum (AP). Detailed observation revealed that Picea, Abies and Alnus were slightly reduced, while Pinus and Betula increased, also slightly, which suggests early succession in open spaces. Increased light intensity after tree-felling could also explain the high numbers of Melampyrum and small increases in Filipendula, Lysimachia vulgaris type and Potentilla type. All the described changes in vegetation were small scale and minor, which accords with the results of the archaeological research (Dreslerová et al., 2020).

The charcoal record suggests earlier human impact marked by a distinct deposition of charcoal during an archaeologically undetected phase of human occupation dated to 2588 cal yr BP (the Hallstatt period) (Figure 9). A step-like increase in Betula and Melampyrum pollen is indicative of patches of early successional stages in the vegetation. A period of pronounced burning is dated to the Late Bronze Age (–3100 cal yr BP; Figure 9). From the same sequence, a single grain of Avena type and Rumex, a peak in Melampyrum and Calluna, an increase in Pinus and a decrease in Alnus independently indicate local human activity, perhaps similar to that from the La Tène period.

Primary anthropogenic indicators: The question of local farming

Pollen records with discontinuous and isolated finds of PAI during pre-Medieval times suggest no arable farming on the plateaus of central Šumava, although in principle agriculture is possible here; written sources show that at the end of the 19th century cereals were grown with relatively good yields up to an elevation of 1000 m (Zeithammer, 1902). Van der Knaap et al. (2020) documented a rare occurrence of Cerealia type pollen at Rachelsee dated to ca. 3500 cal yr BP and interpreted this as an indication of intensified land use (possibly arable fields) in the valley bottoms. Interestingly, multiple-site pollen records from Šumava in which PAI were occasionally detected during the Neolithic and Bronze Age (zone Z1 and Z2, Figures 6 and 7) show a distinct spatial pattern. PAI are frequent at Stráženská, Rybárenská and Rokytecká slať, which are all located on possible long-distance trade routes (see Figure 1). The evidence points to two possible scenarios: small fields managed by local people alongside other activities (related to trading?), or raw grain or food brought in by people moving along the trade routes. Unfortunately, the explanation of the presence of PAI in other profiles is not so straightforward and provides a significant challenge to archaeological research.

The beginning of Medieval arable farming is marked by the continual presence of PAI in pollen spectra (Figure 8). At Mrtvý luh Chlum Secale is continually present from around 1900 cal yr BP until the present day. The onset of Secale at Březník A is dated to ca. 1400 cal yr BP, and Rybárenská slať registers a peak of Secale from a similar date. Until ca. 1300 cal yr BP, cereal pollen appears sporadically but is relatively frequent in the whole dataset. After 1300 cal yr BP, PAI are increasingly present in all pollen profiles. Interestingly, their appearance precedes the first written evidence of the founding of local villages by an interval of 200–600 years (Sklářské Valley ca. 500 years earlier, Chalupská slať 550 years, Stará Jímka 350 years, Rokytecká slať 600 years and Šmauzy 200 years; Figure 8). This is consistent with the results of Fanta et al. (2020), who concluded that Medieval written sources systematically postdate the establishment of many villages in Bohemia by around 250 years. Pollen records therefore strongly suggest a much earlier history of settlement in central Šumava.

Human activity in the wider region

The most distinctive trend across the pollen sites in Šumava is moderately high NAP (5%) in the Middle Holocene; minimum levels of 2% NAP between ca. 5000 and 3300 cal yr BP, followed by a progressive increase up to a level of 15% in the last 1200 years (Figure 5). We interpret this trajectory as a natural succession towards closed-canopy forests which was fully developed during the Middle/Late-Holocene transition and later affected by human-induced deforestation (Fyfe et al., 2015). Similarly, there is a marked increase in pollen percentage of pioneering trees such as Pinus and Betula, implying extensive forest opening from which these trees benefitted. Starting from ca. 3300 cal yr BP, we observed an increase in the frequency and abundance of AI across the majority of pollen records (Z3, Figure 7). Moreover, available charcoal records reveal several periods of increased fire activity, dating back to the Early and Late Bronze Ages, which is likely to have been anthropogenic burning. Human impact is recognisable at more pollen sites, is often longlasting and is usually stronger, as indicated by more SAI taxa. A similar pattern has been reported from the lake records at Prášilské jezero (Carter et al., 2018b), Rachelsee and Stangenfilz (van der Knaap et al., 2020), suggesting that human land use in Šumava/Bavarian Forest intensified during the Bronze Age (see Figure 2). The suggested climate deterioration at ca. 2800 cal yr BP (Speranza et al., 2003) is likely to have had an impact on human activity in the less favourable mountain environment. However, neither the abundance of SAI nor changes in AP/NAP ratio (i.e. reforestation) indicate any pronounced abandonment of the landscape at this time. This is consistent with observations from NW Europe that climate deterioration during the Bronze/Iron Age transition did not prompt people to abandon upland landscapes (Dark, 2006).

The position of several pollen sites, situated far from the present-day settlement zone and which record human impact, seems to correspond to supposed long-distance access routes connecting Bohemia with the Alpine Foreland. The transport of raw material from Bavaria for use in the chipped stone industry could explain the early SAI occurrences at Rybárenská and Rokytecká slať (Neolithic; Eigner et al., 2017; Vondrovský et al., 2018) and Březník B (Iron Age) located on a ‘smuggling branch’ of the Golden Trail. This followed the route towards Bavaria in the Middle Ages, which suggests the terrain was at least passable. The presence of AI at Mrtvý luh, Mrtvý luh Chlum, Malá niva and Stráženská slať can also be connected to the long-distance access routes known later as the Prachatice and Vimperk branches of the Golden Trail. The finds of the Late Bronze Age daggers at Stožec and Volary (Zavřel et al., 2017) suggest that the Prachatice branch already existed at that time (Figure 1). PAI along these routes could represent either small fields (possibly for the supply of camps on the access route) or the importing of grain from elsewhere.

A comparison with the Bavarian Forest

Human impact in the pollen records from both Šumava and the Bavarian Forest is indicated by the same group of SAI, from which Artemisia is the most common (Stalling, 1987). Human impact in the Bavarian Forest is less evident, especially before the Bronze Age. This is an important observation as settlement on the Bavarian side came closest to the mountains during the Eneolithic (Cham culture 3200-2800 cal yr BP; Figure 2; (Sommer, 2006), and it suggests limited pollen transport from the lowlands and the local character of the pollen records. The general trend of an increase in Pinus and Betula from the Bronze Age onwards is not registered in Bavaria. Finds of cereal pollen are extremely rare here before the Subatlanticum (Firbas, 1949). Bavarian pollen data also suggest that lower altitudes for pollen sites do not necessarily mean stronger human impact. In Kulz (Oberes Moss) at 483 m, SAI are infrequent before the Subatlanticum (Firbas, 1949), and in Haidmühle at 835 m the continuous presence of Artemisia and Plantago l is registered from the Bronze Age (Stalling, 1987). A general increase in human impact pollen signal is registered at the end of the Subboreal and the beginning of the Subatlanticum (Late Bronze and Iron Ages). A large increase in Calluna in the youngest sections, representing the period following Medieval colonisation, is usually detected in pollen sites below 750 m and is indicative of forest pasturing and soil degradation (Behre, 1981; Gaillard et al., 1994). This feature is not registered in Šumava, probably because of the absence of pollen sites at such altitudes.

Anthropogenic use of fire

Trends in regional fire activity complement the pollen record and reveal several periods of increased biomass burning (Figure 9). The charcoal composite curve clearly indicates the occurrence of fires dating back to the Bronze Age, namely in the periods 3300–3000 and 4000–3700 cal yr BP. This pattern contradicts the known fire history reconstructions from mountainous areas in Bohemia (Bobek et al., 2018, 2019; Carter et al., 2018b), which show almost no fire activity from –6200 cal yr BP up to the Early Medieval. As the occurrence of fire is driven mainly by the high proportion of conifers in the vegetation (Feurdean et al., 2020), we can reasonably expect low fire activity in the periods when the zonal vegetation was dominated by broadleaved trees. Contrary to this expectation, we found pronounced fire activity in periods with abundant Fagus, which could indicate that such activity was driven by humans. People might have set fires for a variety of purposes. Pastoral fires are commonly used in mountainous areas of Europe where transhumance systems have developed. The intentional use of fire to maintain high-altitude pastures is clearly documented in the nearby region of the Alps from at least the Bronze Age (Gilck and Poschlod, 2019; Gobet et al., 2003) or even as early as the Neolithic (Dietre et al., 2020; Hafner and Schwörer, 2018; Pini et al., 2017). Unlike the Alps, the absence of a treeless zone in Šumava makes such activity less likely or at least limited to open vegetation on peat bogs. Remarkably, we did not find reliable pollen or spore evidence for pastoralism in Šumava, although periods of increased charcoal deposition indicate anthropogenic burning. There might therefore have been other reasons for the use of fire, such as clearing the forest to improve success in hunting or trapping animals. Such a practice is unlikely to be detected by means of AI as it is not linked to the creation of unique vegetation types. However, a noticeable correspondence between the peak in fire activity around 3200 cal BP in the remote mountain area of Šumava and the Late Bronze Age peak of population density in Bohemia (Demján and Dreslerová, 2016) suggests that hunting or trapping might have been of some importance. At the time, there was a great need for luxury goods and raw materials such as salt and copper. These items and materials came to Bohemia via the south and west of the region, usually across Šumava (Jiráň, 2013). It is not clear which commodities were used as a means of payment, but furs and antlers (or buckhorn?) must have been among them. The hunting of bear, beaver, lynx, marten and other fur-bearing animals abundant in the primieval Šumava forest might have benefitted from a ‘shorter’ transportation route towards the Danubian region and the Mediteranean, from which imports came northwards. It would therefore make sense that the main hunting grounds were along the access routes.

Human impact in other mid-mountains

In the Czech mountains, an increase in human impact from the Bronze Age and onwards into the Iron Age and the High and post-Medieval periods is known from the pollen record of the Lusatian Mountains, although here the archaeological evidence around the single pollen site, situated at 400 m, is poor (Kozáková et al., 2015). In the highest regions of the Jeseníky Mountains, pollen and charcoal data suggest the formation of high mountain pastures in the Iron Age (Dudová et al., 2018; Novák et al., 2010), but these data are associated with treeless or parkland subalpine vegetation on summits, a landscape which more closely resembles the Alps than it does Šumava. The subalpine zone is also developed in the Krkonoše, where two pollen sequences show human impact only from the Early Medieval period, ca. 1400 cal yr BP (Speranza et al., 2000) or from ca. 1000 cal yr BP (Malkiewicz et al., 2016). These cases are thought to represent the transport of pollen grains by airstreams passing over the summits. This is clearly confirmed by finds of exotic pollen types in the Jeseníky samples (Dudová et al., 2018). The possibility of the extra-local origin of AI in high-elevation records must therefore be admitted, and this substantially limits the possibility of identifying small-scale human impact.

There are a number of similarities between Šumava/Bavarian Forest and the southern Black Forest: granite and gneiss bedrock, altitude, wooded peaks, the existence of the Rhine lowland (like the Danube lowland) just below the mountains, and scant pre-Medieval archaeological evidence in the mountains. In the Black Forest, archaeological finds suggest human activity in the Neolithic and the La Tène period (Fingerlin, 2006; Humpert, 1991; Schmid, 1992). Detailed pollen records (Elzhof at 940 m and Steerenmoos at 1000 m) show significant human impact as early as the Neolithic (Henkner et al., 2018; Rösch, 2000). In addition to cereal pollen, SAI are much in evidence. Plantago l occurs continuously from the Bronze Age; other common SAI taxa are more abundant than in the Šumava/Bavarian Forest. Rösch (2000) believes the long-distance transportation of AI from the lowlands to be impossible because of the small area of the pollen sites and the degree to which the pollen transport of herbs is limited in the wooded landscape. Local human impact also confirms the occurrence of colluvial deposits dated to the younger Neolithic (Henkner et al., 2018). Cereal pollen appears and SAI increase in the pollen records in the Bronze and Iron Ages (Henkner et al., 2018; Rösch, 2000). Pre-Medieval human activity in the Black Forest is interpreted as small-scale seasonal settlement dominated by pastoralism and the exploitation of wood and bedrock and also associated with trade routes (Henkner et al., 2018). The considerably lower frequency of cereal pollen and SAI (especially Plantago l) in the pre-Medieval period in Šumava/Bavarian Forest represents the most distinct difference between the pollen records of the two regions. Farming, which some researchers assume to have taken place in the Black Forest since the Neolithic (e.g. Rösch, 2000), may have been made possible by a more oceanic climate in this region. However, it is not currently possible to offer a confident explanation about differences in the nature and intensity of human impact between the two regions because of a lack of proxy and archaeological data.

Conclusions

In the forested mountains of Šumava, prehistoric human impact is reflected in pollen records by the most common anthropogenic indicators: Artemisia, Calluna vulgaris, Plantago lanceolata, Plantago major/media, Rumex acetosa type, Urtica, Chenopodiaceae, Secale cereale and Cerealia type. Apart from the cereals, these taxa have no specific indicative value.

All the secondary anthropogenic indicators typical of pollen sites in Šumava are registered since the early Holocene and therefore belong to the original vegetation. The increase in these indicators can be explained by disturbances, either anthropogenic or natural. With sparse archaeological evidence, it is not possible, from a single pollen record, to make a reliable connection between human activity and anthropogenic pollen indicators; a more robust set of data is required.

We detect an increase in NAP, Betula, Pinus and secondary anthropogenic indicators in pollen records from ca. 3300 cal yr BP. Plantago lanceolata and Chenopodiaceae are much more frequent from this time onwards. These changes in vegetation may be a consequence of increased human activity associated with a higher population density in other parts of the country (Demján-Dreslerová, 2016); it is not reasonable to assume a sudden series of simultaneous small-scale natural disturbances (such as storms) across the whole of the Šumava region.

Increased regional fire activity during the Bronze Age indicates the frequent antropogenic use of fires. We did not confirm the use of fire as a tool linked to pastoral activity. Fire was used rather for small-scale forest clearing in connection with hunting and the maintenance of access routes.

Human activity in Šumava during prehistory caused small-scale disturbances of the vegetation but did not affect the composition of natural species in the forest. The disturbances nonetheless supported the growth of Betula and Pinus and the survival and abundance of herbs. In this sense, prehistoric human activity clearly had a positive effect on the diversity of the vegetation.

In comparison to some other Šumava pollen sites, human impact in the La Tène period in the Sklářské Valley is inconspicuous. With no archaeological data, and ignoring a single Secale pollen grain, changes in vegetation in the Sklářské Valley could be explained by small-scale tree-fall during or after a storm. This explanation cannot be applied to the whole data, which suggests a trend of increasing human impact in the late Bronze and Iron Ages. We therefore believe that pollen records, particularly those with a significant human impact signal between 3000 and 2000 cal yr BP, reflect local human activity and present a great challenge to archaeological prospection.

Supplemental Material

Supplementary – Supplemental material for The prehistory and early history of the Šumava Mountains (Czech Republic) as seen through anthropogenic pollen indicators and charcoal data

Supplemental material, Supplementary for The prehistory and early history of the Šumava Mountains (Czech Republic) as seen through anthropogenic pollen indicators and charcoal data by Radka Kozáková, Přemysl Bobek, Dagmar Dreslerová, Vojtěch Abraham and Helena Svobodová-Svitavská in The Holocene

Footnotes

Acknowledgements

We are grateful to Čeněk Čišecký for creating the maps and we thank Tim Morgan for language corrections.

Data availability

The data are available at .

The R code used in this study is available at

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work reported here was funded by the Czech Science Foundation (GACR), grant no. 17-17909S ‘Hidden human prehistoric activities in the mountains. Archaeological and pollen evidence from the Šumava Mountains’. Authors affiliated to the Institute of Botany were further supported by long-term research development project RVO 67985939 (Czech Academy of Sciences).

ORCID iDs

Přemysl Bobek

Vojtěch Abraham

Supplemental material

Supplemental material for this article is available online.

References

Behre

K-E

(1981) The Interpretation of Anthropogenic Indicators in Pollen Diagrams. Pollen et Spores XXIII(2): 225–245.

Behre

K-E

Kucan

(1986) Die Reflektion archäologisch bekannter Siedlungen in Pollendiagrammen verschiedener Entfernung—Beispiele aus der Siedlungskammer Flögeln, Nordwestdeutschland. In: Behre

K-E

(ed.) Anthropogenic Indicators in Pollen Diagrams. Rotterdam, Boston: A.A. Balkema, pp. 95–114.

Beug

(2004) Leitfaden Der Pollenbestimmung Für Mitteleuropa Und Angrenzende Gebiete. München: Verlag Dr. Friedrich Pfeil.

Blaauw

(2010) Methods and code for ’classical’age-modelling of radiocarbon sequences. Quaternary Geochronology 5(5): 512–518.

Blaauw

Christeny

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6(3): 457–474.

Blarquez

Vannière

Marlon

, et al. (2014) Paleofire: An R package to analyse sedimentary charcoal records from the Global Charcoal Database to reconstruct past biomass burning. Computers and Geosciences 72: 255–261.

Bobek

Šamonil

Jamrichová

(2018) Biotic controls on Holocene fire frequency in a temperate mountain forest, Czech Republic. Journal of Quaternary Science 33(8): 892–904.

Bobek

Svobodová-Svitavská

Pokorný

, et al. (2019) Divergent fire history trajectories in Central European temperate forests revealed a pronounced influence of broadleaved trees on fire dynamics. Quaternary Science Reviews 222: 105865.

Broström

Gaillard

Ihse

, et al. (1998) Pollen-landscape relationships in modern analogues of ancient cultural landscapes in southern Sweden—a first step towards quantification of vegetation openness in the past. Vegetation History and Archaeobotany 7(4): 189–201.

10.

Brun

(2011) Anthropogenic indicators in pollen diagrams in eastern France: A critical review. Vegetation History and Archaeobotany 20(2): 135–142.

11.

Carter

Chiverrell

Clear

, et al. (2018a) Quantitative palynology informing conservation ecology in the bohemian/bavarian forests of central europe. Frontiers in Plant Science 8: 2268.

12.

Carter

Moravcová

Chiverrell

, et al. (2018b) Holocene-scale fire dynamics of central European temperate spruce-beech forests. Quaternary Science Reviews 191: 15–30.

13.

Chytrý

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

14.

Connor

Vannière

Colombaroli

, et al. (2019) Humans take control of fire-driven diversity changes in Mediterranean Iberia’s vegetation during the mid–late Holocene. The Holocene 29(5): 886–901.

15.

Court-Picon

Buttler

de Beaulieu

J-L

(2006) Modern pollen/vegetation/land-use relationships in mountain environments: An example from the Champsaur valley (French Alps). Vegetation History and Archaeobotany 15(3): 151–168.

16.

Culek

(ed.) (1996) Biogeografické Členění České Republiky. Praha: Enigma.

17.

Dark

(2006) Climate deterioration and land-use change in the first millennium BC: Perspectives from the British palynological record. Journal of Archaeological Science 33(10): 1381–1395.

18.

Demján

Dreslerová

(2016) Modelling distribution of archaeological settlement evidence based on heterogeneous spatial and temporal data. Journal of Archaeological Science 69: 100–109.

19.

Deza-Araujo

Morales-Molino

Tinner

, et al. (2020) A critical assessment of human-impact indices based on anthropogenic pollen indicators. Quaternary Science Reviews 236: 106291.

20.

Dietre

Walser

Lambers

, et al. (2014) Palaeoecological evidence for Mesolithic to Medieval climatic change and anthropogenic impact on the Alpine flora and vegetation of the Silvretta Massif (Switzerland/Austria). Quaternary International 353: 3–16.

21.

Dietre

Reitmaier

Walser

, et al. (2020) Steady transformation of primeval forest into subalpine pasture during the Late Neolithic to Early Bronze Age (2300−1700 BC) in the Silvretta Alps, Switzerland. The Holocene 30(3): 355–368.

22.

Dietze

Theuerkauf

Bloom

, et al. (2018) Holocene fire activity during low-natural flammability periods reveals scale-dependent cultural human-fire relationships in Europe. Quaternary Science Reviews 201: 44–56.

23.

Dreslerová

(2015) Pravěká transhumance a salašnické pastevectví na území České republiky: možnosti a pochybnosti Dagmar Dreslerova [Prehistoric transhumance and summer farming in the Czech Republic: Possibilities and doubts]. Archeologicke Rozhledy 67: 109–130.

24.

Dreslerová

Kozáková

Metlička

, et al. (2020) Seeking the meaning of a unique mountain site through a multidisciplinary approach. The Late La Tène site at Sklářské Valley, Šumava Mountains, Czech Republic. Quaternary International 542: 88–108.

25.

Dudová

Hájek

Petr

, et al. (2018) Holocene vegetation history of the Jeseníky Mts: Deepening elevational contrast in pollen assemblages since late prehistory. Journal of Vegetation Science 29(3): 371–381.

26.

Eigner

Kapustka

Parkman

, et al. (2017) Mezolitické osídlení Šumavy pohledem studia surovin kamenných artefaktů z lokalit Javoří Pila 1 a Nová Pec [Mesolithic occupation of the Bohemian Forest according to the study of raw materials for lithic production from the sites of Javoří Pila 1 and Nová. Silva Gabreta 23: 33–44.

27.

Faegri

Iversen

Kaland

, et al. (1989) Textbook of Pollen Analysis. 4th Ed. Chichester: John Wiley and Sons.

28.

Fanta

Zouhar

Beneš

, et al. (2020) How old are the towns and villages in Central Europe? Archaeological data reveal the size of bias in dating obtained from traditional historical sources. Journal of Archaeological Science 113: 105044.

29.

Feurdean

Gałka

Tanţău

, et al. (2016) Tree and timberline shifts in the northern Romanian Carpathians during the Holocene and the responses to environmental changes. Quaternary Science Reviews 134: 100–113.

30.

Feurdean

Vannière

Finsinger

, et al. (2020) Fire risk modulation by long-term dynamics in land cover and dominant forest type in Eastern and Central Europe. Biogeosciences 17: 1213–1230.

31.

Fingerlin

(2006) Vom Oberrhein zur jungen Donau: Die Straße durch den südlichen Schwarzwald in keltischer, römischer und frühalamannischer Zeit. Archäeologische Nachrichten aus Baden 72/73: 62–73.

32.

Firbas

(1949) Spät- Und Nacheiszeitliche Waldgeschichte Mitteleuropas Nördlich Der Alpen, Erstrer Band: Allgemeine Waldgeschichte. Jena: Verlag von Gustav Fischer in Jena.

33.

Fyfe

Woodbridge

Roberts

(2015) From forest to farmland: Pollen-inferred land cover change across Europe using the pseudobiomization approach. Global Change Biology 21(3): 1197–1212.

34.

Gaillard

M-J

Birks

HJB

Emanuelsson

, et al. (1994) Application of modern pollen/land-use relationships to the interpretation of pollen diagrams—reconstructions of land-use history in south Sweden, 3000-0 BP. Review of Palaeobotany and Palynology 82(1–2): 47–73.

35.

Gilck

Poschlod

(2019) The origin of alpine farming: A review of archaeological, linguistic and archaeobotanical studies in the Alps. Holocene 29(9): 1503–1511.

36.

Gobet

Tinner

Hochuli

, et al. (2003) Middle to Late-Holocene vegetation history of the Upper Engadine (Swiss Alps): The role of man and fire. Vegetation History and Archaeobotany 12(3): 143–163.

37.

Grimm

(2004) TILIA and TGView. Springfield: Illinois State Museum, Research and Collection Center.

38.

Hafner

Schwörer

(2018) Vertical mobility around the high-alpine Schnidejoch Pass. Indications of Neolithic and bronze age pastoralism in the Swiss Alps from paleoecological and archaeological sources. Quaternary International 484: 3–18.

39.

Henkner

Ahlrichs

Fischer

, et al. (2018) Land use dynamics derived from colluvial deposits and bogs in the Black Forest, Germany. Journal of Plant Nutrition and Soil Science 181(2): 240–260.

40.

Houfková

Horák

Bešta

, et al. (2019) The dynamics of a non-forested stand in the Krušné Mts.: The effect of a short-lived medieval village on the local environment Article. Vegetation History and Archaeobotany 28: 607–621.

41.

Humpert

(1991) Eine römische Straße durch den südlichen Schwarzwald. Archäeologische Nachrichten aus Baden 45: 19–32.

42.

Jankovská

(2006) Late Glacial and Holocene history of Plešné Lake and its surrounding landscape based on pollen and palaeoalgological analyses. Biologia 61(S20): S371–S385.

43.

Jiráň

(ed.) (2013) Prehistory of Bohemia 4. The Bronze Age. Prague: Institute of Archaeology, Czech Academy of Sciences.

44.

Jiráň

Venclová

(eds) (2013) The Prehistory of Bohemia 2-7. Prague: Institute of Archaeology, Czech Academy of Sciences.

45.

Juggins

(2017) rioja: Analysis of quaternary science data, R package. 0.9-21.

46.

Kapustka

Walls

Eigner

(2019) Beginnings of mountain settlement in Czech Republic - a case study from the Bohemian Forest. In: Pelisiak

Nowak

Astaloş

(eds) People in the Mountains – Current Approaches to the Archaeology of Mountainous Landscapes. Oxford: Archaeopress Archaeology, pp. 185–195.

47.

Kozáková

Pokorný

Peša

, et al. (2015) Prehistoric human impact in the mountains of Bohemia. Do pollen and archaeological data support the traditional scenario of a prehistoric “wilderness”? Review of Palaeobotany and Palynology 220: 29–43.

48.

Kuneš

Abraham

Kovářík

, et al. (2009) Czech Quaternary palynological database – PALYCZ. Preslia 81(3): 209–238.

49.

Malkiewicz

Waroszewski

Bojko

, et al. (2016) Holocene vegetation history and soil development reflected in the lake sediments of the Karkonosze Mountains (Poland). The Holocene 26(6): 890–905.

50.

Marlon

Bartlein

Carcaillet

, et al. (2008) Climate and human influences on global biomass burning over the past two millennia. Nature Geoscience 1(10): 697–702.

51.

Miola

(2012) Tools for non-pollen palynomorphs (NPPs) analysis: A list of Quaternary NPP types and reference literature in English language (1972-2011). Review of Palaeobotany and Palynology 186: 142–161.

52.

Novák

Petr

Treml

(2010) Late-Holocene human-induced changes to the extent of alpine areas in the East Sudetes, Central Europe. The Holocene 20(6): 895–905.

53.

Odgaard

(2013) Reconstructing past biodiversity development. In: Elias

(ed.) The Encyclopedia of Quaternary Science. Oxford: Elsevier, pp. 816–820.

54.

Pals

Van Geel

Delfos

(1980) Paleoecological studies in the Klokkeweel bog near hoogkarspel (prov. of Noord-Holland). Review of Palaeobotany and Palynology 30: 371–418.

55.

Pini

Ravazzi

Raiteri

, et al. (2017) From pristine forests to high-altitude pastures: An ecological approach to prehistoric human impact on vegetation and landscapes in the western Italian Alps. Journal of Ecology 105(6): 1580–1597.

56.

Power

Marlon

Bartlein

, et al. (2010) Fire history and the global charcoal database: A new tool for hypothesis testing and data exploration. Palaeogeography, Palaeoclimatology, Palaeoecology 291(1–2): 52–59.

57.

Punt

(ed.) (1976) The Northwest European Pollen Flora I. Amsterdam-New York-Oxford: Elsevier.

58.

Punt

Blackmore

(eds) (1991) The Northwest European Pollen Flora VI. Amsterdam-New York-Oxford-Tokyo: Elsevier.

59.

Punt

Clarke

GCS

(eds) (1980) The Northwest European Pollen Flora II. Amsterdam-New York-Oxford: Elsevier.

60.

Punt

Clarke

GCS

(eds) (1981) The Northwest European Pollen Flora III. Amsterdam-New York-Oxford: Elsevier.

61.

Punt

Clarke

GCS

(eds) (1984) The Northwest European Pollen Flora IV. Amsterdam-New York-Oxford-Tokyo: Elsevier.

62.

Punt

Hoen

(2009) The Northwest European Pollen Flora, 70. Review of Palaeobotany and Palynology 157(1–2): 22–183.

63.

Punt

Blackmore

Clarke

GCS

(eds) (1988) The Northwest European Pollen Flora V. Amsterdam-Oxford-New York-Tokyo: Elsevier.

64.

Punt

Blackmore

Hoen

(eds) (1995) The Northwest European Pollen Flora VII. Amsterdam; Lausanne; New York: Elsevier Science B.V.

65.

R Development Core Team (2017) R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

66.

Ramsey

(2009) Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360.

67.

Reille

(1992) Pollen et Spores d’Europe et d’Afrique Du Nord. Marseille: Laboratoire de Botanique historique et Palynologie.

68.

Reimer

Bard

Bayliss

, et al. (2013) IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP. Radiocarbon 55(04): 1869–1887.

69.

Rius

Vannière

Galop

, et al. (2011) Holocene fire regime changes from multiple-site sedimentary charcoal analyses in the Lourdes basin (Pyrenees, France). Quaternary Science Reviews 30(13–14): 1696–1709.

70.

Robin

Bork

H-R

Nadeau

M-J

, et al. (2013) Fire and forest history of central European low mountain forest sites based on soil charcoal analysis: The case of the eastern Harz. The Holocene 24(1): 35–47.

71.

Rösch

(2000) Long-term human impact as registered in an upland pollen profile from the southern Black Forest, south-western Germany. Vegetation History and Archaeobotany 9(4): 205–218.

72.

Rösch

Stojakowits

Friedmann

(2020) Does site elevation determine the start and intensity of human impact? Pollen evidence from southern Germany. Vegetation History and Archaeobotany. DOI: 10.1007/s00334-020-00780-4.

73.

Rüther

Nelle

(2006) Waldgeschichte, Waldnutzungen und Waldvegetation des Vorderen Bayerischen Waldes – ein multimethodischer Ansatz zum regionalen und lokalen Landschaftswandel seit der letzten Eiszeit. Deggendorfer Geschichtsblätter 28: 43–78.

74.

Schwörer

Colombaroli

Kaltenrieder

, et al. (2015) Early human impact (5000-3000 BC) affects mountain forest dynamics in the Alps. Journal of Ecology 103(2): 281–295.

75.

Seddon

AWR

Macias-Fauria

Willis

(2015) Climate and abrupt vegetation change in Northern Europe since the last deglaciation. Holocene 25(1): 25–36.

76.

Schmid

(1992) Die Urgeschichtlichen Funde Und Fundstellen Der Baar: Eine Auswertung Des Bestandes. Band 2: Katalog. Rheinfelden: Schäuble Verlag.

77.

Sommer

(ed.) (2006) Archäologie in Bayern - Fenster Zur Vergangenheit. Regensburg: Verlag Friedrich Pustet.

78.

Soukupová

(1996) Developmental diversity of peatlands in the Biosphere reserve Šumava. Silva Gabreta 1: 99–107.

79.

Speranza

Hanke

van Geel

, et al. (2000) Late-Holocene human impact and peat development in the Černá Hora bog, Krkonoše Mountains, Czech Republic. The Holocene 10(5): 575–585.

80.

Speranza

Van Geel

Van der Plicht

(2003) Evidence for solar forcing of climate change at ca. 850 cal BC from a Czech peat sequence. Global and Planetary Change 35(1–2): 51–65.

81.

Stalling

(1987) Untersunchungen zur spät-und postglazialen Vegetationsgeschichte im Bayerischen Wald. Göttingen : Georg-August-Universität.

82.

Stopp

(2015) Animal husbandry and hunting activities in the Late Bronze Age Circum-Alpine Region. In: Menotti

Francesco

(ed.) The End of the Lake-Dwellings in the Circum-Alpine Region. Oxford: Oxbow Books, pp. 179–210.

83.

Svobodová

Reille

Goeury

(2001) Past vegetation dynamics of Vltavský luh, upper Vltava river valley in the Šumava mountains. Czech Republic. Vegetation History and Archaeobotany 10(4): 185–199.

84.

Svobodová

Soukupová

Reille

(2002) Diversified development of mountain mires, Bohemian Forest, Central Europe, in the last 13,000 years. Quaternary International 91(1): 123–135.

85.

van der Knaap

van Leeuwen

JFN

Fahse

, et al. (2020) Vegetation and disturbance history of the Bavarian Forest National Park, Germany. Vegetation History and Archaeobotany 29(2): 277–295.

86.

van Geel

(1978) A palaeoecological study of Holocene peat bog sections in Germany and The Netherlands, based on the analysis of pollen, spores and macro- and microscopic remains of fungi, algae, cormophytes and animals. Review of Palaeobotany and Palynology 25(1): 1–120.

87.

van Geel

Klink

Pals

, et al. (1986) An upper Eemian lake deposit from Twente, eastern Netherlands. Review of Palaeobotany and Palynology 47(1–2): 31–61.

88.

van Geel

Coope

Van Der Hammen

(1989) Palaeoecology and stratigraphy of the lateglacial type section at Usselo (the Netherlands). Review of Palaeobotany and Palynology 60(1–2): 25–129.

89.

Vincze

Orbán

Birks

, et al. (2017) Holocene treeline and timberline changes in the South Carpathians (Romania): Climatic and anthropogenic drivers on the southern slopes of the Retezat Mountains. The Holocene 27(11): 1613–1630.

90.

Vondrovský

Bumerl

Šída

, et al. (2018) Nové poznatky o neolitickém osídlení jižních Čech [New information on the Neolithic occupation of south Bohemia]. Archeologicke Rozhledy 70(2): 147–194.

91.

Wacnik

Nalepka

Granoszewski

, et al. (2015) Development of modern forest zones in the Beskid Niski Mts. and adjacent area ( Western Carpathians ) in the late-Holocene : A palaeobotanical perspective. Quaternary International 415: 303–324.

92.

Zavřel

Anděra

(eds) (2003) Šumava – Příroda, Historie, Život. Praha: Baset.

93.

Zavřel

Kubů

(2007a) Zlatá Stezka. Historický a Archeologický Výzkum Středověké Obchodní Cesty 2. Úsek Vimperk - Státní Hranice. České Budějovice: Jihočeské muzeum v Českých Budějovicích.

94.

Zavřel

Kubů

(2007b) Zlatá Stezka. Historický a Archeologický Výzkum Významné Středověké Obchodní Cesty. 1. Úsek Prachatice – Státní Hranice. České Budějovice: Jihočeské muzeum v Českých Budějovicích.

95.

Zavřel

Chvojka

John

(2017) Otázka existence pravěkých stezek v jihočeské části Šumavy. Praehistorica 33(1): 479–485.

96.

Zeithammer

(1902) Šumava, Kraj a Lid [Bohemian Forest, Landscape and People]. České Budějovice: Self-published.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.20 MB