Introduction
The spread of farming from centres of domestication into new biogeographic zones may have led to a loss of the original crop diversity. While the first southwest Asian farmers cultivated thirteen crop species (Zohary et al. Reference Zohary, Hopf and Weiss2012), the Linearbandkeramik (LBK, 5550–4900 BC) communities in Central Europe relied on five staple crops: two hulled wheats—einkorn (Triticum monococcum) and emmer (Triticum dicoccum), two legumes—lentil (Lens culinaris) and common pea (Pisum sativum), and one fibre/oil plant—flax (Linum usitatissimum) (Kreuz et al. Reference Kreuz, Marinova, Schäfer and Wiethold2005; Colledge and Conolly Reference Colledge and Conolly2007; Conolly et al. Reference Conolly, Colledge and Shennan2008; Kreuz and Marinova Reference Kreuz and Marinova2017; Ivanova Reference Ivanova, Gron, Sørensen and Rowley-Conwy2020). The uniformity of Central European Early Neolithic farming may not be surprising if considered in the sociocultural context. The LBK communities are well recognized for their cultural conservatism, characterized by a high degree of uniformity in artefact production, architecture, and economic practices, especially in the early phase, which played a significant role in their swift and widespread expansion across a large territory (Shennan Reference Shennan2018). However, it is crucial to consider the ecological context when examining archaeobotanical data. Since the LBK populations predominantly settled in fertile lowlands of Central Europe (Brigand et al. Reference Brigand, Dubouloz and Weller2022), the limited assortment might be a result of a deliberate selection of crops being well adapted to given conditions. Here we address the question of whether the uniform character of LBK agriculture signifies only a cultural homogenisation or whether it rather represents a deliberate selection of crops well adapted to conditions of fertile lowlands.
To test this hypothesis, we need to study the early farming implementation in less productive regions of Central Europe, mainly in those with ecologically demanding conditions for crop cultivation. South Bohemia (Czechia) provides an ideal study setting because of a diverse landscape mosaic with elevations of above 350 metres a.s.l., less favourable soils (mostly cambic and pseudogley; Chábera et al. Reference Chábera, Demek, Hlavác, Kríž, Malecha, Novák, Odehnal, Suk, Tomášek and Zuska1985), cooler climate (when compared to neighbouring lowlands), strong seasonality of temperatures, and rather moist summers (Supplementary Text 1). In the Early Neolithic, the region stood out as a distinctive (inner) periphery of the LBK settlement area with significantly less intensive occupation (Figure 1) and delayed colonization. While the initial wave of LBK diaspora reached the surrounding lowlands in the 54th century BC (Jakucs et al. Reference Jakucs, Bánffy, Oross, Voicsek, Bronk Ramsey, Dunbar, Kromer, Bayliss, Hofmann, Marshall and Whittle2016; Shennan Reference Shennan2018), it was postponed here ca. 150 years (Vondrovský et al. Reference Vondrovský, Ptáková, Šída, Bumerl, Pták, Kovačiková, Prach, Novák and Budilováin press). The settlements here were far apart and isolated from the surrounding lowland zones (see Figure 1). Towards the end of the LBK period, there was a steep decline in the already sparse occupation, with very few settlements persisting beyond 5000 BC when the region witnessed the emergence of the Middle Neolithic Stichbandkeramik culture (SBK, 5000–4500 BC).
The settlement context of the archaeobotanical record from South Bohemia. (a) kernel density estimation of LBK settlement sites. (b) LBK and SBK sites in the study area. Green dots represent archaeobotanically investigated sites.
Our study bridges a knowledge gap caused by a scarcity of archaeobotanical data from peripheral regions of the Central European LBK settlement zone (Vondrovský and Chvojka Reference Vondrovský and Chvojka2021). Before our project, hardly any archaeobotanical and radiocarbon data on the Neolithic in the region were available. The situation has changed through our numerous archaeological prospection surveys followed by targeted small-scale excavations (Supplementary Text 2). Given the dry conditions of the sites where only charred plant remains are preserved, a low density of archaeological plant material was expected. To overcome the obstacle of obtaining only a limited and/or unreliable dataset we applied an extremely intensive archaeobotanical sampling strategy.
Methods
Excavations and sampling
We chose small-scale excavation with very intensive sampling and detailed recording of spatial and contextual data for all artefacts and samples. The chronology of each archaeological feature was determined based on material culture and/or radiocarbon dating of plant macrofossils. Features were dated and classified to the LBK, SBK, and Neolithic. The latter category includes either LBK or SBK cultures, without the possibility of further differentiation. Detailed descriptions of the sites are provided in Supplementary Text 2.
In Radčice 1 and Mažice sites, features were sampled totally, meaning that 100 % of their infill was taken for archaeobotanical analysis. The samples (analytical units) were collected in a 50 × 50 cm grid and 5 cm arbitrary spits. After the evaluation of the approach’s effectiveness, the strategy was modified. At Radčice 2, Dehtáře, and Horní Bukovsko 10 L of sediment were collected from each 10 cm thick layer within the same grid. An exception is the Horusice site excavated as a rescue campaign triggered by motorway construction. Samples of 10 to 20 L were collected in a 1 × 1 m grid and 10 cm thick spits.
Recovery from deposits, taxa identification, and counting of the plant remains
A total of 2162 samples (25,572 L of sediment) were processed using a flotation machine, with a 0.25 mm mesh for floating remains and a 1 mm mesh to capture heavy fractions (cf. Pearsall Reference Pearsall2015). Both, light and heavy fractions yielded charred plant remains. All were sorted and identified using low-power stereomicroscopes, identification keys, and reference collections available at the Laboratory of Archaeobotany and Palaeoecology at the University of South Bohemia and the Department of Archaeology at Constantine the Philosopher University in Nitra. M. Ptáková, V. Komárková, T. Šálková, and M. Hajnalová carried out the identification of plant macrofossils. Images were taken with a Keyence VHX 7000 digital microscope.
Due to suboptimal preservation conditions, each cereal fragment (grain or chaff) was attributed a numerical value of one. Determined taxa denoted as “cf.” were merged with the corresponding taxonomic categories for further calculations or statistical evaluation. The relative importance of individual crops is evaluated by considering their abundance (number of remains per analytical unit) and frequency (percentage of analytical units where they are found).
Radiocarbon dating of the plant remains
In total, 22 charred plant macrofossils were dated by the AMS method in the Czech Radiocarbon Laboratory (CRL) and Poznan Radiocarbon Laboratory (Poz). Samples were pretreated with an acid-alkali-acid method, followed by combustion and graphitisation. The resulting dates were corrected for isotopic fractionation using δ13C and calibrated in OxCal v.4.4 software using the IntCal20 calibration curve (Bronk Ramsey Reference Bronk Ramsey2009; Reimer et al. Reference Reimer, Austin, Bard, Bayliss, Blackwell and Bronk Ramsey2020).
Results and discussion
We secured an assemblage of 2162 flotation samples, i.e., 25,572 L of archaeological sediment from six Early and Early-to-Middle Neolithic settlement sites and resulted in the recovery of 8605 carbonized determined plant macrofossils. Of these, 92.4% (n = 7955) are remains of domesticated plants (Supplementary Tables 1 and 2).
Diversity in the archaeobotanical crop dataset from 6 Neolithic sites in South Bohemia (see Supplementary Table 1 for further details on the dataset). a, determined crop taxa, their quantity and frequencies in samples
Overview of samples dated by the radiocarbon (AMS) method. Results were calibrated in OxCal v.4.4 software using the IntCal20 calibration curve (Bronk Ramsey Reference Bronk Ramsey2009; Reimer et al. Reference Reimer, Austin, Bard, Bayliss, Blackwell and Bronk Ramsey2020)
The recovered plant macrofossils are extremely fragmented and distorted due to the high charring temperatures. Most finds represent undeterminable fragments of cereal grains (Table 1). Among identified specimens, the majority belong to einkorn (Triticum monococcum) and emmer (Triticum dicoccum), occurring in similar abundances but are also accompanied by barley (Hordeum vulgare) and “new type” wheat Triticum timopheevii. Additional crop diversity is provided by lentil (Lens culinaris), pea (Pisum sativum), and flax (Linum usitatissimum) (Table 1). Cereal spectra are the same for LBK and SBK, only the proportions of individual crops vary slightly. Due to the state of preservation differentiation between hulled (Hordeum vulgare var. vulgare) and naked (Hordeum vulgare var. nudum) forms of barley was problematic; nevertheless, both are present, although the naked variety is sporadic (Figures 2 and 3).
Relative quantities of determined cereal taxa for the LBK and SBK contexts (left); relative quantities of determined cereal taxa for the individual sites (right).
Archaeobotanical remains and radiocarbon dating. (a) Multiplot of calibrated radiocarbon measurements. Barley grains (red), Timopheev’s wheat grains (green). Calibrated using OxCal v.4.4 software and the IntCal20 calibration curve (Table 2). (b) charred grain of Hordeum vulgare var. vulgare. (c) charred grain of Hordeum vulgare var. nudum (d) charred grain of Triticum timopheevii. (e) charred spikelet fork of Triticum timopheevii.
The plant remains come from the fills of sunken features (representing secondary or tertiary contexts) and must have been charred elsewhere. The assemblage exhibits a predominance of cereal grains, the extremely low density of finds with an average value of 0.3 items (0.7 after excluding negative samples) per liter of archaeological deposit, and the prevalence of weed seeds associated with the final stages of crop processing (Hillman Reference Hillman, van Zeist and Caspaire1984; Jones Reference Jones, van Zeist and Caspaire1984). Chaff and weed seeds eliminated in early processing stages are extremely rare. Although they are less likely to survive charring (Boardman and Jones Reference Boardman and Jones1990) and/or redeposition, they might have followed different taphonomic pathways and did not enter archaeological deposits in charred form (Fuller et al. Reference Fuller, Stevens, McClatchie, Madella, Lancelotti and Savard2014)
Twenty-two well-preserved cereal grains were radiocarbon dated by the AMS radiocarbon method. The results confirmed the previous chronology based on artefacts and stratigraphy (Table 2, Figure 3). Nineteen measurements are in accordance with the absolute chronology of the LBK culture in Bohemia set between ca. 5400 and 4950 cal BC (Jakucs et al. Reference Jakucs, Bánffy, Oross, Voicsek, Bronk Ramsey, Dunbar, Kromer, Bayliss, Hofmann, Marshall and Whittle2016; Riedhammer Reference Riedhammer2018). The three measurements from the Radčice 2 site are in accordance with the established chronological timeframe of the SBK period in Bohemia, estimated from ca. 4950 to 4500 cal BC (Řídký et al. Reference Řídký, Květina, Limburský, Končelová, Burgert and Šumberová2019; Riedhammer Reference Riedhammer2018). Most importantly, of these, six caryopses of Hordeum vulgare and three caryopses of Triticum timopheevii were directly dated and respectively confirm their Early and Early-to-Middle Neolithic origin. Furthermore, we assume that these crops were part of the same economy as einkorn and emmer. The chronological homogeneity of fourteen measurements related to caryopses from LBK contexts determined to species level was tested by consistency test and outlier analysis using the OxCal notation Outlier_Model(“General”,T(5),U(0,4), “t”). At the 5% significance level, the results proved the statistical consistency (T = 19.5; T(5%) = 22.4; df = 13) and no outliers in the dataset. Test sensitivity is, however, limited by the radiocarbon curve plateau at 5200–5000 cal BC.
These two species—barley and Timopheev’s wheat—provide unusual diversity in the otherwise conventional crop spectrum. Both occur in most of the studied sites (Figure 2). Their occurrence in assemblages of cleaned crops, with proportions surpassing ten per cent of identified cereal finds at one location for barley and three locations for Timopheev’s wheat, suggests that they were intentionally grown as crops.
Barley was one of the Neolithic founder crops and a key component of the Neolithic package of domesticated crops through which the new economy spread (not only) to Europe. However, the archaeological evidence of barley cultivation in Early Neolithic Central Europe is very sporadic, and its status as an LBK crop in its own right has been surrounded by ambiguity. Consequently, its finds are often viewed as either intrusive or grown unintentionally as a weed admixture in wheat fields (Bogaard Reference Bogaard2004; Bogaard et al. Reference Bogaard, Jacomet, Schibler, Bickle, Cummings, Hofmann and Pollard2017; Filipović et al. Reference Filipović, Kroll, Kirleis, Furholt, Cheben, Müller, Bistáková, Wunderlich and Müller-Scheeßel2020; Ivanova Reference Ivanova, Gron, Sørensen and Rowley-Conwy2020; Ivanova et al. Reference Ivanova, De Cupere, Ethier and Marinova2018; Kočár and Dreslerová Reference Kočár and Dreslerová2010; Kreuz Reference Kreuz2012; Kreuz et al. Reference Kreuz, Marinova, Schäfer and Wiethold2005; Salavert Reference Salavert2011;). Although barley became a significant staple in Central Europe in the Middle and Late Neolithic, its occasional presence increased during the late phase of the LBK (Herbig et al. Reference Herbig, Maier, Stäuble and Elburg2013; Kreuz and Marinova Reference Kreuz and Marinova2017). This delayed adaptation of barley is not yet fully understood (Kreuz et al. Reference Kreuz, Marinova, Schäfer and Wiethold2005). However, the peripheries might have played a significant role as a potential source of crop diversity for their subsequent expansion.
Based on our directly dated finds, we argue that barley cultivation in the LBK of Central Europe might have been an adaptation to higher altitude, lower quality soils, and/or harsher climatic conditions of peripheral regions settled during the second wave of neolithisation. Barley seems a suitable choice since it is highly resilient and capable of thriving in challenging environments where other crops struggle to grow (Newton et al. Reference Newton, Flavell, George, Mullholland, Ramsay, Revoredo-Giha, Russell, Steffenson, Swanston, Thomas, Waugh, White and Bingham2011). According to genetic evidence, Central and Northern Europe is an area where a variety of barley that does not respond to changes in daylight length, allowing it to flower later and avoid late frosts, better suited for cultivation in northern latitudes, is presently predominant. Since no direct genetic evidence is available from archaeological material, it is impossible to pinpoint the date when photoperiod non-responsive barley genotypes were introduced to Central Europe (Jones et al. Reference Jones, Jones, Charles, Jones, Colledge, Leigh, Lister, Smith, Powell and Brown2012, Reference Jones, Charles, Jones, Colledge, Leigh, Lister, Smith, Powell, Brown and Jones2013; Lister et al. Reference Lister, Thaw, Bower, Jones, Charles, Jones, Smith, Howe, Brown and Jones2009).
Timopheev’s wheat, also known as “new” glume wheat, is a distinct prehistoric cereal crop with not yet fully understood history (Czajkowska et al. Reference Czajkowska, Bogaard, Charles, Jones, Kohler-Schneider, Mueller-Bieniek and Brown2020). Since its first identification (Jones et al. Reference Jones, Valamoti and Charles2000) many finds have been reported from Europe and the Near East, however, finds in Neolithic Central Europe are still rare (Bieniek Reference Bieniek, Colledge and Conolly2007; Bogaard Reference Bogaard2011; Hajnalová Reference Hajnalová, College and Conolly2007; Herbig et al. Reference Herbig, Maier, Stäuble and Elburg2013; Kenéz et al. Reference Kenéz, Pető and Gyulai2014; Kohler-Schneider Reference Kohler-Schneider2003; Toulemonde et al. Reference Toulemonde, Durand, Berrio, Bonnaire, Daoulas and Wiethold2015) and it tends to represent a (scarce) contaminant/admixture of other cereal crops. The material presented here is the first in Czechia and one of the oldest in Central Europe. Although Timopheev’s wheat might be underestimated in earlier works due to misidentification (Filipović et al. Reference Filipović, Jones, Kirleis, Bogaard, Ballantyne, de Vareilles, Ergun, Gkatzogia, Holguin, Hristova, Karathanou, Kapcia, Knežić, Kotzamani, Lathiras, Livarda, Marinova, Michou, Mosulishvili, Mueller-Bieniek, Obradović, Padgett and Paraskevopoulou2023), it is plausible that (similar to barley) its cultivation in the Early Neolithic was promoted in specific areas. This species is known to be a suitable choice for poor soils and higher altitudes and is valued for its high immunity to diseases and pests (Filipović et al. Reference Filipović, Jones, Kirleis, Bogaard, Ballantyne, de Vareilles, Ergun, Gkatzogia, Holguin, Hristova, Karathanou, Kapcia, Knežić, Kotzamani, Lathiras, Livarda, Marinova, Michou, Mosulishvili, Mueller-Bieniek, Obradović, Padgett and Paraskevopoulou2023; Jorjadze et al. Reference Jorjadze, Berishvili and Shatberashvili2013).
Although the LBK crop economy in the region appears to deviate from the well-described LBK uniformity, the SBK assemblage from South Bohemia is in line with the general broadening of crop spectra from the Middle Neolithic onwards in Central Europe (e.g. Bogaard Reference Bogaard2004). However, the increased diversity and proportion of each cereal species varies across the study sites and may have been developed at the site level. It is also important to bear in mind the limited scope of excavation, which makes it impossible to detect variability within the settlement and explore overal patterns. To verify these trends, more extensive excavations and studies from other peripheral areas of Central Europe are required.
Conclusions
The increased variety of cultivated species in the peripheral region of South Bohemia indicates a resilient farming strategy, where planting various types of crops with different growth patterns and abilities to withstand varying conditions offers an effective buffering mechanism to overcome environmental constraints and allows for the reduction of the risk of crop failure or collapse. By use of additional cereal taxa, pioneer Central European farmers stepped out of firmly embedded cultural patterns very early on. The continuity of the same crop spectra from the Early to Middle Neolithic also suggests the establishment of local traditions in farming/culinary practices. Finally, this region and dataset provide valuable insights into the role of agricultural peripheries in the early dispersion of domesticated crops, early farmers’ economies, and the multifaceted relationship between people and changing climates and environments.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/RDC.2024.84
Acknowledgments
We are grateful to Professor Glynis Jones and the anonymous reviewers, whose valuable and helpful comments and suggestions allowed us to improve the original manuscript.
Funding statement
This work was supported by the Czech Science Foundation under the project “At the fringe of the neolithization: strategies of the first farmers of South Bohemia (21-16614S)”.
Competing interests declaration
The authors declare that there are no competing interests associated with this paper.
