A palaeoenvironmental reconstruction of the rampart construction of the medieval ring-fort in Rozprza, Central Poland

Archaeological and Anthropological Sciences, Jan 2019

During archaeological excavations of a medieval stronghold in Rozprza, a buried thick deposit of deep black (Dark Earth type) soil was discovered. A multianalytical (sedimentological, geochemical and archaeobotanical) study was carried out in order to identify traits the Rozprza Dark Earth. The analyses demonstrated that the soil was formed as an effect of surface accumulation of organic deposits from swampy areas and waste materials with rich admixtures of organic materials. The organic carbon content of the soil of the Rozprza Dark Earth was twice as high, and the total concentration of P was many times higher as compared with the adjacent soil outside the stronghold. Plant macroremains which were recorded within the buried soil and a cultural layer are evidence for human activity, mainly wood gathering and agriculture. In the Early Middle Ages, summer crops could be cultivated there with the use of tilling methods characteristic for root crops or gardens. The accumulation of the Rozprza Dark Earth commenced in the second half of the tenth century AD. In the period between the eleventh and thirteenth century, a ring-fort was established there. The deep black soil is partly covered by the ring-fort’s rampart. The rampart was built with the use of re-deposited earlier cultural layers and sand of the subsoil. It was then clad with sod bricks. Such a construction of a medieval rampart has been recorded for the first time in the territory of Poland. A new interpretation of archaeological structures and cultural layers offers a basis for new conclusions concerning the chronology and the development stages of the medieval settlement and ring-fort in Rozprza.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://link.springer.com/content/pdf/10.1007%2Fs12520-018-0753-0.pdf

A palaeoenvironmental reconstruction of the rampart construction of the medieval ring-fort in Rozprza, Central Poland

Archaeological and Anthropological Sciences pp 1–33 | Cite as A palaeoenvironmental reconstruction of the rampart construction of the medieval ring-fort in Rozprza, Central Poland AuthorsAuthors and affiliations Jerzy SikoraPiotr KittelMarcin FrączekZbigniew GłąbAlexandra GolyevaAldona Mueller-BieniekJens SchneeweißZofia TomczyńskaKrystyna WasylikowaKatja Wiedner Open Access Original Paper First Online: 24 January 2019 2 Shares 268 Downloads Abstract During archaeological excavations of a medieval stronghold in Rozprza, a buried thick deposit of deep black (Dark Earth type) soil was discovered. A multianalytical (sedimentological, geochemical and archaeobotanical) study was carried out in order to identify traits the Rozprza Dark Earth. The analyses demonstrated that the soil was formed as an effect of surface accumulation of organic deposits from swampy areas and waste materials with rich admixtures of organic materials. The organic carbon content of the soil of the Rozprza Dark Earth was twice as high, and the total concentration of P was many times higher as compared with the adjacent soil outside the stronghold. Plant macroremains which were recorded within the buried soil and a cultural layer are evidence for human activity, mainly wood gathering and agriculture. In the Early Middle Ages, summer crops could be cultivated there with the use of tilling methods characteristic for root crops or gardens. The accumulation of the Rozprza Dark Earth commenced in the second half of the tenth century AD. In the period between the eleventh and thirteenth century, a ring-fort was established there. The deep black soil is partly covered by the ring-fort’s rampart. The rampart was built with the use of re-deposited earlier cultural layers and sand of the subsoil. It was then clad with sod bricks. Such a construction of a medieval rampart has been recorded for the first time in the territory of Poland. A new interpretation of archaeological structures and cultural layers offers a basis for new conclusions concerning the chronology and the development stages of the medieval settlement and ring-fort in Rozprza. KeywordsBuried soil Nordic Dark Earth Sod bricks Ring-fort Rampart Middle Ages  Electronic supplementary material The online version of this article ( https://doi.org/10.1007/s12520-018-0753-0) contains supplementary material, which is available to authorized users. Introduction Remains of the defensive system of the medieval ring-fort in Rozprza are still clearly visible in the terrain relief. The site is currently covered by meadows and fallow fields, whereas shrubs and trees cover the main mound relicts. Due to the poor state of preservation of the stronghold remnants, an archaeological trench from 1963 was re-opened and re-examined with the use of new methods. The re-excavated trench uncovered full sections of what remained of the rampart. The trenching allowed for a detailed recognition of archaeological structures and cultural layers of the stronghold’s rampart on the basis of methods of modern environmental archaeology. Sediments from both the core of the rampart as well as from deposits covered by the rampart were examined with the use of sedimentological, geochemical, archaeobotanical and geochronological analyses. Our results allowed for an identification of deposits and structures which have not been recorded previously in the territory of Poland, such as a “Dark Earth” soil and “sod bricks” used for the construction of the early medieval rampart. The use of various methods of environmental archaeology also enabled us to re-interpret former research results from the 1960s. Our studies are focused on the depositional history of the medieval buried soil of “Nordic Dark Earth” type (as defined by Wiedner et al. 2015) which was discovered at the site of Rozprza. The aim of the research is the reconstruction of environmental conditions of the “Dark Earth” soil formation at Rozprza, including an intense human impact on this process. The soil is covered by the rampart of the medieval ring-fort from the eleventh to thirteenth century AD. This circumstance ensured the preservation of the early medieval buried soil and allowed to examine its original traits. The construction of the rampart is also unusual in the territory of Poland in that period. Therefore, we intended to reconstruct the building process of the rampart where “sod bricks” were used, with special stress on the environmental condition which accompanied the building activity. The use of sod bricks from heather, grass or turf was one of the key elements in early medieval building traditions of north-western Europe. It is well known from Scandinavia and north-western Germany. Sod bricks were used for the construction of ramparts of Saxonian fortifications from the ninth to eleventh century AD (Lemm 2013) as well as in several ring-forts in the nearby Slavic territories in the same period (Struve 1975; Donat 1984; Brandt and Schneeweiss 2017). However, this building technique has not been recorded so far in the territory of medieval Poland. In the Eastern Europe Forest Zone, numerous early medieval settlements are characterised by rarely noticeable deep black cultural layers (e.g. Ambrosiani 2013; Domschke and Wolff 1960). A recent study of Wiedner et al. (2015) suggests that these horizons might be a result of early medieval subsistence economy. However, agricultural areas close to Viking or Slavic settlements are not known and thus intensive horticulture cannot be excluded. Wiedner et al. (2015) demonstrated that the soil within the Slavic settlement of Brünkendorf was comparable to the most fertile man-made soil, the so-called Terra Preta de Índio in tropical Amazonia (Brazil). Further systematic interdisciplinary research is needed in order to gain a deeper insight into the development of comparable soils in temperate Europe. Horizontal layers of dark, strongly humic deposits were more or less frequently found at medieval sites in Poland (e.g. Niewęgłowski 1984; Olczak and Krasnodębski 2002; Buko and Dzieńkowski 2008; Sokołowski et al. 2008; Pluskowski et al. 2014). Difficulties in the interpretation of such deposits can be demonstrated by the example from Bonikowo (com. Kościan), where a layer described as “black”, “clay humus” was discovered below embankments of the ring-fort ramparts. It was 0.2–0.65 m thick and contained a large amount of animal and plant remains, as well as fragments of stones bearing traces of burnout. This stratum was initially recognised as a remnant of the older open settlement (Hołowińska 1956; Hilczerówna 1967). Recently, however, the idea of interpreting this layer as a remnant of the sacrifice and cult activity appeared (Kara 2009). So far, dark, strongly humic horizons have not been regarded as traces of former agricultural or horticultural activity. Therefore, they have not been discussed in reference to the “Nordic Dark Earth” phenomenon, either. We therefore investigated the remains of the defensive system of the medieval settlement in Rozprza, which are also characterised by the presence of deep black soil. Thus, the aims of our study are as follows: (i) Identification of basic properties of the deep black Rozprza soil   (ii) Reconstruction of the chronology of the deep black Rozprza soil development   (iii) Reconstruction of the conditions and chronology of the medieval ring-fort’s rampart construction with the use of “sod bricks”   (iv) Re-interpretation of the traits, chronology and development stages of the medieval settlement and ring-fort at Rozprza   Study area The site (51° 18′ 07″ N; 19° 40′ 04″ E; 182–183 m a.s.l.) is situated in Central Poland ca. 60 km south of Łódź. The remnants of the ring-fort in Rozprza are located in the middle sector of the Luciąża River (third-order river in the Vistula River basin) valley. The valley morphology is crucial for the conditions of the settlement’s location and for functioning of the moat. Three river terraces were identified in the Luciąża River valley in the surroundings of Rozprza: the highest Wartanian glacifluvial, the high Pleni-Weichselian and the low Late Weichselian fluvial terrace (Wachecka-Kotkowska 2004). The ring-fort remnants are situated in the central part of the Luciąża River valley floor on the Pleni-Weichselian residual terrace, which adjoins a floodplain formed in the Late Vistulian and the Holocene. The age of terrace deposits was assessed by optically stimulated luminescence (OSL) dating: 24.7 ± 3.7 ka BP (UJK-OSL-72). The valley floor is strongly expanded in the ring-fort area. An overbank deposit cover with a thickness up to 1 m was identified there. It can be dated back to the Roman Period(?) and the Modern Period (Kittel et al. 2015, 2018a; Sikora et al. 2018) (Fig. 1). Open image in new window Fig. 1 Location of the site Contemporary climatic conditions in the area are highly variable because of the influence of oceanic and continental air masses. The average annual temperature is from 8.6 to 8.9 °C and the average annual precipitation varies from 537 to 626 mm per year (http://www.tutiempo.net). The mean temperature of the warmest month (July) is 18 °C and the mean temperature of the coldest month (January) is − 3.3 °C for the Łódź region (Kłysik 2001). Potential natural plant communities of the Luciąża River valley near Rozprza would be mostly composed of lime–oak–hornbeam forests representing the Tilio-Carpinetum association and to a lesser extent the Potentillo albae–Quercetum typicum. East and south of Rozprza, Vaccinio uliginosi–Pinetum, Leucobryo–Pinetum and Querco–Pinetum would occur in smaller areas. Lowland ash-alder forests and alder of Fraxino–Alnetum and Carici elongatae–Alnetum associations would grow along river channels (Matuszkiewicz 2008). Study of the site Historical data According to written sources, the stronghold in Rozprza was one of the most important medieval ring-forts in Central Poland in the period from the eleventh to thirteenth century, along with Łęczyca, Sieradz and Spycimierz (Kamińska 1953, 1971; Chmielowska 1975; Kajzer 2007; Sikora 2007, 2009). For the first time, the stronghold (castrum) is mentioned in the “Mogilno Forgery” (1065 AD). In the “Gniezno Bull” (1136 AD), it is referred to as a one of the most important ducal forts (castellum) in Central Poland, paying tribute to such ecclesiastical institutions as the monastery in Mogilno and the Archbishopric of Gniezno. On the basis of the knowledge of the chronology of the First Piasts Monarchy, we can assume that Rozprza was incorporated to the Polish state in the second part of the tenth century. In the patrimonial monarchy, it was certainly a ducal property. It was governed by officials who in the thirteenth century began to be called “castellans” and who are known from written sources (Zajączkowski and Zajączkowski 1970; Chmielowska 1966, 1982). One of the best identified castellans was Florian. He served under Duke Leszek Czarny and was the castellan in Rozprza in the 1250s and 1260s. Szyszka (2007) assumes that Florian could be an initiator of redevelopment of the stronghold. A necessity for redevelopment of the fort could be related to supposed military operations in the close vicinity of Rozprza in the mid-thirteenth century in the course of the Mongol invasion in 1241 AD. In 1247 AD, Duke Kazimierz of Kuyavia also had to use force to gain control over the fort. In the Middle Ages, the name “Rozprza” can be assigned to both the fort and a neighbouring town. The town was situated about 1 km to the west of the stronghold. The fort is undoubtedly earlier than the town, which was probably founded in the second half of the thirteenth century (Zajączkowski 1961; Wlaźlak 2009). The fort and the town could have different owners at the same time (Zajączkowski 1961)—a duke and a private person. Most scholars point out that the stronghold came into private hands in the second half of the thirteenth century or the first half of the fourteenth century (Zajączkowski 1961; Chmielowska 1982; Szyszka 2007). The town of Rozprza became private property ca. 1324 AD (DFKR). The new owners of the stronghold and the town (the Nagodzice-Jelitczycy clan) (Szyszka 2007) most probably decided to rebuild their residence (fort). In 1344 AD, King Kazimierz Wielki abolished customs in Rozprza. This implies that the town was not royal property anymore (Zajączkowski 1961; Wlaźlak 2009), or the king merely legitimised the actual status quo (Szyszka 2007). It is likely that castellans did not reside in Rozprza in the fourteenth century, and from that time, the town began to lose its position to neighbouring Piotrków (Zajączkowski 1961). Probably in the fifteenth century, the owners ceased to invest in the fort. In the sixteenth century, this area belonged to a local Catholic church and was used as a meadow (Łaski 1521; DFKR 1521; ADKPwR 1588; KDMiPR 1754; KH 1885). In 1685 (KDMiPR), it was also mentioned as a meadow under a local name “kopiec” (mound—motte). Previous archaeological data The ring-fort was archaeologically excavated for the first time in the 1960s by Chmielowska (1966, 1982). This researcher distinguished four phases of the stronghold’s development between the ninth and fourteenth century AD. The last phase was dated to the thirteenth and fourteenth century. In this period, the stronghold was supposed to be a noble family’s private residence, probably of the motte type. It was built at the site of the earlier early medieval ring-fort. Chmielowska (1982) dated the first phase of the feature to a broad framework between the sixth and ninth century AD. A distinctive layer of “burned material with a distinctive dome-shaped elevation on the edge of it” was identified as remains of this settlement phase. The earliest ring-fort had a diameter of about 25 m, with the rampart being 2.5 to 3.0 m wide. The diameter of the inner yard was about 18–20 m. The height of the preserved rampart would be only a few dozen centimetres (!) (Fig. 2). Open image in new window Fig. 2 Cross-section of the Rozprza stronghold rampart after Chmielowska (1982). Original drawing of the southern wall of trench 1/1963 was digitally processed and mirrored for a better compression with Fig. 4 (J. Sikora). Description after Chmielowska (1982)—and interpretation after the authors: 1. “present-day humus”—stratigraphic unit (S.U.) 1: contemporary humic sand of topsoil. 2. “clay”—S.U. 258: humic sand mixed with clay (slope deposit). 3. “daub”—S.U. 265, 266: humic sands (slope deposits), and S.U. 262: clay. 4. “dark grey earth”—several S.Us. of the late medieval motte relics. 5. “Burnt timbers”—S.U. 263: humic sand and S.U. 270: sod bricks. 6. “Burnt earth”—S.U. 273A and S.U. 273B: humic sand of the “Dark Earth” soil. 7. and 8. “Vertical posts” and “horizontal beams”—several S.Us. of the late medieval moat slope timber construction. 9. “Yellow-greyish sand”—S.U. 271, S.U. 272: sandy core of the ring-fort rampart In the second phase, dated to the ninth century, the ring-work had to be rebuilt after a previous “total disaster”. The aforementioned layer of burned material was believed to be a result of this event (according to Chmielowska 1982). It was assumed that the rampart of the second phase was reconstructed with the use of earth and timbers. It was supposed to be about 7 m wide and was reinforced with “two diagonally arranged piles of beams lying along the rampart” (Chmielowska 1982). An axe of the Great Moravian bradatica type was found “on the outer slope of the rampart”. This artefact is considered as a confirmation of the chronology of this phase (Chmielowska 1966, 1982; Sikora 2009; Kotowicz 2014). The third phase of the stronghold was assumed to be related to the Piasts’ State. Such an assumption implies that this phase should be dated to the second half of the tenth century. The most significant period of the stronghold’s existence in this phase was dated to the twelfth–thirteenth century. The rampart had to be enlarged on the top of the earlier phase by adding timber boxes filled with rocks and by constructing an additional wooden front structure. This way, the entire premise reached a width of approximately 11 m. Outside the rampart, a moat with a width of 12 m was dug, and an external embankment of approximately 8 m width was made (Chmielowska 1982). The last phase extends from the mid-thirteenth to the second half of the fourteenth century. A further development of the fortifications took place and it is even possible that “they were shaped in the conical (motte-type) form”. Remains of four open settlements were believed to have been discovered around the stronghold relics. These settlements were arranged on “hillocks” (Chmielowska 1982) (in fact, it was one elevated surface of a terrace remnant). The results of archaeological research have been closely linked with the written sources and works of historians, who indicated the importance of Rozprza in the early medieval settlement network. The aforementioned references in the Mogilno Forgery and the Gniezno Bull from 1136 AD are usually interpreted as lists of administrative centres of the Piasts’ monarchy (Zajączkowski 1961; Chmielowska 1966; Lalik 1967; Rosin 1970; Sikora 2009). References to the castellans in later sources (thirteenth–fourteenth century) sources were also interpreted as a proof for a relatively high importance of Rozprza in the administrative system of the medieval Polish state. It seemed that the results of archaeological research “confirmed” these observations (Chmielowska 1966, 1982). The aforementioned results of research carried out by A. Chmielowska were widely accepted and quoted in the literature of the subject—see: Kamińska (1971) and Chmielowska and Marosik (1989). However, Sikora (2009) questioned the reconstruction of the earliest phase proposed by Chmielowska (1966, 1982). Sikora rather suggested the existence of an open settlement. He also paid attention to the fact that layers of the rampart’s second phase contained potsherds which may not be dated back to the period before the mid-tenth century AD in Central Poland. This ceramic assemblage with a number of wheel-made vessels could not appear there before the mid-tenth century. Its presence indicated a need to correct the chronology of the origins of the fort. Furthermore, in the opinion of Sikora (2009), the last phase of the fortifications (the fourth phase after Chmielowska 1982) was rather related to the fourteenth century. It is confirmed also by the results of detailed research on the moat fill (Kittel et al. 2018a, b). The result of our study is a re-interpretation of previous conclusions about the chronology, shape and development stages of the ring-fort in Rozprza (see “Stratigraphy—a new archaeological re-interpretation” section). Materials and methods The interdisciplinary approach combined several geoarchaeological and bioarchaeological methods in order to gain a comprehensive insight into the environmental history of the medieval ring-fort in Rozprza and the impact of intensive land use by man. Field work In 2013–2015, a non-destructive survey of the ring-fort surroundings was carried out at Rozprza. The research included analytical field walking, aerial photography and geochemical and geophysical prospection (magnetic gradiometry and earth electrical resistance), along with detailed geological mapping with the use of a dense network of coring. The combined results provided new data which is useful in the reconstruction of the distribution of the moat system, ramparts and other features of the ring-fort complex (Kittel et al. 2015; Sikora et al. 2015a, b). In 2015–2016, an intense field work was undertaken with the use of archaeological trenches (Fig. 3). This allowed for recording of archaeological structures and cultural layers of preserved rampart features as well as moat fills (Kittel et al. 2018a, b). The trench marked as “1/1963” (Fig. 4), 3 m width and 27 m long, crossed the preserved part of the stronghold’s earthwork. It was the aforementioned re-opened trench from the field work of Chmielowska in 1963. The aim was to re-document it with the use of modern digital tools (such as photogrammetry), to re-analyse the stratigraphy of the remnants of the earthwork and to collect finds and samples for establishing the chronology of subsequent phases of the stronghold’s development. First, the northern wall of the trench was analysed and documented. Only the lowest contexts, preserved in the “steps” left in 1963 in order to secure the stability of the sandy sections underwent exploration according to the stratigraphic units (S.U.) order. Open image in new window Fig. 3 Plan of the location of trenches and research profiles Open image in new window Fig. 4 Trenches 1/1963 and 6/2016: section of the Rozprza stronghold’s rampart (orthophotoplan of the northern wall of the exposure after Sikora 2015). Research profiles RP WW, RP M and RP M2 and radiocarbon dates are marked Sampling strategy After documentation of the trench wall, three profiles of samples taken every 5 cm were collected both from the rampart layers and underlying units (Fig. 4; Table 1). The main profile (RP WW) was situated in the central part of the rampart’s remnants. It covered the deep black “Dark Earth” buried soil (partly re-deposited) (S.U. 273A and 273B) and underlying layers (S.U. 274, 275, PLV). The samples were collected for plant macroremains and phytoliths, as well as for textural and geochemical analyses. For the analysis of plant macroremains, two separate samples were also taken from humic sand layers within the rampart’s earthwork, i.e. within S.U. 272 (samples 272a and 272b). Also, one separate sample for the plant macroremains analysis and six samples for phytoliths were collected from “imbricatus” structures of humic sands identified as relicts of sod bricks fastening the rampart surface S.U. 270 (samples T1, T2, T3—Fig. 4). Profile RP M was situated below the inner slope of the rampart of the first phase. The third profile RP M2 was located below the preserved rampart relicts in the courtyard area. Both profiles RP M and RP M2 covered the “Dark Earth” (S.U. 273B), underlying layers of yellowish sand and brownish soil (S.U. 274 and 275), as well as basement deposits (S.U. PLV). Next, one sample for plant macroremains and six samples for phytoliths analyses were collected. Table 1 Deposits profiles studied by archaeobotanical, sedimentological and geochemical analyses at Rozprza Symbol of profile (for the location, see Figs. 3 and 4) Stratigraphical unit (S.U.) analysed (see Figs. 6, 7, 8, 9 and 10) The number of samples for plant macroremain analysis The number of samples for phytolith analysis The number of samples for textural analysis The number of samples for geochemical analysis RP WW PLV, 275, 274, 273B, 273A, 272, 270 18 + 3 17 + 6 22 22 RP M PLV, 275, 274, 273B 12   16 16 RP M2 PLV, 275, 274, 273B 13   17 17 + 8 bulk samples RP RM PLV, 165, 164, 163, 162 16   22 22 RP GE     12 12 + 8 bulk samples Two referential profiles were also prepared: the first one (RP RM) from the area outside the main moat of the ring-fort and the second one (RP GE) from the isolated terrace remnant 150 m NE of the stronghold. In profile RP RM, samples were collected from buried soils and basement deposits for plant macroremains and textural and geochemical analyses. Textural and geochemical analyses were carried out for profile RP GE. Bulk soil samples from profiles RP M2 and RP GE were analysed in terms of pH, soil organic matter composition (C and N), isotope signature (δ15N) and total P. In profile RP M2, bulk soil samples were taken at 10 cm increments down to the depth of 80 cm from the “Dark Earth” (S.U. 273B). In profile RP GE, bulk soil samples were taken from the adjacent reference soil with no visible traces of anthropogenic influence, also at 10 cm intervals. Stratigraphic, sedimentological and geochemical analysesStratigraphic analysis The basic stratigraphic layout consisted of several subsequent layers, which were analysed in terms of their archaeological and geological characteristics as well as their mutual stratigraphic relationships with the use of the Harris (1979) method, with further improvements (Herzog 2004). The S.U. were defined, described and numbered (Fig. 4). Their stratigraphical relationships were defined during the field work. They were recorded with the use of basic field description (using Strati5 and Stratify software (Sikora et al. 2016) as well as by means of 2D and 3D digital photogrammetry methods with subsequent digital post-processing in GIS and vector drawing software (Sikora and Kittel 2018). The stratigraphy of the site which was analysed with these tools helped to establish the relative chronology based on the sequence of the following units (or contexts): deposits; cuts; structures, such as timber constructions; as well as key contacts between them. Sediment texture analysis For profiles RP WW, RP M, RP M2, RP RM and RP GE, the analysis of particle size distribution for silty sediments was made with the use of a Malvern Instruments Mastersizer 3000 Particle Analyzer. This device allows for measurements of the particle size in the range of 0.01–3500 μm. For the sieve analysis of sandy deposits, a 200 × 25-mm sieve set was used, in accordance with DIN ISO 3310-1 and BS 410-1 norms (sieve size 63–2800 μm). The sieve was coupled with a Retsh AS 200 basic Vibratory Sieve Shaker. All samples had the same weight (100 g), and they were measured with an amplitude of 0.8 for 10 min without the interval function. And then the textural features were evaluated using Folk and Ward (1957) coefficients. GRANULOM software (with some modifications) was used for the graphic presentation of results. The relationship between the mean grain size and the sorting indexes (the so-called co-ordinate system) was analysed with the use of methods proposed by Mycielska-Dowgiałło (2007). Sedimentological analysis is crucial for the recognition of depositional environments and geomorphologic processes responsible for accumulation of deposits, especially sediments with massive structure. Geochemical properties of soil and sediment pH For profiles RP WW, RP M, RP M2, RP RM and RP GE and for bulks from profiles RP M2 and RP GE, soil pH was measured using suspensions with distilled water (1:2.5) and 0.01 M CaCl2 (1:5) for determining absorbable forms of magnesium according to the Schachtschabel method. Soil pH for all samples was determined in the supernatant, after at least 1 h of shaking in a low-speed shaker and sedimentation of solids for at least 1 h at room temperature. Carbonate Calcium carbonate content (both for profiles RP WW, RP M, RP M2, RP RM and RP GE and bulk soil samples from profiles RP M2 and RP GE) was determined by volumetric methods. Scheibler’s calcimeter (Jackson 1958; Bengtsson and Enell 1986; Woszczyk and Szczepaniak 2008) was used for the calculation of organic C stocks. Elemental composition (C and N) and δ 15N The organic matter was measured by the loss-on-ignition method (LOI). Samples were first dried in 105 °C with constant weight. Then, they were roasted at the temperature of 500–550 °C for at least 5 h (Bengtsson and Enell 1986). The next stages were 3 h, and the last one consisted in roasting the samples for 1 h. In the case of the weight loss after the third burning, analyses were repeated until a constant sample weight was achieved. A SNOL muffle furnace was used for the loss-on-ignition analysis. Total organic C, total N and δ15N of bulk soil samples (RP M2 and RP GE) were measured using a EURO EA Elemental Analyzer (EuroVector, Hekatech, Germany) coupled via a Conflo III Interface to an isotope ratio mass spectrometer (IRMS; Finnigen Delta V Advantage, Thermo Scientific, Bremen, Germany). Total P content For total P extraction, 3 g of dried and sieved bulk soil (RP M2 and RP GE) was weighed in glass tubes and digested with 21 mL HCl and 7 mL HNO3 for 16 h without heating and then for 90 min at 120 °C. After cooling (1 h), samples were filtered into a 100-mL volumetric flask. Volumetric flasks were filled with deionised water. Diluted samples were analysed using an inductively coupled plasma optical emission spectrometer (ICP-OES, ULTIMA 2, HORIBA Scientific S.A.S, Jubin-Yvon, France). Archaeobotanical analysesPlant macroremains The samples for the analysis of plant macrofossils were collected from three profiles RP WW (21 samples), RP M (12 samples) and RP M2 (13 samples) situated below the rampart and one profile RP RM (16 samples) outside the rampart. One sample RP T1 was taken from a sod brick used for fastening of the rampart (Fig. 4). In the laboratory, each soil sample of known volume was soaked in water and the floating fraction was poured through sieves with 0.5 mm mesh. Heavy fraction was sieved with the use of ca. 2 mm mesh sieves in order to collect mineralised plant remains and other heavy artefacts (potsherds, bones, etc.). Both charred and uncharred plant remains were sorted out. The greatest total volume of examined soil originated from profile RP WW (65.1 L). Next, there were profiles RP M (9.5 L), RP M2 (3.96 L), RP RM (4.2 L) and RP T1 (0.8 L). Macrofossils were identified with the use of keys and atlases (e.g. Greguss 1945; Grosser 1977; Berggren 1969; Cappers et al. 2006; Kats et al. 1965; Schweingruber 1978; Schweingruber et al. 2011; Velichkevich and Zastawniak 2006, 2008), as well as the reference collection of modern seeds, fruits, wood and charcoal and the collection of fossil floras stored in the Palaeobotanical Museum of the W. Szafer Institute of Botany, Polish Academy of Sciences, Kraków. Diagrams were prepared with the use of POLPAL software (Nalepka and Walanus 2003). Phytoliths and other biogenic microparticles The main method involves a consecutive microscopic study of different types of biomorphs. Samples of 50 g each are treated with a 30% solution of hydrogen peroxide. They are then separated from quartz and other mineral grains by flotation in a heavy liquid, usually a mixed cadmium iodide and potassium iodide solution with a specific gravity of 2.3. After centrifugation, the floating siliceous and organic microbiomorphs are collected and washed several times with distilled water. Next, they are immersed in oils (glycerine) for optical microscope study at magnifications ranging from ×200 to ×900. The entire assemblage of microbiomorphs is identified and counted including phytoliths, diatoms, spicules, detritus and others. Nomenclature of microbiomorphs is given after Golyeva (2007). Quantitative assessment allows for comparisons of their distribution between samples. At Rozprza, a total number of 23 samples were studied from profile RP WW and from sod bricks. Geochronology Four samples of selected plant macroremains from the RP WW core and the humic layer within S.U. 272 (Fig. 4) were dated with the radiocarbon (14C) method, using accelerator mass spectrometry technique (AMS) (see Goslar et al. 2004 and Krąpiec and Walanus 2011 for details). The OxCal ver. 4.2.3 calibration software (Bronk Ramsey 2009) was used for the calibration of radiocarbon dates, using atmospheric data provided by Reimer et al. (2013). The absolute chronology determinations were carried out in the Laboratory of Absolute Dating in Skała (Poland) in cooperation with the Accelerator Mass Spectrometry Laboratory in Seattle (D-AMS signature) (see Zoppi et al. 2007 for details). Results Absolute chronology of the buried soil and the rampart earthwork Two radiocarbon dates available from profile RP WW indicate that plant remains recovered from layers 273A and 273B are of more or less the same age (Fig. 4; Table 2). A charred wheat grain (Triticum aestivum/spelta) from the top of layer 273B was dated to 690–872 AD with a probability of 68.2% and 659–962 AD with a probability of 95.4% (D-AMS 015984). Charred millet grains (Panicum miliaceum) from the top of layer 273A yielded the date of 677–868 AD (68.2%) and 649–968 AD (95.4%) (D-AMS 015983). The date of 568–679 AD (68.2%) and 433–770 AD (95.4%) (D-AMS 015985) was obtained from wood charcoal from the bottom of layer 275. It probably documents (as a terminus post quem) the very beginning of medieval occupation of the area. Table 2 Radiocarbon data set from trench 1/1963 Stratigraphical units (S.U.) Dated material Depth in the WW profile [cm] 14C age years BP Laboratory code Calibrated age (cal AD, prob. 68.2%) Calibrated age (cal AD, prob. 95.4%) Remarks 272 Picea abies (charred leafs)   1185 ± 55 MKL-3505A 770–941 AD 690–975 AD Sample 272a (see Table 3) 273A (top) Panicum miliaceum (4 charred grains) 0–5 1250 ± 80 D-AMS 015983 677–868 AD 649–968 AD   273B (top) Triticum aestivum/spelta (charred grain) 20–25 1236 ± 71 D-AMS 015984 690–872 AD 659–962 AD   275 charcoal 75–80 1402 ± 71 D-AMS 015985 568–679 AD 433–770 AD   It must be emphasised that potsherds from layers 273A and 273B were of a later age than the radiocarbon data set. The pottery is dated to the period between the mid-tenth and the beginning of the twelfth century AD. We cannot exclude that the soil (S.U. 273B) which formed before the construction of the ring-fort was mixed with later artefacts. However, these potsherds were recorded within the whole thickness of the buried soil. Therefore, the buried soil most probably formed in the tenth–eleventh century AD with admixture of deposits containing (partly) earlier ecofacts, such as charred cereal grains. Alternatively, we can relate the age of the grains to the mid-tenth century, as confirmed by 95.4% probability of the 14C data range. The radiocarbon date of 770–941 AD (68.2%) and 690–975 AD (95.4%) (MKL-3505A) for charred spruce leafs from a humic lamination within layer 272 (Fig. 4) is too early if compared with the chronology of the archaeological material from this layer (see “Archaeological chronology of stratigraphic units” section). Therefore, what was actually dated was an effect of a re-deposition of sediments containing charred plant macroremains. Archaeological chronology of stratigraphic units Typological and technological analyses of potsherds were also used for establishing the chronology of the archaeological contexts. Almost 30,000 (29,553 finds) potsherds were excavated in the years 2015 and 2016 at the site in Rozprza; 386 potsherds were found altogether in layer 273B. Parts of this assemblage were hand-made vessels, with only upper parts having been processed with the use of wheel. They were ornamented with carved waves, which are typical for the ninth and tenth centuries. Among them, a certain number could be related to type Menkendorf, which was popular in north-western Slav areas (Schuldt 1956; Łosiński and Rogosz 1986; Cnotliwy et al. 1983). On the other hand, a large number of wheel-made potsherds which were ornamented with horizontal lines were typical for the period after the mid-tenth century. The entire assemblage is typical for the so-called phase D of development of early medieval pottery in Greater Poland (Hensel 1948; Hilczerówna 1967; Dzieduszycki 1982). Therefore, it should be dated to between the mid-tenth and mid-eleventh century. The chronology of later layers of the ring-fort rampart is not clear. One hundred and four potsherds were found in layers 273A and 272. Some of them are similar to those from layer 273B and should be considered as re-deposited during the development of the earthwork. Several potsherds of later wheel-made vessels were also found. These can be dated to the eleventh–twelfth century. It should be mentioned that in S.U. 272 a very small number of potsherds which are datable to the thirteenth century was found. These are two pieces which are typical rather for the so-called group A according to Kajzer (1991). Taking into account a possible re-deposition caused by soil fauna, whose activity traces were noticed in some parts of the layer, all we can suggest is a wide chronological framework between the eleventh and thirteenth century. In the layers of the motte mound and the bottom strata of the moat (in trenches 1/1963, 1/2015 and 3/2015), late medieval pottery was found, including “traditional” as well as more advanced black or grey vessels. Grey and black pots were fired in reducing atmosphere, and such pottery appears in the territory of Poland territory in the thirteenth century (Kruppe 1981). The whole assemblage of pottery from the moat and motte mound should be dated to the second half of the thirteenth and to the fourteenth century. Stratigraphy—a new archaeological re-interpretation The analysis of features uncovered in 2015 in trench 1/1963 led to a need for important changes in the interpretation of stratigraphy, as well as in the chronology of the rampart’s development, as compared with observations proposed by Chmielowska (1982) (see “Previous archaeological data” section; Fig. 5; Table 3). Open image in new window Fig. 5 Stratigraphy and chronology of stratigraphic units (S.U.) in trench 1/1963: top: after Chmielowska (1982): 1. first phase; 2. second phase; 3. third Phase; 4. fourth phase (according to Chmielowska); 5. parts destroyed by animal burrows. Bottom: re-interpreted by the authors. 1. Prehistoric or early medieval cultural layers (between the fifth and eighth century AD); 2. “Dark Earth” soil (second half of the tenth to first half of the eleventh century AD); 3. rampart of the medieval ring-fort (between the second half of the eleventh and thirteenth century AD); 4. slope deposits from the destruction of the rampart; 5. fill of the late medieval moat and the motte mound (fourteenth century AD); 6. timber constructions of the late medieval moat (ca. 1330 AD); 7. pits which remained after the demolition of the foundation of the late medieval manor house or a keep; 8. slope deposits—erosion of the late medieval mound; 9. early modern posts (sixteenth century AD); 10. parts destroyed by animal burrows Table 3 Re-interpretation of chronology and description of layers of the rampart remnants at Rozprza Description of layers after Chmielowska (1982) No. of layers after Chmielowska (1982) (see Fig. 2) Chronology of layers after Chmielowska (1982) Stratigraphical unit (S.U.) no. after authors (see Figs. 4 and 5) Description of layers after authors Chronology of layers after authors Present-day humus 1 Modern period 1 Contemporary humic sand of topsoil Modern period Clay 2 Mid-thirteenth to second half of the fourteenth century AD 258 Humic sand mixed with clay (slope deposit) Modern period Daub 3 Mid-thirteenth to second half of the fourteenth century AD 265, 266 262 Humic sands (slope deposits) clay Modern period Dark grey earth 4 Second half of the tenth to mid-thirteenth century AD 257, 259, 261, 304, 306, 313–317, 330–335 Humic sands of the Late Medieval motte remnants 20s–30s of the fourteenth to the fifteenth century AD Burnt timbers 5 Ninth century AD 263 270 Humic sand sod bricks Second half of the eleventh/twelfth to end of thirteenth century AD Burnt earth 6 Sixth–ninth century AD 273A 273B Re-deposited “Dark Earth” Deep black humic sand of the “Dark Earth” soil Second half of the eleventh/twelfth to end of the thirteenth century AD Second half of the tenth to half of the eleventh century AD Vertical posts 7 Second half of tenth to mid-thirteenth century AD Several S.U. Vertical posts of the Late Medieval moat slope timber construction 20s–30s of fourteenth century AD Horizontal beams 8 Second half of tenth to mid-thirteenth century AD Several S.U. Horizontal beams of the Late Medieval moat slope timber construction 20s–30s of fourteenth century AD Yellow-greyish sand 9 Ninth century AD 271, 272 and also: 261, 262, 263, 267, 268, 269, 309, 310, 333, 336, 337, 338 Yellow-greyish sand of the rampart core interbedded humic, brownish and yellowish sands sandy of the rampart core Yellow-greyish laminated humic sands and solids of clay (slope deposits) Second half of the eleventh–end of the thirteenth century AD End of the thirteenth century AD (?) to 20s–30s of fourteenth century AD Yellow-greyish sand 9   274 Deluvial yellow-greyish sand Ca. tenth/eleventh century AD (?) Yellow-greyish sand 9   275 Brownish poorly humic sand (buried soil with cultural layer) Prehistory and Early Middle Ages (before tenth/eleventh century AD) Yellow-greyish sand 9   PLV Pleni-Weichselian lightly yellow sand of terrace remnant ca. 25 ka BP The section of the rampart obtained after the unveiling of the trench fill from the 1960s generally corresponds to the situation documented by Chmielowska (Chmielowska 1982: Fig. 2—see Fig. 2). However, there are some differences, which were a result of inaccurate field documentation in the 1960s. Uncovering the section allowed for making necessary corrections in the interpretation of several stratigraphical units. Additional difficulties in the observation and interpretation of the layers resulted from bioturbation associated with the activity of wild animals (such as fox and badger), which disturbed the medieval stratigraphic systems. Moreover, it was possible to acquire artefacts and ecofacts from key contexts, including the layer of “burned material”, discussed by Chmielowska (1966, 1982). It turned out that it was deep black and strongly humic sand with admixture of charcoal, wood and other plant macroremains, as well as with sand and clay additions. This layer was recognised in the field as “Dark Earth” buried soil and was marked as S.U. 273B. The layer which was considered by Chmielowska (1982) as a remnant of the earliest stronghold can in no case be regarded as a relic of the defensive rampart. The nature and origin of the “buried material” layer (S.U. 273) is one of the main issues discussed in this paper. In terms of stratigraphy, it should be separated (in profile RP WW) into two levels: the horizontal layer 273B, which is a vestige of the settlement occupation processes, and layer 273A, which is a small dome-shaped mound of a base width of 2.3 m and a height of 0.25 m. Layer 273A layer is composed of material of layer 273B, which was re-deposited in the process of building the rampart embankment. This process was identified as the second phase of the stronghold by Chmielowska. The second phase of the stronghold as proposed by Chmielowska should also be interpreted differently. The development of the rampart’s structure can be divided into two stages. The original face of this construction was not a clay layer (S.U. 262) but a level that was interpreted as a diagonally arranged stack of “beams lying along the course of rampart” (S.U. 270) (Chmielowska 1982). These structures were called “roofing tile-shaped” or “placed in a roofing tile manner” and were believed to have been made up of logs which were not longer than a half of a metre. However, the discussed layer (S.U. 270) is certainly not a remnant of mineralised and/or burnt pieces of wood. This layer consists of structures whose shapes in their cross-sections are similar to deformed rectangles, triangles or trapezoids. These structures are composed of strongly humic various-grained sands mixed with charcoal and single plant macroremains. Already at the stage of field research, we hypothesised that we were dealing here with relics of cubes made of turf or sod together with the top layer of the soil humus horizon. These cubes were used for facing the slopes of the rampart, both external (S.U. 270) and internal (S.U. 339) ones. Relics of such “sod bricks” were perfectly recognisable during the exploration of “steps” left in 1963. “Sod bricks” facing of the rampart is also manifested by slight post-depositional deformations, consisting of micro-scale crawling of their edges on the surface of the rampart slope. The core of the rampart at this stage of development was a sandy embankment, with clearly legible interbeds of darker, humus and brawn ferriferous sand. It was therefore identified as a reversed and slightly impaired sequence of original (i.e. before the construction of the rampart) layers of subsurface deposits. The embankment of the rampart (S.U. 273A, 272 and 271) arose as a result of digging of a moat or ditch and a successive re-deposition of earlier strata. These consisted of early medieval dark humic sands horizon containing numerous early medieval artefacts and ecofacts, mainly bones of livestock and plant macroremains (S.U. 273B); early medieval (?) deluvium (S.U. 274); brownish poorly humic sand of prehistoric and/or early medieval buried soil with a cultural layer (S.U. 275); and lightly yellow sand of the terrace remnant (S.U. PLV). The re-deposited layers forming the core of the rampart are characterised by a typical structure spotted in the embankments, with clearly visible layers arranged in parallel to the shape of the rampart’s surface. Well-legible structures are created during micro-scale gravitational landslides of loose layers of the successively formed embankment. Layer 273B strata can be recognised as the “Dark Earth” buried soil dated to the tenth and not later than the mid-eleventh century, on the basis of the analysis of potsherds. It was covered by the rampart embankment not later than in the fourteenth century AD. A different interpretation should also be proposed on the basis of the analysis of the relics which were linked by Chmielowska (1982) to the third phase of the development of the fortifications. This phase was supposed to be represented by a sandy unit (S.U. 261). According to Chmielowska (1982), sandy covers (S.U. 269 and 267) and also thin levels of humic sand (S.U. 263) and clay (S.U. 262) formed the embankment of the second phase of the rampart. The presence of a 15-cm-thick layer of clay was interpreted as a kind of face of the extended rampart (Chmielowska 1982). However, the sand cover (S.U. 269, 267 and 261) is an effect of slope processes, which resulted in erosion associated with the destruction of the already existing rampart of the first phase. The clay layer (S.U. 262) has a varied thickness of only a few millimetres up to 15 or even 20 cm. Fingerprints were observed on larger fragments of clay, which may indicate the covering of the surface with clay. On the other hand, these relics and bulks of humic sands (S.U. 268 and 263) could be interpreted as elements of the earlier rampart construction. They were most probably re-deposited by slope wash processes. A similar stratigraphy was also recorded at the inner slope deposits of the rampart—laminated humic sandy deluvia (S.U. 338, 336, 337, 310 and 309) and solids of clay (S.U. 333) were observed there. A number of stratigraphic units that covered earlier structures were interpreted by Chmielowska (1982) as the fourth phase of the stronghold development. However, “the mantle of clay” covering the rampart remains from outside has not been identified by the authors of this paper. It can probably be identified with sands mixed with clay and tiny fragments of charcoal and burnt clay (this unit was marked as S.U. 258). This layer is inclined and should be rather regarded as a result of erosion of the embankment (slope deposits). Numerous stones mentioned by Chmielowska (1982), laid in the top part of the rampart relics, may be remains of some defensive structures, maybe even the fill of timber boxes (“izbice”) which did not survive. The new re-interpretation of the stratigraphic structure studied in trench 1/1963 (Fig. 5) allows for the definition of the following phases of the development of the settlement structures at the site of the stronghold in Rozprza (Sikora et al. 2018): 1. Prehistoric and/or early medieval (?) open settlement—S.U. 275 and 274(?).   2. Early medieval (?) open settlement which functioned between the second half of the tenth and the first half of the eleventh century AD—S.U. 274(?) and 273B (accumulation of the “Dark Earth” soil).   3. Ring-fort which was constructed in the period between the second half of the eleventh or the beginning of the twelfth and the end of the thirteenth century AD—S.U. 273A, 272, 271 (re-deposition of earlier units) and 270 (sod bricks on the rampart slope surface).   4. Destruction of the ring-fort’s rampart—several S.U.: 261, 262, 263, 267, 268 and 269 on the outer slope and 309, 310, 333, 336, 337 and 338 on the inner slope. These can be interpreted as slope deposits. It can point to the abandonment of the stronghold.   5. Motte-and-bailey with a system of moats. This phase could last from the 1320s–1330s to the fifteenth century AD.   Sedimentological traits of deposits The lowest samples of all studied profiles (RP WW, RP M, RP M2, RP RM and RP GE) represent substratum deposits of the Pleni-Weichselian terrace (PLV). Twnety-five samples of fluvial Pleni-Weichselian sediments were analysed altogether. These river sands consist mostly of a medium-grained sand fraction between 0.5 and 0.25 mm. The percentage of coarse-grained sand and gravel (i.e. > 0.5 mm; < 1.00 φ) ranges from 17.9 to 31.7% (only in two samples there was more than 40%) and of mud (i.e. < 0.1 mm; > 3.32 φ)—from 1.0 to 3.1%. These deposits are characterised by a mean grain size of 1.02–1.58 Phi (φ) (i.e. 0.49–0.34 mm) and a sorting index of 0.64–0.78 (i.e. moderately well-sorted and moderately sorted) (Fig. 6). The relationships between the mean grain size and the sorting indexes, the skewness and the mean grain size, as well as between the sorting index and the skewness, demonstrate that the features of the studied sediments are situated on the border between channel deposits and overbank deposits as defined by Mycielska-Dowgiałło (2007). In the C-M diagram after Passega and Byramjee (1969), these samples can be placed within class I and within segment O-P according to Passega (1964). This demonstrates that the analysed sediments have been transported by rolling and saltation (Passega and Byramjee 1969; Mycielska-Dowgiałło and Ludwikowska-Kędzia 2011). Open image in new window Open image in new window Open image in new window Fig. 6 Stratigraphic diagrams of the lithological and geochemical composition of research profiles. A. Lithologic profile. B. Folk and Ward coefficients: mm—mean grain size [mm], Mz—mean grain size [Phi], δ1—sorting index (standard deviation), Sk1—skewness, Kg—kurtosis. C. Geochemical parameters: Corg—content of organic carbon (%), LOI—loss-on-ignition (%), CaCO3—content of carbonate (%), pH—pH of deposits. a Profile RP WW. b Profile RP M. c Profile RP M2. d Profile RP RM. e Profile RP GE Traits which are similar to the substratum deposits can also be seen in the brownish buried soil (S.U. 275) (Fig. 6a–e). However, the percentage of coarse-grained sand and gravel ranges from 21.2 to 28.8% and that of mud—from 1.6 to 3.6%. The mean grain size is 1.41–1.58 Phi (0.38–0.34 mm) and a sorting index is 0.73–0.78. This demonstrates that this topsoil was formed on the surface of the Pleni-Weichselian terrace. The layer covering the brownish buried soil is constituted by yellow-greyish sands (S.U. 274), characterised by a mean grain size of 1.53–1.67 Phi (φ) (i.e. 0.35–0.31 mm) and a sorting index of 0.78–0.89 (i.e. moderately sorted). The relation between the mean grain size and the sorting indexes for S.U. 274 represents the second co-ordinate system after Mycielska-Dowgiałło (2007). It is characteristic for overbank alluvia but also for slope-wash deposits (deluvium) (Twardy 2000, 2008; Smolska 2008; Mycielska-Dowgiałło 2007; Kittel 2014). The admixture of fine-grained fractions can reach up to 6% for these deposits, but it can be an effect of in-wash from the covering layer. Both textural and geochemical (see below) traits of the “Dark Earth” layer (S.U. 273B) differ distinctly from the characteristics of underlying deposits. This stratum is characterised by a mean grain size from 1.73 to 1.99 Phi (φ) (i.e. 0.25–0.26 mm) and a moderate and poor sorting index of 0.87–1.36. A high percentage of coarse-grained sand and gravel (from 19.4 to 26.6%) and of fine-grained fractions (from 3.1 to 16.1%) is very significant. An increase in the amount of very coarse sands, gravels and mud fraction is clearly visible for this layer (Fig. 6a–e). This trait strongly differs S.U. 273B from the underlying deposits. It is a proof for (at least partly) artificial origin of the earthwork of the “Dark Earth” soil deposits at Rozprza. It means that they were accumulated (at least partly) as a result of intentional and direct human activity. Geochemical traits of depositsProfiles RP WW, RP M, RP M2, RP RM and RP GE The pH values for the substratum deposits of the Pleni-Weichselian terrace are from 5.8 to 6.6 and no organic carbon stocks were detected (Fig. 6a–e). The LOI values are below 1% and Corg—0.12%. These traits testify to a trivial admixture of organic matter within fluvial terrace sediments. The brownish buried soil (S.U. 275) is characterised by pH from 6.1 to 6.8 and the content of carbonates is very low—up to 0.04%. An admixture of organic matter (LOI) reaches up to 1.2% and Corg—0.25%. It could be however (partially) caused by the presence of iron compounds. As confirmed by archaeobotanical research, charcoal fragments were found within this layer. Geochemical traits of the deluvial deposits (S.U. 274) differ in the studied profiles. The pH values range from 5.5 to 7.3—the lowest are in profile RP M2 and the highest in profile RP WW (i.e. below the rampart earthwork). The organic content also differs in selected profiles—the highest is in profile RP WW (LOI 0.8–7.4; Corg 0.2–2.1%) and the lowest in profile RP M (LOI 0.6–1.0; mCorg 0.2–0.3%). Very numerous charcoal fragments were recorded in profile RP M2. The layer recognised as the “Dark Earth” soil (S.U. 273B) is characterised by a distinct increase of the amount of carbonates (up to 4.8%) in the studied profiles. An increase of organic matter admixture, both LOI (7.6–13.0%) and Corg (3.1–4.9%), is evident in all profiles. Only pH values differ from 5.4 to 5.7 in profile RP M2 up to 6.7–7.3 in profile RP WW. Bulk soil samples from profiles RP M2 and RP GE The pH value of profile RP M2 ranges from 5.5 (0–10 cm) to 6.4 (70–80 cm). In contrast, pH values of the reference profile (RP GE) are more acidic, ranging from 3.8 (0–10) to 5.3 (70–80 cm) (Table 4). Table 4 Geochemical traits of bulk soil samples from profiles RP M2 and RP GE profiles Soil depth [cm] Corg [g kg−1] Corg [Mg ha−1] N [g kg−1] d15N [‰] P [mg kg−1] C/N pH [CaCl2] Bulk density [g cm−3] Coloura Textureb Horizonc Stratigraphical units RP M2  0–10 7.7 10.4 0.5 8.1 1231.7 15.1 5.5 1.3 10YR 3/1 Sandy loam Ap 273B  10–20 46.2 51.7 2.1 8.6 4318.2 21.9 5.9 1.1 2.5YR 2.5/1 Silt loam   273B  20–30 43.9 56.9 1.9 8.3 4307.6 23.0 5.9 1.3 2.5YR 2.5/1 Silt loam Ah 273B  30–40 27.6 38.1 1.1 8.6 3796.9 24.3 5.9 1.4 10YR 2/1 Sandy loam   274  40–50 7.0 9.7 0.4 8.1 2939.6 18.3 6.1 1.4 10YR 3/1 Sandy loam   274  50–60 3.3 4.7 0.3 5.1 3294.1 11.7 6.3 1.4 10YR 3/2 Loamy sand Bw 275  60–70 0.8 1.1 0.1 7.5 851.1 12.2 6.4 1.5 10YR 4/1 Loamy sand   275  70–80 0.5 0.7 nd nd 327.2 nd 6.4 1.4 10YR 5/2 Loamy sand   PLV RP GE  0–10 87.9 61.2 8.0 4.9 1470.5 11.1 3.8 0.7 10YR 3/1 Sandy loam Ap    10–20 18.1 27.4 1.8 5.3 208.9 10.5 4.3 1.5 10YR 4/1 Sandy loam     20–30 3.3 5.4 0.3 6.3 89.4 10.9 4.5 1.6 10YR 4/2 Sandy loam Bw    30–40 1.8 3.0 0.2 5.1 19.4 12.1 4.6 1.6 10YR 4/1 Loamy sand   PLV  40–50 0.3 0.5 nd nd nd nd 4.6 1.6 10YR 5/2 Sand C PLV  50–60 nd nd nd nd nd nd 4.6 1.7 10YR 6/2 Sand   PLV  60–70 0.4 0.5 nd nd nd nd 4.7 1.5 10YR 6/2 Sand   PLV  70–80 nd nd nd nd nd nd 5.3 1.7 10YR 6/3 Sand    aMunsell Soil Colour Chart bAccording to Jahn et al. (2006) cWorld Reference Base The total N concentration in profile RP M2 and in the reference soil (profile RP GE) displays a high variability and the same can be said about differences within both profiles (Table 4). The total N concentration in profile RP M2 ranges between 0.1 g kg−1 (60–70 cm) and 2.1 g kg−1 (10–20 cm), whereas the content of N was below the detection limit at the depth of 70–80 cm. Total N concentrations in the reference soil (RP GE) are 8 g kg−1 in the upper 10 cm with a rapid decrease to 0.2 g kg−1 at the depth of 30–40 cm. The content of N was not detectable between the depths of 40 and 80 cm in the reference soil. The δ15N values are highly enriched in profile RP M2 and the same can be seen in the reference soil (Table 4). However, δ15N values of RP M2 are throughout highly enriched, ranging from + 5.1‰ (50–60 cm) to + 8.6‰ (30–40 cm). In comparison, δ15N values of RP GE are the highest at the depth of 20–30 cm (+ 6.3‰). Soil organic carbon (SOC) stocks of profile RP M2 are the highest between the depths of 10 and 40 cm, ranging between 27.6 and 46.2 g kg−1. In contrast, the highest SOC concentrations in the reference soil were detected in the upper 0–10 cm (87.9 g kg−1). Hectare soil organic carbon stocks show high differences between the reference soil (98 Mg ha−1) and the “Dark Earth” (173.5 Mg ha−1). High total P concentrations up to 4318.2 mg kg−1 (10–20 cm) were detected in profile RP M2. In reference profile RP GE, the content of P was only detectable in the upper 40 cm, ranging from 1470.5 mg kg−1 (0–10 cm) to 19.4 mg kg−1. Archaeobotany of the buried soil and the rampart earthworkPlant macroremains The main components of plant macrofossils from profiles RP WW, RP M, RP M2 and RP RM were wood charcoal and charred fruits and seeds. Uncharred diaspores were much less numerous, and very few specimens were mineralised. The occurrence of charred and uncharred specimens in one and the same archaeological feature or layer provokes a question about the origin and age of these two different types of remains. In deposits lying above the ground water level, uncharred specimens usually quickly decay and their relation to a given archaeological context may be doubtful (Lityńska-Zając et al. 2014). At Rozprza, these two types could correspond to two different “histories” of remnants of the same age. Namely, charred specimens could be evidence for intentional human activity and uncharred ones could be accidental admixtures. On the other hand, numerous traces of animal activity observed in several trenches indicate that post-depositional downward transportation of uncharred material from the surface could take place. Therefore, the uncharred specimens are considered here to be secondary intrusions and they are not taken into consideration in the discussion below (All data, including ucharred, are listed in Supplementary Tables 1–4 while in diagrames, Figs 7, 8, 9, only charred remains are considered). The spectra of plant macroremains recovered from geoarchaeological layers are very similar in profiles RP WW and RP M (Tables 5 and 6; Figs. 7 and 8). Bottom layers (S.U. 275 and 274) contain only a small number of wood charcoal pieces. The cultural layer 273B (and 273A in RP WW), recognised as the “Dark Earth” soil, is characterised by a great amount and taxonomic diversity of wood charcoal. Furthermore, the appearance of charred cereals and wild herbaceous plants is noticeable. Profile RP M2 (Table 7; Fig. 9) differs from the former two in the composition of layer 274. This layer contains a higher amount of wood charcoal, one millet grain and a few species of wild herbs. In layer 273, the plant assemblage is similar to that from layer 273A in RP WW. The poorest plant material which is differently distributed in the samples comes from profile RP RM situated outside the rampart (Table 8). A few cereal grains appear as early as layer W3 (corresponding to layer 274 in the other profiles). Their number does not increase in layer W2 which should correspond to layer 273B. Very few plant remnants were found in the sample from a sod brick (S.U. 270) (sample T1, Table 9). Table 5 Charred plant remains from profile RP WW Open image in new window w—wood charcoal, r—rhizome, st—fruit stone, c—grain, s—seed, f—fruit, b—bark, l—leaf, sc—scale, scl—sclerotium, tub—tuber, veg—vegetative, mi—mineralised Table 6 Charred plant remains from profile RP M Open image in new window w—wood charcoal, r—rhizome, st—fruit stone, c—grain, s—seed, f—fruit, b—bark, l—leaf, sc—scale, scl—sclerotium, tub—tuber, veg—vegetative, mi—mineralised Open image in new window Fig. 7 Simplified plant macrofossil diagram from profile RP WW Open image in new window Fig. 8 Simplified plant macrofossil diagram from profile RP M Table 7 Charred plant remains from profile RP M2   Trees and shrubs Cereals Herbs (weeds, ruderals, grassland plants) Ecologically undetermined Depth, cm Volume, dm3 Layer (S.U.) Quercus sp. Pinus sylvestris Picea abies/Larix sp. Alnus sp. Betula sp. Fraxinus excelsior Carpinus betulus Fagus sylvatica Acer sp. Populus sp. Corylus avellana Cornus sanguinea Indeterminata Corylus avellana Picea excelsa Panicum miliaceum Cerealia Galium spurium Polygonum lapathifolium Polygonum persicaria Chenopodium album Galium sylvaticum Chenopodium ficifolium Fallopia convolvulus Cerastium caespitosum Rumex acetosa Veronica officinalis Poaceae Chenopodium sp. Polygonum sp. Galium sp. Veronica sp. Indeterminata     w w w w w w w w w w w w w f l c c f f f s f s f s f f c s f f f f/s 10–15 0.42 273 27 4 7 11 3 3       2 13   1 9 3         1 7 1   1    1 3 15–20 0.35 273 17 5 32 27   6        20    6 2    1     1           20–25 0.35 273 22 18 42 16 3 8    9     28    3 9       1           2 25–30 0.25 273 26 6 21 14 4 3 1     1   14 1   3 2    1            1   2 30–35 0.4 273 103 19 29 10      5     20    4 1      1        1 1     35–40 0.3 273/274 33 4 8 4 3   1 1   4    6        1 1              40–45 0.32 274 251 8 17   3         29      2 1          2       45–50 0.32 274 15 2   11          6    1                   50–55 0.25 274/275 10 1            3                      55–60 0.22 275 8             2                      60–65 0.3 275 1 3 1                                65–70 0.24 275 2   2 1          1                      70–75 0.24 275/PLV 2 4                                 w—wood charcoal, r—rhizome, st—fruit stone, c—grain, s—seed, f—fruit, b—bark, l—leaf, sc—scale, scl—sclerotium, tub—tuber, veg—vegetative, mi—mineralised Open image in new window Fig. 9 Simplified plant macrofossil diagram from profile RP M2 Table 8 Charred plant remains from profile RP RM   Trees and shrubs Cereals Herbs (weeds, ruderals, grassland plants) Ecologically undetermined Depth, cm Volume, dm3 Layer (S.U.) Correlation with S.U. in profile RP WW Picea abies/Larix sp. Pinus sylvestris Quercus sp. Alnus sp. Betula sp. Carpinus betulus Fraxinus excelsior Indeterminata Panicum miliaceum Triticum aestivum s.l. Triticum dicoccon/spelta Triticum sp. Secale cereale Cerealia indet. Agrostemma githago? Galium boreale cf. Plantago media Rumex acetosella agg. Melandrium rubrum Setaria sp. Poaceae Indeterminata Indeterminata      w w w w w w w w c c c c c c s f s f s c c s/f veg 0–5 0.3 W.1    2       5 1 1     2           5–10 0.35 W.1    2       12 3     1        1     10–15 0.35 W.1   2 2 1    1   6 2   3         1    1   15–20 0.3 W.1   4 1 4    1   9 2    1        1      20–25 0.35 W.2 (163) 273B 3 3 1      10 3      1    1      1   25–30 0.2 W.2 (163) 273B   5       16 3      1      1   1    30–35 0.2 W 2/3 (163/164) 273B/274 3 7 3 4     + 3 1             7   35–40 0.2 W.3 (164) 274 2    2 1   1 7       1   1   1     2 1 40–45 0.2 W.3 (164) 274 2 5       6 1 1     1 1        1   45–50 0.3 W.3 (164) 274                         50–55 0.2 W.4 (165) 275 2 2       3                 55–60 0.2 W.4 (165) 275    1     4 2                 60–65 0.3 W.4 (165) 275    3 1     3 1                65–70 0.25 W.4 (165) 275 1 1       1                 70–75 0.2 W.4 (165) bottom 275                         75–80 0.3 W.5 PLV 2                1         w—wood charcoal, r—rhizome, st—fruit stone, c—grain, s—seed, f—fruit, b—bark, l—leaf, sc—scale, scl—sclerotium, tub—tuber, veg—vegetative, mi—mineralised Table 9 Charred plant remains from sample RP T1 Profile RP T1 Trees and shrubs Cultivated plants   Volume, dm3 Layer (S.U.) Picea/Larix Pinus sylvestris Quercus sp. Betula sp. Fraxinus excelsior Populus sp. Indeterminata Sambucus nigra, nch Panicum miliaceum Pisum sp. Indeterminata    w w w w w w w st c s s/f 0.8 270 13 11 5 2 1 2 5 1 3 1 3 w—wood charcoal, r—rhizome, st—fruit stone, c—grain, s—seed, f—fruit, b—bark, l—leaf, sc—scale, scl—sclerotium, tub—tuber, veg—vegetative, mi—mineralised Phytoliths and other biogenic microparticles Practically, all microparticles in all samples from profile RP WW (nos. 7–23) are charred (Table 10). It means that fires often took place in the area or some sort of a hearth (of a kitchen?) was located here in the past. A secondary dump of deposits with charred remains cannot be however ruled out too. The last five samples (nos. 19–23; 65–85 cm; S.U. 274 and 275) have a natural-like phytolith distribution, with a maximum at the top and a decrease with the depth. All the other part of the column is of artificial origin—it is a cultural layer or (most probably) several cultural layers one by one. This confirms the artificial origin of the earthwork of the “Dark Earth” deposits. All samples from this part of the profile contain lot of phytoliths, and their content is higher than at the top of the modern soil. It is possible only in artificial concentration of organic material here. The quantitative content of phytoliths is similar but it is possible to divide them into three parts: samples 7–9 (5–15 cm; S.U. 273), 10–16 (20–50 cm; S.U. 273A, 273B) and 17–18 (55–60 cm; S.U. 273B). We can see the different contents of mosses and reed phytoliths. Samples 11–13 (25–35 cm; S.U. 273B) contain no indicator forms of grasses of steppe land habitats and cereals. Such differences could be random but perhaps they testify to some differences in the history of this site. Sample 23 (85 cm; S.U. 275) can be used as a diagnostic one for the original landscape before it came to an intense human impact connected with the “Dark Earth” soil development. The site had been covered with a broad-leaved forest with admixture of coniferous trees, before people came there. Table 10 Semiquantitative contents of different types of microbiomorphs Sample no. Object Depth, cm Non-siliceous vegetable indicators1 Non-vegetable siliceous indicator types1 Plant silica1 Note Detritus Diatoms Phytolith 1 T1 Top +++ Single, broken +++ Aboreal large detritus predominant, soil fungi. All particles are charred 2 Bottom +++ Single, broken ++ 3 T2 Top +++ − + 4 Bottom +++ − +++ 5 T3 Top +++ − +++ 6 Bottom +++ Single, broken +++ 7 WW 273 A 5 +++ − +++ Charcoal particles of grasses and trees. All particles are charred 8   A 10 +++ Single, unbroken +++ 9   A 15 +++ Single, unbroken +++ 10   A 20 +++ − +++ 11   AB 25 +++ − +++ 12   B 30 +++ Single, broken and unbroken +++ 13   B 35 +++ − +++ 14   B 40 +++ − +++ 15   B 45 +++ − +++ 16   B 50 +++ Single, unbroken +++ 17   B 55 +++ − +++ 18 273B/274 60 +++ Single, broken +++ 19 274 65 ++ − +++ – 20 274 70 ++ − ++   21 275 75 + − + – 22 275 80 ++ − + All particles are charred 23 275 85 + − + 1+++ (many): over 100 units; ++ (middle): 40–100 units; + (little): 5–40 units; ± (single): 1–4 units; (absent): − All particles from the sod bricks—i.e. T1 (nos. 1–2), T2 (nos. 3–4) and T3 (nos. 5–6)—are charred (Table 11). There are no differences between top and bottom samples: all are rich in aboreal large detritus and soil fungi. The content of phytoliths differs but there are no rules for it. For example, in feature T1, we can see more phytoliths in the top sample (no. 1), but a reverse situation can be seen in feature T2. Furthermore, there are no quantitative differences in phytoliths in feature T3. The phytolith assemblages are practically the same in all samples except for the top of T2 (no. 3). Samples are rich in mosses, forms of meadow grasses and herbs. They contain phytoliths of reed and cultural cereals (except for the top and bottom of T2). Three samples (T1’s top and bottom and T3’s bottom—nos. 1 and 6) contain fragments of diatoms (i.a. Hantzschia amphioxys and Pinnularia borealis—M. Rzodkiewicz pers. com.). There are no pollen grains in any samples. Such a microbiomorphical content is typical for meadow–forest litter, which was wet but not bog. Phytoliths of cereals allow to assume that there was some human impact in this area. The charred nature of the samples can be a result of fire. Table 11 Silica microbiomorphs (unit/%) and distribution of phytolith forms (%) Sample Total Diatoms Phytolith total Phytolith distribution Herbs a1 b1 c1 d1 e1 i1 Reed Mosses 1 107/100 3/3 104/97 35 8 8 15 4 2 – 4 24 2 85/100 1/1 84/99 51 7 6 19 – 1 – 6 10 3 36/100 – 36/100 67 8 3 8 – – – – 14 4 294/100 – 294/100 39 3 5 32 3 – – 7 11 5 309/100 – 309/100 36 1 1 38 7 1 – 2 14 6 196/100 1/*2 195/100 46 2 1 26 6 5 – 2 12 7 198/100 – 198/100 36 2 3 44 5 2 – 2 6 8 265/100 1/* 265/100 24 1 5 51 7 1 – 3 8 9 352/100 1/* 351/100 30 1 5 39 8 4 – 4 9 10 392/100 – 392/100 34 3 4 38 – 1 – 2 18 11 228/100 – 228/100 35 2 2 38 – – – – 23 12 350/100 2/1 349/99 32 5 3 37 – – – – 23 13 311/100 – 311/100 37 4 4 29 – – – – 26 14 464/100 – 464/100 39 3 6 20 3 1 – – 28 15 436/100 – 436/100 39 3 8 24 1 1 – – 24 16 455/100 3/1 452/99 36 2 7 24 2 3 – – 26 17 236/100 – 236/100 39 2 8 22 2 – – – 27 18 461/100 1/* 460/100 37 3 8 23 2 * – – 27 19 104/100 – 104/100 42 17 4 25 6 – – 3 3 20 66/100 – 66/100 45 12 6 20 2 – 2 8 5 21 46/100 – 46/100 33 11 13 28 11 – 2 2 – 22 27/100 – 27/100 51 15 7 19 – – – 4 4 23 20/100 – 20/100 70 15 10 – – – – – 5 1a: indicator types of coniferous species; b: trichomes of grasses of forest habitats; c: trichomes of grasses of meadow-type habitats; d: indicator forms of grasses of steppeland habitats; e: dendritic forms—indicator of cultural cereals; i: trichomes of grasses of ruderal habitats 2* particles of less than 1% Discussion Landscape reconstruction of the Rozprza Dark Earth development Archaeobotanical interpretation is based on the plant material found in layers 273B, 273A and 273, which were interpreted as cultural layers (“Dark Earth”) or re-worked cultural layers. The most distinctive trait of the “Dark Earth” horizon as compared with the underlying layers is an increased amount and taxonomic richness of wood charcoal, as well as the appearance of charred seeds and fruits of cereals and wild herbs which are typical for anthropogenic habitats. The analysis of this material provides some data about plant cultivation and gathering. Indirectly, it also informs us about the vegetation in the surrounding terrains which were exploited by the Early Medieval communities, as indicated by the radiocarbon dates. The sets of arborescent taxa are similar in all profiles (Tables 5, 6, 7 and 8; Figs. 7, 8 and 9). The most abundant are charcoal fragments of spruce or larch Picea abies/Larix sp., oak Quercus sp., pine Pinus sylvestris and alder Alnus sp. Fairly numerous are birch Betula sp., hornbeam Carpinus betulus and ash Fraxinus excelsior, while beech Fagus sylvatica, fir Abies alba, poplar Populus sp., maple Acer sp., lime Tilia sp., elm Ulmus sp. and willow Salix sp. are less abundant. Sporadically, and mainly in profile RP WW, there appear yew Taxus baccata, juniper Juniperus communis, dogwood Cornus sanguinea, mistletoe Viscum album, evonymus Euonymus sp., viburnum Viburnum sp., plum tree Prunus sp. and some undetermined species from Pomoideae and Rosaceae. Proportions of various trees are slightly different in individual profiles but this may be insignificant. All these trees and shrubs were undoubtedly available in the near surroundings, though some of them could have been looked for on purpose. Their ecological demands described by the ecological numbers (Zarzycki et al. 2002) suggest that most of them grew on fresh (soil moisture index W 3) or moist soils (W 4). These soils were fertile (trophy index Tr 4) or moderately poor (Tr 3), neutral (acidity index R 4) or moderately acidic (R 3) (Fig. 10). Open image in new window Fig. 10 Indicators of the most typical habitat conditions of plants found at Rozprza after Zarzycki et al. (2002) for layers (S.U.) 273, 273B and 273A in profiles RP WW, RP M and RB M2. For each category, the absolute number of taxa is given According to the reconstruction of the potential natural vegetation in the surroundings of the settlement at Rozprza (Czyżewska 1982), all trees and shrubs identified among charcoal could be gathered in various forest and thicket stands growing in the area. Spruce, pine, fir and larch could originate from pine and pine–oak forests growing on drier soils in the upland. Oak–hornbeam forests covering fresh soils of the upland could be the source of oak, hornbeam, lime and maple wood. Alder, ash and dogwood could be collected in low-lying wet and moist habitats. Cereal cultivation is documented by numerous grains present in all profiles (Tables 5, 6, 7 and 8; Figs. 7, 8 and 9). Most frequently, there occur millet Panicum miliaceum, bread wheat Triticum aestivum and rye Secale cereale (Fig. 11), which belonged to the most common cereals in the Early Middle Ages in Poland (Lityńska-Zając 2005). The cultivation of other plants is indicated by the occurrence of single seeds of flax Linum usitatissimum, opium poppy Papaver somniferum and perhaps field pea Pisum sp. (sativum?) as well as hulled wheats (Triticum monococcum and T. dicoccum or T. spelta) and oats (Avena sp.) (Tables 5, 6, 7 and 8). Open image in new window Fig. 11 Percentage numbers of cereal species in profiles RP WW and RP M All wild herbaceous plants, among which 27 species were identified (Tables 5, 6, 7 and 8), could grow in anthropogenic habitats, most of them in corn fields, some in ruderal places or in meadows. The largest group of field weeds is nowadays typical for root crop and garden cultivations but it also appears in millet fields (Matuszkiewicz 2001). This weed group includes Digitaria ischaemum, Setaria pumila, S. viridis/verticillata (if viridis), Chenopodium album, Ch. polyspermum, Ch. ficifolium, Polygonum minus, P. lapathifolium, Echinochloa crus-galli, Rumex acetosella, Galium spurium and G. aparine. In cereal fields, Polygonum persicaria, Fallopia convolvulus, Sinapis arvensis, Melandrium album and Galium aparine could also grow. It is worth noticing that typical segetal weeds, which until recently have infested particularly winter crops, are represented by only one species, namely Agrostemma githago. This suggests that summer crops together with garden cultivations played (at least locally) a significant role in the surroundings of Rozprza during the Early Middle Ages. All these species often appear in association with cereal grain at archaeological sites of different ages in Poland (Lityńska-Zając 2005). Some species could also come from other habitats, e.g. Malva neglecta, Urtica urens and Chenopodium ficifolium from ruderal places which are rich in nitrogen; Polygonum persicaria, P. minus and P. lapathifolium from periodically flooded shores of water reservoirs; and Rumex acetosa, Plantago media and Silene nutans from meadows and grasslands. Few species could grow in different forest stands, for instance, Galium sylvaticum in oak–hornbeam forests, Silene nutans in lighted oak forests and G. aparine and Melandrium rubrum in riverside forests. Soil preferences of herbaceous plant species defined by ecological numbers (Zarzycki et al. 2002) indicate that fresh (moisture index W 3), fertile (trophy index Tr 4) and neutral (acidity index R 4) soils were mainly taken under cultivation (Fig. 10). The analysis of ecological demands demonstrates a certain difference between the assemblages of trees and herbs: (1) more herbaceous species preferred fresh (W 3) and dry (W2) soils while more arborescent species grew on fresh (W 3) and moist (W 4) soils; (2) more herbs preferred soils which are rich in nutrients (Tr 4), while many woody species grew on moderately poor (Tr 3) and rich soils (Tr 4); (3) herbs preferred neutral soils (R 4), while woody species grew on neutral (R 4) and moderately acidic (R 3) soils (Fig. 10). This indicates some predictable process of taking richer and drier soils under cultivation, which were located in areas safe from spring and summer floods. The recovered plant material offers some suggestions about plant gathering. The rich and diversified set of wood charcoals indicates the collection of wood (mainly for fuel) in the vicinity of the settlement. Forests and forest edges were also the source of edible fruits represented by hazel nuts Corylus avellana and fruit stones of raspberry Rubus idaeus and elder Sambucus nigra. Among the herbaceous plants, there are several species which were collected in the past (e.g. Chenopodium and Polygonum species), but their sporadic occurrence in the material from Rozprza provides no direct evidence for their use. The Rozprza Dark Earth layer—a man-made soil? The documented geological and geochemical traits of stratigraphic unit 273B, recognised as the deep black (“Dark Earth”) buried soil, differ strongly from those of underlying deposits of brownish buried soil (S.U. 275) developed in the top of the Pleni-Weichselian alluvium on the river terrace remnant and deluvial sands (S.U. 274). It is a result of (at least partly) artificial origin of the earthwork of the deep black deposits, as an effect of human intentional deposition of organic rich sediments on the terrace surface. Intentional anthropogenic accumulation of organic matter on cultivated soils was discussed for an early medieval motte in Belgium by Gebhardt and Langohr (1999). A rich admixture of artefacts (potsherds, burned stones) and ecofacts (animal bones, plant macroremains, charcoal) clearly indicate a soil made by man (Anthrosol) (Pluskowski et al. 2014; Mazurek et al. 2016). Furthermore, the enriched total P and δ15N values of profile RP M2 are typical for soils which were manured or contained excrements (Glaser and Birk 2012; Simpson et al. 1997). This is also clearly visible in the reference soil (RP GE) which is currently used for agriculture and obviously manured. Due to the exposed location of profile RP M2 and the presence of a cover offered by the medieval rampart, modern agricultural cultivation can be excluded. In comparison to this, the properties of profile RP M2 suggest strong similarities with the “Nordic Dark Earth” (NDE) phenomenon known from the Slavic settlement reference site in Brünkendorf in Lower Saxony. The term NDE for medieval settlements (and especially in the Brünkendorf study) was defined on the basis of a comparison with the Amazonian Dark Earth. The Slavic black soil bears a strong similarity to the most fertile man-made soils in the world, that is, the Amazonian Dark Earth (=Terra Preta) (Wiedner et al. 2015). For instance, δ15N values of the NDE at Brünkendorf were similarly enriched (up to + 7.1‰) to the δ15N values of the Rozprza Anthrosol (Table 4). SOM stocks of RP M2 are 173.5 Mg ha−1 and even exceed the NDE stocks (121 Mg ha−1) by 43%. The same can be said for the total P stock of the Rozprza Anthrosol which is five times higher as compared with the NDE (5.3 vs. 27.8 Mg ha−1 at Rozprza). However, it will be necessary to carry out further analyses, such as micromorphology combined with molecular marker for soil organic matter inputs, e.g. derived from faecal or charcoal. In sum, the increased soil organic carbon stocks in profile RP M2 indicate high amounts of organic input materials, which is inherent with the δ15N and total P values. However, neither P nor δ15N can be used as a direct marker for faeces or other SOM inputs indicating its sources, e.g. such as burned residues, organic waste or metabolic by-products, e.g. from faecal matter. Thus, more specific biomarkers such as sterols, bile acids and benzenepolycaboxylic acids (black carbon) are indispensable in order to identify the sources of organic soil inputs and to draw any conclusions regarding a possible similar soil origin as the NDE (Wiedner et al. 2015). Except the Brünkendorf site, the “Nordic Dark Earth” phenomenon is also known from southern Scandinavia. The most famous example is Svarta Jorden (which means Black Earth) from Birka, where this phenomenon became eponymous for the site (Ambrosiani 2013). Micromorphological analysis revealed evidence for coproliths and high amounts of charcoal (Håkansson 1997). “Dark Earth” is also reported from excavations in later medieval towns, such as Brussels. It was interpreted there as a sign of pre-urban pastures and urban agricultural activity and gardening (Devos et al. 2009, 2013). The origin of the thick humus horizon of “Dark Earth” which has recently been discussed for Wrocław (SW Poland) has been identified as resulting from triple- and double-depth digging (Rigosol) of the soil used for gardens and orchards in the Late Middle Ages and the Modern Period (Krupski et al. 2017). Based on the results of archaeobotanical research, we can hypothesise that at Rozprza organic matter which accumulated on the terrace surface was collected from different landscapes. These were varied forest, meadow and water habitats. This reflects the economic activity of the settlers during the Early Middle Ages, who used diverse types of the surrounding environment. An effective deposition of sediments which were rich in organic material (including manure), and which were subsequently mixed with clay, charcoal, burned stones, bones and potsherds, resulted in accumulation of thick artificial fertile topsoil on the sandy surface (see also Sokołowski et al. 2008). It could be important that due to poor natural conditions of the terrace remnant in the valley floor, these deep black highly fertile soils were very useful for agricultural development. The process of the ring-fort’s rampart construction The process of the ring-fort’s rampart development can be reconstructed on the basis of the stratigraphy and chronology of identified stratigraphic units (Fig. 12). The surface on the eve of construction was the top of layer 273B (defined as the early medieval “Dark Earth”). At the very beginning (between the second half of the eleventh or the beginning of the twelfth and the end of the thirteenth century AD), sod bricks collected from the original surface were perhaps removed and deposited nearby. A small embankment was erected (S.U. 273A) from the material of the existing “Dark Earth” (S.U. 273B) layer. It could serve as a marker for the builders defining the course of the planned fortification. Then the core of the rampart was created (S.U. 272) with the use of mixed material from layers 273B, 274 and 275 lying closest to the surface. They were probably re-deposited from the simultaneously dug moat. After the removal of layers 273B, 274 and 275, the builders reached yellow sands of the Pleni-Weichselian residual terrace (PLV). Using this material, they erected the top and outer layer of the rampart. Then, it was covered with previously prepared sod bricks. Open image in new window Fig. 12 The process of the ring-fort’s rampart construction: 1. Existence of the tenth–eleventh century AD garden/agricultural “Dark Earth” soil. 2. Removing of sod, construction of the rampart’s core using the “Dark Earth” soil. 3. Construction of the upper part of the rampart with the use of mixed material from the “Dark Earth” soil, the early medieval cultural layer and the sandy substrate of the terrace remnant. 4. Development of the sandy embankment of the rampart with the use of terrace remnant sands and sod brick faces of the rampart, building of a hypothetical upper timber construction The use of sod bricks in the construction of early medieval defensive ramparts, as well as other buildings in Polish territory, has not been reported by archaeologists so far. Such building techniques were usually connected with areas not rich in timber. In Scandinavia, sod or turf bricks were commonly used as building material in Denmark, some regions of Norway, Faeroe Islands, Iceland and Greenland. This phenomenon is seen as an environmental adaptation in areas poor in timber (Urbańczyk 1999, 2004). However, turf was mostly used for wall construction of houses and even churches, and it is known that in some cases, it was also used for defensive architecture. For example, it was applied as building material for the construction of the motte-type castle of Ulfsborg near Ribe, built in 1147/1148 AD (Søvsø 2014). In northwestern Germany, sod bricks were commonly used for the construction of the ramparts of Saxonian fortifications from the ninth to eleventh century AD, despite the fact that other materials, such as sand, clay or timber, were also regularly used (Lemm 2013). In general, there was a significant variety of building techniques for fortifications in the Saxonian territory. On the other hand, these techniques differed from those used in the nearby Slavic lands (Brandt and Schneeweiss 2017). Slavic strongholds of the ninth/tenth centuries were usually sophisticated timber constructions with additional use of sand, clay and stones. The use of sod bricks was very rare and is mostly known from North-West Slavic territories in the area of today’s east Germany. Examples for additional use of sod bricks in Slavic ramparts are offered by the stronghold of Scharstorf (Schleswig-Holstein, Germany; cf. Struve 1975) or Burg A of the Mecklenburg stronghold (Mecklenburg-Vorpommern, Germany; Donat 1984). It is striking that the different building traditions go hand in hand with different agricultural economic systems. Thus, in a generalised perspective, the use of sod bricks as building material seems to be common in regions with extensive agriculture by sod plugging, manuring and pasture of sheep, goat and cattle, while extensive timber constructions seem to be restricted to regions with small fields and wood pasture (Schneeweiss, 2019). The use of sod bricks in the Rozprza ring-fort differs from the way they were used in house and defensive architecture of North Germany and Scandinavia. In these areas, turf was used to create wide, multi-layered wall-like structures. At Rozprza, the builders wanted to cover only the face of the rampart with a single layer of sod bricks in order to protect it against erosion. The use of sod bricks resulted probably from the necessity of fast rampart construction or it was due to the lack of suitable and sufficient quantity of wood material. However, the sandy embankment was unstable and several layers of slope deposits both in the outer and inner side of the construction were identified. This also suggests that an additional structure initially existed on the top of the embankment. It could be constructed as timber boxes or rather wicker baskets filled with sand and covered with clay. After the decomposition or destruction of the wicker construction, the sandy, clay and partly sod bricks were re-deposited by slope wash processes and mass movements. In this way, partly humic sandy deluvial covers (S.U. 269, 267, 261, 338, 336, 337, 310 and 309) with solids of humic sands (268, 263) and clay (262, 333) were deposited (Figs. 4 and 5; Table 3). These processes can be dated back to the period from the end of the thirteenth century to ca. 1330 AD (i.e. to the construction of the motte). Most probably, ca. 1330 AD, the new owners of the stronghold initiated the next phase of construction works. The whole embankment was covered with sands and clayey sands to form a mound that was a platform to build a timber manor house or a keep. The material came from the levelling of part of earlier structures as well as from the extension of the moat. At the foot of the mound, a timber construction was created, consisting of a double vertical palisade and horizontal beams. It strengthened the internal slope of the moat, which was extended to the width of even 21 m wide. It completely destroyed any remains of the earlier moat. The manor house built on the top of the mound was probably in the shape of a tower. Re-deposited layers of clay, carbonates, stones, humic sands and wooden remains suggest that the building was of timber frame construction with wattle and daub walls. Conclusion New research demonstrates that the area was occupied by open settlement from the second half of the tenth century. The earlier ring-fort was established there in the period between the eleventh and thirteenth century as a local administration and military center of the Piast Monarchy. From the fourteenth to fifteenth century, the stronghold in the form of motte-and-bailey functioned as a seat of a noble family (Sikora et al. 2018; Kittel et al. 2018a, b). Pieces of charcoal of trees and shrubs are probably remnants of wood gathered for fuel in forest stands growing on dry (e.g. pine, spruce, oak), fresh (oak, hornbeam, lime, maple) and moist (alder, ash) lands in the surroundings of the settlement. The local agriculture was based on cultivation of bread wheat, rye and millet, possibly also oat. The list of field weeds suggests that summer crops were cultivated and tilling methods characteristic for root crops or gardens were used in the Early Middle Ages. The differences in dating of charred plant macroremains and potsherds can suggest some use of earlier soil (cultural layer). We cannot exclude that some cultivated plots of early medieval origin were located in the area where the Rozprza stronghold was built a few centuries later. The presence of earlier charred remains could be explained by manuring plots with a cultural layer of early medieval origin which contained remains of ashes from fireplaces and other household activities. In such ashes, potsherds would be few in number or strongly fragmented. It is conceivable that the “Dark Earth” soil and the sod bricks, respectively, from Rozprza indirectly reflect not only different environments (wood vs. open landscape) but also different economic systems with differing impacts on the local natural environment. However, the use of sod bricks at Rozprza, which is far away from the traditional “sod brick territory”, is until now unique in Poland and should be explained. The “Dark Earth” soil, in contrast, could be an evidence of a type of economy which was typical at least for the northern Slavic territory until the eleventh century (cf. Wiedner et al. 2015). Notes Acknowledgments We thank Prof. E. Schychowska-Krąpiec and Prof. M. Krąpiec for 14C and dendrochronological dating; Dr. A Nierychlewska for the analysis of potsherds; Dr. M. Romanow for archaeozoological research; and Dr. W. Tołoczko and W. Piech MSc for a part of geochemical analyses. Sedimentological and geochemical analyses were carried out in: the Scientific-Didactic Laboratory Centre of the Institute of Geography of the Jan Kochanowski University in Kielce; the Institute of Agronomy and Nutritional Sciences, Soil Biogeochemistry, Martin Luther University Halle-Wittenberg; and the Pedological Laboratory of the Sub-Department of Environmental Dynamics and Soil Science, University of Lodz. We thank Leica Geosystems-Polska for providing access to SmartNet Leica 4G services. The authors would like to kindly thank the local government of the Rozprza Commune Office and numerous local residents, as well as the Voivodeship Monument Protection Office in Łódź, Branch in Piotrków Trybunalski, for their help and support during our fieldwork. We would like to thank the reviewers for their very constructive remarks which highly improved the final version of the paper. Funding information The research project has been financed from a grant from “The National Science Centre, Poland” (No. “DEC-2013/11/B/HS3/03785”, project: “Environmental condition of development of the medieval ringfort settlement complex in Rozprza (Central Poland) in the light of multidisciplinary research”). Palaeobotanical studies were also financially supported by the W. Szafer Institute of Botany, Polish Academy of Sciences from statutory funds. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary material 12520_2018_753_MOESM1_ESM.doc (122 kb) Supplementary Table 1 (DOC 122 kb) 12520_2018_753_MOESM2_ESM.doc (60 kb) Supplementary Table 2 (DOC 60 kb) 12520_2018_753_MOESM3_ESM.doc (56 kb) Supplementary Table 3 (DOC 56 kb) 12520_2018_753_MOESM4_ESM.doc (82 kb) Supplementary Table 4 (DOC 82 kb) 12520_2018_753_MOESM5_ESM.kml (1 kb) ESM 1 (KML 518 bytes) 12520_2018_753_MOESM6_ESM.tif (2.5 mb) ESM 2 (KML 2.49 MB) References Ambrosiani B (2013) Stratigraphy: excavations in the Black Earth 1990–1995. Part 1. The site and the shore, the bronze caster’s workshop. Birka Studies 9, Birka Project for Riksantikvarieämbetet, Stockholm, 328 ppGoogle Scholar Bengtsson L, Enell M (1986) Chemical analysis. In: Berglund BE (ed) Handbook of Holocene palaeoecology and palaeohydrology. Wiley, Chichester, pp 423–451Google Scholar Berggren G (1969) Atlas of seeds and small fruits of northwest-European plant species with morphological descriptions. Part 2. Cyperaceae. Swedish Nat. Sci. Res. Council, Stockholm, 69 ppGoogle Scholar Brandt J, Schneeweiss J (2017) Ein neuer Blick auf die “Alte Burg” von Hollenstedt im Lichte aktueller Untersuchungen. Hammaburg NF 17:71–145Google Scholar Bronk Ramsey C (2009) Bayesian analysis of radiocarbon dates. Radiocarbon 51(1):337–360CrossRefGoogle Scholar Buko A, Dzieńkowski T (2008) The early medieval settlements at the borderland between Poland and Rus’: the stronghold in Busówno in recent archaeological exploration. Spraw Archeol 60:325–367Google Scholar Cappers RTJ, Bekker RM, Jans JEA (2006) Digital seed atlas of the Netherlands. Barkhuis/Groningen University Library, Groningen 502 ppGoogle Scholar Chmielowska A (1966) Sprawozdanie z badań wczesnośredniowiecznego kompleksu osadniczego w Rozprzy. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi. Seria Archeol 13:249–265Google Scholar Chmielowska A (1975) Osadnictwo w Polsce Środkowej w okresie od połowy X do połowy XIII w. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi. Seria Archeol 22:331–371Google Scholar Chmielowska A (1982) Rozprza we wczesnym i późnym średniowieczu. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi. Seria Archeol 29:159–206Google Scholar Chmielowska A, Marosik P (1989) Wczesnośredniowieczne budownictwo obronne między Prosną a Pilicą. Państwowe Wydawnictwo Naukowe, Warszawa-Łódź, 162 ppGoogle Scholar Cnotliwy E, Leciejewicz L, Łosiński W (Eds.) (1983) Szczecin we wczesnym średniowieczu: wzgórze zamkowe. Zakład Narodowy im. Ossolińskich, Wrocław-Łódź, 348 ppGoogle Scholar Czyżewska K (1982) Środowisko roślinne osadnictwa średniowiecznego w okolicach Roprzy. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi, Seria Archeologiczna 29:153–158.Google Scholar Devos Y, Vrydaghs L, Degraeve A, Fechner K (2009) An archaeopedological and phytolitarian study of the “Dark Earth” on the site of Rue de Dinant (Brussels, Belgium). Catena 78(3):270–284.  https://doi.org/10.1016/j.catena.2009.02.013 CrossRefGoogle Scholar Devos Y, Nicosia C, Vrydaghs L, Modrie S (2013) Studying urban stratigraphy: Dark Earth and a microstratified sequence on the site of the Court of Hoogstraeten (Brussels, Belgium). Integrating archaeopedology and phytolith analysis. Quat Int 315:147–166.  https://doi.org/10.1016/j.quaint.2013.07.024 CrossRefGoogle Scholar Domschke H, Wolff G (1960) Mikroskopische und chemische Untersuchungen vor- und frühgeschichtlicher Kulturschichten auf dem Matthäi-Kirchhof in Leipzig. Forschungen zur Vor- und Frühgeschichte 4, Stadtkernforschung in Leipzig 1. Die Ausgrabungen auf dem Matthäikirchhof, pp 102–111.Google Scholar Donat P (1984) Die Mecklenburg, eine Hauptburg der Obodriten. In: Schriften zur Ur- und Frühgeschichte 37. Akademie-Verlag, Berlin 187 ppGoogle Scholar Dzieduszycki W (1982) Wczesnomiejska ceramika kruszwicka w okresie od 2 połowy X w. do połowy XIV w. Zakład Narodowy im. Ossolińskich, Wrocław, 205 pp.Google Scholar Folk R, Ward W (1957) Brazos River bar: a study in the significance of grain size parameters. J Sediment Petrol 27(1):3–26CrossRefGoogle Scholar Gebhardt A, Langohr R (1999) Micromorphological study of construction materials and living floors in the medieval motte of Werken (West Flanders, Belgium). Geoarchaeology 14:595–620CrossRefGoogle Scholar Glaser B, Birk JJ (2012) State of the scientific knowledge on properties and genesis of Anthropogenic Dark Earths in Central Amazonia (terra preta de Índio). Geochim Cosmochim Acta 82:39–51.  https://doi.org/10.1016/j.gca.2010.11.029 CrossRefGoogle Scholar Golyeva AA (2007) Various phytolith types as bearers of different kinds of ecological information. In: Madella M, Debora Z (eds) Plant, people and places—recent studies in phytolith analysis. Oxbow, Oxford, pp 196–208Google Scholar Goslar T, Czernik J, Goslar E (2004) Low-energy 14C AMS in Poznan Radiocarbon Laboratory, Poland. Nucl Inst Methods Phys Res B 223–224:5–11CrossRefGoogle Scholar Greguss P (1945) Bestimmung der mitteleuropäischen Laubhölzer und Sträucher auf xylotomischer Grundlage. Verlag des Ungarisches Naturwissenschaftliches Museum, Hungarian Museum of Natural History, Budapest, 330 ppGoogle Scholar Grosser D (1977) Die Hölzer Mitteleuropas. Springer, New York 208 ppCrossRefGoogle Scholar Håkansson T (1997) Soil micromorphology with special reference to ‘Dark Earth’. Birka Studies 4:151–154Google Scholar Harris EC (1979) Principles of archaeological stratigraphy. Academic Press, London 136 ppGoogle Scholar Hensel W (1948) Wstęp do studiów nad osadnictwem Wielkopolski wczesnohistorycznej. Biblioteka Prehistoryczna 5, Polskie Towarzystwo Prehistoryczne, Poznań, 244 ppGoogle Scholar Herzog I (2004) Group and conquer—a method for displaying large stratigraphic data sets. British Archaeological Reports International Series 1227, 423–6. Oxford, Archaeopress. [Online]. Available from: http://www.stratify.org. Accessed: 4 July 2015 Hilczerówna Z (1967) Dorzecze górnej i środkowej Obry od VI do początków XI wieku. Zakład Narodowy im. Ossolińskich, Wrocław, 346 ppGoogle Scholar Jackson ML (1958) Soil chemical analysis. Prentice-Hall, Englewood Cliffs 498 ppGoogle Scholar Hołowińska Z (1956) Wczesnośredniowieczne grodzisko w Bonikowie w powiecie kościańskim. Wyniki badań z lat 1951–1953, PWN, Poznań, 86 ppGoogle Scholar Kajzer L (1991) W sprawie waloryzacji masowych zbiorów ceramiki późnośredniowiecznej i nowożytnej. Kwart Hist Kult Mater 39(4):467–484Google Scholar Kajzer L (2007) Polska Centralna w XIII wieku – wstępna próba charakterystyki archeologicznej. Archaeol Hist Pol 16:19–40Google Scholar Kamińska J (1953) Grody Polski środkowej na tle osadnictwa. Łódzkie Towarzystwo Naukowe, Łódź, 267 ppGoogle Scholar Kamińska J (1971) Grody Polski środkowej w organizacji wczesnopaństwowej. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi. Seria Archeol 18:41–74Google Scholar Kara M (2009) Najstarsze państwo Piastów – rezultat przełomu czy kontynuacji?. Studium archeologiczne. Instytut Archeologii i Etnologii PAN, Poznań, 407 ppGoogle Scholar Kats NYa, Kats SV, Kipiani MG (1965) Atlas i opredelitel’ plodov i semyan vstrechayushchikhsya v chetviertichnykh otlozheniyach SSSR (Atlas and keys of fruits and seeds occuring in the Quaternary deposits of the USSR). Izdat. Nauka, Moskva, 365 ppGoogle Scholar Kittel P (2014) Slope deposits as an indicator of anthropopressure in the light of research in Central Poland. Quat Int 324:34–55CrossRefGoogle Scholar Kittel P, Sikora J, Wroniecki P (2015) The morphology of the Luciąża River valley floor in the vicinity of the Rozprza medieval ring-fort in light of geophysical survey. Bulletin of Geography. Phys Geogr Ser 8:95–106.  https://doi.org/10.2478/7033 CrossRefGoogle Scholar Kittel P, Sikora J, Wroniecki P (2018a) A Late Medieval motte-and-bailey settlement in a lowland river valley landscape of Central Poland. Geoarchaeol Int J. 33(5):558–578.  https://doi.org/10.1002/gea.21676 Kittel P, Sikora J, Antczak O, Brooks SJ, Elias S, Krąpiec M, Luoto TP, Borówka RK, Okupny D, Pawłowski D, Płóciennik M, Rzodkiewicz M, Stachowicz-Rybka R, Wacnik A (2018b) The palaeoecological development of the Late Medieval moat—multiproxy research at Rozprza, Central Poland. Quat Int 482:131–156CrossRefGoogle Scholar Kłysik K (2001) Warunki klimatyczne, In: Liszewski S (ed) Zarys monografii województwa łódzkiego. Funkcja regionalna Łodzi i jej rola w kształtowaniu województwa. Łódzkie Towarzystwo Naukowe, Łódź, pp 68–81Google Scholar Kotowicz P (2014) Topory wczesnośredniowieczne z ziem polskich (PhD thesis). Uniwersytet Rzeszowski, RzeszówGoogle Scholar Krąpiec M, Walanus A (2011) Application of the triple photomultiplier liquid spectrometer Hidex 300SL in radiocarbon dating. Radiocarbon 53:543–550CrossRefGoogle Scholar Kruppe J (1981) Garncarstwo późnośredniowieczne w Polsce. Zakład Narodowy im. Ossolińskich, Wrocław-Warszwa-Kraków, 233 ppGoogle Scholar Krupski M, Kabala C, Sady A, Gliński R, Wojcieszak J (2017) Double- and triple-depth digging and Anthrosol formation in a medieval and modern-era city (Wrocław, SW Poland). Geoarchaeological research on past horticultural practices. Catena 153:9–20.  https://doi.org/10.1016/j.catena.2017.01.028 CrossRefGoogle Scholar Lalik T. (1967) Organizacja grodowo – prowincjonalna w Polsce XI i początków XII wieku. Studia z Dziejów Osadnictwa 5:5–51Google Scholar Lemm, T., 2013. Die frühmittelalterlichen Ringwälle im westlichen und mittleren Holstein. Schriften des Archäologischen Landesmuseums 11, Schleswig, 764 pp.Google Scholar Lityńska-Zając M (2005) Chwasty w uprawach roślinnych w pradziejach i wczesnym średniowieczu. Instytut Archeologii i Etnologii Polskiej Akademii Nauk, Kraków 444 ppGoogle Scholar Lityńska-Zając M, Wasylikowa K, Cywa K, Madeyska E, Tomczyńska Z (2014) Badania archeobotaniczne na stanowiskach 10, 11 i 12 w Targowisku, pow. wielicki. Osady i cmentarzysko kultury łużyckiej, In: Górski J (ed) Kompleks osadniczy kultury łużyckiej w Targowisku, stan 10–12, pow. wielicki. Via archaeologica. Źródła z badań wykopaliskowych na trasie autostrady A4 w Małopolsce. Krakowski Zespół do Badań Autostrad, Kraków, pp 243–276Google Scholar Łosiński W, Rogosz R (1986) Próba periodyzacji ceramiki wczesnośredniowiecznej ze Szczecina, In: Gromnicki J (ed) Problemy chronologii ceramiki wczesnośredniowiecznej na Pomorzu Zachodnim. PKZ, Warszawa, pp 51–61Google Scholar Matuszkiewicz W (2001) Przewodnik do oznaczania zbiorowisk roślinnych Polski. Wydawnictwo Naukowe PWN, Warszawa, 536 ppGoogle Scholar Matuszkiewicz JM (2008) Potencjalna roślinność naturalna Polski (Potential natural vegetation of Poland), IGiPZ PAN, Warszawa http://www.igipz.pan.pl/Roslinnosc-potencjalna-zgik.html. Accessed 10 May 2017 Mazurek R, Kowalska J, Gąsiorek M, Setlak M (2016) Micromorphological and physico-chemical analyses of cultural layers in the urban soil of a medieval city—a case study from Krakow, Poland. Catena 141:73–84CrossRefGoogle Scholar Mycielska-Dowgiałło E (2007) Metody badań cech teksturalnych osadów klastycznych i wartość interpretacyjna wyników, In: Mycielska-Dowgiałło, E., Rutkowski J. (Eds.), Badania cech teksturalnych osadów czwartorzędowych i wybrane metody oznaczania ich wieku. Wydawnictwo Szkoły Wyższej Przymierza Rodzin, Warszawa, pp 95–180Google Scholar Mycielska-Dowgiałło E, Ludwikowska-Kędzia M (2011) Alternative interpretations of grain-size data from quaternary deposits. Geologos 17:189–203CrossRefGoogle Scholar Nalepka D, Walanus A (2003) Data processing in pollen analysis. Acta Palaeobot 43(1):125–134Google Scholar Niewęgłowski A (1984) Grodzisko w Błoniu-Osieku, woj. Warszawa, w świetle badań w latach 1975 i 1976. Sprawozdania Archeologiczne 36:193–212Google Scholar Olczak H, Krasnodębski D (2002) Wczesnośredniowieczne miejsce kultu i osada w Mołoczkach, pow. Bielsk Podlaski, woj. Podlaskie. Sprawozdania Archeologiczne 54:141–171Google Scholar Passega R (1964) Grain size representation by CM patterns as a geological tool. J Sediment Petrol 34:830–847CrossRefGoogle Scholar Passega R, Byramjee R (1969) Grain size image of clastic deposits. Sedimentology 13:830–847CrossRefGoogle Scholar Pluskowski A, Sawicki Z, Shillito L, Badura M, Makowiecki D, Zabilska-Kunek M, Seetah K, Brown A (2014) Biała Góra: the forgotten colony in the medieval Pomeranian-Prussian borderlands. Antiquity 88(341):863–882CrossRefGoogle Scholar Reimer PJ, Bard E, Bayliss A, Beck JW, Blackwell PG, Bronk Ramsey C, Grootes PM, Guilderson TP, Haflidason H, Hajdas I, Hatt C, Heaton TJ, Hoffmann DL, Hogg AG, Hughen KA, Kaiser KF, Kromer B, Manning SW, Niu M, Reimer RW, Richards DA, Scott EM, Southon JR, Staff RA, Turney CSM, van der Plicht J (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55(4)Google Scholar Rosin R. (1970).Rozwój polityczno – terytorialny łęczyckiego – sieradzkiego i wieluńskiego (do przełomu XIV i XV w.). Rocznik Łódzki 14(17):277–304.Google Scholar Schneeweiss J (2019) Zwischen den Welten. Archäologie einer europäischen Grenzregion zwischen Sachsen, Slawen, Franken und Wikingern. Wachholtz-Verlag, Hamburg-NeumünsterGoogle Scholar Schuldt E (1956) Die slawische Keramik in Mecklenburg. Akademie-Verlag, Berlin 74 ppGoogle Scholar Schweingruber FH (1978) Mikroskopische Holzanatomie. Available at Swiss Federal Institute for Forest, Snow and Landscape Research, CH-8903 Birmensdorf, 226 ppGoogle Scholar Schweingruber FH, Börner A, Schulze ED (2011) Atlas of stem anatomy in herbs, shrubs and trees, vol 1. Springer, Berlin 495 ppCrossRefGoogle Scholar Sikora J (2007) Polska Centralna we wczesnym średniowieczu w świetle badań archeologicznych i osadniczych. Slavia Antiqua 48:121–159Google Scholar Sikora J (2009) Ziemie Centralnej Polski we wczesnym średniowieczu. Studium archeologiczno-osadnicze. Instytut Archeologii Uniwersytetu Łódzkiego, Łódź, 238 ppGoogle Scholar Sikora, J., Kittel, P., (2018). Closing a gap with a simple toy. How the use of the tablet affected the documentation workflow during the excavations of the Rozprza ring-fort (Central Poland), CAA ProcedingsGoogle Scholar Sikora J, Kittel P, Wroniecki P (2015a) Nieinwazyjne badania grodzisk wczesnośredniowiecznych Polski Centralnej i ich zaplecza osadniczego: Chełmo, Rękoraj, Rozprza, Stare Skoszewy, Szydłów. Prace i Materiały Muzeum Archeologicznego i Etnograficznego w Łodzi, Seria Archeologiczna 46:253–293.Google Scholar Sikora J, Kittel P, Wroniecki P (2015b) From a point on the map to a shape in the landscape. Non-invasive verification of medieval ring-forts in Central Poland: Rozprza case study. Archaeol Pol 53:510–514Google Scholar Sikora J, Sroka J, Tyszkiewicz J (2016) ‘Strati5 - Open Mobile Software for Harris Matrix’, In: Campana S, Scopigno R, Carpentiero G, Cirillo M (eds) CAA2015. Keep the revolution going. Proceedings of the 43rd annual conference on computer applications and quantitative methods in archaeology, Oxford, Archaeopress Archaeology, pp 1005–1014Google Scholar Sikora J, Kittel P, Wroniecki P (2018 in press) W poszukiwaniu milenijnego mirażu. Nowe badania grodziska w Rozprzy (pow. piotrkowski, woj. łódzkie). ŚwiatowitGoogle Scholar Simpson IA, Bol R, Dockrill SJ, Petzka K-J, Evershed RP (1997) Compound-specific d15N amino acid signals in palaeosols as indicators of early land use: a preliminary study. Archaeol Prospect 4:147–152CrossRefGoogle Scholar Smolska E (2008) Badania sedymentologiczne współczesnych osadów deluwialnych i ich znaczenie w ocenie dynamiki procesów erozji gleby na przykładzie Pojezierza Suwalskiego. Assessment of soil erosion processes based on sedimentological analyses of contemporary deluvial sediments on the example of the Suwałki Lakeland. Przegląd Naukowy Inżynierja i Kształtowanie Środowiska 2(40):202–212Google Scholar Sokołowski T, Wacnik A, Wardas M, Pawlikowski M, Pazdur A, Madeja J, Woronko B, Madej P (2008) Changes of natural environment in Kraków Downtown—its chronology and directions. Case geoarchaeological studies of Krupnicza Street site. Geochronometria 31:7–19CrossRefGoogle Scholar Søvsø M (2014) Riber Ulfs borg, Arkæologi i Slesvig. Archäologie in Schleswig 15:191–203Google Scholar Struve KW (1975) Ziel und Ergebnisse von Untersuchungen auf drei slawischen Burgwällen Ostholsteins, In: Ausgrabungen in Deutschland gefordert von der Deutschen Forschungsgemeinschaft 1950–1975. Teil 3. Römisch-Germanisches Zentralmuseum. Monographien 3, Mainz, pp 98–122Google Scholar Szyszka J (2007) Gród w Rozprzy i najdawniejsza struktura własnościowa opola rozpierskiego. Rocz Hist 73:57–81Google Scholar Twardy J (2000) Deluwia neoholoceńskie - przykłady z Wyżyny Łódzkiej. Acta Geograph Lodz 78:135–173Google Scholar Twardy J (2008) Transformacja rzeźby centralnej części Polski Środkowej w warunkach antropopresji. Wydawnictwo Uniwersytetu Łódzkiego, Łódź, 292 pp.Google Scholar Urbańczyk P (1999) North Atlantic turf architecture as an example of environmental adaptation. Archaeol Pol 37:119–133Google Scholar Urbańczyk P (2004) Zdobywcy północnego Atlantyku. Wydawnictwo Uniwersytetu Wrocławskiego, Wrocław 293 ppGoogle Scholar Velichkevich FY, Zastawniak E (2006) Atlas of Pleistocene vascular plant macroremains of Central and Eastern Europe, Part I—pteridophytes and monocotyledons. W. Szafer Institute of Botany, Polish Academy of Sciences, Krakow 224 ppGoogle Scholar Velichkevich FYu, Zastawniak E (2008) Atlas of vascular plant macroremains from the Pleistocene of central and eastern Europe, part II—herbaceous dicotyledons. W. Szafer Institute of Botany, Polish Academy of Sciences, Krakow, 379 ppGoogle Scholar Wachecka-Kotkowska L (2004) Ewolucja doliny Luciąży - uwarunkowania klimatyczne a lokalne. Acta Geograph Lodz 86:1–161Google Scholar Wiedner K, Schneeweiß J, Dippold MA, Glaser B (2015) Anthropogenic Dark Earth in Northern Germany—the Nordic analogue to terra preta de Índio in Amazonia. Catena 132:114–125CrossRefGoogle Scholar Wlaźlak P (2009) Sanktuarium Najświętszej Maryi Panny Matki Życia w Rozprzy. Wydawnictwo Archidiecezjalne Regina Poloniae, Częstochowa, 74 ppGoogle Scholar Woszczyk M, Szczepaniak M (2008) Reevaluation of the Scheibler method and its usefulness in the analysis of carbonate content in lake sediments. In: Bajkiewicz-Grabowska E, Borowiak D (eds) Anthropogenic and natural transformations of lakes, vol 2. Department of Limnology, University of Gdańsk, Polish Limnological Society, Gdańsk, pp 223–226Google Scholar Zajączkowski S, Zajączkowski SM (1970) Materiały do słownika geograficzno-historycznego dawnych ziem Łęczyckiej i Sieradzkiej do 1400 roku. Część II. Łódzkie Towarzystwo Naukowe, Łódź, 248 ppGoogle Scholar Zajączkowski SM (1961) Uwagi nad najdawniejszymi dziejami Rozprzy. Rocz Łódz 5:197–202Google Scholar Zarzycki K, Trzcińska-Tacik H, Różański W, Szeląg Z, Wołek J, Korzeniak U (2002) Ecological indicator values of vascular plants of Poland. W. Szafer Institute of Botany, Polish Academy of Sciences, Krakow, 183 ppGoogle Scholar Zoppi U, Crye J, Song Q, Arjomand A (2007) Performance evaluation of the new AMS system at Accium Biosciences. Radiocarbon 49:173–182CrossRefGoogle Scholar Written sources ADKPwR (1588) Akta dziekańskie kościoła parafialnego w Rozprzy dekanatu piotrkowskiego, Archiwum Archidiecezji Częstochowskiej im. ks. Walentego Patykiewicza w CzęstochowieGoogle Scholar DFKR (1521) Dokumenta funduszu Kościoła Rozpierskiego zebrane w roku 1803, Archiwum Archidiecezji Częstochowskiej im. ks. Walentego Patykiewicza w Częstochowie.Google Scholar KH (1885) Księgi Hipoteczne, Archiwum Państwowe w Piotrkowie.Google Scholar KDMiPR (1754) Kopiarz dokumentów miasta i parafii Rozprza, Archiwum Archidiecezji Częstochowskiej im. ks. Walentego Patykiewicza w Częstochowie.Google Scholar Łaski J (1521) Liber beneficiorum, Vol. 2, Łukowski, J. (Ed.), Gniezno 1831.Google Scholar Copyright information © The Author(s) 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Authors and Affiliations Jerzy Sikora1View author's OrcID profilePiotr Kittel2Email authorView author's OrcID profileMarcin Frączek3Zbigniew Głąb4Alexandra Golyeva5Aldona Mueller-Bieniek6View author's OrcID profileJens Schneeweiß7Zofia Tomczyńska6Krystyna Wasylikowa6Katja Wiedner81.Department of Historical Archaeology and Weapon Studies, Institute of ArchaeologyUniversity of LodzŁódźPoland2.Department of Geomorphology and Palaeogeography, Faculty of Geographical SciencesUniversity of LodzŁódźPoland3.Institute of GeographyJan Kochanowski University in KielceKielcePoland4.Foundation for HumanitiesPabianicePoland5.Institute of GeographyRussian Academy of SciencesMoscowRussian Federation6.W. Szafer Institute of BotanyPolish Academy of SciencesKrakówPoland7.Institute of History of Material CultureRussian Academy of SciencesSaint PetersburgRussian Federation8.Institute of Agronomy and Nutritional Sciences, Soil BiogeochemistryMartin Luther University Halle-WittenbergHalleGermany


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2Fs12520-018-0753-0.pdf

Jerzy Sikora, Piotr Kittel, Marcin Frączek, Zbigniew Głąb, Alexandra Golyeva, Aldona Mueller-Bieniek, Jens Schneeweiß, Zofia Tomczyńska, Krystyna Wasylikowa, Katja Wiedner. A palaeoenvironmental reconstruction of the rampart construction of the medieval ring-fort in Rozprza, Central Poland, Archaeological and Anthropological Sciences, 2019, 1-33, DOI: 10.1007/s12520-018-0753-0