Skip to navigation – Site map
Case studies

Fluvial dynamics and phases of landscape development in the Bronze Age settlement area of the Sambek valley (northeastern hinterland of the Sea of Azov)

Dynamiques fluviales et mobilité des paysages dans la vallée du Sambek à l’âge du bronze (nord-est de la mer d’Azov)
Marlen Schlöffel, Steffen Schneider and Brigitta Schütt
p. 91-100


We present results of a geomorphological-sedimentological study on the Holocene evolution and the Bronze Age palaeoenvironment of a lowland river valley in the Russian northeastern Azov Sea hinterland. The study investigates the valley of the Sambek, 15 km west of the Don Delta. This area has been colonized since the Bronze Age, as settlements and burial mounds testify. The results show that in the early Holocene the valley floor was captured by a meandering river system and sedimentation was dominated by autochthonous processes in oxbow environments. From the mid-Holocene until about fifty years ago, allochthonous processes in the form of overbank deposition prevailed. After the construction of reservoirs along the river course in the 1960s, the fluvial system changed back to autochthonous depositional processes. The prehistoric changes of the fluvial dynamics do not coincide with settlement phases. We assume they were climatically induced. Since the Bronze Age, siltation has caused a 3–4 m elevation of the valley floor.

Top of page

Full text

This study was supported by the financial, technical and scientific help of the DFG Cluster of Excellence 264 “Topoi”. We thank Leon van Hoof (German Archaeological Institute), PhD Irina Tolochko (Southern Scientific Center of the Russian Academy of Sciences) and Dr. Alexey Porotov (Lomonosov Moscow State University) for advice and fruitful discussions, as well as Stefan Schimpf (Alfred Wegener Institute Potsdam) for his help during field work. We are grateful to Ruth Strunk (University of Basel) for C/N analysis and Dr. Reinder Neef (German Archaeological Institute) for identifying the botanical macro-remains. We are especially thankful to Prof. Dr. Ortwin Dally and his team from the German Archaeological Institute and to the Don Archaeological Society for supporting the fieldwork. Finally, we thank Ginger Diekmann (textetage) for improving the manuscript language.

1The Azovian Plain is a slightly undulating steppe landscape northeast of the Black Sea, between the Donetsk Upland in the north and the Sea of Azov in the south. The region has been settled at least since late Palaeolithic times, but most of the numerous archaeological sites can be attributed to the Bronze Age, the Iron Age, the period of Greek colonization, and the Middle Ages (VAN HOOF and SCHLÖFFEL, 2016; PARZINGER, 2011). The most prominent archaeological feature is the up to 6,000–year–old burial mounds (in Russian, kurgans; RASSAMAKIN, 2004). Whereas the kurgans are most often situated on widely visible ridges, the prehistoric settlements, in contrast, are mostly located in the river valleys (VAN HOOF et al., 2012a).

2On a regional scale, Holocene climate and landscape changes in the steppe zone north of the Black Sea have been intensively studied (GERASIMENKO, 1997; ALEXANDROVSKI and CHICHAGOVA, 1998; KREMENETSKI et al., 1999; SPIRIDONOVA and ALESHINSKAYA, 1999; CHENDEV and IVANOV, 2007; KOTOVA and MAKHORTYKH, 2010). However, the local environmental conditions in the steppe valleys during the Holocene are still largely unknown. How were the fluvial dynamics in the valleys, and what did the valley bottom look like? Did the valleys change under the influence of the early settlers?

3We address these questions by investigating the middle course of a typical river valley of this region, the Sambek valley in the Russian Rostov department (or oblast in Russian). The Sambek is a lowland river draining into the Sea of Azov, disemboguing 15 km west of the Don Delta (Fig. 1), with its headwaters 35 km north of the Taganrog Bay. Its catchment of 680 km² mainly forms part of the Azovian Plain, only in the north it extends into the Donets Range.

Fig. 1 ‑ Topographical and geological settings of the Sambek valley and its surroundings

Fig. 1 ‑ Topographical and geological settings of the Sambek valley and its surroundings

Source: Anonymous, 1966, modified.

4The Sambek valley has been of archaeological interest since the early decades of the twentieth century (VAN HOOF et al., 2012a, 2012b). The valley and its environs were investigated during an extensive archaeological survey in the early 1990s. Prior investigations and excavations concentrated on the lower valley and on the delta area (VAN HOOF et al., 2012a, b). To date no research has been conducted on the Holocene landscape development and the assessment of the prehistoric environmental settlement conditions of the Sambek valley.

5The main questions of our geomorphological and sedimentological study are as follows:

  • What were the environmental settings in the middle Sambek valley before, during and after the first colonization in the Bronze Age in the fourth millennium BP?

  • Is it possible to detect any influence of human colonization of the valley on the landscape?

1 - Regional setting

6The focus of our study is on the middle section of the Sambek valley. This area is located within the Azovian Plain at the southern boundary of the East European Platform. The Precambrian crystalline basement of the plain is covered by a 400 to 500 m thick sedimentary sequence of Cretaceous and Cainozoic marine deposits (clay, sands and limestone) (DOLUKHANOV, 2009; TESAKOV et al., 2010), which are overlain by Pleistocene loess-soil series (VELICHKO et al., 2009a). The Azovian Plain gently dips southwards and determines the orientation of the hydrological network. The river valleys and the U-shaped dry valleys (in Russian, Balki) dissect the terrain into low‑lying plateaus with protruding ridges and summits. The Balki are assumed to represent the final stage of periglacial valley development (SYCHEVA et al., 2003).

7In terms of ecological-geographical zoning, the Azovian Plain forms part of the steppe zone with a temperate continental climate and variable precipitation. The mean annual temperature of the coldest month is about -5 °C (January) and for the warmest month about 23 °C (July). The average annual precipitation is higher than 500 mm. The lowest monthly precipitation occurs in February and March (39–43 mm per month) (KHRUSTALEV et al., 2002). Arid conditions are characteristic of the late summer months, with dry winds enhancing the moisture deficit. The hydrological system directly correlates with the climate. The small rivers are of variable stream flow and dry up locally. Flooding occurs after snow melts and during high precipitation in May and June (56–63 mm per month) (KHRUSTALEV et al., 2002). The predominant soils of the region are ordinary and southern chernozems, which have developed on the calcareous loess and loess-like loam (PESOCHINA, 2010). The natural vegetation is forb-rich grassland dominated by sheep fescue, feather grass and xerophilous herbs. Populations of trees and shrubs occur only along rivers and in gullies or ravines (BERG, 1950; WALTER, 1974). Because of intensive land use dating back to Soviet times, soil and vegetation cover are degraded and river courses are regulated and dammed. After the end of the Soviet Union, the agricultural use has decreased significantly (SUKHODOLOV et al., 2009). However, the region is still one of Russia’s important sources for agricultural goods. About 85 % of the area is arable or pasture land and the region is a leader in keeping and breeding beef and dairy cattle (MINORANSKIJ, 2006).

2 - Methods

8Geomorphic landforms and processes were mapped during a field survey following the methodology of H. LESER and G. STÄBLEIN (1980). Two sediment cores were obtained by means of a Wacker vibracoring device (BHF 30 S) with a 5 cm core diameter in closed plastic tubes (Table 1).

Tab. 1 ‑ Coordinates of the two drilling sites

Tab. 1 ‑ Coordinates of the two drilling sites

9The cores were described, documented and sampled according to lithostratigraphic units or at 10 cm intervals. The grain size composition was described in accordance with the guidelines of the German manual of soil mapping (AD-HOC-AG BODEN, 2005). The sediment colour was determined by means of the Munsell soil color chart. The volume magnetic susceptibility (ĸ) was measured through use of the Bartington Instruments MS2C system at 2 cm intervals on bisected plastic tubes. Prior to chemical analysis, samples were dried at 105 °C and passed through a 2 mm sieve. The electrical conductivity (EC) was measured with a Hanna Checker device (HI 98301), using a ratio of sediment to distilled water of 1:2.5. The loss-on-ignition, total carbon content and total inorganic carbon content were measured as described by S. SCHNEIDER et al., (2014). By means of a LECO CHN 2000 instrument the total carbon and nitrogen contents were measured in order to calculate the organic-carbon-to-total-nitrogen ratio (C/N ratio). Ten samples – organic sediments, pieces of wood and pieces of charred wood – from the two profiles were prepared for AMS dating (14C) and measured at the Poznan Radiocarbon Laboratory (Table 2). The AMS dates were calibrated with the IntCal13 calibration curve (REIMER et al., 2013) in OxCal 4.2.3 (BRONK RAMSEY, 2009).

3 - Results

3.1 - Geomorphic setting and archaeological sites

10The study area is located in the middle course of the Sambek River, about 10 km north of the coast of Taganrog Bay. It is a basin-like widening of the valley where the rivers Sambek and Biryuchya and the Balki Buzikova and Landina meet. The area under investigation is differentiated into denudational plateaus (I), slopes (II) and the fluvial areas of the valleys (III) and comprises several archaeological sites (Fig. 2).

Fig. 2

Fig. 2

a) Geomorphological setting of the study area, b) EW-transect through the Sambek valley, c) The Sambek valley

© M. Schlöffel, July 2009.

11(I) Most of the study area is occupied by undulating loess plateaus, in which the Sambek and its tributaries have incised. The plateaus’ altitudes slightly decrease towards the valleys from 88 to 45 m above sea level. Calcareous chernozems have developed on the loess deposits. The plateaus are under tillage; huge fields of crop and sunflower cover the wide and level areas. Deflation and surface wash are the predominating processes affecting this geomorphic unit. The highest parts of the plateaus are occupied by numerous linearly arranged burial mounds.

12(II) The slopes of the Sambek valley are asymmetric, which is characteristic of the valleys of the lowland rivers in this region (VAN HOOF et al., 2012a, b; Fig. 2b). The gently inclined eastern slope (2–6°) has, in general, a straight longitudinal profile, and is under tillage. Its slope toes are characterized by slumping. The western slopes are steeper (4–18°), typically convex and unused. On the foot slopes, Bronze Age settlement sites and burial mounds of uncertain age have been identified. Clusters of burial mounds also occur on the medial eastern slope (Fig. 2; VAN HOOF et al., 2012a, b).

13(III) At its basin-like widening, the Sambek valley floor has a width of 1 km and a maximum elevation difference to the plateaus of about 35 m. Within the valley widening, the Sambek tends to meandering. The water current is of lentic character, and the floodplain is densely covered with reed. As a result of high flood episodes, numerous chute channels cross the valley floor as well as the toes of the eastern, northern and western slopes. Since the 1960s, the stream flow of the Sambek has been regulated and retained by several dams (MINASHINA, 2009; SUKHODOLOV et al., 2009).

3.2 - Lithostratigraphy

3.2.1 - Sediment profile 1

14Sediment profile 1 was retrieved close to Bronze Age settlement site S1, on the toe of the eastern slope of the Sambek valley (Fig. 2). The site is about 2.5 m above the floodplain and located north of the outlet of the Balka Landina, close to the outer bank of an abandoned meander of the Sambek River.

15The stratigraphy shows a clear bisection of the profile (Fig. 3).

Fig. 3 – Stratigraphy and chemical parameters of Profiles 1 and 2

Fig. 3 – Stratigraphy and chemical parameters of Profiles 1 and 2

TOC: total organic carbon, TIC: total inorganic carbon, C: carbon, N: nitrogen, EC: electric conductivity, Mag. Suscep.: volume magnetic susceptibility.

16The underlying unit 1 (5‑1.9 m depth) has a homogeneous, yellowish-brown, clast-free, calcareous and silt-dominated matrix. Grain sizes alternate between clayey silt and loamy clay. Black, vertical traces indicating root channels appear at a depth of 3.3–2.0 m. The total organic carbon content (TOC) is below 0.9 mass-%. The total inorganic carbon content (TIC) fluctuates between 0.5 and 1.7 mass‑%. The magnetic susceptibility has a mean value of 22*10-5 SI. The overlying unit 2 (1.9–0 m depth) is characterized by dense rooting and high humus accumulation in a dark-brown matrix of clayey silt. The total organic carbon is about 1.3 mass‑%, with a maximum of 2.6 mass‑%, at 0.4–0 m depth. The mean total inorganic carbon is 0.3 mass‑%. The magnetic susceptibility varies between 46 and 70*10-5 SI.

3.2.2 - Sediment profile 2

17Sediment Profile 2 was retrieved 200 m north of Profile 1 in the Sambek floodplain within an abandoned and, at least during the fieldwork in summer, dried-up channel (Fig. 2). Over its depth of 9 m, the profile is divided into ten sediment units (Fig. 3).

18Units 1 (9–8.4 m depth) and 3 (7.8–7.4 m depth) at the base of the profile consist of very dark grey clay with horizontally laminated, thinly interbedded sands and well-preserved plant remains, among them remnants of Phragmites spec., Acorus calamus and Schoenoplectus spec., and fruits of Najas marina. The clast-free matrix is interspersed with debris of molluscs and particles of charcoal. The high content of plant macro remains is well reflected by the total organic carbon content of 2.7 mass-% on average. The C/N ratios range from 15 to 22. Unit 2 (8.4–7.8 m depth) is composed of poorly sorted coarse sand and coarse debris of molluscs with clast-supported sandy detritus at the top of the unit. The sediments are almost free of total organic carbon (0.2 mass-%), whereas their total inorganic carbon content totals 5.8 mass-%. The electric conductivity is below 1,500 µS cm-1. In units 4 to 6, dark-grey clayey sediments and coarse sands alternate. Unit 4 (7.4–7.2 m depth) is homogeneous and interspersed with charred organic particles and fine mollusc debris, whereas sediment Unit 6 (7.0–6.8 m depth) is loamier, containing few fine detritus and a few thinly laminated sands, and an intercalated sand lens of 2 cm. The total organic carbon content totals 0.8 mass-% in unit 4 and decreases upwards; the total inorganic carbon content increases upwards and reaches values around 3 mass-% in unit 6. Unit 5 (7.2–7.0 m depth) is an ungraded sand layer and contains much shell debris but no organic matter (TOC 0 mass-%). The total inorganic carbon content averages 4.8 mass-% and the electric conductivity is about 1150 µS cm-1. Unit 7 (6.8–5 m depth) consists of dark-grey sandy loam, interspersed with fine to coarse shell debris. The matrix is homogeneous and ungraded, except at the base of the unit, where coarse sands with detritus and fine shells are deposited. Black vertical traces pervade unit 7, mainly in the upper meter. The C/N ratio varies between 13 and 42. Units 8–10 have a clast-free matrix and show a fining-upward sequence from clayey loam to loamy clay. Unit 8 (5–2.8 m depth) is dark grey and characterized by a gradual decrease of total inorganic carbon content from the bottom to the top and concurrently an increase of the organic carbon contents found in units 9 (2.9–1.9 m depth) and 10 (1.9–0 m depth). The C/N ratio fluctuates between 12 and 26. A high concentration of iron and manganese concretions occurs at a depth of 4–2.8 m. Units 9 and 10 are characterized by very dark grey to black colour, fine roots, a total organic carbon content of 2 mass-%, constant C/N ratios of on average 12, upward increasing electric conductivity and magnetic susceptibility values higher than 20*10-5 SI.

3.3 - Chronology

19The calibrated ages from seven samples of organic sediment, two samples of indeterminable wood and one sample of indeterminable charred wood cover the Holocene back to c. 12,000 cal. BP. Sample Poz-37325 contained less than 1 mg of carbon and therefore must be evaluated with caution (Table 2). See section 4.2 and Fig. 3 for the interpretation of the datings within their lithostratigraphical context and the discussion of the chronological model.

Tab 2 ‑ Results of radiocarbon dating at the Poznan Radiocarbon Laboratory

Tab 2 ‑ Results of radiocarbon dating at the Poznan Radiocarbon Laboratory

Calibrated with OxCal 4.2.3 (Bronk Ramsey, 2009) and the IntCal13 calibration curve

Reimer et al., 2013.

4 - Discussion

4.1 - Facies interpretation

20As a consequence of the seasonally fluctuating groundwater level, hydromorphic processes have overprinted the primary depositional features (Fig. 3). Therefore, we did not use stratigraphy-discordant features such as carbonate and manganese precipitations or iron and gypsum concretions for the facies interpretation.

21The sediments observed in the two profiles relate to a reworked aeolian system (profile 1) and a fluvial system (profile 2). In profile 1, unit 1 shows some of the typical characteristics of the loess deposits of the Azovian Plain, such as the yellow-brown colour and the silty calcareous matrix (VELICHKO et al., 2009a). However, the geomorphologic position on the slope toe and variations in the grain size composition imply that the loess was not deposited in situ, but was reworked. The rework processes could be of alluvial or fluvial nature. Unit 2 shows the typical features of the A-horizon of the regional chernozems (PESOCHINA, 2010), which have developed in the loess and the loess-like sediments. Compared to the loess-like parent material (unit 1), the A-horizon has higher soil organic matter, is unstratified, almost decalcified and has higher magnetic susceptibility values.

22In profile 2, all units exhibit the character of the facies of an oxbow environment. According to the genetic facies classification developed by K. WOJCICKI (2006), the fine-grained units 1 and 3 fit in with the autochthonous “biochemical facies” [bc], which is well proven by the high content of macro remains of aquatic plants that are typical of habitats of stagnant or slow flowing water of about 1–3 m depth (HANNON and GAILLARD, 1997; ELLENBERG, 2009). Furthermore, the C/N ratios with ranges between 16 and 22 indicate that the organic matter originates from vascular terrestrial plants which show variable C/N ratios higher than 15 (MEYERS and ISHIWATARI, 1995). A. HERCZEG et al. (2001) document that reeds along the riparian zone have C/N values of 10 to 24. Unit 2 and units 7 to 10 have the characteristics of the “overbank facies” [ob], which accumulates during episodic flow events in formerly abandoned channels or during overbank discharges (WOJCICKI, 2006). Units 4–6 combine features of the [bc] facies and the [ob] facies. We assume that these units are a transition between the biochemical and the overbank facies.

4.2 - Geochronological model

23Radiocarbon dating on samples of different material and from different depositional environments was used for the geochronological model (section 3.3, Table 2, Fig. 3). Dating of profile 1 concentrated on the A-horizon (unit 2) of the chernozem. Two samples of soil organic matter were measured (Poz-37324, Poz-37325). Although it has to be taken into account that the chernozem may be intensely bioturbated, both the dating from the base of the A-horizon (7th millennium BP) and the dating from the middle part (c. 1000 BP) represent consistent and plausible ages.

24The chronological model of profile 2 is based on eight datings covering six out of ten sediment units. The absence of datings from unit 7 (6.8–5 m depth) results from the lack of datable material. The datings cluster around four periods. The first age cluster at the base of the profile dates to the second half of the tenth millennium BP and consists of two datings of indeterminable wood (Poz‑33979, Poz‑33935). The pieces of wood were extracted from the in situ deposited biochemical oxbow facies [bc] in units 1 and 3. Because of the autochthonous character of this facies, we consider these datings as relatively secure and accurate. A second cluster of two ages from the dating of a sample of organic sediment (Poz‑37391) and a piece of charred wood (Poz‑33934) from units 5 and 6 points to the 13th to 11th millennium BP, resultant in a time inversion. The inversion appears in a transition between the biochemical and the overbank facies. We conclude that these two ages do not reflect the time of deposition, but are the result of the reworking of older sediments. We excluded them from the chronological model. Two further age clusters appear in the middle section of unit 8 (8th to 7th millennium BP; Poz‑37359, Poz‑37361) and in units 8-10 (4th to 2nd millennium BP; Poz‑37334, Poz‑37335). These four ages result from the dating of organic sediments and are not as reliable and exact as the datings of the pieces of wood from the autochthonous base of the profile. However, they show a consistent chronology and we consider them as sufficiently accurate to contribute to a robust chronological model.

25In conclusion, the clustered ages in combination with the fluvial and pedogenic nature of the sediments in profiles 1 and 2 imply that the sediments are not a continuous and comprehensive geoarchive. Rather, we assume that the profiles contain erosional discordances and that the geochronological model derived from the sediments and soils only represents select phases of the Holocene.

4.3 - Phases of landscape development in the middle Sambek valley

26By combining the facies interpretation and the chronological model, we identified four phases of Holocene landscape development in the middle Sambek valley:

  • During the first phase, from around 9550 to 9250 cal. BP, the fluvial system was determined by the in situ deposition of the biochemical facies [bc]. Plant and wood remains within the sediments document peat formation in oxbow lakes on the valley floor. There is no indication of the deposition of allochthonous sediments from the hinterland of the Sambek River.

  • In the second phase, from around 7200 to 6000 cal. BP, the fluvial system significantly changed. Abandoned channels were no longer filled up by autochthonous peat formation, but by overbank deposition of allochthonous material from the hinterland. The base of the chernozem in profile 1, and therefore possibly the beginning of the soil formation on the foot slopes of the valley, dates to the end of the second phase, around 6100 cal. BP.

  • The third phase of landscape development began around 3350 cal. BP and lasted until about fifty years ago. Overbank deposition was still the dominating process on the valley floor, whereas soil formation occurred on the nearby slopes. As a result of the fluvial aggradation of the valley floor – overall up to 9 m during the first three phases of landscape development – it became more common for high flood events to inundate the slopes adjacent to the floodplain, which in turn led to the development of chute channels crossing the slopes and to the erosion of the slope toes.

  • The fourth phase of landscape development began around fifty years ago and is still ongoing. Reservoirs were built along the rivers and, as a consequence, the discharge of the Sambek River was reduced and became more equable. The frequency of high flood events and overbank sedimentation diminished. As a result, the fluvial aggradation came almost completely to a halt. Today, the valley floor is dominated by wetlands and, therefore, by the depositional environment of the biochemical facies [bc].

27What were the factors that led to the changing depositional environments on the Sambek valley floor and to the onset of soil formation on the adjacent slopes? The first phase of peat formation and the biochemical facies [bc] (c. 9550‑9250 cal. BP) coincides with a humid period when the Azovian Plain was covered by forest-steppe (SPIRIDONOVA and ALESHINSKAYA, 1999). Hence, we conclude that the surface of the Sambek catchment was stabilized by the vegetation. Furthermore, the relatively high infiltration rates in the forest-steppe vegetation must have delimited surface runoff generation, and thus high flood events were rare. As a consequence, siltation due to flooding occurred rarely, whereas in situ oxbow filling by biochemical processes such as peat growth were favoured. The first part of the second phase (c. 7200‑6300 cal. BP) coincided with an aridification and the spread of steppe to almost semi-desert conditions and therefore a lack of stabilizing vegetation cover (SPIRIDONOVA and ALESHINSKAYA, 1999; KOTOVA and MAKHORTYHK, 2010). We thus assume that soil erosion rates increased and that run-off occurred in periodical to ephemeral draining channels with frequent high flood events causing the deposition of overbank sediments. In the second part of phase 2, c. 6100 cal. BP, the climatic conditions became more humid, and forest-steppe spread again (SPIRIDONOVA and ALESHINSKAYA, 1999). We infer that these wetter conditions led to the development of a denser vegetation cover and to a stabilization of the land surface and initiated soil formation. During the third stage, between c. 3350 and 0 cal. BP, climate and vegetation cover were very variable and shifted between arid conditions with steppe cover and more humid conditions with forest-steppe cover (SPIRIDONOVA and ALESHINSKAYA, 1999). Furthermore, human influence grew with time. The Sambek catchment was settled at the latest in the Bronze Age in the fourth millennium BP (VAN HOOF et al., 2012a, b) and underwent phases of higher and lower population density with accompanying variations in land use intensity. Although it is unknown in what intensity land use most likely fostered soil erosion and therefore favoured, as with arid conditions, overbank deposition of allochthonous material in the middle Sambek valley.


28The environmental setting of the valley floor in the middle Sambek valley has changed over the last ten thousand years (Fig. 4).

Fig. 4 ‑ Sketches of the four phases of Holocene landscape development in the Sambek valley

Fig. 4 ‑ Sketches of the four phases of Holocene landscape development in the Sambek valley

29In the tenth millennium BP, the Sambek valley floor was captured by active meandering channels and oxbow lakes. The latter were filled up by in situ peat formation. From the end of the eighth millennium BP at the latest, the fluvial dynamics changed, and in situ sedimentation processes were replaced by the deposition of allochthonous material through overbank deposition. Not only the overbank area, but also abandoned channels were filled up by the overbank deposits. Because there are no signs of settlement activities before the fourth millennium BP, we assume that this change is the result of the mid-Holocene aridification described by A. SPIRIDONOVA and E. ALESHINSKAYA (1999) and N. KOTOVA and S. MAKHORTYKH (2010). The phase of overbank deposition lasted until the strong human impact on the fluvial system over the last fifty years, such as the construction of reservoirs and the straightening of the Sambek River. Since then, the equalized stream flow has almost completely lacked floods. Instead of the deposition of allochthonous detritus, in situ peat formation dominates on the valley floor.

30In conclusion, there was no change within the fluvial system between the fifth and the third millennium BP, although the valley was populated in the fourth millennium BP. Nor was there any detectable change in later settlement phases, such as the Iron Age or the Middle Ages, with the exception of the modern era, which brought a completely modified run-off system following the implementation of hydraulic constructions. Therefore, either there was no anthropogenic landscape change in the Sambek valley, or the change is not reflected by the fluvial geoarchives.

31However, although the fluvial system was stable, the fluvial aggradation led to an elevation of the valley floor in relation to the slopes. Hence, the Bronze Age sites on the foot slopes of the Sambek valley were, at the time of their colonization, about 3 to 4 m higher above the floodplain than today, and in a position clearly less prone to flooding. Furthermore, the rise of the valley floor led to a loss of slope area. The habitable, flood-protected slope area today has a significantly smaller extent than it did four thousand years ago.

Top of page


AD-HOC-AG BODEN, (2005), Bodenkundliche Kartieranleitung, Schweizerbart, Stuttgart, 438 p.

ALEXANDROVSKY A., CHICHAGOVA O., (1998), Radiocarbon age of Holocene paleosols of the East European forest‑steppe zone, Catena, 34, p. 197-207.

ANONYMOUS, (1966), Geologicheskaya Karta Rostovskoy, Bolgogradskoy, Astrakhanskoy oblastey u Kalmytskoy ASSR 1:1000000, Ministerstvo Geologii SSSR, Moscow.

BERG L., (1950), Natural regions of the U.S.S.R., The MacMillan Company, New York, 436 p.

BRONK RAMSEY C., (2009), Bayesian analysis of radiocarbon dates, Radiocarbon, 51, 1, p. 337-360.

CHENDEV Y., IVANOV I., (2007), Dynamics of the soil cover in the southeast of Europe and in the southern Trans-Ural Region during the Subboreal period, Eurasian Soil Science, 40, 11, p. 1298-1308.

DOLUKHANOV P., (2009), Geography of the East European Plain, in DOLUKHANOV P., SARSON G., SHUKUROV A. (eds), The East European Plain on the eve of agriculture, BAR International Series, 1964, p. 9-15.

ELLENBERG H., (2009), Vegetation ecology of Central Europe, Cambridge University Press, Cambridge, 731 p.

GERASIMENKO N., (1997), Environmental changes between 3 and 5ka BP in southeastern Ukraine, in DALFES N., KUKLA G., WEISS H., Third Millennium BC climate change and old world collapse, Springer, Heidelberg, p. 371-399.

HANNON G., GAILLARD M., (1997), The plant-macrofossil record of past lake-level changes, Journal of Paleolimnology, 18, p. 15-28.

HERCZEG A., SMITH A., DIGHTON J., (2001), A 120 year record of changes in nitrogen and carbon cycling in Lake Alexandrina, south Australia: C:N, δ15N and δ13C in sediments, Applied Geochemistry, 16, p. 73-84.

KOTOVA N., MAKHORTYKH S., (2010), Human adaption to past climate changes in the northern Pontic steppe, Quaternary International, 220, p. 88‑94.

KREMENETSKI C., CHICHAGOVA O., SHISHLINA N., (1999), Palaeoecological evidence for Holocene vegetation, climate and land-use change in the low Don basin and Kalmuk area, southern Russia, Vegetation History and Archaeobotany, 8, p. 233-246.

KHRUSTALEV YU., VASILENKO V., SVISYUKA J. et al., (2002), Klimat i agrogklimaticheskie resursy rostovskoy oblasti, Batayskoe iknizhnoe izdatelstvo, Rostov-na-Donu, 184 p.

LESER H., STÄBLEIN G., (1980), Legende der geomorphologischen Karte 1:25 000 (GMK 25), in BARSCH D., LIEDTKE H. (eds), Methoden und Anwendbarkeit geomorphologischer Detailkarten, Berliner Geographische Abhandlungen, 31, p. 91-100.

MEYERS P., ISHIWATARI R., (1995), Organic matter accumulation records in lake sediments, in LERMAN A., IMBODEN D., GAT J. (eds), Physics and Chemistry of Lakes, Springer, Berlin, p. 279-328.

MINASHINA N., (2009), Irrigation of steppe soils in the south of Russia: Problems and solutions (Analysis of irrigation practices in 1950-1990), Eurasian Soil Science, 42, 7, p. 807‑815.

MINORANSKIJ B., (2006), Rostov Oblast, Wetlands in Russia, 6, p. 10-12.

PARZINGER H., (2011), Die frühen Völker Eurasiens. Vom Neolithikum zum Mittelalter, C. H. Beck, München, 1045 p.

PESOCHINA L., (2010), The regularities of the late Holocene soil formation in the Lower Don Steppes, Eurasian Soil Science, 43, 13, p. 1-10.

RASSAMAKIN Yu., (2004), Die nordpontische Steppe in der Kupferzeit: Gräber aus der Mitte des 5. Jts. bis Ende des 4. Jts. v.Chr., Archäologie in Eurasien, 17, p. 1-234.

REIMER P. J., BARD E., BAYLISS A. et al., (2013), IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal. BP, Radiocarbon, 55, p. 1869‑1887.

SCHNEIDER S., MATTHAEI A., BEBERMEIER W. et al., (2014), Late Holocene human-environmental interactions in the Eastern Mediterranean: settlement history and paleogeography of an ancient Aegean hill-top settlement, Quaternary International, 324, p. 84-98.

SPIRIDONOVA E., ALESHINSKAYA A., (1999), Periodizatsiya neolita-eneolita evropeyskoy rossii po dannym palinologicheskovo analiza, Rossiyskaya Arkheologiya, 1, p. 23‑33.

SUKHODOLOV A., LOBODA N., KATOLIKOV V. et al., (2009), Western steppic rivers, in TOCKNER K., UEHLINGER U., ROBINSON C. (eds), Rivers of Europe, Academic Press, London, p. 497-523.

SYCHEVA S., GLASKO M., CHICHAGOVA O., (2003), Holocene rhythms of soil formation and sedimentation in the Central Russian Upland, Quaternary International, 106-107, p. 203-213.

TESAKOV A., TITOV V., LEONOVA N. et al., (2010), Quaternary stratigraphy and paleontology of the Southern Russia: connections between Europe, Africa and Asia, Southern Scientific Centre, Rostov-on-Don, 48 p.

VAN HOOF L., SCHÜTT B., SCHLÖFFEL M., (2012a), Untersuchungen zur naturräumlichen und siedlungsgeschichtlichen Entwicklung am Beispiel des Sambek-Tals, in DALLY O., VAN HOOF L., HUY S. et al. (eds), Deutsch-russische Ausgrabungen am Don. Ergebnisse der Kampagnen 2008-2010, Archäologischer Anzeiger, 1, p. 140-152.

VAN HOOF L., DALLY O., SCHLÖFFEL M., (2012b), Staying home or staying with your cattle? Different reactions to environmental changes in the Late Bronze Age of the Lower Don area (Southern Russia), eTopoi, 3, p. 71-75.

VAN HOOF L., SCHLÖFFEL M., (2016), Kurgans in the northeastern Azov Sea region. Proposals for a Geo-archaeological Research Program, in KELP U., HENRY O. (eds), Tumulus as Sema. Space, politics, culture and religion in the first millennium BC, De Gruyter, Berlin, p. 677‑702.

VELICHKO A., CATTO N., KONONOV M. et al., (2009a), Progressively cooler, drier interglacials in southern Russia through the Quaternary: Evidence from the Sea of Azov region, Quaternary International, 198, p. 204–219.

VELICHKO A., CATTO N., TESAKOV A. et al., (2009b), Structural specificity of Pleistocene loess and soil formation of the Southern Russian Plain according to materials of eastern Priazovie, Doklady Earth Sciences, 429, 8, p. 1364-1368.

WALTER H., (1974), Die Vegetation Osteuropas, Nord- und Zentralasiens, Gustav Fischer, Stuttgart, 452 p.

WOJCICKI K., (2006), The oxbow sedimentary subenvironment: its value in palaeogeographical studies as illustrated by selected fluvial systems in the Upper Odra catchment, southern Poland, The Holocene, 16, 4, p. 589‑603.

Top of page

List of illustrations

Title Fig. 1 ‑ Topographical and geological settings of the Sambek valley and its surroundings
Credits Source: Anonymous, 1966, modified.
File image/jpeg, 320k
Title Tab. 1 ‑ Coordinates of the two drilling sites
File image/jpeg, 40k
Title Fig. 2
Caption a) Geomorphological setting of the study area, b) EW-transect through the Sambek valley, c) The Sambek valley
File image/jpeg, 548k
Title Fig. 3 – Stratigraphy and chemical parameters of Profiles 1 and 2
Caption TOC: total organic carbon, TIC: total inorganic carbon, C: carbon, N: nitrogen, EC: electric conductivity, Mag. Suscep.: volume magnetic susceptibility.
File image/jpeg, 452k
Title Tab 2 ‑ Results of radiocarbon dating at the Poznan Radiocarbon Laboratory
Caption Calibrated with OxCal 4.2.3 (Bronk Ramsey, 2009) and the IntCal13 calibration curve
Credits Reimer et al., 2013.
File image/jpeg, 160k
Title Fig. 4 ‑ Sketches of the four phases of Holocene landscape development in the Sambek valley
File image/jpeg, 580k
Top of page


Bibliographical reference

Marlen Schlöffel, Steffen Schneider and Brigitta Schütt, « Fluvial dynamics and phases of landscape development in the Bronze Age settlement area of the Sambek valley (northeastern hinterland of the Sea of Azov) », Méditerranée, 126 | 2016, 91-100.

Electronic reference

Marlen Schlöffel, Steffen Schneider and Brigitta Schütt, « Fluvial dynamics and phases of landscape development in the Bronze Age settlement area of the Sambek valley (northeastern hinterland of the Sea of Azov) », Méditerranée [Online], 126 | 2016, Online since 01 January 2018, connection on 19 April 2018. URL : ; DOI : 10.4000/mediterranee.8265

Top of page

About the authors

Marlen Schlöffel

Institute of Geography/Palaeoecology and Geoarchaeology, Universität Osnabrück, Osnabrück, Germany,

Steffen Schneider

Institute of Geography/Palaeoecology and Geoarchaeology, Universität Osnabrück, Osnabrück, Germany,

Brigitta Schütt

Department of Earth Sciences/Physical Geography, Freie Universität Berlin, Berlin, Germany,

Top of page


Tous droits réservés

Top of page
  • Logo Presses Universitaires de Provence
  • OpenEdition Journals