Version classiqueVersion mobile

La géoarchéologie française au xxie siècle

Nathalie Carcaud
Gilles Arnaud-Fassetta

Partie I. Paléoenvironnements, biogéographie et paysages/Section I. Palaeoenvironments, biogeography and landscapes

Chapter 3. Geoarchaeological and palaeoenvironmental studies along the LGV (High-speed rail) in Alsace (France)

A contribution to landscape history

Nathalie Schneider, Damien Ertlen, Frédérique Durand, Noémie Nocus, Anne Gebhardt, Yohann Thomas, Matthieu Michler, François Schneikert et Éric Boës

Texte intégral


1The erosion of cultivated soils is a major problem in silty landscapes. Relief, even on moderate slopes, can be very sensitive to crusting and surface runoff, particularly in agricultural regions where the cultivated soils have been leftfallow during part of the winter and the spring (Auzet et al., 1995). This results in substantial degradation of the soils and the movement of fine materials. Apart from the degrading of pedological horizons and of agronomic potential, the processes of surface runoff can sometimes become highly concentrated muddy floods (Evrard et al., 2007), causing damage in urban areas situated at the outlets of basic catchments (Heitz et al., 2009). These erosional phenomena can be observed throughout the Holocene (Dotterweich, 2008; Houben, 2008), and from rescue archaeological operations conducted in Alsace (Ertlen et al., 2012). Evidence of erosion is often found under thick layers of colluvium at the bottom of catchments while other vestiges lie under the surface or exposed on summits and upper catchments (Leopold and Völkel, 2007). In order to better understand the origins of this process, it is necessary to know as accurately as possible the history of land use and its dynamics. The aim of this work is to measure the local impact of agricultural societies on the environment since prehistory. The study of agricultural productivity and of agrarian erosion is the entry point for retracing our human imprint (Leopold and Völkel, 2007). This gives a measure of the relative impact of different societies and thus offers new data on some previously poorly understood civilisations. Archaeology, carpology (study of fruits and seeds), anthracology (charcoal), palynology and geomorphology all need to be mobilised in order to reconstruct the formation and dynamics of the agrarian landscape.

2This research was conducted within the context of an archaeological evaluation in advance of the construction of the Baudrecourt (Moselle, Lorraine) to Strasbourg (Bas-Rhin, Alsace) section of the LGV Est (high-speed rail network extension). The studies have provided numerous data along a 35 km segment in the Alsatian sector of the project. In order to position the results within a research framework aimed at the understanding past erosion (Delmas et al., 2012), a basic catchment, the Gingsheimerbaechel, was chosen to be the focus of the study. When, however, there was insufficient or missing evidence for a data type or for a period, results from nearby catchments were used. To the west of Strasbourg, the evaluated zone runs through the hilly region of Kochersberg (Fig. 1). This area is characterised by sizeable Pleistocene loess deposits. These wind-blown silts have been substantially altered in such a way that the mosaic of superficial formations is much more complex than the original layered deposit (Antoine and Lautridou, 2008; Schneider and Ertlen, 2010). Thick layers of colluvium, sometimes intercalated with alluvium, cover some of the slopes and the valley bottoms (Ertlen et al., 2012). Tertiary formations reach the surface in places, on slopes less well-covered or more sensitive to erosion. Finally, a mosaic of soils, recognisable by their degree of evolution, has developed on these various formations. Generally these are fertile soils that are fairly easy to work. They have been attractive for agriculture since the Neolithic period.

Fig. 1. The area of the evaluation and the archaeological discoveries between Gougenheim and Mittelhausen, Kochersberg, Bas-Rhin (67), Alsace, France.

The Gingsheimerbaechel catchment

3The catchment at Gingsheimerbaechel is located to the south of the municipality of Gingsheim (Bas-Rhin). his is a sub-catchment of the Zorn, a let-bank tributary of the Rhine River, with an area of ca. 1500 m2 (Fig. 1). The study area, the head of the catchment, forms an amphitheatre of approximately 600 m long by 800 m wide. At the foot of this amphitheatre, the Gingsheimerbaechel stream flows intermittently, entrenched to a depth of about one metre in the landscape at the bottom of the slope. he outlet used for this study lies at the northern edge of the path by the future LGV (Fig. 1). he entire area is devoted to cereal farming and features very large open fields. he archaeological evaluation crossed the catchment from east to west, furnishing a wide transect across the thalweg and two asymmetrical valley sides. Geotechnical data (Velasque and Blanc, 2009) complete the overall picture of the catchment geology and geomorphology. Tertiary formations are covered with loess layers, arranged by the faults in the fracture zone of Saverne (BRGM, 1979). On some of the eastern and southern slopes, the processes of surface runof and colluviation have resulted in thinner loess layers. On the western slope, the downwind accumulations of undisturbed loess reach a thickness of greater than 10 m in some locations.

Land use since the Neolithic period

4Little evidence of human occupation had been recovered before the rescue project along the future rail path (Zehner and Zefner, 1993 a and b; Flotté and Fuchs, 2000). he archaeological and palaeoenvironmental data from sites spread along the route now ofers a comprehensive picture of human impact and occupation on the landscape.

The archaeological remains

5The preserved remains correspond largely to the excavated structures, which belong to several occupation sites. At high points these structures have been eroded and only deep pits are preserved, such as silos, silt extraction pits and various pits associated with occupation. In some of the excavated areas at the bottoms of slopes (Fig. 1, sites 5 and 6) rather more ephemeral remains have been found, including postholes for buildings, sometimes associated with floor layers. Funerary evidence indicates deposits of human remains in domestic circular or silo pits, found at the periphery of occupation areas; no cemeteries have been found so far.

6As a result of extensive surface stripping, two excavations on the perimeter of the Gingsheimerbaechel valley, at Gingsheimer Feld and Steinbrunnen in Gougenheim (sites 2 and 4), have enabled the recording of the presence of human occupation covering long periods from the Neolithic through the late Iron Age. The site at Gingsheimer Feld (site 4) occupies the summit and the upper slopes of a small hill. This seems to be associated with a pass between two valleys, on which human occupation was established from the early Neolithic period (5218-5209 BC;14C, with probability of 1 σ) until an early phase of the middle Iron Age (392-235 BC). The numerous structures that were revealed across a surface area of four hectares belong to several settlements including an unusual presence during the late Neolithic period (Réveillas et al., in press). The neighbouring site at Steinbrunnen, Gougenheim (site 2) has furnished, apart from a late Neolithic settlement, evidence of human presence in the early and middle Neolithic periods (Schneikert, 2011). The Bronze Age is represented by some rare structures, to which must be added the data recorded in the area during the evaluation phase of the project (Thomas et al., 2008; Peytremann, 2009) and during the excavation of site 3 (Michler, 2011). Bronze Age occupation in the valley and its environs extends from the end of the early Bronze Age to the beginning of the late Bronze Age. Following an interruption in the archaeological sequence of evidence from the 12th to the 11th c. BC, early Iron Age occupation begins to appear, as it does at the Kochersberg assemblage. This site at Steinbrunnen indicates relative continuity in occupation, lasting between the 8th and the 5th centuries BC. Human occupation during the end of the Iron Age is fairly patchy. But along with the sparse late Iron Age data from the evaluation (site 6)14 C results for some ephemeral structures (site 5) probably belonging to the middle Iron Age (394-234 BC). The investigations carried out in the area do not offer much information on the nature of Roman-period occupation although some isolated pits found on two sites (2 and 7) can probably be associated with poorly-defined settlements belonging to the second half of the 1st c. AD (Peytremann 2009) or the second half of the 2nd c. AD (Thomas et al., 2012).

The agro-systems

7The carpological (fruit and seeds) and soil micromorphological data recorded in the catchment and its surrounding area, coupled with the regional palynological data, provide an opportunity to describe the pattern of agricultural practices in the Gingsheimerbaechel. The sparse carpological data for the Neolithic period only partially represent the agricultural production in the Gingsheimerbaechel basin. Further east, evidence of cereal cultivation is better preserved at the middle Neolithic sites at Duntzenheim (Véber et al., 2012). Einkorn wheat appeared and came to dominate the carpological spectrum. Emblematic of northern France (Ruas and Marinval, 1991), it became the staple foodstuff. Correspondingly, the discovery of the pea and perhaps also bitter vetch indicate the cultivation of legumes. Within the archaeological data the succeeding legumes associated with the Bronze Age are poorly represented. There is, however, an interesting assemblage from a fire pit in the Gingsheimerbaechel (site 3). Dated to the beginning of the late Bronze Age (1411-1212 BC), this site has produced 165 burnt seeds. Several species have been identified, revealing diversified grain cultivation. These seeds are the result of treating harvests by a form of sieving (Ertlen et al., 2012; Michler, in press). Although it is an isolated assemblage, it offers evidence of agricultural activity for a period that is only fleetingly represented in the area. These traces of agriculture are complemented by the micro-morphological analysis of the sequence at Mittelhausen Kellen (site 6; Fig. 4). In Alsace, apart from some indications of proto-historic tillage recorded at both macroscopic and microscopic scales (Landolt et al., 2007; Schwartz et al., 2009; Gebhardt, pers. com.), this is the first evidence of Bronze Age agriculture.

8For the Iron Age, the works of Schaal (2009) and Wiethold (2010) have established the first inventory of cultivated species in Alsace. Barley was the most common crop, followed by proto-millet. The region can be distinguished from the rest of northwestern France by the notable presence of free-threshing wheat (Wiethold, 2010). For the early Iron Age, the sites at Gougenheim Steinbrunnen and Gingsheimer Feld (sites 2 and 4) have furnished assemblages representing daily life, composed of the remains of consumption and of the processing of harvests. The numerous wild seeds indicate the flora of the fields and grasslands, some being particularly useful as forage, such as lucerne, ryegrass and phleum (Lecoq, 1844). Wetland species, growing alongside streams or in wet grasslands, are also well represented. A palynological study has shown evidence of pastures in the bottom of the valley of the Ungerbruchgraben during the early Iron Age (wet grasslands, coprophilous mushrooms; Thomas et al., 2010). This data, coming from the carpological identifications, suggests the exploitation of wetlands for livestock farming, but harvesting reed plants, as a raw material for the roof construction or for basketwork, is also likely (Lachivier, 1997). At the Halstatt period site at Duntzenheim (Michler, in press), cultivation was predominantly hulled cereals, among which millets, which are short-cycle summer crops, are well represented. The building of agriculture around diversified grain production and mainly species with hulled grains is a development that has been observed for the early Iron Age throughout France (Matterne, 2001; Bouby, 2010; Cabanis et al., 2010; Wiethold, 2010). At the site at Mittelhausen Langmatt (site 5; Thomas et al., 2012), cereals, predominantly polystic barley, were accompanied by legumes such as peas (Pisum sativum), hairy vetch (Vicia sp.), and possibly lentils (Lens sp.).

9For the later Iron Age, there is evidence only on the periphery of the catchment. Agricultural activity has been recorded, such as sieving harvests, but also culinary activity, represented by fragments of a burnt galette, a form of flat round cake (Véber, in press). The diverse range of cereals present underlines the importance of grain cultivation. Einkorn wheat appears frequently in the samples, which is unusual since this form of wheat had been in decline since the Bronze Age (Marinval, 1988). This reappearance of einkorn wheat is surprising in an area characterised by the high quality of its soils. While such a regional specificity could be the result of a shiftin the status of einkorn wheat towards a forage or fortuitously exploited species, it may also belong to an agricultural tradition (Wiethold 2008).

The eco-systems

10The first anthracological (charcoal) data recorded in the Alsatian plain, along with the data from palynological and carpological studies, provide an initial overview of vegetation in the Kochersberg region. In the Gingsheimerbaechel, just two samples have furnished data for the Neolithic and Bronze Age periods (sites 2 and 3). These results should therefore be interpreted with great caution, although they are complemented by other regional data. West of the Kochersberg, the more developed corpus of data from the site at Ingenheim (Nocus, 2010) provides a first insight into the vegetation of the late Neolithic period. Oak, a post-pioneer species, is the most frequent tree, accompanied by ash, maple and by trees of the Maloideae subfamily (apple, pear, hawthorn, etc.). The anthracological spectrum suggests the presence of a forest in an advanced phase, consisting mainly of deciduous oak. The other more heliophilous tree types belong to formations along the edge of the forest or at more open spaces seem to have been rare at this site. Riparian species are almost nonexistent. Consistent with the palynological data for the Rhine region (Lechner, 2005), this appears to have been a closed-forest environment during this period. In the Bronze Age, the anthracological spectrum (site 2) emphasises the predominance of beech and oak. The palynological study undertaken in the bottom of the Ungerbruchgraben valley, corroborates the hypothesis of an environment as yet little influenced by man, despite the poor preservation of pollen grains (Thomas et al., 2010).

11In the early Iron Age, the palynological data (Thomas et al., 2010) reveal several indicators of human activity, including the clearing of the forest and the appearance of cereals and microfossils associated with livestock. At the site of Gougenheim Steinbrunnen (site 2), oak and beech are the most common taxa in the anthracological spectrum, accompanied by hornbeam. These taxa suggest the presence of groves or small woodland stands of oak, oak and beech, and oak and hornbeam. In more open environments, the maples, the Amygdaloideae and the hazels can form, the edges of forests, thickets or hedges. The distribution of the taxa in the spectra from the different structures can vary enormously. These variations may suggest a mosaic of wood formations near the settlement or, as has been proposed for the Paris basin (Pernaud, 1997), they may indicate fluctuations in wood supply over time (random collection of wood, woodland management, or a cycle of wood exploitation). Indications of human activity increase during the course of the Iron Age. The heliophilous taxa in some anthracological spectra (site 5) and some nitrophilous species, such as elderberry, also increase (Michler, in press). This coincides with grasslands and the cultures described by the carpological studies and the associated phases of erosion. Riparian forest, always under-represented, seems often to be absent, even at the site at the bottom of the valley (site 5). In the carpological assemblages, particularly at Gougenheim Steinbrunnen (site 2), two vegetation groupings are observed: the flora of fields and plants alongside watercourses, and, rather less importantly, the grasslands, ruderal plants and wooded areas. The abundant occurrence of hydrophilous plants is a reminder that the Gingsheimerbaechel flows in close proximity to the site. These taxa indicate the nearby presence of fields and/or of a zone where these plants were exploited.

12As with the regional data furnished by the palynology, the later Iron Age is marked by the beginning of the clearing of the woodland areas (site 7; Véber et al., 2012; Véber, in press). The mature woods epitomised by the oak, the beech, the hornbeam and the Scots pine dominate the anthracological spectra, whilst the edges of the forests and the open woodlands decline. Riparian forest remains poorly represented. During the later Iron Age, the anthropogenic impact diminishes and some areas seem to have been re-colonised by forest. All the same, we should not lose sight of the broad ecological importance of most of the taxa, and of their ubiquitous development. It is possible that the substantial percentages of oaks are not only correlated with mature forest as the the ‘nomadic’ post-pioneer taxa can play role of pioneers in open space (e.g., the pedunculate oak) and can be found in the edges of woods, in hedges and in riparian forests. For all of the chronological periods, several hypotheses could explain the modest importance of riparian species: either the woods alongside streams were little used for irewood, or there were anthropogenic or natural reasons for their partial forest cover. For the early Iron Age, the presence of wet grasslands concurrently with riparian forest is underlined by the carpology.

Surface formations

13Surface depositional formations are predominantly loess strata modiied by luvial processes. hese processes are weak under forest cover, but can become intense when the soils are uncovered. As a result, deposits that accumulate at the head of a valley or at the foot of catchments (Fig. 2) constitute a good indicator of the opening of the landscape by human activity. Local stratigraphy is thus dependant on cycles of occupation and abandonment of the land.

Erosion in the Neolithic period

14On the right bank of the Gingsheimerbaechel, the much-eroded catchment reveals, under a variable thickness of colluvium (< 1 m), a marly-calcareous geological substratum (Fig. 2). Transverse crosssections have unveiled an ancient drainage channel (Michler, 2011; site 2) and a bedded periglacial scree, composed of angular limestone blocks of decimeter size, indicates the limit between the palaeochannel and its ill. his conirms that the channel was already functioning prior to the Holocene. he geomorphology and the micromorphology show that the palaeochannel was artiicially illed. he density of the Neolithic remains upstream (site 4) and the signs of cultivation recorded by thin section of the colluvium of the adjacent catchments (sites 3 and 6) suggest a irst phase of land clearance. he core sampling carried out at the outlet of the catchment has ofered the opportunity to study a sedimentary sequence beginning in the Bronze Age. he Neolithic deposits had undoubtedly been removed or deposited further downstream. Other studies along the line of the LGV corroborate the occurrence of an erosive crisis during the late Neolithic period (Ertlen et al., 2012; Lefranc et al., 2012). his last record from the bottom of the catchment indicates several cycles of pedogenesis and erosion, under which there remains a Neolithic soil (3741-3372 BC). The presence of numerous micro-fragments of charcoal, of fragments of crust from the top of horizon Bt (Fig. 4, M4) and powdery clay surfaces at the bottom (M3) mark the irst clearance of the soil. hese elements illustrate the anthropogenic pressure on the Kochersberg landscape, although a clear chronological framework has yet to be established.

Fig. 2. Cartography of the surface formations in the Gingsheimerbaechel catchment.

Morphodynamic stability in the Bronze Age

15The stratigraphy at the foot of the Gingsheimerbaechel (Fig. 3) includes at its base a peaty level the end of which is dated to the late Bronze Age (1030-830 BC). On the slope, the illing of the palaeochannel (site 3) developed a soil due to a period of great morphodynamic stability. he cutting of this level by a ire pit (1411-1212 BC) conirms that this soil was already well developed by the late Bronze Age. Alongside the Vierbruckgraben (site 6), the upper level of the sequence, dated to the late Bronze Age (1212-1217 BC), is a horizon previously cultivated by humans (Fig. 4, M5). These results underline the weak prevalent erosional process despite some indications of cultivation (see below).

Erosion in the Iron Age

16At the bottom of the Gingsheimerbaechel, the cycle of peaty accumulation was succeeded by a cycle of intense colluviation reaching a thickness of almost one metre, the base of which is dated to the early Iron Age (800-547 BC). he upper limit of the sequence could not be dated. his accumulation is evident in most of the loessial valley bottoms cut through for the LGV. It diminishes and then ends in the later Iron Age. On the slope, the channel shifts and divides, and lows on both sides of the palaeochannel (site 3). hick colluvium is noticeable in the Vierbruckgraben sequence (Fig. 4) and it contains numerous residual micro-fragments of ceramics. Erosion of the slopes becomes widespread in the Iron Age.

Fig. 3. Stratigraphy of the sedimentary accumulations in the Gingsheimerbaechel catchment.

Fig. 4. The pedo-sedimentary sequence at Mittelhausen Kellen, site 6.

Historical colluviation

17After the Iron Age, the stratigraphy across all the sites becomes more uniform and chronological markers become sparser. Without well-identified sites, Roman-period and Mediaeval sherds only partially improve the dating. An assemblage of very homogeneous loessial silts is only interrupted by some more coarse intercalations of the marly-calcareous substratum, thus indicating that locally the surface formations and the soil on the slopes was in a state of advanced erosion. Where not eroded, the colluvium regularly reaches a thickness of more than 2 m.

Landscape dynamics

18The comparison of the archaeological and palaeoenvironmental data allows us to identify four stages marking the evolution of the landscape of the Gingsheimerbaechel (Fig. 5).

19For the Neolithic, there are numerous signs of human activity. The collection of wood takes place in an almost untouched forest. Cereal farming intrudes into the landscape. The valley thalweg does not contain colluvium, despite the important settlement at the top of the basin (site 4). With the current state of knowledge, it is difficult to specify whether this lacuna has resulted from the removal of the depositional units or from low sediment contributions from the slopes. The absence is partly explained by the equilibrium of the landscape mosaic, with open spaces that are unconnected or are poorly linked with the hydrographical network. Erosion was thus limited to the level of the parcel of land, without contributing to sedimentary movement at the scale of the catchment.

20In the Bronze Age, the weak influence of humans on the land led to regeneration by the forest. The various indicators combine to reveal the morphodynamic stability of the slopes. Man, however, is not completely absent, and some catchments have experienced different trends. This is shown by the ecological assemblages from the carpology of the Bronze Age silos that are typical of cereal farming. Also, this apparently weak influence must be qualified by the taphonomy, which may have led to differential preservation of the archaeology. The well-preserved sites of the early Neolithic and the Iron Age benefitted from their elevated location. On the slopes, the most exposed remains were grazed. This hypothesis is confirmed by the presence of Bronze Age sherds in some of the colluvial sequences. Nevertheless, these aspects do not entirely explain the obvious contrast in density between the Bronze Age and the Iron Age.

Fig. 5. Landscape dynamics in the Gingsheimerbaechel catchment.

21A series of geoarchaeological studies have established that a major erosive crisis occurs in the Iron Age. The catchment records a clear retreat of the forest, in response to agricultural and pastoral exploitation. A signiicant demographic expansion revealed by the archaeology coincides with a climactic crisis (Holzauzer et al., 2005; van Geel and Berglund, 2010). he mastery of new tools, such as the ard, the plough and the scythe, leads to the rapid increase in arable land (Malrain, 2000). As in Normandy (Lespez, 2012) or in Lorraine (Gebhardt, pers. com.), the land clearance and the destruction of the mosaic landscape reduce the bufer role of the forest and triggers a generalised erosion of catchment soils. During this period, agricultural impact is ampliied by climatic change.

22At the top of the stratigraphy, major colluvial sequences lie discontinuously above the Iron Age sequence. In the best cases, some evidence indicates that these accumulations began in the early Mediaeval period (Schneider and Ertlen, 2010). Ater this period, there is a lack of chronological markers, such that it is difficult to quantify the respective proportions of Mediaeval and post-Mediaeval colluvium. A recent discovery of ceramics in-situ has, for the first time, provided an estimated colluvial thickness of 1 m of deposition since the 18th c., giving a rate of sedimentation of approximately 3 mm yr-1 (Dabek and Schneider, 2012). his emphasises the first accentuation of erosive processes in the Mediaeval period–a phenomenon that reaches its apogee in the 20th c. with mechanisation and land consolidation. However, quantitative studies in future should begin to clarify this question.


23At the level of the topographic catchment, analysis of the erosion process affirms the correlation between the period of occupation and the instability of the slopes. The same finding can be applied to the loessrich sector of the Kochersberg cut by the LGV. There are, however, some subtleties in these patterns. For example, for the Roman period, erosion was limited to the edges of some villae that were found within the path of the rail line. Also, at the scale of the Rhine valley, the rules of the population record change. The Bronze Age cultures are well represented in the valleys and in the river plain. For the Iron Age, there is a generalised phenomenon of demographic expansion. The sensitivity of erosion on loess-covered hills accentuates the process of erosion, transportation and then accumulation at the foot of the slope and makes this terrain conducive to geoarchaeological studies. Nevertheless, each of the studies involved has its limitations, and often includes discontinuous records, which tend to hinder precise dating. These limitations can be overcome through inter-disciplinary comparisons that can lead to a series of convergent arguments, put forward to explain the landscape and its evolution. In the 1970s, the construction of the A4 motorway, running along a parallel axis to the LGV a few kilometres to the north, was used to support the development of the national geological map (BRGM, 1979). The LGV line between Strasbourg and Saverne has served to advance our knowledge of the deposition of Holocene sedimentary units and their close relationship with cycles of occupation, agriculture and erosion.



Antoine P., Lautridou J.-P., «Les formations lœssiques en milieux périglaciaires», in Dewolf Y., Bourrié G. (éds.), Les formations superficielles: genèse, typologie, classification, paysages et environnement, ressources et risques, Paris, Ellipses, 2008.

Auzet A.-V., Boiffin J., Ludwig B., «Concentrated flow erosion in cultivated catchments: Influence of soil surface state», Earth surface processes and landforms, no 20, 1995, p. 759-767.

Bouby L., «Agriculture dans le bassin du Rhône du Bronze final à l’Antiquité. Agrobiodiversité, économie, cultures», Thèse de doctorat, École des Hautes Études en Sciences Sociales, 2010.

BRGM, «Carte géologique de la France au 1/50 000e, feuille de Saverne», Orléans, BRGM, 1979.

Cabanis M., Mennessier-Jouannet C., Bouby L., Hajnalova M., Wiethold J., «Economie végétale en basse Auvergne à l’âge du Bronze et au premier âge du Fer», in Delhon C., Théry-Parisot I., Thiébault S. (éds.), Des hommes et des plantes, Exploitation du milieu et gestion des ressources végétales de la Préhistoire à nos jours, Actes des XXXe rencontres internationales d’archéologie et d’histoire d’Antibes, 2011.

Dabek P., Schneider N., «Lampertheim (Bas-Rhin), Strendfeld, Occupations diachroniques du début de la Tène jusqu’à la fin de la période romaine», Rapport de diagnostic, INRAP, Grand Est Sud, Strasbourg, 2012.

Delmas M., Pak L.T., Cerdan O., Souchère V., Le Bissonnais Y., Couturier A., Sorel L., «Erosion and sediment budget across scale: A case study in a catchment of the European loess belt», Journal of Hydrology, no 420-421, 2012, p. 255-263.

Dotterweich M., «The history of soil erosion and fluvial deposits in small catchments of central Europe: deciphering the long-term interaction between human and the environment. A review», Geomorphology, no 101, 2008, p. 192-208.

Ertlen D., Gebhardt A., Schneider N., Durand F., Thomas Y., Michler M., Schneikert F., Boës E., Schwartz D., «Anthropisation et érosion agraire dans un paysage lœssique (Bas-Rhin, France)», in Bertoncello F., Braemer F. (éds.), Variabilités environnementales, mutations sociales. Nature, Intensités, Échelles et Temporalités des changements, XXXIIe rencontres internationales d’archéologie et d’histoire, Antibes, 2012.

Evrard O., Bielders C., Vandaele K., van Wesemael B., «Spatial and temporal variation of muddy floods in central Belgium, off-site impacts and potential control measures», Catena, no 70, 2007, p. 443-454.

Flotte P., Fuchs M., «Carte archéologique de la Gaule: le Bas-Rhin», Paris, MSH, Académie des inscriptions et Belles-Lettres, 2000.

Heitz C., Spaeter S., Auzet A.-V., Glatron S., «Local stakeholders’perception of muddy flood risk and implications for management approaches: A case study in Alsace (France)», Land Use Policy, no 26, 2009, p. 443-451.

Holzhauser H., Magny M., Zumbuhl H., «Glacier and lake level variations in west-central Europe over the last 3500 years», The Holocene, no 15, 2005, p. 789-801.

Houben P., «Scale linkage and contingency effects of field-scale and hillslope-scale controls of long-term soil erosion: Anthropogeomorphic sediment flux in agricultural loess watersheds of Southern Germany», Geomorphology, no 101, 2008, p. 172-191.

Lachiver M., Dictionnaire du monde rural, les mots du passé, Paris, Fayard, 1997.

Landolt M., Croutsch C., Flotté P., Schneider N., Putelat O., «Entzheim-Geispolsheim (Alsace, Bas-Rhin) Aéroparc (LIDL-CUS)», Rapport de fouille préventive, no 31, Pôle d’Archéologie Interdépartemental Rhénan, Séléstat, 2007, p. 326-329.

Lechner A., «Paläoökologische Beiträge zur Rekonstruktion der holozänen Vegetations-, Moorund Flussauenentwicklung im Oberrheintiefland», Thèse de doctorat, Freiburg, 2005.

Lecoq H., Traité sur les plantes fourragères ou flore des praires naturelles et artificielles de France, Paris, Edition H. Lecousin, 1844.

Lefranc P., Braguier S, van Es M., Gebhardt A., Schneider N., «Mittelhausen (67), Kellen, Neun Quellen, tronçon H, site LGV 9-4, Occupations du Néolithique ancien et de l’âge du Fer», Rapport de fouille archéologique, INRAP Grand Est-Sud, Strasbourg, 2012, 99 p.

Leopold M., Völkel J., «Quantifying prehistoric soil erosion. A review of soil loss methods and their application to a Celtic square enclosure (Viereckschanze) in Southern Germany», Geoarchaeology, no 22, 2007, p. 873-889.

Lespez L., «Le temps des paysages de l’environnement et des plaines littorales bas-normandes», in Lespez L. (éd.), Paysages et gestion de l’eau, Bibliothèque du Pôle Rural no 3, 2012, p. 199-234.

Malrain F., «Fonctionnement et hiérarchies des fermes dans la société gauloise du iiie siècle à la période romaine: l’apport des sites de la moyenne vallée de l’Oise», Thèse de doctorat, université Panthéon-Sorbonne (Paris 1), 2000.

Marinval P., «Cueillette, Agriculture et alimentation végétale de l’Epi-paléolithique jusqu’au deuxième âge du Fer en France Méridionale. Apports paléthnographiques de la carpologie», Thèse de doctorat, EHESS, 1988.

Matterne V., Agriculture et alimentation végétale durant l’âge du Fer et l’époque gallo-romaine en France septentrionale, Montignac, Editions Monique Mergoil, 2001.

Michler M., «Gingsheim, Bas-Rhin, Steinbrünnen, Aschenbuckel LGV EE, site 9-1. Un paléo-vallon sur la rive droite du Gingsheimerbaechel», Rapport de fouille archéologique, INRAP Grand Est-Sud, Strasbourg, 2011.

Michler M., «Duntzenheim, Bas-Rhin, Frauenabwand, Ebenheit, LGV EE - Stockage no 2, Duntzenheim nord. Entre Rohrbach et Weingartenberg, des occupations du Néolithique ancien au Hallstatt final», Rapport de fouille archéologique, INRAP Grand Est-Sud, Strasbourg, sous presse.

Nocus N., «Analyse anthracologique du site néolithique récent d’Ingenheim», Mémoire de Master 2, université Panthéon-Sorbonne (Paris 1), 2010.

Pernaud J.-M., «Paléoenvironnements végétaux et sociétés à l’Holocène dans le nord du Bassin-Parisien», Thèse de doctorat, université de Lille, 1997.

Peytremann E., «Gougenheim-Gingsheim, Steinbrunnen, Katzenellebogen, Aschenbuckel. Indices d’occupation du Bronze final ou/et du Hallstatt C et gallo-romain. Document intermédiaire de diagnostic archéologique, LGV Est Européenne», INRAP Grand Est-Sud, Strasbourg, 2009.

Réveillas H., Thomas Y., Lefranc P., «Premières données sur les pratiques mortuaires du site néolithique récent de Gougenheim (Bas-Rhin, Alsace)», Bulletins et Mémoires de la Société d’Anthropologie de Paris, Publication du Colloque annuel de la Société d’Anthropologie de Paris, 1837e réunion scientifique, 25-27 janvier 2012, sous presse.

Ruas M.-P., Marinval P., «Alimentation végétale et agriculture d’après les semences archéologiques (9000 av. J.-C. au xve siècle)», in Guilaine J. (éd.), Pour une archéologie agraire, Paris, Armand Colin, 1991.

Schaal C., «Les plantes», in Landolt M., van Es M., Putelat O., Bouquin D., Schaal C., Boës E., Baccharetti E., Bevilaqua D., Déchanez-Clerc I. (éds.), Entzheim-Geispolsheim (Alsace, Bas-Rhin) Aéroparc (LIDL-CUS), Rapport de fouille préventive, vol. 3 Les occupations protohistoriques, Pôle d’Archéologie Interdépartemental Rhénan, Séléstat, 2009, p. 326-329.

Schneider N., Ertlen D., «Diagnostic géoarchéologique: Rapport de synthèse, LGV Est Européenne», INRAP Grand Est-Sud, Strasbourg, 2010.

Schneikert F., «Gougenheim stockage. Document intermédiaire de diagnostic archéologique, LGV Est Européenne», INRAP Grand Est-Sud, Strasbourg, SRA Alsace, 2011.

Schwartz D., Gebhardt A., Ertlen D.,Trautmann M.,Zumbrunn O.,Boes E., «Découverte de traces de paléolabours (protohistoriques) sous un luvisol à Gottesheim (67), Étude pédologique, conséquences sur la formation des luvisols», Actes des 10e journées de l’Association Française pour l’Études des Sols, Strasbourg, mai 2009.

Thomas Y., Véber C., Ertlen D., «Gingsheim Steinbrunnenn, Asschenbuckel. Indices d’occupation à l’âge du Bronze», Document intermédiaire de diagnostic archéologique, LGV Est Européenne, INRAP Grand Est-Sud, Strasbourg, 2008.

Thomas Y., Ertlen D., Schneider N., Richard H., Wiethold J., Croutsch C., van Es M., Fleischer F., Véber C., «Le terroir du Kochersberg au premier âge du Fer. Premiers résultats de fouilles et d’études paléoenvironnementales», Poster au 34e colloque international de l’AFEAF, Aschaffenburg, 2010.

Thomas Y., Braguier S., Chenal F., Cicutta H., Durand F., van Es M., Jodry F., Méloche C., Millet E., Minni D., Nocus N., Schneider N., «Mittelhausen (Bas-Rhin) Vorderen Berg (LGV-Est), Des vestiges d’habitat de La Tène A2/B1a et des fosses d’époque romaine», Rapport d’opération Fouille archéologique, INRAP Grand Est-Sud, Strasbourg, 2012.

van Geel B., Berglund B.E., «A causal link between a climatic deterioration around 850 cal BC and a subsequent rise in human population density in NW-Europe?», Terra Nostra, no 7, 2010, p. 126-130.

Véber C., «Duntzenheim site 8.4», Rapport de fouille, INRAP Grand Est-Sud, Strasbourg, sous presse.

Véber C., Boisseau F., Braguier S., Durand F., Van ES M., Gebardt A., Jodry F., Méloche C., Nocus M., Schneider N., Simonin O., «Duntzenheim, Bas-Rhin, Schlittweg, Weingartenberg LGV EE Site 8-2, Une occupation du Bronze moyen, des silos de La Tène ancienne et un chemin médiéval», Rapport de fouille, INRAP Grand Est-Sud, Strasbourg, 2012.

Velasque P.-C., Blanc O., «Maquettes géotechniques LGV EE, 2e phase, tronçon Danne et Quatre Vents – Vendenheim», Réseau Ferré de France, Maîtrise d’œuvre génie civil, SETEC, 2009.

Wiethold J., «Eckbolsheim (Bas-Rhin), Le Zenith. Agriculture et alimentation végétale déduite d’analyse carpologique: Les macro-restes végétaux provenant des fosses, silos et trous de poteaux datés du Bronze final, du La Tène D et de l’époque Gallo-romaine», Rapport d’étude carpologique, INRAP Grand Est Nord, Metz, 2008.

Wiethold J., «Les Trois Domaines, Meuse (55), La Hachie. Macrorestes végétaux carbonises provenant d’occupations protohistoriques et médiévales», Rapport archéobotanique non publié, INRAP Grand Est Nord, Metz, 2010.

Zehner M., Zefner M., «TGV EST, A.P.S., Étude d’impact archéologique sommaire», Alsace. Strasbourg, SRA, 1993a.

Zehner M., Zefner M., «TGV EST, A.P.S. Étude d’impact archéologique sommaire, Alsace», Rapport complémentaire, Strasbourg, SRA, 1993b.

Table des illustrations

Légende Fig. 1. The area of the evaluation and the archaeological discoveries between Gougenheim and Mittelhausen, Kochersberg, Bas-Rhin (67), Alsace, France.
Fichier image/jpeg, 262k
Légende Fig. 2. Cartography of the surface formations in the Gingsheimerbaechel catchment.
Fichier image/jpeg, 189k
Légende Fig. 3. Stratigraphy of the sedimentary accumulations in the Gingsheimerbaechel catchment.
Fichier image/jpeg, 206k
Légende Fig. 4. The pedo-sedimentary sequence at Mittelhausen Kellen, site 6.
Fichier image/jpeg, 285k
Légende Fig. 5. Landscape dynamics in the Gingsheimerbaechel catchment.
Fichier image/jpeg, 546k


Research Engineer, National Institute for Preventive Archaeological Research (INRAP Grand Est-Sud), Mixed Research Unit (UMR 7362) CNRS/University of Strasbourg (Image Lab, City, Environment–LIVE), Strasbourg, France (

Archaeologist carpologist, National Institute for Preventive Archaeological Research (INRAP Grand Est-Sud), Mixed Research Unit (UMR 5608) CNRS/University of Toulouse 2/EHESS/Ministry of Culture and Communication/INRAP (Works and Archaeological Research on Cultures, Spaces and Societies – TRACES), Toulouse, France (

Ph. D Student, National Museum of National History, Mixed Research Unit (UMR 7209) CNRS/MNHN (Archaeozoology, Archaeobotany: Societies, Practices and Environment – AASPE), Paris, France (

Site Supervisor, National Institute for Preventive Archaeological Research (INRAP Grand Est-Sud), Mixed Research Unit (UMR 6298) CNRS/University of Bourgogne/Ministry of Culture and Communication (Archaeology, Earth, History and Societies – ARTEHIS), Dijon, France (

Project Manager, National Institute for Preventive Archaeological Research (INRAP Grand Est-Sud), Mixed Research Unit (UMR 7044) CNRS/Universities of Strasbourg & Haute-Alsace/Ministry of Culture and Communication, SRA DRAC Alsace & Lorraine (Study of Civilisations of Antiquity: from Prehistory to Byzantium), Strasbourg, France (

© CNRS Éditions, 2015

Licence OpenEdition Books

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search