Navigation – Plan du site

AccueilNumérosvol. 27 - n° 2Geoarchaeology of the middle Hera...

Geoarchaeology of the middle Herault valley (southern France) since the Bronze Age: cartographic approach

Géoarchéologie de la moyenne vallée de l’Hérault depuis l’Âge du Bronze : approche cartographique
Ambrine Bouchène et Benoît Devillers
p. 159-170

Résumés

Alors que le concept de l’Anthropocène fait encore débat, il apparait crucial d’approfondir nos connaissances des interactions s’opérant entre les sociétés humaines et leur environnement. Occupée depuis au moins l’Âge du Bronze, la moyenne vallée de l’Hérault de par la richesse des données archéologiques mises au jour et leur proximité avec l’activité fluviale, offre un territoire propice à l’analyse de ces interactions. Pour ce faire, nous avons déterminé quelle était la nature de l'évolution hydromorphologique de l'Hérault et son impact sur la morphologie de la plaine, à travers une analyse comparative des données cartographiques, permettant d'identifier les capacités morphogéniques du fleuve. En croisant les résultats d'analyses géomorphologiques, hydromorphométriques, paléogéographiques et archéologiques, nous avons pu définir trois tronçons de la plaine inondable, chacune caractérisée par sa propre dynamique. Le premier tronçon se distingue par l'écoulement de la rivière dans un chenal profond et le faible développement de la terrasse holocène. L'analyse paléogéographique y a révélé une relative conservation de l'hydrographie fluviale. Par conséquent, sur cette section, le fonctionnement hydrologique du fleuve a peu d'impact sur l'évolution du paysage alluvial. Le deuxième tronçon se démarque particulièrement du reste de la vallée suite à l'analyse des données hydromorphométriques. La fréquence et l'amplitude des méandres sur cette section sont 3,7 fois supérieures à la moyenne observée sur l'ensemble de la moyenne vallée. Par conséquent, le paysage fluvial de ce tronçon se singularise par une plus forte instabilité. Enfin, le troisième tronçon a été caractérisé par le rétrécissement du chenal, une diminution de sa profondeur et une vaste expansion de la terrasse holocène, reflétant une forte dynamique d'alluvionnement.

Haut de page

Notes de la rédaction

Manuscript received on May 15, 2020, revised version received on October 21, 2020, accepted on, November 20, 2020

Texte intégral

This work benefited from the support of the LabEx ARCHIMEDE under the "Investissement d'Avenir" ANR-11-LABX-0032-01. We thank Dr. Stéphane Mauné (UMR5140) and two anonymous reviewers for their very helpful remarks and suggestions.

1. Introduction

1Although still debated, the notion of the Anthropocene proposes to recognised humanity as a geological force, due to the consequent impact of human activities on the evolution of the Earth system (Malm and Hornborg, 2014). Among the clues that led to this conclusion, we find in particular the major developments that the rivers have given rise to, both for the exploitation of hydraulic energy, the prevention of flood risks, navigation, the cultivation of alluvial terraces and even fisheries (Crutzen, 2002).

2Recent research in fluvial geomorphology has all converged on demonstrating a close relationship between the hydromorphology of a stream and the functioning of its riparian habitats (Schmidt et al., 2013). Fluvial hydromorphology has consequences on the morphologies of the minor bed, the banks and the major bed; spaces that form the focal points for the development of an ecological system that is highly attractive to human populations (Brierley and Fryirs, 2004). As a result, river systems are privileged areas for observing society-environment interactions and the influence of natural and anthropogenic forcings on the morphological evolution of the alluvial plain (Brierley, 2006).

3While recent work carried out in the Mediterranean southeast France, has restored the evolution of Herault River coastal landscape over 8,000 years (Devillers et al., 2019), revealing the combined impacts of natural and anthropogenic factors, the history of its middle valley is still poorly understood.

4However, excavations and archaeological studies carried out for almost 30 years on this territory have brought to light a constant, although contrasting, occupation of the valley since the Neolithic (Garcia, 1993). These material vestiges are those of an inhabited place from the floodplain to the slope (Mauné, 2001), exploited for the fertility of its alluvial terraces or the stream power of its river (Phalip, 1992) and crossed by trade networks (Ropiot, 2007).

5In order to clarify the nature of the interactions between ancient societies and the floodplain, we propose to initiate a restitution of the palaeohydrography of the Herault River, through a cartographic approach. It will be a question here of presenting the first results of a thesis in progress through the combined analysis of the geological, historical and archaeological maps. This study will aim to offer initial line of thought on the morphogenic capacities of the river by characterising the dynamics at work in the floodplain and restoring the recent development of the river recorded by the cartographic archives.

6In the current state of research, our knowledge of the evolution of the floodplain remains fragmentary. What hydromorphological changes has the Herault River experienced? How have these developments impacted the alluvial landscape? This general problem must be answered through the multiplication of research in interdisciplinarity and interventions in the field. In a first step, we have chosen here to propose an initial response, through geoarchaeological analysis, in order to first determine the evolutionary capacities of the river and the recorded evolutions of its hydrography. In a second step, we made initial hypotheses as to the origin of these mutations and their impact on the landscape. To meet these objectives, it is cartographic analysis for its qualities of synthesis, which was privileged through the solicitation of geological, historical and archaeological maps.

2. Presentation of the middle valley of the Herault River

2.1. Hydrology and climate

7According to the French public structure EPTB (territorial public basin establishment), from Saint-Jean-de-Fos to Saint-Thibéry, the river flows for 48 km. The catchment basin of the middle Herault valley extends over 1,305 km² (fig. 1). It is characterised by a significant asymmetry of the tributaries, of which the 4 main ones are on the right bank, namely the Lergue River (catchment basin of 428 km²), the Boyne River (catchment basin of 90 km²), the Peyne River (catchment basin of 122 km²) and the Thongue River (catchment basin of 158 km²). Between Saint-Jean-de-Fos and Paulhan, the slope of the Herault River is 1 m/km, then 0.6 m/km to Saint-Thibéry.

8The middle Herault valley benefits from a Mediterranean-type climate (Briola and Larrey, 2011). It is characterised by mild fall and winter seasons, but interspersed with heavy rainfall. These are the Cevennes episodes. The spring seasons are temperate and marked by occasional thunderstorms. On the other hand, the summers are dry and mark a period of pluviometric deficit at the origin of low flows for the rivers. These climatic vagaries specific to the region particularly influence the activity of the river. According to data from the French Hydrological Bank, the average monthly flow of the Herault River in February 2017 in Gignac was 79.8 m3/s, in May 2017 it was 24.10 m3/s and it drops to 16.42 m3/s in July 2017.

Fig. 1 – Location maps of the Herault Valley.
Fig. 1 – Cartes de localisation de la vallée de l’Hérault.

Fig. 1 – Location maps of the Herault Valley.Fig. 1 – Cartes de localisation de la vallée de l’Hérault.

A: National scale; B: Local scale; C: Section 1; D: Section 2; E: Section 3. 1. Herault Valley; 2. Adjacent catchment basin; 3. Lake and ponds; 4. Sea ; 5. River ; 6. Tributaries ; 7. Departmental boundaries.
A : Échelle nationale ; B : Échelle départementale ; C : Tronçon 1 ; D : Tronçon 2 ; E : Tronçon 3. 1. Vallée de l’Hérault ; 2. Bassins versants adjacents ; 3. Lac et étangs ; 4. Mer ; 5. Fleuve; 6. Affluents; 7. Limites départementales.

2.3. Geological setting

9In its high valley, the river cuts the Jurassic limestone to form the Herault gorges. The minor bed is cashed there. Its slope is steep and the gully pronounced (Ambert, 2004). We leave the high valley beyond the normal east-west oriented Arboras faults, at the southern edge of the Causse du Larzac, where the limestone and dolomitic slopes of the mountains of Saint-Guilhem le Désert are formed to open on the floodplain of the middle valley. This floodplain is settled in the hollow of more malleable Miocene formations in yellow and blue marl and sandy molasses. They favoured alluvial erosion and the installation, according to the glacial and interglacial cycles of the Quaternary, of a set of alluvial terraces in sands and pebbles. These alluvial terraces are demarcated in the west by a group of low-altitude peaks, culminating at 100 m a.s.l. on average. These peaks are contiguous to the normal fault of the Cevennes oriented north-south. Beyond this fault stands the Black Mountain (Montagne Noire), which owes its name to its substrate in gneiss and granite (Alabouvette, 1982). East of the river, we found low-lying Miocene mountains in blue and yellow marls. They are overtaken by a set of faults and are blocked by limestone rocks for the Eocene and gray dolomite for the Middle Jurassic south of the town of Saint-Bauzille-de-la-Sylve (Berger et al., 1981). For this study, we have chosen to place the southern limit of the middle valley at the level of the commune of Saint-Thibéry, where the last geological layers of the continental Pliocene plunge. Onwards, the geological formations and the geomorphology of the valley are influenced by the coastal dynamics, we are then in the lower valley.

2.3. River dynamics and geomorphology

10Understanding the evolution of the alluvial landscape require understanding the nature of the river dynamic’s impact on the geomorphology of the floodplain. Due to its bank erosion capacity and aggradation, rivers have constantly reshaped the soil cover of alluvial terraces since the beginning of the Holocene (Bravard and Petit, 2000). Under the combined influence of the geomorphology of old terraces, climatic hazard and anthropogenic forcing, the hydrological responses that will result, such as overbank flow, avulsions or natural levee deposits (Arnaud-Fassetta and Landuré, 2015), will shape the organisation of the floodplain (Arnaud-Fassetta and Suc, 2015).

11Thus, from the Middle Neolithic onwards, human activities characterised by agro-pastoralism had a direct impact, through the opening up of the forest landscape, and indirectly through the promotion of colluvial dynamics, on the evolution of the landscape (Desbat and Lascoux, 1999). In the middle Herault Valley, this anthropogenic impact remained low until the recent Neolithic (3rd millennium BC). It was during the Age of Metals (2200-52 BC) that a correlation was observed between the acceleration of slope erosion and human activity (Devillers and Provansal, 2003). We are in an interaction phase during which the impact of the anthropogenic factor on the geomorphology of the valley remains contrasted (Desbat and Lascoux, 1999). During antiquity, the sparse establishment of cultivation terraces on the slopes and the installation of drainage ditches gave rise to a phase of pedo-sedimentary equilibrium (Devillers and Provansal, 2003). Since late antiquity (3rd-4th centuries AD) and during the early Middle Ages, human activities have generated a real destabilisation of the environment. Deforestation will increase the erodibility of soils, promoting instability in colluvial and alluvial dynamics.

2.4. Human occupation

12The middle Herault Valley is not only a geodynamic territory, it is also a territory occupied and exploited by human populations since at least the Bronze Age (fig. 2).

13The signs of occupation in the middle Herault Valley during the Bronze Age are concentrated in the Late Bronze Age IIIb (900-775 BC). In the current state of knowledge, no occupation has been attested in the Herault floodplain. The sites uncovered are mainly perched installations such as the Pioch-du-Télégraphe site (fig. 1-2) in Aumes (Mauné, 2002). Located on the Miocene blue marls of the slopes, it overlooks the floodplain at an altitude of 80 m a.s.l.

Fig. 2 – Distribution of archaeological sites from the Bronze Age to Roman Antiquity, in connection with the reliefs of the middle Herault Valley.
Fig. 2 – Répartition des sites archéologiques de l’Âge du Bronze jusqu’à l’Antiquité Romaine, en lien avec les reliefs de la moyenne vallée de l’Hérault.

Fig. 2 – Distribution of archaeological sites from the Bronze Age to Roman Antiquity, in connection with the reliefs of the middle Herault Valley.Fig. 2 – Répartition des sites archéologiques de l’Âge du Bronze jusqu’à l’Antiquité Romaine, en lien avec les reliefs de la moyenne vallée de l’Hérault.

1. Altitudes in NGF meter; 2. Main streams; 3. Secondary streams; 4. Contour line every 50m; 5. Archaeological sites (a: Pioch du Télégraphe, b: Cessero Oppidum, c: Roman bridge d: Roqueloupie, e: Causse-Est, f: Belarga bridge, g: The Estagnola warehouse, h: The potter's workshop of Estagnola, i: Mas de Fraysse).
1. Altitudes en metre NGFinférieures à 50 m ; 2. Cours d’eau principaux ; 3. Cours d’eau secondaires ; 4. Courbe de niveau tous les 50 m ; 5. Sites archéologiques (a : Pioch du Télégraphe, b : Oppidum de Cessero, c : Pont Romain, d : Roqueloupie, e : Causse-Est, f : Pont de Bélarga, g : L’entrepôt de l’Estagnola, h : L’atelier de potier de l’Estagnola, i : Mas de Fraysse).

14During the First Iron Age, the human occupation is characterised, first of all, by the sustainability of Bronze Age habitats. In addition, thanks to preventive excavation (Chatrain et al., 1996), the archaeological data are more numerous. The nature of the sites uncovered is therefore more heterogeneous, ranging from simple signs of frequentation by the presence of ceramic fragments as on the Causse-Est (fig. 2) in Valros (Feugère and Mauné, 1995), to the current attestation of occupation as on the Roqueloupie site (fig. 2, 4) in Castelnau de Guers (Ropiot, 2007).

15Regarding the layout of the sites, we identify two sites in the floodplain at Campagnan. The first site, known as the Peyralous (fig. 2) (Feugère and Mauné, 1995), is a small isolated habitat from the middle of the 6th century BC. Located on the average Fyb terrace, it is located in a non-floodable area 500 m from the Herault riverbed. The second site known as the Bélarga Bridge (fig. 2, 6) is a burial from the second half of the 7th century BC found in the current west bank of the Herault River (Arnal, 1963).

16For Second Iron Age, the archaeological signs of occupation are rarer. It is around the town of Aumes that archaeological finds have been most numerous. However, none of these establishments are located in the floodplain, as occupations are overlooking the river as the Cessero oppidum (fig. 2) witnesses, rising at 80 m a.s.l. on the basaltic heights of the Fort (Coulouma and Claustre, 1943).

17During Roman antiquity, again, the human settlement on the edge of the Herault River is generally little attested. It is around the city of Aspiran that archaeological excavations have multiplied and have brought to light eight sites with unprecedented proximity to the river (Chatrain et al., 1996). Indeed, these eight sites are all located on an alluvial terrace. The Estagnola sites (fig. 2), a wine amphora production workshop and a warehouse (Mauné and Desbonnets, 2017), have the particularity of being located a few tens of metres from the river on its major bed, revealing in this place a reduced impact of river activity on the geomorphology of the terrace since the High Empire. In Tressan, another wine amphora production workshop (fig. 2) has been discovered within 35 m of the Herault River (Mauné, 2001). These closer relations with the Herault River invite us to consider the interactions that may have existed between the respective evolution of the river and human occupations, as well as their consequence on the evolution of the landscape.

3. Methodology

3.1. Geomorphological analysis of the floodplain

18The geological map of the Bureau of Geological and Mining Research (BRGM) at 1:25,000 has been adapted to the subject of our study. For this, the pre-Quaternary structures have been grouped by epoch or period in order to highlight the detail of the Holocene alluvium. The main objective of this harmonisation was to shed light on the geomorphological organisation of the floodplain, and in a second step, to facilitate the characterisation of the river sections according to their erosion and alluvial dynamics. These first degrees of reading made it possible to single out the areas in which the river dynamics were the most important. Linked to the organisation of archaeological sites, these results offer the possibility of submitting correlation hypotheses between the geomorphology of the valley, the strategies for establishing habitat and the conservation of the sites.

3.2. Palaeogeography reconstruction?

19Historical maps compose a privileged corpus of archives in restoring the ancient landscape. Confronted with each other through a chronological approach, they allowed us to retrace the recent evolution of the hydrography of the Herault River. The maps requested were those of the Cassini (AD 1747-1815), the General Staff (AD 1820-1866) and the Napoleonic cadastre (AD 1808-1850). The maps were georeferenced with a spline transformation. This method of rubberised stretching optimises local accuracy but not overall accuracy.

20The Cassini map, produced at 1:86,400 suffers from deformations resulting from its projection system, known as the Cassini-Soldner. Consequently, if the hydrography of the river is carefully represented there, since it presents the width of the minor bed, the vegetated islets and the side channels, the geographical position of these is not exact. The elements relating to this card will therefore be approached with caution. Regarding the Napoleonic cadastre, 41 plates dated from AD 1821 to 1836 were collected. As for the General Staff, the sheets covering the territory of the middle Herault Valley, which we have used, are the two sheets of Bédarieux, the sheet of Narbonne and that of Montpellier. They were produced in 1:40,000 scale and are dated 1866. Here, the chosen projection system allows us to envisage with more precision the hydrographic evolution of the Herault River.

21In order to gather, assemble and extract the data from these documents, each map or cadastre has been georeferenced in a GIS (Geographic Information System). Once these maps were georeferenced, the layout of the Herault River was manually vectorised for each of them. Finally, each ancient river hydrography has been superimposed on the current one, to extract the main differences. In this way, sixteen palaeochannels were located, seven of which were recognised on topographic surveys by the presence of depressions.

3.3. Hydromorphometry

22In order to analyse the evolution of the hydrography of the Herault River, it appeared necessary to clarify its hydromorphological functioning. For this, we carried out a morphometric analysis from the IGN (National Geographic Institute) topographic map at 1:25,000.

23We thus determined W, the width at full edge from the inflection points between two sinuosities. The width was measured every 50 m along the valley, then an average value was deduced from it. We then determined the meandering envelope (EM) corresponding to the meander development area (Malavoi and Bravard, 2010). It allows estimating the amplitude of evolution of the river hydrography. We then took the dimensions L (dev), L (EM), L (inflex) and L (arc) corresponding to the developed length of the watercourse, the length of the meandering envelope, the length of the meandering envelope passing through the points of inflection of the meanders and the arc length of a meander, respectively. To determine the dynamics of meandering we have identified Rc, the value in metres of the radius of curvature of a meander. Divided by W, we get a relative value to assess the degree of maturity of a meander. The meandering is also characterised by its amplitude A and its wavelength ʎ. Reduced to a relative value by dividing them by W, they make it possible to specify the meandering character of a watercourse and to characterise its geodynamic activity (Malavoi and Bravard, 2010). The sinuosity coefficient SI used to classify the river according to the development of its meanders was calculated by dividing the developed length of the river by the length of the meandering envelope (Malavoi and Bravard, 2010).

24Finally, we were able to determine the sinuosity coefficient SI, corresponding to the quotient of L (dev) by L (EM). It allows the hydrography of the river to be defined metrically by determining the degree of sinuosity. All of these data will allow us to identify and discriminate sections in the middle valley according to their morphodynamic character. The margins of error are the difference between the average measurements after they have all been taken three times.

3.4. Geomorphology mapping

25The results of the geological, hydromorphometric and palaeogeographic analyses were synthesised in a geomorphological map. This aims to offer a dynamic vision of the middle Herault valley through six themes. First of all, a topographic background of the valley presents the current rivers, vineyards and wooded areas. Secondly, the alluvial terraces and the Miocene slopes are represented according to their lithological nature in order to highlight the banks most subject to river erosion. Conversely, this information makes it possible to identify the geological layers that most constrain hydrography. Third, information regarding geomorphological dynamics has been collected. Thus, areas with significant sedimentary deposits were identified by crossing information from historical maps and observations from aerial photos of the Herault River. To this were added field observations highlighting geological ridges marking the staging of the old alluvium. Fluvial erosion zones have been characterised according to the lithological nature of the layers in order to highlight the impact of the sedimentary texture of the alluvial terraces on the evolution of the fluvial hydrography. To this information was added the notion of risk. For this, we have integrated data from the PPRI (Flood Risk Prevention Plan) presenting the areas at high or moderate risk of water overflow. In addition, by crossing the results bringing together the erosive activity of the river, the attestations of avulsions and the presence of archaeological sites, we were able to discriminate the sections for which the fluvial dynamics was at the origin of evolution in the forms of the landscape, of those who have less suffered from this influence.

26Thus, this geomorphological map had the primary interest of showing all the palaeochannels extracted from the processing of historical maps. To these is added a palaeochannel identified during an archaeological excavation campaign scheduled for 2017, on the Estagnola site in Aspiran (Mauné and Desbonnets, 2017). The dates of these channels are not absolute but correspond to those of the production of the maps from which they were extracted. Finally, the archaeological sites of the floodplain have been integrated in order to testify to the conservation of the river route or the anthropogenic forcing causing avulsions.

4. Results

4.1. Geomorphology of an inhabited territory

27The geomorphological analysis of the middle Herault Valley through the geological map allowed us to divide the floodplain into three sections.

28The first section is between Saint-Jean-de-Fos and Saint-André-de-Sangonis (fig. 3). It is characterised by the presence of an alluvial sandy-silty terrace (Fz) of 250 m wide on average. The profile of the Herault River is encased there with a depth of about 15 m for an average full-edge width of 59 m. The average slope on this section is about 0.14 m/km, it is disturbed by the presence of three contemporary structures which maintain threshold levels. On the left bank, the alluvial deposits of sand and gravel from the average Fy terrace face the Miocene slope of yellow marl and sandstone. On the right bank, there is a succession of strath terraces. The average intermediate Fy2 terrace consists of sand and gravel. It is dominated by the old alluvium of the high Fx terrace composed of sands, pebbles and cryoclastic material. This terrace is connected to the limestone structure of Saint-Guilhem-Arboras to form the spreading of limestone gravel from the FPx terrace. The morphology and the low development of the Holocene alluvium and the encasement of the Herault River in this section testify to an incision dynamic of taking precedence over the dynamics of alluviation. Thus, at this location, it is the geomorphology of the old terraces that constrains the hydrography of the Herault River. The margin of evolution of the river is reduced there, revealing both the low probability of major changes in the fluvial route during the Holocene, and a reduced impact of alluvial phenomena on the modification of the landscape.

Fig. 3 – Geology of the first section and human occupation of the middle Herault Valley during the Bronze Age.
Fig. 3 – Géologie du premier tronçon et occupation humaine de la moyenne vallée de l’Hérault durant l’Âge du Bronze.

Fig. 3 – Geology of the first section and human occupation of the middle Herault Valley during the Bronze Age.Fig. 3 – Géologie du premier tronçon et occupation humaine de la moyenne vallée de l’Hérault durant l’Âge du Bronze.

1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium terrace (Fy); 4. Cryoclastic glacis of accumulation, high terrace (FPx); 5. Old alluvium of high terrace (Fx); 6. Colluvium (C); 7. Scree (E); 8. Pouring basalt (β); 9. Miocene formation (m1-3); 10. Eocene formation (e1-6); 11. Jurassic training (l1-8); 12. Permian formation; 13. Archaeological site; 14. Herault River; 15. Appointment of geological layers.
1. Alluvions récentes (Fz) ; 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse (Fy) ; 4. Glacis d’accumulation cryoclastique, haute terrasse (FPx) ; 5. Alluvions anciennes de haute terrasse (Fx) ; 6. Colluvions (C) ; 7. Éboulis (E) ; 8. Basalte de coulée (β) ; 9. Formations miocènes (m1-3) ; 10. Formations éocènes (e1-6) ; 11. Formations jurassiques (l1-8) ; 12. Formations permiennes ; 13. Site archéologique ; 14. Fleuve Hérault ; 15. Dénomination des couches géologiques.

29From the Bronze Age to Roman antiquity, the archaeological discoveries are few and are concentrated mainly around the municipalities of Saint-Jean-de-Fos and Aniane with the notable presence of the slope habitat of the Bois de Brousses for the Final Bronze III (Garcia, 1993). These dwelling took advantage of the narrowness of the floodplain to settle in the direct vicinity of the river, benefiting from privileged access to fresh water, while exploiting the natural shelters offered by the Jurassic slopes. The second section extends from the municipalities of Pouzols to that of Aspiran (fig. 4). The development of the Fz alluvium is more significant with an average width of 400 m. The river channel is quite enclosed with a depth ranging from 5 to 10 m for a full-edge width of 61 m. The average slope on this section is about 0.8 m/km, it is disturbed by the presence of two contemporary structures. The slight incision at the bottom of the minor bed goes hand in hand with the interlocking of the alluvial terraces. Thus, the Fz, Fy2 and Fyb terraces border the river on each bank. Unlike the previous section, the dynamics of sediment deposition are more important here than those of the incision. Consequently, the margin of evolution of the river is wider and results in a succession of meanders constrained by the old Fy and Fx terraces. It is on this alluvial territory that, as we mentioned earlier, we observe the most important set of sites in the floodplain for the Roman period (fig. 4).

Fig. 4 – Geology of the second section and occupation of the Herault floodplain during Roman antiquity.
Fig. 4 – Géologie du deuxième tronçon et occupation de la plaine alluviale de l’Hérault durant l'Antiquité romaine.

Fig. 4 – Geology of the second section and occupation of the Herault floodplain during Roman antiquity.Fig. 4 – Géologie du deuxième tronçon et occupation de la plaine alluviale de l’Hérault durant l'Antiquité romaine.

1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium upper terrace (Fyb); 4. Alluvium of medium terrace (Fy); 5. Old alluvium of high terrace (Fx); 6. Villafranchian gravel (Fv); 7. Lacustrine silt and clay (FL); 8. Wind silt and sand (N); 9. Miocene formation (m1-3); 10. Archaeological site; 11. Herault River ; 12. Appointment of geological layers.
1. Alluvions récentes (Fz) ; 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse inférieure (Fyb) ; 4. Alluvions de moyenne terrasse (Fy) ; 5. Alluvions anciennes de haute terrasse (Fx) ; 6. Cailloutis villafranchiens (Fv) ; 7. Limons et argiles lacustres (FL) ; 8. Limons et sables éoliens (N) ; 9. Formations miocènes (m1-3) ; 10. Site archéologique ; 11. Fleuve Hérault ; 12. Dénomination des couches géologiques.

30Furthermore, no sign of occupation was reported on this section, in the floodplain, during the Iron Ages (fig. 5). In fact, the occupations uncovered focus on the slopes. This absence of vestiges for the pre-Roman periods, in the floodplain, could be explained by a more important alluvial dynamic in this sector as we will demonstrate, favouring the frequent reworking of the terraces, thus causing the disturbance of the stratigraphy. As a result, the conservation of tenuous remains is compromised, contrary to what will be observed in the third section.

Fig. 5 – Geology of the second and third sections and occupation of the middle valley of the Herault River during the First Iron Age.
Fig. 5 – Géologie des deuxième et troisième tronçons et occupation de la moyenne vallée de l’Hérault durant le Premier Âge du Fer.

Fig. 5 – Geology of the second and third sections and occupation of the middle valley of the Herault River during the First Iron Age.Fig. 5 – Géologie des deuxième et troisième tronçons et occupation de la moyenne vallée de l’Hérault durant le Premier Âge du Fer.

1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium upper terrace (Fyb); 4. Alluvium of medium terrace (Fy); 5. Alluvium from the upper terrace (Fya); 6. Old alluvium of high terrace (Fx); 7. Villafranchian gravel (Fv); 8. Colluvium (C); 9. Lacustrine silt and clay (FL); 10. Volcano-detrital sediment (Vs); 11. Basalt (β); 12. Miocene formation (m1-3); 13. Oligocene formation (g2-3); 14. Eocene formation (e1-6); 15. Archaeological site; 16. Herault River; 17. Appointment of geological layers.
1. Alluvions récentes (Fz); 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse supérieure (Fyb) ; 4. Alluvions de moyenne terrasse (Fy) ; 5. Alluvions de terrasse supérieure (Fya) ; 6. Alluvions anciennes de haute terrasse (Fx) ; 7. Cailloutis villafranchiens (Fv) ; 8. Colluvions (C) ; 9. Limons et argiles lacustres (FL) ; 10. Sédiments volcano-détritiques (Vs) ; 11. Basalte de coulée (β) ; 12. Formations miocènes (m1-3) ; 13. Formations oligocènes (g2-3) ; 14. Formations éocènes (e1-6) ; 15. Site archéologique ; 16. Fleuve Hérault ; 17. Dénomination des couches géologiques.

31From Bélarga to Saint-Thibéry the widening of the Fz terrace is pronounced to reach an expansion of 1,480 m on average. The full width of the channel narrows around an average value of 47 m for 3 to 6 m deep (fig. 6). The average slope on this section is about 0.6 m/km. The Fz and Fy terraces are filled and separated by a morphological leap drawing the middle Fya terrace, which form stepped terraces. In this section, the risk of flooding is greater. It is reflected in particular from an archaeological point of view by the establishment of perched habitat overlooking the floodplain. One can thus cite the oppidum of Puech-du-Télégraphe in Aumes (Mauné, 2002), and the occupation of the First Iron Age of Roqueloupie in Castelnau de Guers (Ropiot, 2007), overlooking an ancient ford on the meanders of the Herault River. In Saint-Thibéry, the contact zone between the middle and the lower valley, the same phenomenon can be observed. The significant development of the Holocene terrace made up of easily mobilised silty sands makes the multiplication of avulsions possible.

Fig. 6 – Geology of the third section and occupation of the middle valley of the Herault River during the Second Iron Age.
Fig. 6 – Géologie du troisième tronçon et occupation de la moyenne vallée de l’Hérault durant le Second Âge du Fer.

Fig. 6 – Geology of the third section and occupation of the middle valley of the Herault River during the Second Iron Age.Fig. 6 – Géologie du troisième tronçon et occupation de la moyenne vallée de l’Hérault durant le Second Âge du Fer.

1. Recent alluvium (Fz); 2. Alluvium of medium terrace (Fy); 3. Old alluvium of high terrace (Fx); 4. Villafranchian gravel (Fv); 5. Colluvium (C); 6. Scree (E); 7. Basalt (β); 8. Continental Pliocene (Pc); 9. Miocene formation (m1-3); 10. Eocene formation (e1-6); 11. Cretaceous formation (c2-7); 15. Archaeological site; 16. Herault River; 17. Appointment of geological layers.
1. Alluvions récentes (Fz); 2. Alluvions de moyenne terrasse (Fy) ; 3. Alluvions anciennes de haute terrasse (Fx) ; 4. Cailloutis villafranchiens (Fv) ; 5. Colluvions (C) ; 6. Éboulis (E) ; 7. Basalte de coulée (β) ; 8. Pliocène continental (Pc) ; 9. Formations miocènes (m1-3) ; 10. Formations éocènes (e1-6) ; 11. Formations crétacées (c2-7) ; 15. Site archéologique ; 16. Fleuve Hérault ; 17. Dénomination des couches géologiques.

4.2. Hydromorphological evolution

4.2.1. Hydromorphometry

32Hydromorphometric data have shed light on the river dynamics. Thus, in the middle valley, the average width of the full-edge channel (W) is 51.3 m for a developed length of the river of 47.2 km (tab. 1). The sinuosity coefficient obtained with a value of 1.20 reflects the sinuous nature of the river. However, this overall interpretation should be qualified since the fluvial section between Pouzols and Aspiran was distinguished by the development of large meanders whose sinuosity coefficient reaching 1.5, corresponding to the minimum meandering coefficient (tab. 1). However, the radius of curvature Rc, quantifying the degree of maturity of the meander, has an average relative value of 6 W. This equivalence can be explained by the heterogeneity of the meanders in the Pouzols-Aspiran section. Indeed, if seven meanders have a Rc less than 400 m for an average value of 200 m or 3.9 W, three meanders have an average Rc of 561 m or 11 W. In other words, the two-thirds of the meanders of the Pouzols-Aspiran section have an advanced degree of maturity testifying to an erosion capacity higher than the average observed in the rest of the valley, where the radii of curvature are 12 W for section 1 and 8 W for section 3 (tab. 1). However, this value, compared to the reference values (Malavoi and Bravard, 2010), indicates a low erosion capacity for these meanders. This result is explained by the fact that the meanders are constrained and not free in this section. The singularity of these meanders is also characterised by the importance of their amplitude, since they are in comparison with the average observed, 3.7 times larger. Although important, this 26 W amplitude is characteristic of meandering rivers that are not very active. Likewise, the wavelength indicating the frequency of meander oscillations is 33 W in the middle valley when it is 9 W in this section. The meanders are therefore 3.7 times closer to each other between Pouzols and Aspiran.

33To sum up, the geomorphological tripartition of the floodplain is also reflected in the morphometric organisation of the Herault River. It shows the heterogeneity of the geomorphological and hydrological dynamics within the middle valley. This heterogeneity implies dissimilar evolutionary responses. Consequently, the hydrography of the Herault and the morphology of its alluvial landscape have not known a single evolution scenario, which would translate a global response to different forcing, as evidenced by the comparative analysis and the georeferencing of the old maps.

Tab. 1 – Hydromorphometric data of the middle Herault Valley and the three river sections.
Tab. 1 – Données hydromorphométriques de la moyenne vallée de l’Hérault et des trois tronçons fluviaux.

Tab. 1 – Hydromorphometric data of the middle Herault Valley and the three river sections.Tab. 1 – Données hydromorphométriques de la moyenne vallée de l’Hérault et des trois tronçons fluviaux.

4.2.2. Palaeogeography and hydromorphodynamics

34The analysis of the historical maps and their georeferencing confirmed the observations made after studying the cartographic data relating to the geomorphology of the middle valley.

35The first section, singled out by a narrow Fz terrace, shows a relative conservation of the hydrography of the Herault River since the middle of the 18th century AD, since no palaeochannel was identified on the old maps (fig. 7). In the upper valley, upstream from Saint-Jean-de-Fos, the slope of the river channel is on average four times greater than in the middle valley with a value of around 4.4 m / km. The Saint-Jean-de-Fos sector was originally a deposit sector located on the alluvial cone of the upstream Hérault. However, due to contemporary developments, there is only a small surface area of alluvial deposits at the confluence of the upper and middle valleys (SMBFH, 2012). Due to aggregate extraction operations, banned after 1992-1993, the minor bed has been incised. Thus, between 1967 and 1991 the EPTB hydrological surveys showing the level of the water line showed an average decrease of 1.71 m. Consequently, it is risky on this section to bring out a river dynamic which would make it possible to understand the ancient functioning of the river. On the other hand, one notes a reduced mobility of the Hérault River and constrained by the geomorphology of the Pleistocene terraces. The hydromorphology of the river in this sector therefore has a more restricted development perimeter than in sections 2 and 3.

36Conversely, we observed in the second section a higher alluvial capacity and a reduced minor bed incision. This reversal of hydrodynamics can be explained by the spill of the Lergue into the Herault River at the commune of Pouzols. This tributary taking its source in the Causses du Larzac constitutes the first major tributary that the Herault River meets in its middle valley. This confluence results in a significant external sediment supply in the main river (SMBFH, 2012). This natural sediment input has allowed this sector to maintain a natural dynamic, little impacted by contemporary aggregate extractions. This phenomenon is reflected in particular by the presence of large sandy banks in convexities of the meanders between Pouzols and Canet. Subsequently, during flood events, the river will remobilise these fine sediment deposits. It is notably this phenomenon as well as that relating to the natural evolution of the meanders, which will explain the significant changes in the river route on this section. On the last section, the inflow of external sediments is increasing since the Herault River is joined by two large tributaries on the right bank. The alluvial phenomenon is then more significant. However, as the slope of the Herault River eases (Bouchène, 2019) the propensity of the river to meander decreases. As a result, scattered palaeochannels could be identified, without, however, marking as strong hydrographic transitions as those observed between Pouzols and Aspiran.

Fig. 7 – Geomorphological map of the first section of the middle Herault Valley.
Fig. 7 – Carte géomorphologique du premier tronçon de la moyenne vallée de l’Hérault.

Fig. 7 – Geomorphological map of the first section of the middle Herault Valley.Fig. 7 – Carte géomorphologique du premier tronçon de la moyenne vallée de l’Hérault.

1. Administrative fund; 2. Vineyard; 3. Forest, wood, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Sandy alluvium; 7. Alluvial deposit of sand and pebbles; 8. Cryoclastic deposit; 9. Alluvial deposit of sand, pebbles and gravel; 10. Limestone slope; 11. Strong erosion zone; 12. Moderate erosion zone; 13. Zone of alluvial deposition; 14. Spot height; 15. Archaeological site; 16. Mediaeval mills; 17. Palaeohydrography of the Herault River according to the Napoleonic cadastre.
1. Fond administratif; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Alluvions sableuses ; 7. Alluvions de sables et galets ; 8. Dépôts cryoclastiques ; 9. Alluvions de sables, galets et cailloutis ; 10. Versant calcaire ; 11. Zone de forte érosion ; 12. Zone d’érosion modérée ; 13. Zone d’accumulation sédimentaire ; 14. Point coté (en m) ; 15. Site archéologique ; 16. Moulin médiéval ; 17 Paléohydrographie de l’Hérault d’après le Cadastre napoléonien.

4.3. Palaeohydrography

37The first palaeochannel identified is located near Pouzols. It reflects a lateral displacement of the minor bed (fig. 8). The imprecision of the Cassini map led to a georeferencing of this channel on the average Fy terrace. The realisation of topographic profiles at the level of the location of this channel made it possible to identify a thalweg, further east, on the Fz terrace, taking up the hydrography of the minor bed indicated on the map of AD 1740. However, only observations in the field would allow us to assess the nature of this depression.

Fig. 8 – Geomorphological map of the second section of the middle Herault Valley.
Fig. 8 – Carte géomorphologique du deuxième tronçon de la moyenne vallée de l’Hérault.

Fig. 8 – Geomorphological map of the second section of the middle Herault Valley.Fig. 8 – Carte géomorphologique du deuxième tronçon de la moyenne vallée de l’Hérault.

1. Administrative fund; 2. Vineyard; 3. Forest, wood, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Area at moderate risk of flooding 7. Sandy alluvium; 8. Alluvial deposit of sand and pebbles; 9. Gravel; 10. Alluvial deposits of sand, pebbles and gravel; 11. Limestone slope; 12. Strong erosion zone; 13. Moderate erosion zone; 14. Zone of alluvial deposition; 15. Spot height; 16. Bank; 17. Archaeological site; 18. Mediaeval mills; 19. Palaeochannel according to the Cassini’s map; 20. Palaeochannel prior to AD 70 at the Estagnola site; 21. Dating from historical maps.
1. Fond administratif; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Zone à risque d’inondation modéré ; 7. Alluvions sableuses ; 8. Alluvions de sables et galets ; 9. Dépôts cryoclastiques ; 10. Alluvions de sables, galets et cailloutis ; 11. Versant calcaire ; 12. Zone de forte érosion ; 13. Zone d’érosion modérée ; 14. Zone d’accumulation sédimentaire ; 15. Point coté (en m) ; 16. Talus ; 17. Site archéologique ; 18. Moulin médiéval ; 19 Paléochenal d’après la carte de Cassini ; 20. Paléochenal antérieur à 70 apr. J.-C. sur le site de l’Estagnola ; 21. Datation d’après les cartes historiques.

38Downstream of the confluence between the rivers Lergue and Herault, we observe the most important evolution of the fluvial hydrography, consequence of a hydrological activity disturbed by an important sedimentary supply. Thus, in AD 1740, according to the Cassini map, the river flows in a straight line and its right bank is cultivated with the establishment of vines on the average Fy terrace. There too, topographic surveys revealed the presence of depressions from 2 to 3.5 deep whose morphology and location evoke those of a palaeochannel.

39On the plan by mass of culture of the Napoleonic cadastre published in AD 1804, the rectilinear layout gave way to important meanders. A pasture area is demarcated on the Fz terrace while the Fy terrace hosts ploughed land and vine plants. In AD 1866, on the General Staff map, the hydrography remained unchanged. However, the Fz terrace is partially overgrown with water to form a pond. Wine farms are multiplying on the alluvium of low, medium and high terraces. In 1950, the basin dried up and gave way again to the establishment of culture. Today, the pond is in water and forms a privileged area for the accumulation of sediments carried by the Lergue.

40From Aspiran to Usclas-d'Hérault, we do not observe any major evolution of the river hydrography, as evidenced by the conservation of a warehouse of the High Roman Empire on the tenement of the Estagnola in Aspiran, 40 from the current channel, or the discovery in Bélarga of a burial from the second half of the 7th century AD (Arnal 1963) located on the bank of the Herault River (fig. 9). Our hypothesis is that released upstream of its excess solid load, the functioning of the river is stabilised. Consequently, we observe a relative sustainability of the river hydrography on this section.

41Between Cazouls d'Hérault and Saint Thibéry, as we have seen, the Herault River is joined by two major tributaries: the Peyne River and the Thongue River. As the slope of the bed decreases as it approaches the lower valley, these two tributaries disturb the river balance. In order to remove both sediment and liquid overload, the river dynamics will be characterised by a significant alluvial flux, favouring heterogeneous avulsion phenomena. Thus, we observe both the appearance of secondary channels at the origin of the islet formation, as well as lateral displacements of the main channel.

Fig. 9 – Geomorphological map of the third section of the middle Herault Valley.
Fig. 9 – Carte géomorphologique du troisième tronçon de la moyenne vallée de l’Hérault.

Fig. 9 – Geomorphological map of the third section of the middle Herault Valley.Fig. 9 – Carte géomorphologique du troisième tronçon de la moyenne vallée de l’Hérault.

1. Administrative fund; 2. Vineyard; 3. Forests, woods, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Area at moderate risk of flooding; 7. Sandy alluvium; 8. Alluvial deposit of sand and pebbles; 9. Gravel; 10. Alluvial deposit of sand, pebbles and gravel; 11. Limestone slope; 12. Strong erosion zone; 13. Moderate erosion zone; 14. Zone of alluvial deposition; 15. Spot height; 16. Bank; 17. Archaeological site; 18. Mediaeval mills; 19. Palaeochannel according to the Cassini map; 20. Palaeochannel according to the Napoleonic cadastre; 21. Palaeochannel according to the General Staff map; 22. Dating from historical maps.
1. Fond administratif ; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Zone à risque d’inondation modéré ; 7. Alluvions sableuses ; 8. Alluvions de sables et galets ; 9. Dépôts cryoclastiques ; 10. Alluvions de sables, galets et cailloutis ; 11. Versant calcaire ; 12. Zone de forte érosion ; 13. Zone d’érosion modérée ; 14. Zone d’accumulation sédimentaire ; 15. Point coté (en m) ; 16. Talus ; 17. Site archéologique ; 18. Moulin médiéval ; 19. Paléochenal d’après la carte de Cassini ; 20. Paléochenal d’après le cadastre napoléonien ; 21. Paléochenal d’après la carte d’État-Major ; 22. Datation d’après les cartes historiques.

4. Discussion

42These analyses highlighted a heterogeneous evolution of the Herault floodplain (fig. 10). This evolution was conditioned at the same time by the geomorphological features of the alluvial terraces, the hydrological dynamics of the river, and the anthropogenic activities.

43First, the geomorphology of the alluvial terraces constrained the hydromorphological functioning of the Herault River at the beginning of the floodplain, favouring the incision of the channel rather than the erosion of the banks and the alluvium. As demonstrated during research carried out on the Rhone delta (Arnaud-Fassetta and Landuré, 2015), the geometry of the channel will condition the maximum absorption capacity of flood flows. Consequently, due to a steep channel, a reduced Holocene terrace and hydrography controlled by the morphology of the Pleistocene alluvium, geomorphological expressions such as avulsions or natural levee deposits were not observed in this section.

Fig. 10 – Heterogeneous evolutionary of hydromorphological process.
Fig. 10 – Processus d'évolution hydromorphologique hétérogène.

Fig. 10 – Heterogeneous evolutionary of hydromorphological process.Fig. 10 – Processus d'évolution hydromorphologique hétérogène.

44In the second section, the sedimentary flows carried by the Lergue River lead to a significant alluvial dynamic, which, due to the creation of mobilisable alluvium, favoured the meandering of the river. The study of mutations in the Rhone floodplain highlights the morphogenic potential of the meandering sections with low banks (Provansal et al., 1999). As we have been able to observe, it is in this type of stretch that the hydromorphological evolution is able to have most modified the alluvial landscape. If the links between climatic variations and responses of the fluvial style are still poorly understood, a study of sedimentary deposits of floods would allow us to characterise periods of hydrological calm or crisis (Arnaud-Fassetta and Landuré, 1997).

45The impact of human activities on the evolution of the Herault hydrography in its middle valley remains to be quantified and characterised. In correlation with the conclusions made on the morphodynamics of the slopes (Devillers and Provansal, 2003), which showed little destabilisation of the environment during the Metal Ages, we do not observe any correlation between anthropic activities, from the Age of the Bronze Age and the Iron Ages, and the evolution of the river route. On the contrary, indications such as the conservation of the burial tomb of the Bélarga Bridge in the river bank (Arnal, 1963), suggest that the hydrography should be preserved on the least winding sections. Furthermore, during the excavations programmed on the Estagnola site in Aspiran, the realisation of deep trenches transverse to the Herault River made it possible to uncover the spreading of openwork pebbles, sometimes arranged in tiling, below the Roman levels (Mauné and Desbonnets, 2017; Bouchene, 2019). Due to their size, certain blocks reaching 20 to 60 cm long, as well as their arrangement these deposits could suggest a fluvial metamorphosis, which occurred before the High Roman Empire, passing from a braid or meander style to the sinuous style of the current Herault River. However, only the multiplication of sediment samples and their dating will allow us to confirm or not this hypothesis.

46Concerning Roman antiquity, archaeological discoveries characterised by the development of an artisanal activity in the floodplain (Mauné, 2001), suggested a relationship with the river diverging from previous centuries. If the dynamics were at the installation in elevation on the slopes, we observe for this period a commercial exploitation of the floodplain. The floodplain then becomes an occupied landscape. These installations then raise the question of the responses provided by ancient societies to the risk of flooding. Indeed, if floods pose a low risk for cultivated land, they can have a devastating impact on masonry structures (Provansal et al., 1999). It is possible, as in the case of what has been observed in the lower Rhone Valley on the ancient grouped habitats of La Capélière or Le Carrelet (Landuré, 2000; Crichton and Siboni, 2001), developments directly affecting the river route, such as bank embankments or riprap, have been carried out. Again, only the excavation of new sites in the floodplain will allow us to assess the content of these potential developments.

47Finally, it is in the Middle Ages that anthropogenic forcing is likely to have been the most important because the relationship between society and the river evolves towards an overinvestment of minor and major beds, as evidenced by research carried out on the Loire River (Burnouf, 2009). Indeed, the stream power of the Herault River will be exploited through the establishment of mills along the middle valley, sixteen of which have been preserved in the immediate vicinity of the river (Phalip, 1992). The combined analysis of their location and their layout, with regard to the palaeochannels that we have mapped, may or may not reveal a correlation between the establishment of these structures and the phenomena of avulsion.

6. Conclusion

48By crossing the results of geomorphological, hydromorphometric, palaeogeographic and archaeological analysis, we were able to define three sections in the floodplain, each characterised by its own dynamics. The first section from Saint-Jean-de-Fos to Saint-André-de-Sangonis is distinguished by the flow of the river in a deep channel and the poor development of the Holocene terrace. The second section defined from the municipalities of Pouzols to Aspiran, particularly stood out from the rest of the valley following the analysis of the hydromorphometric data. Finally, the third section, stretching from Bélarga to Saint-Thibéry, was characterised by the narrowing of the channel, a decrease in its depth and a vast expansion of the Holocene terrace, reflecting a strong dynamic of alluviation. This dynamic was confirmed by the identification of nine palaeochannels. These first results need to be deepened by field observations. Thus, carrying out prospecting along the river would make it possible to characterise with greater precision the geomorphological dynamics of the floodplain and would take into account the probable ancient agricultural exploitation of the alluvial terraces and their impact on the development of the valley. Also, coring in the palaeochannels would make it possible to account for the activity of the river and its impact on the fluvial morphology and the distribution of human settlements.

*Auteur correspondant : Tel : +33 (0)4 67 14 58 10 ambrine.bouchene@etu.univ-montp3.fr (Bouchène A.)

Haut de page

Bibliographie

Alabouvette B. (1982) – Notice explicative de la feuille de Lodève au 1/50 000ème, s.l., BRGM, 51 p.

Ambert A. (2004) – Hérault, miroir de la Terre, BRGM éditions, Montpellier, 157 p.

Arnal J. (1963) – Un torques hallstattico inedito (Belarga, Hérault, Francia). Ampuria, 15, 203- 205.

Arnaud-Fassetta G., Landuré C. (1997) – Occupation du sol et contraintes fluviales dans le delta du Rhône (France du Sud). In Burnouf J., Bravard J.-P., Chouquer P. (Eds.) : Dynamique des paysages protohistoriques antiques, médiévaux et modernes. APDCA, 17, 285-308.

Arnaud-Fassetta G., Landuré C. (2015) – Fluvial risk in rural areas from the Greek period to the early Middle Ages In Carcaud, N., Arnaud-Fassetta, G. (Eds.): La géoarchéologie française au XXIème siècle. CNRS Éditions. 215-236.

DOI : 10.4000/books.editionscnrs.22158

Arnaud-Fassetta G., Suc J.-P. (2015) – Dynamique hydrogéomorphologique et diversité végétale dans le delta du Rhône (France) de -10 000 ans à demain. In Reynard E., Evéquoz-Daven M., Borel G. (Eds.) : Le Rhône, entre nature et société, Cahiers de Vallesia, 29, 63-98.

Berger G., Feist R., Freytet P. (1981) – Notice explicative de la feuille de Pézenas au 1/50 000ème, s.l., BRGM, 40 p.

Bouchène A. (2019) – Géoarchéologie fluviale de la moyenne vallée de l’Hérault de l’Âge du Bronze à l’Antiquité romaine. Mémoire de master, Université Montpellier III, 167 p.

Bravard J.P., Petit F. (2000) – Les cours d’eau, dynamique du système fluvial. Armand Colin, Paris, 222 p.

Brierley G. (2008) – Geomorphology and river management. Kemanusiaan, 15, 13-26.

Brierley G., Fryirs A. (2004) – Geomorphology and river management: applications of the river styles framework. Blackwell publishing, Malden, 405 p.

Burnouf J. (2009) – Vulnérabilité des sociétés médiévales aux aléas météorologiques et climatiques. Archéologie du Midi Médiéval, 27, 249-254.

DOI : 10.3406/amime.2009.1902

Chatrain A., Kotarba J., Mauné S. (1996) – Gazoduc « Artère du Midi » – Moyenne vallée de l'Hérault. ADLFI. Archéologie de la France – Informations, Languedoc-Roussillon, 8 p.

Coulouma J., Claustre G., (1943) – L’oppidum de Cessero (Hérault). Gallia, 1(2), 1-18.

Crichton B., Siboni M., (2001) – La reconstitution paléoydrologique dans les deltas : l’exemple du Rhône d’Ulmet sur les sites du Pont Noir, de La Capelière et de La Tour du Valat (Camargue). Mémoire de Maîtrise, université Paris-Diderot.

Crutzen P.-J. (2002) – Geology of mankind. Nature, 415 (23).

DOI : 10.1038/415023a

Desbat A., Lascoux J.P. (1999) – Le Rhône et la Saône à Lyon à l'époque romaine. Bilan archéologique. Gallia, 56, 45-69.

Devillers B., Provansal M. (2003) – La morphogenèse d'un géosystème cultivé depuis le Néolithique récent : les petits bassins versants de la moyenne vallée de l'Hérault (France). Géomorphologie : Relief, Processus, Environnement, 9 (2), 83-98.

DOI : 10.3406/morfo.2003.1171

Devillers B., Bony G., Degeai J.-P., Gascò J., Lachenal T., Bruneton H., Yung F., Oueslati H., Thierry A. (2019) – Holocene coastal environmental changes and human occupation of the lower Hérault River, southern France. Quaternary Science Reviews, 222, 105912.

DOI : 10.1016/j.quascirev.2019.105912

Feugère M., Mauné S. (1995) – L’occupation du sol du VIIe au Ve siècle avant notre ère dans la moyenne vallée de l’Hérault. Documents d’Archéologie Méridionale, 18, 95-103.

Garcia D. (1993) – Entre Ibères et Ligures, Lodévois et moyenne vallée de l’Hérault protohistorique. CNRS éditions, Paris. 355 p.

Landuré C. (2000) - La Capelière, un habitat fluvial en Camargue. In Chausserie-Lapree J. (Eds.) : Le temps des Gaulois en Provence, Marseille Images en Manœuvre Editions.

Malavoi J.-R., Bravard J.-P. (2010) – Eléments d’hydromorphologie fluviale. Onema, Lyon, 224 p.

Malm A., Hornborg A. (2014) – The geology of mankind? A critique of the Anthropocene narrative. The Anthropocene Review, 1 (1), 62-69.

DOI : 10.1177/2053019613516291

Mauné S. (2001) – Les ateliers de potiers d'Aspiran dans l'Antiquité (Ier-IIIe apr. J.-C.). Bilan et perspectives. In Laubenheimer F. (Ed.) : 20 ans de recherches à Sallèles d'Aude, Institut des Sciences et Techniques de l'Antiquité, Besançon, 163-198.

Mauné S. (2002) – L’oppidum d’Aumes (Hérault). In Fiches J.-L. (Ed.) : Les agglomérations gallo-romaines en Languedoc-Roussillon, CNRS éditions, Lattes, 317-332.

Mauné S., Desbonnets Q. (2017) – L'Estagnola (Aspiran, Hérault). Un vaste entrepôt d'époque romaine sur la berge de l'Hérault (fin du Ier s.-IIe s. apr. J.-C.) 154 p. (non publié).

Phalip B. (1992) – Le moulin à eau médiéval. Problème et apport de la documentation languedocienne. [Bassins de l'Hérault, Orb et Vidourle]. Archéologie du Midi Médiéval, 10, 63-96.

Provansal M., Berger J.-F., Bravard J.-P., Salvador P.-G., Arnaud-Fassetta G., Bruneton H., Vérot-Bourrély A. (1999) – Le régime du Rhône dans l'Antiquité et au Haut Moyen Âge. Gallia, 56, 13-32.

DOI : 10.3406/galia.1999.3241

Ropiot V. (2007) – Peuplement et circulation dans les bassins fluviaux du Languedoc Occidental, du Roussillon et de l’Ampourdan du IXe s. au début IIe s. av. n. è. Thèse de doctorat, Université de Besançon, 402 p.

Schmidt L., Bravard J.-J., Rey F. (2013) – Maîtriser les évolutions du lit des cours d’eau (incision, atterrissement, …) et mieux gérer les formes fluviales. In Chocat B. (Eds.) : Ingénierie écologique appliquée aux milieux aquatiques : Pourquoi ? Comment ?, 84-93.

SMBFH, Syndicat mixte du Bassin du fleuve Hérault (2012) – Etude du transport solide du fleuve Hérault. Note de synthèse de l’étude. (non publié).

Haut de page

Annexe

Version française abrégée

Le territoire de la moyenne vallée de l’Hérault (fig. 1) s’est façonné au rythme des aléas de l'activité fluviale tout au long de l'ère quaternaire. Malgré une occupation humaine de la vallée depuis à minima l’Âge du Bronze (fig. 2), nos connaissances concernant l'évolution de la plaine alluviale et la nature des interactions entre les sociétés anciennes et ce territoire restent fragmentaires. Quels changements hydromorphologiques a connu l'Hérault ? Comment ces développements ont-ils impacté la morphologie de la plaine ? Quels liens peut-on établir entre l’organisation spatiale des données archéologiques et l’évolution du tracé fluvial ?

À travers une analyse combinée des cartes géologiques, historiques et archéologiques, nous avons cartographié les évolutions récentes majeures du tracé fluvial de l’Hérault et évalué leur impact sur l’évolution morphologique de la plaine. Les résultats des analyses géomorphologique, hydromorphométrique et paléogéographique ont convergé vers une tripartition de la moyenne vallée (fig. 2).

Ainsi, le premier tronçon de Saint-Jean-de-Fos à Saint-André-de-Sangonis, est singularisé par l’écoulement du fleuve dans un chenal encaissé et un faible développement de la couche d’alluvions holocènes (fig. 3). L’analyse paléogéographique a révélé des déplacements latéraux du chenal inférieurs à 100 m, conséquence d’une hydrodynamique marquée par une faible érosion des berges au profit d’une forte incision du lit mineur. Par conséquent, sur ce tronçon, le fonctionnement hydrologique du fleuve a une incidence modérée sur l’évolution de la morphologie de la plaine alluviale.

Le deuxième tronçon défini des communes de Pouzols à Aspiran (fig. 4), s’est particulièrement détaché du reste de la vallée à la suite de l’analyse des données hydromorphométriques. La fréquence et l’amplitude des méandres sur ce tronçon est 3,7 fois plus importante que la moyenne observée sur toute la moyenne vallée. Le tracé actuel de style méandriforme apparait peu sinueux sur la carte des Cassini (XVIIIème s. ap. J.-C.). À Aspiran, les fouilles archéologiques ont permis d’identifier au moins un paléochenal antérieur au Haut Empire romain. Cette capacité morphogène pourrait s’expliquer par la confluence de l’Hérault avec la Lergue, à l’aval de laquelle l’on observe une sédimentation plus prononcée que dans le tronçon 1. Par conséquent, les abords du fleuve dans ce tronçon font l’objet d’un remaniement régulier des alluvions holocènes, phénomène observable à l’échelle de la décennie sur les photos aériennes de 1950-1965. Ce phénomène soulève la question, dans ce secteur, d’un possible enfouissement en profondeur des vestiges des Âges des Métaux ou d’une disparition des traces d’occupation les plus ténues.

Enfin, le troisième tronçon, de Bélarga à Saint-Thibéry (fig. 5-6), est caractérisé par le rétrécissement du chenal, une diminution de sa profondeur et une large expansion des alluvions holocènes. A cette dynamique d’alluvionnement s’est ajoutée l’identification de neuf paléochenaux (fig. 7-9). La multiplication de ces défluviations peut être mise en relation avec la présence de deux importants affluents et l’implantation de plusieurs moulins hydrauliques médiévaux, dont la localisation converge parfois avec l’origine d’un déplacement latéral du lit mineur.

Les résultats de cette analyse nous invitent à explorer la question des facteurs majeurs et mineurs à l’origine de la sectorisation des dynamiques fluviales (fig. 10) et mettre en évidence une chronologie de l’évolution de ces dynamiques à travers une analyse des archives sédimentaires prises dans les paléochenaux. Ils permettent également d’ouvrir une réflexion autour de l’impact de l’évolution de l’activité fluviale sur les implantations humaines anciennes et réciproquement, l’impact hydromorphologique des aménagements anthropiques sur les berges de l’Hérault.

Haut de page

Table des illustrations

Titre Fig. 1 – Location maps of the Herault Valley.Fig. 1 – Cartes de localisation de la vallée de l’Hérault.
Légende A: National scale; B: Local scale; C: Section 1; D: Section 2; E: Section 3. 1. Herault Valley; 2. Adjacent catchment basin; 3. Lake and ponds; 4. Sea ; 5. River ; 6. Tributaries ; 7. Departmental boundaries.A : Échelle nationale ; B : Échelle départementale ; C : Tronçon 1 ; D : Tronçon 2 ; E : Tronçon 3. 1. Vallée de l’Hérault ; 2. Bassins versants adjacents ; 3. Lac et étangs ; 4. Mer ; 5. Fleuve; 6. Affluents; 7. Limites départementales.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-1.jpg
Fichier image/jpeg, 932k
Titre Fig. 2 – Distribution of archaeological sites from the Bronze Age to Roman Antiquity, in connection with the reliefs of the middle Herault Valley.Fig. 2 – Répartition des sites archéologiques de l’Âge du Bronze jusqu’à l’Antiquité Romaine, en lien avec les reliefs de la moyenne vallée de l’Hérault.
Légende 1. Altitudes in NGF meter; 2. Main streams; 3. Secondary streams; 4. Contour line every 50m; 5. Archaeological sites (a: Pioch du Télégraphe, b: Cessero Oppidum, c: Roman bridge d: Roqueloupie, e: Causse-Est, f: Belarga bridge, g: The Estagnola warehouse, h: The potter's workshop of Estagnola, i: Mas de Fraysse).1. Altitudes en metre NGFinférieures à 50 m ; 2. Cours d’eau principaux ; 3. Cours d’eau secondaires ; 4. Courbe de niveau tous les 50 m ; 5. Sites archéologiques (a : Pioch du Télégraphe, b : Oppidum de Cessero, c : Pont Romain, d : Roqueloupie, e : Causse-Est, f : Pont de Bélarga, g : L’entrepôt de l’Estagnola, h : L’atelier de potier de l’Estagnola, i : Mas de Fraysse).
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-2.jpg
Fichier image/jpeg, 1,6M
Titre Fig. 3 – Geology of the first section and human occupation of the middle Herault Valley during the Bronze Age.Fig. 3 – Géologie du premier tronçon et occupation humaine de la moyenne vallée de l’Hérault durant l’Âge du Bronze.
Légende 1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium terrace (Fy); 4. Cryoclastic glacis of accumulation, high terrace (FPx); 5. Old alluvium of high terrace (Fx); 6. Colluvium (C); 7. Scree (E); 8. Pouring basalt (β); 9. Miocene formation (m1-3); 10. Eocene formation (e1-6); 11. Jurassic training (l1-8); 12. Permian formation; 13. Archaeological site; 14. Herault River; 15. Appointment of geological layers.1. Alluvions récentes (Fz) ; 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse (Fy) ; 4. Glacis d’accumulation cryoclastique, haute terrasse (FPx) ; 5. Alluvions anciennes de haute terrasse (Fx) ; 6. Colluvions (C) ; 7. Éboulis (E) ; 8. Basalte de coulée (β) ; 9. Formations miocènes (m1-3) ; 10. Formations éocènes (e1-6) ; 11. Formations jurassiques (l1-8) ; 12. Formations permiennes ; 13. Site archéologique ; 14. Fleuve Hérault ; 15. Dénomination des couches géologiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-3.jpg
Fichier image/jpeg, 448k
Titre Fig. 4 – Geology of the second section and occupation of the Herault floodplain during Roman antiquity.Fig. 4 – Géologie du deuxième tronçon et occupation de la plaine alluviale de l’Hérault durant l'Antiquité romaine.
Légende 1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium upper terrace (Fyb); 4. Alluvium of medium terrace (Fy); 5. Old alluvium of high terrace (Fx); 6. Villafranchian gravel (Fv); 7. Lacustrine silt and clay (FL); 8. Wind silt and sand (N); 9. Miocene formation (m1-3); 10. Archaeological site; 11. Herault River ; 12. Appointment of geological layers.1. Alluvions récentes (Fz) ; 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse inférieure (Fyb) ; 4. Alluvions de moyenne terrasse (Fy) ; 5. Alluvions anciennes de haute terrasse (Fx) ; 6. Cailloutis villafranchiens (Fv) ; 7. Limons et argiles lacustres (FL) ; 8. Limons et sables éoliens (N) ; 9. Formations miocènes (m1-3) ; 10. Site archéologique ; 11. Fleuve Hérault ; 12. Dénomination des couches géologiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-4.jpg
Fichier image/jpeg, 208k
Titre Fig. 5 – Geology of the second and third sections and occupation of the middle valley of the Herault River during the First Iron Age.Fig. 5 – Géologie des deuxième et troisième tronçons et occupation de la moyenne vallée de l’Hérault durant le Premier Âge du Fer.
Légende 1. Recent alluvium (Fz); 2. Alluvium of medium intermediate terrace (Fy2); 3. Alluvium of medium upper terrace (Fyb); 4. Alluvium of medium terrace (Fy); 5. Alluvium from the upper terrace (Fya); 6. Old alluvium of high terrace (Fx); 7. Villafranchian gravel (Fv); 8. Colluvium (C); 9. Lacustrine silt and clay (FL); 10. Volcano-detrital sediment (Vs); 11. Basalt (β); 12. Miocene formation (m1-3); 13. Oligocene formation (g2-3); 14. Eocene formation (e1-6); 15. Archaeological site; 16. Herault River; 17. Appointment of geological layers.1. Alluvions récentes (Fz); 2. Alluvions de moyenne terrasse intermédiaire (Fy2) ; 3. Alluvions de moyenne terrasse supérieure (Fyb) ; 4. Alluvions de moyenne terrasse (Fy) ; 5. Alluvions de terrasse supérieure (Fya) ; 6. Alluvions anciennes de haute terrasse (Fx) ; 7. Cailloutis villafranchiens (Fv) ; 8. Colluvions (C) ; 9. Limons et argiles lacustres (FL) ; 10. Sédiments volcano-détritiques (Vs) ; 11. Basalte de coulée (β) ; 12. Formations miocènes (m1-3) ; 13. Formations oligocènes (g2-3) ; 14. Formations éocènes (e1-6) ; 15. Site archéologique ; 16. Fleuve Hérault ; 17. Dénomination des couches géologiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-5.jpg
Fichier image/jpeg, 460k
Titre Fig. 6 – Geology of the third section and occupation of the middle valley of the Herault River during the Second Iron Age.Fig. 6 – Géologie du troisième tronçon et occupation de la moyenne vallée de l’Hérault durant le Second Âge du Fer.
Légende 1. Recent alluvium (Fz); 2. Alluvium of medium terrace (Fy); 3. Old alluvium of high terrace (Fx); 4. Villafranchian gravel (Fv); 5. Colluvium (C); 6. Scree (E); 7. Basalt (β); 8. Continental Pliocene (Pc); 9. Miocene formation (m1-3); 10. Eocene formation (e1-6); 11. Cretaceous formation (c2-7); 15. Archaeological site; 16. Herault River; 17. Appointment of geological layers.1. Alluvions récentes (Fz); 2. Alluvions de moyenne terrasse (Fy) ; 3. Alluvions anciennes de haute terrasse (Fx) ; 4. Cailloutis villafranchiens (Fv) ; 5. Colluvions (C) ; 6. Éboulis (E) ; 7. Basalte de coulée (β) ; 8. Pliocène continental (Pc) ; 9. Formations miocènes (m1-3) ; 10. Formations éocènes (e1-6) ; 11. Formations crétacées (c2-7) ; 15. Site archéologique ; 16. Fleuve Hérault ; 17. Dénomination des couches géologiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-6.jpg
Fichier image/jpeg, 248k
Titre Tab. 1 – Hydromorphometric data of the middle Herault Valley and the three river sections.Tab. 1 – Données hydromorphométriques de la moyenne vallée de l’Hérault et des trois tronçons fluviaux.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-7.jpg
Fichier image/jpeg, 352k
Titre Fig. 7 – Geomorphological map of the first section of the middle Herault Valley.Fig. 7 – Carte géomorphologique du premier tronçon de la moyenne vallée de l’Hérault.
Légende 1. Administrative fund; 2. Vineyard; 3. Forest, wood, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Sandy alluvium; 7. Alluvial deposit of sand and pebbles; 8. Cryoclastic deposit; 9. Alluvial deposit of sand, pebbles and gravel; 10. Limestone slope; 11. Strong erosion zone; 12. Moderate erosion zone; 13. Zone of alluvial deposition; 14. Spot height; 15. Archaeological site; 16. Mediaeval mills; 17. Palaeohydrography of the Herault River according to the Napoleonic cadastre.1. Fond administratif; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Alluvions sableuses ; 7. Alluvions de sables et galets ; 8. Dépôts cryoclastiques ; 9. Alluvions de sables, galets et cailloutis ; 10. Versant calcaire ; 11. Zone de forte érosion ; 12. Zone d’érosion modérée ; 13. Zone d’accumulation sédimentaire ; 14. Point coté (en m) ; 15. Site archéologique ; 16. Moulin médiéval ; 17 Paléohydrographie de l’Hérault d’après le Cadastre napoléonien.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-8.jpg
Fichier image/jpeg, 900k
Titre Fig. 8 – Geomorphological map of the second section of the middle Herault Valley.Fig. 8 – Carte géomorphologique du deuxième tronçon de la moyenne vallée de l’Hérault.
Légende 1. Administrative fund; 2. Vineyard; 3. Forest, wood, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Area at moderate risk of flooding 7. Sandy alluvium; 8. Alluvial deposit of sand and pebbles; 9. Gravel; 10. Alluvial deposits of sand, pebbles and gravel; 11. Limestone slope; 12. Strong erosion zone; 13. Moderate erosion zone; 14. Zone of alluvial deposition; 15. Spot height; 16. Bank; 17. Archaeological site; 18. Mediaeval mills; 19. Palaeochannel according to the Cassini’s map; 20. Palaeochannel prior to AD 70 at the Estagnola site; 21. Dating from historical maps.1. Fond administratif; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Zone à risque d’inondation modéré ; 7. Alluvions sableuses ; 8. Alluvions de sables et galets ; 9. Dépôts cryoclastiques ; 10. Alluvions de sables, galets et cailloutis ; 11. Versant calcaire ; 12. Zone de forte érosion ; 13. Zone d’érosion modérée ; 14. Zone d’accumulation sédimentaire ; 15. Point coté (en m) ; 16. Talus ; 17. Site archéologique ; 18. Moulin médiéval ; 19 Paléochenal d’après la carte de Cassini ; 20. Paléochenal antérieur à 70 apr. J.-C. sur le site de l’Estagnola ; 21. Datation d’après les cartes historiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-9.jpg
Fichier image/jpeg, 924k
Titre Fig. 9 – Geomorphological map of the third section of the middle Herault Valley.Fig. 9 – Carte géomorphologique du troisième tronçon de la moyenne vallée de l’Hérault.
Légende 1. Administrative fund; 2. Vineyard; 3. Forests, woods, scrubland; 4. Water; 5. Area at high risk of flooding; 6. Area at moderate risk of flooding; 7. Sandy alluvium; 8. Alluvial deposit of sand and pebbles; 9. Gravel; 10. Alluvial deposit of sand, pebbles and gravel; 11. Limestone slope; 12. Strong erosion zone; 13. Moderate erosion zone; 14. Zone of alluvial deposition; 15. Spot height; 16. Bank; 17. Archaeological site; 18. Mediaeval mills; 19. Palaeochannel according to the Cassini map; 20. Palaeochannel according to the Napoleonic cadastre; 21. Palaeochannel according to the General Staff map; 22. Dating from historical maps.1. Fond administratif ; 2. Vignoble ; 3. Forêts, bois, garrigues ; 4. Eau ; 5. Zone à risque d’inondation élevé ; 6. Zone à risque d’inondation modéré ; 7. Alluvions sableuses ; 8. Alluvions de sables et galets ; 9. Dépôts cryoclastiques ; 10. Alluvions de sables, galets et cailloutis ; 11. Versant calcaire ; 12. Zone de forte érosion ; 13. Zone d’érosion modérée ; 14. Zone d’accumulation sédimentaire ; 15. Point coté (en m) ; 16. Talus ; 17. Site archéologique ; 18. Moulin médiéval ; 19. Paléochenal d’après la carte de Cassini ; 20. Paléochenal d’après le cadastre napoléonien ; 21. Paléochenal d’après la carte d’État-Major ; 22. Datation d’après les cartes historiques.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-10.jpg
Fichier image/jpeg, 2,0M
Titre Fig. 10 – Heterogeneous evolutionary of hydromorphological process.Fig. 10 – Processus d'évolution hydromorphologique hétérogène.
URL http://journals.openedition.org/geomorphologie/docannexe/image/15510/img-11.jpg
Fichier image/jpeg, 306k
Haut de page

Pour citer cet article

Référence papier

Ambrine Bouchène et Benoît Devillers, « Geoarchaeology of the middle Herault valley (southern France) since the Bronze Age: cartographic approach », Géomorphologie : relief, processus, environnement, vol. 27 - n° 2 | 2021, 159-170.

Référence électronique

Ambrine Bouchène et Benoît Devillers, « Geoarchaeology of the middle Herault valley (southern France) since the Bronze Age: cartographic approach », Géomorphologie : relief, processus, environnement [En ligne], vol. 27 - n° 2 | 2021, mis en ligne le 19 mai 2021, consulté le 28 juillet 2021. URL : http://journals.openedition.org/geomorphologie/15510 ; DOI : https://doi.org/10.4000/geomorphologie.15510

Haut de page

Auteurs

Ambrine Bouchène

Laboratoire d’Archéologie des Sociétés Méditerranéennes (ASM), UMR 5140 CNRS, Université Paul Valéry - Montpellier 3, route de Mende, 34 199 Montpellier, France

Benoît Devillers

Laboratoire d’Archéologie des Sociétés Méditerranéennes (ASM), UMR 5140 CNRS, Université Paul Valéry - Montpellier 3, route de Mende, 34 199 Montpellier, France

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • OpenEdition Journals
Search OpenEdition Search

You will be redirected to OpenEdition Search