Navigation – Plan du site

Settlement potential and constraints on the lower Médoc coastline: results of the Litaq project and considerations on coastal palaeo‑risks in the protohistoric periods

Potentialités et contraintes d’occupation du littoral du Bas-Médoc : bilan du projet Litaq et réflexions sur les paléo-risques durant les périodes protohistoriques
Frédéric Bertrand, Florence Verdin, Frédérique Eynaud, Gilles Arnaud-Fassetta, Pierre Stéphan, Stéphane Costa et Serge Suanez

Résumés

La question de l’adaptation des systèmes territoriaux au changement climatique en cours conduit à s’interroger sur les capacités des populations anciennes à faire face, à « rebondir » ou à se transformer dans le contexte des changements environnementaux passés. Les résultats chronostratigraphiques et archéologiques du projet LITAQ permettent de mieux connaître les modifications d’un système littoral aujourd’hui de « bord de mer » dont l’occupation intense, dès le Néolithique, est liée à la valorisation de potentialités spécifiques (sel, pâturage), inféodées à un système littoral « d’embouchure » fossilisé sous le massif dunaire actuel. Une des problématiques que soulèvent ces résultats est celle du repli de l’activité salicole au cours de toute la période du Bronze, entre une période de croissance de cette activité au Néolithique (dont on était loin d’avoir pris toute la mesure) et le premier Âge du Fer durant lequel la production du sel apparaît comme la raison principale de la fréquentation et de l’aménagement des marais littoraux. Or, l’écart chronologique d’environ treize siècles entre le début du Bronze ancien (~2200 av. J.-C.) et le Bronze final (~900 av. J.-C.) empêche de faire des fluctuations climatiques autour du premier millénaire av. J.-C. le facteur de contrôle direct et unique des changements d’occupation de la péninsule médocaine. La complexité des rythmes qui accompagnent les changements d’occupation du sol au cours de cette période et de la période suivante de l’Âge du Fer invitent à s’interroger sur les facteurs de vulnérabilité territoriale liés aux changements de configuration hydrogéomorphologique du littoral et aux transformations des composantes du milieu naturel impliquées dans le processus de production du sel. Les modalités de déploiement spatial de cette activité, telles que déduites de l’analyse des vestiges inventoriés, dans des milieux à échange tidal restreint (estuaire barré) témoignent de l’adaptabilité des populations protohistoriques du littoral médocain face à une perturbation lente et progressive du système littoral ainsi que de la résilience passée, dans son sens systémique, d’un territoire aujourd’hui excentré.

Haut de page

Texte intégral

1 - Introduction

1Coastal dune and estuarine system are widely spread along the European Atlantic coast from the Andalusian coast (Ménanteau, 1979) to the North-West of the Jutland peninsula (Clemmensen et al., 2001), via the coasts of the British Isles (Pye & Allen, 2000). Case-studies from the coastal zones of the North Sea (Van de Noort, 2004) of the English (Rippon, 2002) and the French Channel (Meurisse-Fort et al., 2009; Lespez et al., 2010) and of the Bay of Biscay (Roussot-Larroque, 1994; Mellallieu et al., 2000) illustrate the high potential of geoarchaeology and landscape archaeology of these coastal environments for palaeogeographical reconstructions over the past 7000 years.

2Among an abundant literature and contrary to the geoarchaeological studies conducted around the Mediterranean area (Morhange & Marriner, 2009), very few geomorphological, stratigraphic and geoarchaeological data were analyzed for arguing and discussing coastal risk in the past. Similarly, it seems that the exploration of the link between coastal dynamics and human societies along the European Atlantic coast in terms of coastal risk was of little interest for coastal palaeoenvironmental reconstruction. Only the southern part of the European Atlantic facade has been truly investigated in terms of palaeorisk in relation with the historical records of the famous 1755 Lisbon earthquake and the identification of palaeotsunami deposits in the region of Algarve, Southern Portugal (Font et al., 2013). Thus the aim of this paper is to demonstrate the relevance of combining a geoarchaeological approach and a risk approach to better characterize human societies and landscape evolution in European Atlantic sub-region.

2 - Source datas, methods and concepts

2.1 - Geographical and chronostratigraphic settings

3Recent research as part of the LITAQ project has targeted an area between Soulac-sur-Mer and Grayan-et-l’Hôpital (fig. 1) in order to establish a detailed chronostratigraphy supported by multiple dates. The results provide insight into the changes that occurred in what is now a coastal environment. They cover all the periods of human settlement in this unusual milieu between the end of the Pleistocene and late Antiquity when the present-day dune massif formed.

4The earliest sedimentary formations (Pleistocene) crop out south of the study area between the headland of La Négade and L’Amélie. The strata tilts northward and most of the outcrops on the beaches of L’Amélie and La Glaneuse date from the Holocene (fig. 2A). The lowermost strata are the “Argiles du Gurp” that can be attributed to a brine pond or estuary channel type of depositional setting. The upper part of this formation corresponds to the gradual decline of marine influences and the transition from a tidal environment to a freshwater pond (Bosq et al., 2019) in an interglacial climatic context. It was in this environment that the elephant Paleoloxodon antiquus lived, numerous remains of which have been reported since the nineteenth century. The discovery by beachcombers during the course of the program of isolated molar, tusk and vertebra provided an opportunity to revise the dating of the clay formations from which they came from (Beauval et al., 2019). New IR-RF and ESR dates for the formations in the sector suggest the “Argiles du Gurp” date to the Holsteinian or MIS 11 (424‑374 ka). However, certain inconsistent datings in the stratigraphic series rule out any firm and final conclusions as yet.

5At the end of this interglacial period (MIS 11-9-7 or MIS 5), the “Sables du Gurp” (or “Sables de l’Amélie”) progressively filled the marshland in a periglacial, low sea-level, estuarine-deltaic setting. The base of the formation displays convolutions and cracks that are attributed to cryoturbation. Seismic phenomena might also have come into play and caused such deformations by sand liquefaction.

6Further north, at L’Amélie, the clays are overlain by the “Grès de l’Amélie”, a formation of indurated clayey sands (fig. 3A). These are overlain by bluish clays of the same nature and very similar depositional facies. In addition to ligneous debris, they include archaeological material and mollusc shells in life position that date the L’Amélie formations to the late Holocene (Stephan et al., 2019). These initial settlements of the middle and recent Neolithic were located on both sides of a tidal channel occupying the centre of the beach and that began to fill with sediment in the fifth millennium BC. This tidal channel evolved into an estuarine marshland protected by a line of dunes and underwent two phases of sedimentation: one between 3000 and 1500 BC (recent Neolithic-Bronze Age) and one between 800 BC (early Iron Age) and AD 300 (late Empire). These two periods of marshland generation were separated by an erosional phase between 1500 and 1000 BC (middle and late Bronze Age). From AD 300 to 700, the marshland was gradually covered by aeolian deposits that formed the base of the present-day dune system (fig. 3B).

7On the La Lède du Gurp site, just south of the local authority boundaries of Soulac-sur-Mer and Grayan-et-l’Hôpital (fig. 4A), a residual butte is currently visible in the intertidal zone (Faye et al., 2019). It is the last remaining part of the archaeological site excavated in the 1990s. New sedimentological investigations as part of the LITAQ project (and the subsequent FAST LITAQ project between 2016 and 2018) have revealed an exceptionally thick series of peat to clay-peat deposits of the last 10,000 years that may be of karstic to thermokarstic origin (Faye et al., 2019) (fig. 4B). This material also exhibits a very rich collection of archaeological artefacts indicative of still discreet occupation in the Mesolithic and early Neolithic (8th-6th millennium BC) and that grew denser from the middle Neolithic until the middle Bronze Age (5th-2nd millennium BC) (Verdin et al., 2018; Verdin et al., 2019).

Fig. 1: Location maps of the study site.

Fig. 1: Location maps of the study site.

A/ Location of studied sites along the west coast of the Médoc peninsula (Sources: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, Swisstopo and the GIS User Community). B/ Detail of the historical map “Carte du premier quarré de la Generalle de Médoc et d’une partie de la Guienne et de la Saintonge”, surveyed and drawn by C. Masse in 1707 (Source: Institut National de l’Information Géographique et Forestière).

Fig. 2: Location map of archaeological sites and sedimentary outcrops along North-Amélie (A) and Google Earth vertical image providing a map‑like view of local coastal features (B).

Fig. 2: Location map of archaeological sites and sedimentary outcrops along North-Amélie (A) and Google Earth vertical image providing a map‑like view of local coastal features (B).

Description of units U1 to U5 (according to Stephan et al., 2019). Unit 1: base of the sedimentary sequence. Unit 1a/ 2 m-thick compact dark-green clay attributed to the “Argiles du Gurp” Pleistocene formation. Unit 1b/ Gravelly sandstone (colluvial material) locally referred to as “Grès de l’Amélie”. Unit 2: first stage of the Holocene estuarine sediment infilling of L’Amélie and La Glaneuse beaches. Unit 2a/ Medium to coarse sand deposit, dated to ca. 7290 cal. a BP and interpreted as the gradual infilling of a large tidal channel named the ‘Soulac channel’. Unit 2b/ Silty sand layer containing rootlets dated to 7420-7180 cal. a BP (early Neolithic). Unit 3: intertidal estuarine sediments interpreted as a first generation of estuarine deposits (late Neolithic to the early Iron Age). Unit 3a/ Sandy-silt at the base of the unit to a clayey-silt formed in an intertidal mudflat environment behind a progradational dune barrier. Unit 3b/ Dark organic-rich clayey-silt to silty sand corresponding to a late Neolithic high marsh environment. Unit 3c/ Humate-impregnated black sandstone (‘alios’). Unit 3d/ Thin layer of reworked lithoclasts (‘alios gravels’) with similar sedimentological characteristics to the unit 3c. Unit 4: clayey-silt deposit interpreted as a second generation of estuarine mudflats (ca. 2700 to 1250 cal. a BP). Unit 5: organic-rich clayey-silt, interpreted as a salt marsh deposit (ca. 2500 to 1070 cal. a BP). This unit was buried by aeolian sand supplies.

Fig. 3: Coastal morphology and features along North-Amélie beach.

Fig. 3: Coastal morphology and features along North-Amélie beach.

A/ Longitudinal view of the beach profile (north oriented) showing the Pleistocene sand terraces (unit 1b, locally referred to as “Grès de l’Amélie”) that dominate a thick compact dark-green clay attributed to an estuarine mudflat (unit 1a). A significant retreat of the overlooking dune front escarpment made possible the description of 28 sections (here between section 22 et 25) and the subdivision of the dune sequence into five aeolian sands units, separated by three layers of palaeosols or organic-rich sediment (dark lines) (photo: Y. Ceinturet, April 9, 20). B/ Cross-sectional view of the upper foreshore of AML-N-001 where the sandstone sands of the Amélie (unit 1b) are exposed at the foot of the dune. Oval archaeological structure (ST1) in the process of clearing its clay-sand filling horizon (unit 6) (photo: M. Seutin, March 22, 2015).

Fig. 4: Residual mound of La Lède du Gurp detached from the dune cliff after the complete uncovering of the underlying coastal marsh deposits by long-term erosion.

Fig. 4: Residual mound of La Lède du Gurp detached from the dune cliff after the complete uncovering of the underlying coastal marsh deposits by long-term erosion.

A.  Aerial oblique back view northwest oriented showing the remnants of coastal marsh deposits after the recurrent winter storms of 2013-2014 (Photo: F. Didierjean, March 17, 2014). B. Front view towards the inland showing the Mid Holocene sedimentary filling sequence (from about 10 to 3 ka BP) of a circular depression partially connected to a former arm of the Gironde estuary. The background dune scarp exhibits thick aeolian sand and fossilized palaeosols deposits with very little evidence testifying for alternating periods (from 1275-1260 BC to the present-day) of aeolian activity and dune stability. Trace of resettlement after major environmental changes at the turn of the first millennium are rare (Photo: F. Bertrand, June 16, 2014).

2.2 - A strong correlation between human occupations and environmental changes

8Archaeological investigations have also made it possible to go beyond the chronological definition of the main phases of occupation by linking them to environmental changes and explaining the intense use of these wetlands through the exploitation of specific resources. The periods when archaeological evidence is most numerous correspond to the two phases of sedimentation of the dune barrier marshes, the first phase extending from the Middle Neolithic to the Middle Bronze Age (ca. 5000/1200 BC), the second from the Iron Age to Roman Antiquity (ca. 800 BC/AD 200).

9The exploitation of sea salt appeared to be the predominant activity both in the Neolithic and in the early Iron Age. As early as the 4th millennium BC, many pits containing wooden structures were used as water catchment wells, settling tanks or brine storage tanks (Verdin et al., 2019, figs. 7-10, 14-16). The brine was then evaporated or heated to reduce it to salt breads. These discoveries are part of a context of salt production well known since the Neolithic period throughout Europe (Cassen & Weller, 2013). On the Atlantic coast, the presence of many fragments of salt moulds discarded in the Neolithic enclosures around the Marais Poitevin shows the importance of the production of this resource within the framework of social competition between the communities (Ard & Weller, 2012). Neolithic capture and storage structures located as close as possible to the source of supply have so far been rarely identified because they are often sealed under thick sediment fillings. The hypothesis of the use of Lower Médoc wooden and wattle constructions for salt production is supported by the discovery of similar remains in Iron Age or Roman period salt workers’ workshops in estuarine environments, either in France (Weller & Desfossés, 2002) or in the United Kingdom (Cranstone et al., 2018).

10The second phase of intensive exploitation of the salt resource corresponds to the Early Iron Age (700-400 BC). Techniques change and brine is now heated in kilns whose destruction leaves a lot of briquetage debris (“augets”, pillars and others clay pieces) (Verdin et al., 2019, figs. 19-20). The blooming of these workshops in the Iron Age is well documented along the Atlantic coast, especially in ecosystems similar to those of the Lower Médoc, as ancient marshes of Rochefort, Brouage, or the Marais Poitevin (Maguer et al., 2011). The landscape is also subject to development, as shown by the existence of a bridge or pontoon crossing a palaeochannel in the Amélie (Verdin et al., 2019, figs. 17-18).

11Between these two periods, the Bronze Age constitutes a break in the process since salt production seems to have been interrupted. Sedimentary formations remain difficult to perceive because of the scarcity of archaeological evidence. However, some deposits of bronze axes and ceramic remains testify to a continuous use of the marshes in the Early and Middle Bronze Age (2300-1200 BC), while the remains of the Late Bronze Age are non-existent (1200-800 BC). This raises the question of the impact that environmental changes could have on human activities. The same applies to the second Iron Age (400-50 BC), a period when all archaeological indicators disappear.

12The Roman period (50 BC-AD 200) again appears as a phase of intensive use and frequentation of the marshes. Salt is probably produced using other techniques that have not left any traces (salt marshes?; Daire, 2003). The presence of oyster deposits in several palaeochannels suggests that they were collected for food. These molluscs were highly appreciated and widely consumed at all sites during this period (Bardot-Cambon, 2013).

2.3 - A framework to provide a template suitable for palaeo-risk analysis

13This multidisciplinary research finally provides the possibility of testing the direct correlation between the pattern of human settlement and palaeoenvironmental evolution. While human pressure on these wetlands was almost constant, phases of change in sedimentary dynamics and chronological gaps in the archaeological evidence raise questions about human and biophysical conditions (dynamics factors) that determined the sensitivity of the coastal coupled human–environment system to any set of exposures (Turner et al., 2003) (fig. 5).

14In that context sensitivity refers to the degree to which the coastal system undergoes changes due to natural forces – tides, waves, floods and storms – following human interference whereas exposure is the extent to which the system is physically in harm’s way by processes operating at different spatiotemporal scales: in place (e.g. coastal erosion), beyond place (e.g. atmospheric and/or oceanic circulation) and cross-scale (e.g. climate worsening). The two concepts of sensitivity and exposure have been widely recognized and used as key components of risk analysis in both the vulnerability and resilience literatures. While vulnerability and resilience researches continue to appear as complementary and sometimes conflicting concepts (Miller et al., 2010), the concept of adaptive capacity that describes in simple terms, the ability to adapt, has begun to receive more attention in both literatures (Carpenter & Brock, 2008; Pahl-Wostl, 2009; Engle, 2011). Because of its unique position as being shaped by human actions and as influencing both the biophysical and social elements of a system, adaptive capacity is considered critical for reducing vulnerability (Eakin & Luers, 2006). Figure 5 illustrates how adaptive capacity plays a critical role in determining vulnerability by moderating exposure and sensitivity (Yohe & Tol, 2002; Adger et al., 2007).

15This work takes a fresh critical look, then, at the dynamics of the use of space and exploitation of natural resources in the northern Médoc through environmental change in the past focusing on assessing adaptive capacity and resilience within one particular coastal system. This conclusive and prospective paper marks out a path for future research on the Aquitaine coastline and the challenges facing it.

Fig. 5: Vulnerability framework with emphasis placed on adaptive capacity and its role in influencing vulnerability through modulating exposure and sensitivity.

Fig. 5: Vulnerability framework with emphasis placed on adaptive capacity and its role in influencing vulnerability through modulating exposure and sensitivity.

Figure adapted from Turner et al. (2003) and Engle (2011). The components of vulnerability are linked to factors external to the studied system and interact at various spatial and temporal scales.

3 - The natural risk-based approach
for providing complementary information to the
palaeoenvironmental investigation methods

3.1  Attractiveness outweighing environmental constraints

16The Médoc peninsula is peripheral to the present-day settlement centres of the Aquitaine Basin and to the major national and international movements of people and goods. This has not always been the case and the discovery – from the early 1960s (Roussot-Larroque, 1994) until recent finds as part of the LITAQ project (Verdin, 2015; Verdin et al., 2019) – of a rich archaeological heritage attests to recurring if not uninterrupted settlement from the Mesolithic (La Lède du Gurp site) until late Antiquity (L’Amélie nord site). One of the noteworthy contributions of the LITAQ project clarified the chronology of settlement of the peninsula in conjunction with the traces of salt-producing activity already glimpsed in the L’Amélie beach sector at Soulac-sur-Mer (Boudet et al., 1990) but undated (Verdin, 2015). The dating of archaeological and sedimentary material yielded by erosion of the Médoc coastal dunes (fig. 2B), both in the intertidal zone – where Neolithic levels are directly exposed (northern half of La Glaneuse beach) – and on the cliff face – where settlement levels appear both perched on Pleistocene sand terraces and fossilized beneath modern dunes (southern half of La Glaneuse beach, residual butte of La Lède du Gurp) – now reveals the middle to recent Neolithic (fifth and fourth millennia) to have been one of the protohistorical periods when settlements were most populous.

17Accordingly, one of the issues raised by the LITAQ project findings is that of the decline of salt production throughout the Bronze Age, between a period of growth of this activity in the Neolithic, which we were far from having fully gauged, and the early Iron Age when salt production was plainly the main reason for the coastal marshes being frequented and developed (pontoon bridge). The main point of our interrogation concerns the explanation of the chronological gap between, on the one hand, the disappearance of traces of salt production – with its repercussion on traces of other production activities – from the early Bronze Age and, on the other hand, the climatic worsening of the late Bronze Age – characterized by climatic changes towards moister, more Atlantic conditions with lower summer temperatures and less severe winters – that affected all Europe and led to all Médoc sites to being abandoned, including that of La Lède du Gurp which had been occupied since Mesolithic times (Roussot-Larroque, 2007; Verdin et al., 2018).

18The chronological gap of some thirteen centuries between the early and late Bronze Ages strictly prevents us from interpreting climatic fluctuations around the first millennium as the main factor controlling changes in settlements of the Médoc peninsula. Therefore, the role of geoarchaeologists is to elucidate the natural forcings that allow a contextualization of the archaeological sites at spatial and temporal scales (Morhange et al., 2014). This prompts us to analyse the relations between societies and climate through the double filter of behaviour with respect to environmental constraints and methods of resource exploitation (Allinne, 2006; Arnaud-Fassetta & Carcaud, 2015). This geoarchaeological approach will lead us to analyse environmental modifications in terms of natural hazards and then to identify changes induced in the Médoc coastal environment by regional climate changes and the rise in relative sea level in the Bay of Biscay.

19The density and spatial continuity of traces of ancient settlements along the Médoc coastline, especially between the beach at La Glaneuse (to the north) and the beach at La Lède de Gurp (to the south) (fig. 4B) are evidence, if not of uninterrupted polarization of activities and populations, at least of the attractiveness of the northern part of the Aquitaine coastline in protohistoric times. However, if the long-term attractiveness of an area is linked to its ability to draw and hold mobile production factors and/or populations (Poirot & Girardin, 2010), the attractiveness of the peninsula cannot be gauged simply from it having been settled in early times, as far back as the Neolithic, because of the opportunities for salt production offered by the coastal environment of the Gironde and its central position along Europe’s Atlantic coast which was suitable for exchanges between regions. The succession of relatively sustained economic activity – early to middle Bronze Age, early Iron Age, Antiquity – and less dynamic periods – late Iron Age – is now well established (Verdin et al., 2015) to ponder the modes of decline and then resumption of salt production between these various periods. In addition to reconstructing the palaeoenvironmental evolution of the northern coast of the Médoc, it is important to investigate the scale of coastal sites in order to precize how ancien hazards and local populations did impact upon the coastal area (Morhange & Marriner, 2009).

3.2 - Transposition of the hazard‑vunerability‑risk‑adaptation scheme

20As the know-how of salt workers has always rested on the optimal exploitation of natural (climatic, edaphic and maritime) conditions, here along France’s Atlantic coastline (Hocquet, 1994; Cassen et al., 2006; Gouletquer & Weller, 2010) as in other parts of the world (Geslin, 1999), it seems opportune to focus first on the variability of natural factors that might have boosted or restricted the activity. It shall thus be considered that the opportunities provided by the Médoc coastal environment may have been transformed into constraints or even threats to salt production.

21Investigating palaeoenvironmental changes through the prism of natural risks involves looking through the archaeological and sedimentary records for traces of high-intensity events that induce dangers and changes in technical and territorial contexts as it has been possible to reconstruct them under the LITAQ programme (Stéphan et al., 2019, figs. 15-16). In this way, the discovery in marsh levels unearthed in the northern part of the beach at L’Amélie of nodules (alios) remobilized by high-energy waves or currents points to locally more intense agitation of the environment in conjunction with the general climatic change of the Bronze Age and the intensification of potentially high-impact morphogenic storms. However, these depositional records are insufficient to define and describe a situation of risk. Risk necessarily implies the presence of populations and activities that are either directly – through the destructive effects of extreme events – or indirectly – through changes in the regional bio-physical context – exposed to natural phenomena. Accordingly, in order to avoid a vision of risk situations that is overly centred on hazard, dictated by the absence of data about the representation of risks by the populations involved, it seems expedient to address these risk situations through the double prism of palaeo-hazards and palaeo-vulnerability of the territory (Arnaud-Fassetta & Landuré, 1997, 2003, 2015). Territorial palaeo-vulnerability is to be understood as a transposition to past times of the notion of territorial vulnerability, understood for all the Médoc coastal sites that are sensitive to variations in environmental operating parameters and whose reactivity to these agents may have disrupted the salt production system (D’Ercole & Metzger, 2009). The scientific challenge is to free ourselves from the reductive opposition between what appertains to nature (hazard) and what appertains to societies (vulnerability), and to allow ourselves to think about the interaction between the two terms and, in doing so, to acknowledge that protohistoric populations had a capacity to adapt that is often denied to them on the pretext that the proof of their adaptability, or adaptive capacity, often rests on an a posteriori interpretation of the synchronism between the environmental change and the social change in the broad sense (Antoine & Carozza, 2010).

22But adaptive capacity is not only difficult to gauge because of its latent nature. Related to the latency issue, another limitation involves adaptive assessment, inspired either by the vulnerability framework or the resilience framework (Engle, 2011). Assessments of adaptive capacity in the vulnerability framework are often performed at a single spatial scale and/or are merely ‘snapshots’ in time (Vincent, 2007). Furthermore, descriptive case studies inspired by the resilience framework present methodological difficulties for measuring and characterizing adaptive capacity (Ostrom et al., 2007; Carpenter & Brock, 2008). In accordance with what can be expected from an earth-science approach (Morhange & Marriner, 2009), the LITAQ project permitted us (1) to demonstrate that diagnostic litho-and bio-stratigraphies, consistent with meteorological hazards and man-modified environments, are recorded in the geological record and (2) to date the transition from natural to anthropogenic environnements in the northern part of the Médoc peninsula. Nevertheless, the answer to the question “How did ancient hazards and local populations impact upon coastal zones?” (Morhange & Marriner, 2009) remains to be asked in light of a social sciences approach. Considering that the role of geoarchaeologists is to elucidate the natural forcings that allow a dynamic contextualization of the archaeological sites at different spatial and temporal scales (Morhange et al., 2014), we will favour an approach, pulling initially from the resilience framework, that address the dynamic nature of coupled human/environnement systems (Turner et al., 2003; Engle, 2011). Thus, beyond its proposal for a synthesis this article is an attempt to better understanding the coastal archaeological records by measuring/characterizing adaptive capacity. It will involve looking at multiple periods in time leading up, during, and after a climatic event to evaluate whether systems prepared for and/or adapted to the stress.

3.3 - Societal responses between adaptation and collapse

23The chronological refinement of settlement dynamics of the Médoc peninsula during the Bronze Age (2200-800 BC) sheds new light on the much debated question of the adaptability of protohistoric populations to climate change. Although the break in sedimentary and archaeological records of the end of the first millennium, confirmed by the LITAQ project, seems to indicate populations were unable to adapt in situ to a changing environment brought about by climatic constraints (cooler and wetter climate) or the choice of more suitable habitat, changes in land use in the course of the preceding early Bronze Age and above all the middle Bronze Age (the development of agriculture to complement or replace salt production, the practice of depositing objects in water and wetlands) are indicative on the contrary of a genuine capacity to implement varied actions to attenuate their vulnerability in a context of transformation of the natural environment that may have compelled the abandonment of Neolithic sites which were buried by dunes between 750 BC and AD 300 (Stéphan et al., 2019, fig. 16B). Do the techniques and practices associated with these new land uses not bear witness to the capacity of Médoc populations to innovate under pressure from environmental change and so attest to great resilience in the face of disruptive meteorological events?

24In any event, the complexity of the rhythms that accompany the changes in settlement of the Médoc peninsula between the two phases of growth of salt production in the Neolithic and the early Iron Age are evidence of graduated responses over time to environmental change, with societal responses to environmental change having fluctuated in all likelihood, depending on the variability of risk factors and their interactions, between the two extremes of adaptation (Waters, 1992) and collapse (Diamond, 2005). Among the factors liable to prompt adaptation or cause collapse of populations (Pelletier, 2011), changes in the landscape of the peninsula as it has been possible to reconstruct them (Stephan et al., 2019, figs. 15-16) seem best able to illuminate the development and abandonment of salt production around the sites recorded. Without excluding the early appearance of technical innovation before the Iron Age during which the heating of salt water or brine in recipients over a hearth (“ignigenous” or fire-source salt) is well established, such palaeogeographical changes plainly entailed changes in the state of natural resources – water, air, land, vegetation – jointly involved in the salt production system and in the relations between the coastal populations and the area they settled (Hilhorst & Bankoff, 2004).

4 - Dynamics in the use of médoc space and variations in the state of its natural resources

4.1 - the climatic downturn of the late Bronze Age and its consequences

25Described as a phase of climatic downturn even before the scale and chronology of salt production around the Médoc peninsula was established, the late Bronze Age is characterized both by a shift to wetter conditions due to the deterioration of climatic conditions – cooler and wetter (prejudicial to farming and fishing developed over several centuries) – and by morphological changes which likely impacted salt production by significantly reducing evaporation and a change from tidal-dominated to fluvial and marine processes. Beyond the greater number of potentially destructive storm events, attested to in particular by deposits with concentrations of nodules (alios) up to ten or so centimetres in size, it may be thought that the conditions for exploitation of Pleistocene clay terraces (unit 1a) and estuary mudflats from the first filling phase (unit 3b) were largely altered further to the opening of breeches through the line of dunes and the widening of two tidal inlets under the combined action of waves, tides and the river (Stéphan et al., 2019, fig. 15D).

26The insularization of the Médoc peninsula associated, in our hypothesis, with increased flow of the Gironde estuary and the relative strengthening of fluvial actions, in all likelihood entailed a change in the hydrodynamic regime and the mode of exposure of the coastal marshes. Although, as analysis of sedimentary records has confirmed, the periods of economic dynamism correspond to the main phases of sedimentation of the marshes which then form environments amenable to the development of varied human activities, we are entitled to infer that the break in the sedimentary and archaeological records of the end of the first millennium resulted from the advent of conditions of fluvial-marine agitation (semi-agitated mode) that may have eroded and submerged early and middle Bronze Age sites that had developed at relatively low elevations during the earlier sheltered period (calm mode). A radical change in the hydrodynamic regime that can be attributed to the transformation of the mouth of the Gironde from a single-arm into a multiple-arm outlet of the “estuarine-delta” type as in oceanographic terminology (Mihajlov in Carré, 2001), might explain the regression of the Neolithic back-barrier marshes and the associated activities, even before the formation between 2575-1420 BP of the narrow elongate sandbar parallel to the estuary known as the cordon de Richard (Pontee et al., 1998) that led to progressive closure of the ancient outlets from these marshes into the estuary (Stéphan et al., 2019, figs. 3-4). According to the proposed pattern of change, the challenge to the socio-economic system based in the early and middle Bronze Age on the exploitation of environments in unstable equilibrium with climatic conditions that were probably already changing, would seem to stem more from a progressive change in the physiographic setting of the estuary marshes than from the occurrence of extreme climatic events with highly destructive potential (catastrophic events).

4.2 - Extrapolation to the beginning and the end of Protohistory

27While it is possible, within the limits of the accessible sedimentary and archaeological sources, to identify a break between Médoc societies and their coastal environment in the late Bronze Age, the scanter sedimentary records and archaeological evidence of the periods encompassing the Bronze Age, make it more difficult to interpret the environmental changes that occurred before, on the boundary of the Mesolithic and early Neolithic, and after, around the late Iron Age. The large chronological gap in settlement of the site of La Lède du Gurp in the late Bronze Age is correlated, as at the site of L’Amélie, with a phase of climatic downturn that brought about far-reaching changes in human societies (Verdin et al., 2019). But unlike at the L’Amélie site, where resettlement in the early Iron Age occurred in a very different sedimentary context from that of the previous period of abandonment, the resettlement of the site of La Lède au Gurp in the middle Neolithic (Chassean) and its intensification at the end of the fifth millennium (4400 and 4000 BC) took place in the same sedimentary context that this previous period, namely in clay-sand inputs with a high peat content, containing many plant remains, that are visually much like the Mesolithic strata (Verdin et al., 2018; Faye et al., 2019).

28The chronological gap of two centuries (4th-3rd centuries BC) which, at both the L’Amélie and La Lède du Gurp sites and similarly to many places in Gaul, separates the early Iron Age settlement (8th–5th century BC) from the very tenuous settlement of the late Iron Age (2nd–1st century BC; Verdin et al., 2019) does not coincide, unlike the gaps at the hinge point of the Meso-Neolithic and late Bronze Age, with a phase of climatic downturn. On the contrary, this phase of abandonment of human settlement occurred in a period of relative warming marked by soil formation in a reducing environment of sand levels (unit 6) that previously covered the southern edges of the arm of the Gironde (palaeosol 0 dated between 40 BC and AD 300; see Stéphan et al., 2019, fig. 16B) before being fossilized in turn by the northward shift of the first dunes (units 7 and 8, see Stéphan et al., 2019, fig. 16C). A determining constraint here seems therefore not to have been climate but the transformation of a part of the back-barrier marshland, until then with surface connections to the ocean, into wet back-dune depressions (“lèdes” in Gascon dialect), subjected to a groundwater regime that was far less amenable to the development of coastal activities. Pending palynological analyses to validate the probable hypothesis that these depressions were home to freshwater vegetation, it is difficult to specify how and when the setting began its transformation. It will simply be noted that similar movements of dune systems, from the end of the Iron Age on, and that entailed the transformation of salt marsh into fresh-water marsh, have been observed elsewhere around the Bay of Biscay (Visset & Bernard, 2006). The origin of these changes is subject to debate if it is considered that no significant fall in sea level – a factor many commentators have proposed to explain the change to fresh-water of lagoon marshes fed by streams (Cassen, 1987) – is recorded during the Holocene along the Atlantic coast of France (Stéphan & Goslin, 2014). The breaks in settlement of the Médoc peninsula in protohistoric times are therefore due to the processes of destabilization of the conditions of exploitation of the environment and the vulnerabilization of activities which are far from systematically determined by climate. While the concept of climatic downturn very broadly holds for the period at the turn of the first millennium during which the deterioration of climatic conditions may have exceeded populations’ capacities to adapt, it is more debateable for the Mesolithic/early Neolithic boundary, which is poorly identified in the sedimentary record, and it seems inappropriate for characterizing the changes to the environment at the dawn of the late Iron Age. More directly than changes in climate, it is the relation between morphogenesis and land use, by way of the state of natural resources that should be examined to clarify the forms of adaptation of protohistorical populations to environmental change and wider socio-eco-cultural changes as well.

4.3 - Salt production and its adaptation to environmental changes

29The identification of correlations between the abandonment of the coastal sites of L’Amélie and La Lède de Gurp and the episodes of global atmospheric cooling known as “climatic downturn” (the “downturn” at the end of the 1st millennium BC) or “rapid climatic events” (“the 8.2 ka BP event”) has rather overshadowed the co-occurrence between on the one hand the two major periods of development of salt production, in the late Neolithic and the early Iron Age, and on the other hand the phases of intense morphodynamic instability of the dune cordon. Although the profile across the tidal mouth of the “Soulac channel” (Allen et al., 1974; Pontee et al., 1988) between 5000 and 3000 BC has been identified from the elevation of the coarse marine sands marking the bottom of a vast tidal channel (unit 2a, see Stéphan et al., 2019, fig.7), its closure under the levelling action of oceanic swell, although it remains to be dated, probably marks the end of a phase of great instability of landforms, against which populations seem to have defended themselves by settling preferentially on the inner side of the inlet (see Stéphan et al., 2019, fig. 16B). The resettlement from the early Iron Age (8th-6th centuries BC) of back-barrier marshes formed in the lee of the sandbar, in marshland that was successively occupied and then abandoned during the Bronze Age (see above), fits into a context of instability that is well known throughout north-western Europe (Sorrel et al., 2012) where it is correlated with an episode of cooling and increased storminess around 1000-700 BP (Van Geel et al., 1996). This demonstrates that the signification of signals of hydroclimatic and hydrosedimentary changes in the stratigraphic sequences of the Médoc coast is far from unequivocal and that populations at certain times managed to adapt to alterations in the environment induced by climate change.

30The co-occurrence between the two main periods of development of salt production and the two marked phases of transformation of the coastal landscapes suggests we should take into account the geographical dimension of salt production by considering more closely how this space-consuming activity fitted in with settings that were being spatially re-shaped. The correspondence emphasized by F. Verdin et al. (2019) between periods of economic dynamism and the main phases of sedimentation of the back-barrier marshes is no chance matter. It underscores a deliberate choice of production sites providing not just protection against the action of swell but above all vast intertidal areas with possibilities of exploiting salt pans, as has been shown for the coastal bays of southern Armorica (Cassen et al., 2006) and for tropical mangrove estuaries (Bertrand, 1993; Geslin, 1999).

31The fact is that in the context of a wet and comparatively warm climate in the protohistoric period under consideration, salt production could not have been feasible without abundant resources of salt land. From this point of view, one of the most noteworthy contributions of the LITAQ project is most certainly to have enabled the dating of a series of pits lined with stakes, planks or wickerwork (beach at L’Amélie: ST1, ST4, AML-N-004), known until then to be abundant on the beaches of Nord-Médoc (Verdin et al., 2015) but not associated with any series of operations until now. The peaty backfill of the pits and their truncation by marine erosion (fig. 3A) suggests that it cannot be directly determined for what they were excavated for. But the proximity of many other types of remains (heating structures, areas for discarded items), linked with salt production, now shows that these structures were part of the salt production operations, taking advantage of the complementary features of the different landscape elements of the setting, in particular the salt pans developed on tidal salt marshes.

32These structures are reminiscent of the circular excavations interpreted in several geographical areas of western Europe such as washing and/or filtering pits (Cassen et al., 2006). They may be considered to be a technical answer to local constraints on the production of a commodity of high trade value representing a major economic challenge in the context of trade networks for rare goods over the long term (Carozza et al., 2009). The structures attest more generally to the capacity of adaptation and/or the opportunism of Médoc populations to adapt to the hydrogeomorphological changes induced by climate change through their understanding of the variations in the sediment budget which were sometimes favorable (accumulation) and sometimes unfavorable (excavation) to saline activity. When set in the chronostratigraphic and palaeoenvironmental frameworks of the Médoc peninsula, salt production seems to have been used in the Neolithic and early Iron Age as a form of adaptation to long-term morphological change in the Gironde Estuary rather than to hazardous events in the true (short-term) meaning of the term.

4.4 - The working of the salt production features in question

33The probable disappearance of many of the remains (ceramics, miscellaneous developments) associated with the essential stage of production consisting in obtaining brine, prevents any reconstruction of the precise spatial layout of the salt-makers’ workshops. The distribution of vegetation resources required to boil and crystallize the brine is very hypothetical.

34Judging from the varied nature of the sedimentary horizons on which the salt production structures rest, the exact functions -settling or collection- remains to be clarified. Their location seems comparatively independent with respect to the energy sources required to evaporate the natural brine (sea water) or artificial brine (from lixiviation of sand). The connections between the resettlement of the site at La Lède de Gurp (fig. 4A) in the early Iron Age (6th century BC), the making of fire-source salt and the reactivation of the peat bog that may have supplied fuel, do not apply to the site of L’Amélie-Nord. Indeed, the discarded briquetage may, as at La Lède de Gurp, have been transported from nearby production zones whereas the oven “in position” was found in peat not clay horizons. In a sedimentary production context that is less difficult to characterize in the Neolithic, pits and wickerwork observed in the middle part of the stratigraphic sequence of La Lède du Gurp and in the central part of the beach at L’Amélie (ST 4) were positioned on different types of bedrock: a peat-sand interface at La Lède du Gurp, and a clay-sand interface at L’Amélie implying in the latter case a supply of timber for fuel from nearby woods.

35The highly variable nature of the sedimentary underpinnings of the salt manufacturing sites in the narrow sense leads to the placing of this decisive phase of the salt production process within operational sequences that vary with site and period depending on the evolution of the environment and the relative arrangement of the various landscape features (tidal channel, intertidal flat, dry land wood stands) involved in production. Our understanding of the way the various developments associated with salt production worked depends closely on the reconstruction of the environment that we are able to propose. The pits with wickerwork or timber walls may have been the bottom of collection structures. But it must be ascertained whether the ground water was fresh – and so for drinking or washing sand particles – or salted – and so directly used for brine production. From what levels was the water drained off? From a higher dune level or from a level close to the present day one, and so flush with the marsh (schorre or slikke)? If these structures were open in wetlands, were they used seasonally when the saltmarsh dried or were they filled with brine in a mudflat context? These pits may also be interpreted as storage tanks for fresh or salt water, which raises exactly the same questions. Lastly, if some are similar to sand washing structures, it has to be imagined they were sited on the edges of marshes. All of these questions recall the decisive role of availability of water in salt production systems and underscore the value of being able to reconstruct the hydrological parameters of past fluvial systems as well as freshwater supply and marine influence (Baker, 2008). It will be observed here that one of the original features of the Médoc hydro-geosystem is that it benefited from a very early stage from the capacity for storage (“water tower” effect) and resupply of a complex dune system, the impermeable character of some of its “layers” (Gurp clays, perched nodular (alios) palaeosols) gave rise to, as can be seen from the current cliff retreat, seepage and resurgence, characteristic of the dune hygrosere (Blanchard & de Cacqueray, 2010). While the past dynamics of resurgence and seepage seem impossible to reconstruct precisely given the intermingling of the factors involved (stratigraphy, sedimentary record, landslips) and the disappearance of the perched layers that are impermeable to water there is every reason to believe that the various compartments of the dune hygrosere have multiplied the zones for drawing fresh water and rinsing sand, as well as for water consumption and human settlements.

5 - Conclusions

36Expressing both the climate change “experienced” by societies and the contribution of the long time scale to the construction of the regional economic structure, the dense and unbroken scatter of archaeological remains along the beach of L’Amélie or concentrated in the formerly hanging basin of La Lède du Gurp supplements the climatic data recorded by and in the exceptional stratigraphic sequences of these two sites. The renewed abundance of signs of settlements in the periods of the early/middle Bronze Age and the early Iron Age confirmed by a contrario reasoning the correlation between the deterioration in climatic conditions and human abandonment of the coastal environment over the intermediate period of the late Bronze Age. But the many material traces of change in the form of settlements of the area during the periods of intense morphodynamic instability attest also to the capacity of protohistoric populations to adapt to the effects induced by climate change. The risk-based approach to settlement dynamics of the Médoc coast in protohistoric times thus supplements the approach using analysis of sedimentary sequences, palaeoenvironmental data and archaeological remains emphasizing the varying degrees of sensitivity and resilience to change of natural environments and societies (fig. 5). The distinction between these two concepts seems essential to any transposition of the concept of vulnerability to past times. Indeed, the amplitude to which a coastal system or a system component (marsh) can experience losses as a consequence of disturbances (sensitivity) does not pre-judge the tendency of the system to return or not to an equilibrium position after a disturbance (resilience), or in other words its “dynamic” stability (Bertrand, 2008). Indeed, human can manage social-ecological interactions in a manner that maintains the system state (i.e. the more adaptive capacity within a system, the greater the likehood that the system will be resilient in face of climate stress) (Engle, 2011). Vulnerability to climate can be represented by a lower boundary, or critical threshold below which negative impacts might be tolerated or not, depending on the range of climate variability (fig. 6). Thus it seems that the Médoc coupled human/environmental system tolerated some large part of the range of climate variability in the Northern Hemisphere from the Latest Mesolithic to the Early and Middle Bronze Age normally experienced, i.e. when the difference with respect to the long-term average temperature (the temperature anomaly) was less than -0.2°C or 0.2°C. Within this range of variability, or coping range according to Hewitt and Burton (1971), climate variable, either primary or compound variable dependent on primary variables, might have varied from tolerable to beneficial. Following this scheme, the negative anomaly -0.2°C represents the upper boundary beyond which adaptation might have become a necessity considering the inacceptable levels of harm.

37Thus, the resettlement from the early Iron Age (8th-6th century BC) of back-barrier marshes formed in the shelter of the line of dunes, in a regional context of cooling and increased storminess, is evidence both of great sensitivity and great resilience of the Médoc coastal system unlike the late Bronze Age when populations’ capacity to adapt seems to have been insufficient to counter the constraints induced by a cooler and wetter climate. In the regional context of change between the Neolithic and Late Antiquity, the expansion of salt production enabled coastal populations to reduce their vulnerability by providing the means to operate, at certain times. Regardless, salt production thus comes across as a factor of medium and long-term resilience. Flexible adaptation strategies like salt production were successful in reducing vulnerability during the early Iron Age (8th-6th century BC) not long after the period of Bronze Age crisis. Conversely, the longer duration of the 8200 BP “event” compared to the Late Bronze period and the rapidly changing climate from the Late Mesolithic to the Early Neolithic compared to the climate instability of the Iron Age may have impeded the ability of older communities to cope with the consequences of climate deterioration or to take advantage of opportunities.

Fig. 6: Vulnerability, coping range and adaptation of coastal communities to climate variability and change along of the North Médoc peninsula over the last 12000 years.

Fig. 6: Vulnerability, coping range and adaptation of coastal communities to climate variability and change along of the North Médoc peninsula over the last 12000 years.

The figure, adapted from Willows et al. (2003), indicates the relationship between coping range, critical threshold, vulnerability and temperature considered as a key indicator of changes in the climate throughout the course of the Holocene. The coping range represents the tolerable climate (in both value and variability) that a system can endure at a given time given its vulnerability and resilience characteristics (Quenaut, 2014). The coping range boundaries lie about 0.2°C below the average value of reconstructions temperature proxies in the Northern hemisphere (Davis et al., 2003; Moberg et al., 2005).

Haut de page

Bibliographie

ADGER W.N., AGRAWALA S., MIRZA M.M.Q., CONDÉ C., O’BRIEN K., PUHLIN J., PULWARTHY R., SMIT B. & TAKAHASHI K., 2007 - Assessment of adaptation practices, options, constraints and capacity. In M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden & C.E. Hanson (eds.), Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the IPCC. Cambridge University Press, Cambridge (UK), 717-743.

ALLEN G.P., BOUCHET J.-M., CARBONEL P., CASTAING P., GAYET J., GONTHIER E., JOUANNEAU J.-M., KLINGEBIEL A., LATOUCHE C, LEGIGAN P., ORGERON C, PUJOS M., TESSON M. & VERNETTE G., 1974 - Environnements et processus sédimentaires sur le littoral Nord-Aquitain. Bulletin de l’Institut de Géologie du Bassin d’Aquitaine, 15, 3-183.

ALLINNE C., 2006 - L’évolution du climat à l’époque romaine en Méditerranée occidentale : aperçu historiographique et nouvelles approches. In E. Hermon (ed.), Vers une gestion intégrée de l’eau dans l’empire romain. Atlante Tematico di Topografia Antica Supplemento (Testo stampato), 16. L’Erma di Bretschneider, Roma, 89-99.

ANTOINE J.-M. & CAROZZA J.-M., 2011 - Aléas, vulnérabilités et adaptations des sociétés du passé. Sud-Ouest européen, 32, 1-3.

ARD V. & WELLER O., 2012 - La céramique. 2. Les vases de « type Champ-Durand » : témoins d’une exploitation du sel au Néolithique récent dans le Marais poitevin. In R. Joussaume (dir.), L’enceinte néolithique de Champ-Durand à Nieul-sur-l’Autise (Vendée). Mémoire - Société de Recherches Archéologiques de Chauvigny, XLIV. Association des Publications Chauvinoises, Chauvigny, 319-343.

ARNAUD-FASSETTA G. & CARCAUD N. (eds.), 2015 - French Geoarchaeology in the 21st Century. CNRS Editions, Paris, 620 p.

ARNAUD-FASSETTA G. & LANDURÉ C., 1997 - Occupation du sol et contraintes fluviales dans le delta du Rhône (France du Sud). In J. Burnouf, J.-P. Bravard & G. Chouquer (eds.), La dynamique des paysages protohistoriques, antiques, médiévaux et modernes. XVIIe rencontres internationales d’archéologie et d’histoire d’Antibes : actes des rencontres, 19-20-21 octobre 1996. Éditions APDCA, Sophia-Antipolis, 285-308.

ARNAUD-FASSETTA G. & LANDURÉ C., 2003 - Hydroclimatic hazards, vulnerability of societies and fluvial risk in the Rhône Delta (Mediterranean France) from the Greek period to the Early Middle Ages. In E. Fouache (ed.), The Mediterranean World Environment and History. IAG Working Group on Geo-archeology Symposium Proceedings: Environmental Dynamics and History in Mediterranean areas: Paris, Université de Paris-Sorbonne, 24-26 avril 2002. Elsevier, Paris, 51-76.

ARNAUD-FASSETTA G. & LANDURÉ C., 2015 - Fluvial risk in rural areas from the Greek period to the early Middle Ages: The case of the Rhône delta (France). In G. Arnaud-Fassetta & N. Carcaud (eds.), French Geoarchaeology in the 21st Century. CNRS Editions, Paris, 213-234.

BAKER V.R., 2008 - Paleoflood hydrology: Origin, progress, prospects. Geomorphology, 101, 1-13.

BARDOT-CAMBOT A., 2013 - Les coquillages marins en Gaule romaine. Approche socio-économique et socio-culturelle. BAR. International Series, 2481. Archaeopress, Oxford, 2013.

BEAUVAL C., MICHEL P., BOSQ M., BERTRAN P., LACRAMPE-CUYAUBÈRE F. & PASTY S., 2019 - A new straight-tusked elephant tooth from Soulac-sur-Mer (Gironde, France). Review of elephant discoveries in the Northern Médoc. Quaternaire, 30, 21-30.

BERTRAND F., 1993 - Contribution à l’étude de l’environnement et de la dynamique des mangroves de Guinée : données de terrain et apport de la télédétection. Etudes et Thèses. ORSTOM, Paris, 201 p.

BERTRAND F., 2008 - La vulnérabilité des milieux lagunaires : fondements, représentations spatiales et évaluations dans l’aire atlantique. Thèse d’Habilitation à Diriger des Recherches, Université Paris 1 Panthéon-Sorbonne, Paris, 351 p.

BLANCHARD F. & DE CACQUERAY M., 2010 - Premiers éléments typologiques et phytoécologiques sur l’hygrosère dunaire. Patience des rochers (Rumex rupestris) des falaises meubles à paléosols du littoral médocain (Gironde). Rapport Office National des Forêts - Conservatoire Botanique National Sud Atlantique, 66 p.

BOSQ M., BERTRAN P., BEAUVAL C., KREUTZER S., DUVAL M., BARTZ M., MERCIER N., SITZIA L. & STÉPHAN P., 2019 - Stratigraphy and chronology of Pleistocene coastal deposits in northern Aquitaine, France: a reinvestigation. Quaternaire, 30, 5-20.

BOUDET R., FRUGIER G. & MOREAU J., 1990 - Les sauneries de l’âge du Fer du littoral médocain (Gironde). Cahiers Médulliens, 14, 1-14.

CAROZZA L., MARCIGNY C. & TALON M., 2009 - Ordres et désordres de l’économie des sociétés durant l’âge du Bronze. In M. Bartelheim & H. Stäuble (eds.), The Economic Foundations of the European Bronze Age. Forschungen zur Archäometrie und Altertumswissenschaft, 4. M. Leidorf, Rahden, 23-65.

CARPENTER S.R. & BROCK W.A., 2008 - Adaptive capacity and traps. Ecology and Society, 13 (2): 40. http://www.ecologyandsociety.org/vol13/iss2/art40/

CARRÉ F., 2001 - XLVIIIe chronique océanographique. Les embouchures fluviales et leur classification. Norois, 188, 123-127.

CASSEN S., 1987 - Le Centre-Ouest de la France au IVème millénaire av. J.-C. BAR. International Series, 342. B.A.R., Oxford, 390 p.

CASSEN S. & WELLER O., 2013 - Idées et faits relatifs à la production des sels marins et terrestres en Europe, du VIe au IIIe millénaire. In J. Soares (ed.), Pre-history of Wetlands. Landscapes of Salt. Setúbal Arqueológica, 14. Museu de Arqueologia e Etnografia do Distrito de Setúbal & Assembleia Distrital de Setúbal, Setúbal, 255-304.

CASSEN S., DE LABRIFFE P.-A. & MÉNANTEAU L., 2006 - Le sel « chauffé » des baies marines en Armorique-sud durant les Ve et IVe millénaires av. J.-C. : à la recherche (ouest-européenne) de croyances et de faits techniques. In J.-C. Hocquet & J.-L. Sarrazin (dir.), Le sel de la baie : Histoire, archéologie, ethnologie des sels atlantiques. Presses universitaires de Rennes, Rennes, 33-54.

CLEMMENSEN L.B., ANDREASEN F., HEINEMEIER J. & MURRAY A. 2001 - Sedimentology, stratigraphy and landscape evolution of a Holocene coastal dune system, Lodbjerg, NW Jutland, Denmark. Sedimentology, 4 (1), 3-27.

CRANSTONE D., BRISBANE M., FIELDING A., MENANTEAU L. & HUESO KORTEKAAS K., 2018 - Production du sel et géohistoire des paysages salicoles atlantiques : Royaume-Uni et quelques exemples continentaux. In L. Ménanteau (dir.), Sels et salines de l’Europe atlantique. Presses universitaires de Rennes, Rennes, 22-61.

DAIRE M.-Y., 2003 - Le sel des Gaulois. Editions Errance, Paris, 152 p.

DAVIS B.A.S., BREWER S., STEVENSON A.C., GUIOT J. & Data Contributors, 2003 - The temperature of Europe during the Holocene reconstructed from pollen data. Quaternary Science Reviews, 22, 1701-1716.

D’ERCOLE R. & METZGER P., 2009 - La vulnérabilité territoriale : une nouvelle approche des risques en milieu urbain. Cybergeo: European Journal of Geography, Dossiers, document 447. https://journals.openedition.org/cybergeo/22022. DOI: 10.4000/cybergeo.22022

DIAMOND J., 2005 - Collapse: How Societies Choose to Fail or Succeed. Penguin Books, New York, 575 p.

EAKIN H. & LUERS A.L., 2006 - Assessing the vulnerability of social-environmental systems. Annual Review of Environment and Resources, 31, 365-394.

ENGLE N.L., 2011 - Adaptive capacity and its assessment. Global Environmental Change, 21, 647-656.

FAYE S., EYNAUD F., BOSQ M., LAMBERT C., VERDIN F., VEQUAUD P., LODYGA O., DERIENNIC H., LEBLEU P., BUJAN S., BILLY I., MARTIN B. & ROUSSOT-LARROQUE J., 2019 - Holocene palaeoenvironmental evolution of the Médoc peninsula (SW France): insights from the sedimentological study of the “Lède du Gurp” archaeological site. Quaternaire, 30, 31-46.

FONT E., VEIGA-PIRES C., POZO M., NAVE S., COSTAS S., RUIZ-MUÑOZ F., ABAD M., SIMÕES N., DUARTE S. & RODRIGUEZ-VIDAL J., 2013 - Benchmarks and sediment source(s) of the 1755 Lisbon tsunami deposit at Boca do Rio Estuary. Marine Geology, 343, 1-14.

GESLIN P., 1999 - L’Apprentissage des mondes : Une anthropologie appliquée aux transferts de technologies. Octarès, Toulouse & Éditions de la Maison des Sciences de l’Homme, Paris, 254 p.

GOULETQUER P. & WELLER O., 2010 - Continuités et discontinuités dans l’exploitation du sel sur la côte atlantique de la Bretagne. Haute-Normandie archéologique, 14, 95-105.

HEWITT K. & BURTON I., 1971 - The Hazardousness of a Place: A Regional Ecology and Damaging Events. University of Toronto Press, Toronto, 154 p.

HILHORST D. & BANKOFF G., 2004 - Complexity and diversity: unlocking social domains of disaster response. In G. Bankoff, G. Frerks & D. Hilhorst (eds.), Mapping Vulnerability: Disasters, Development and People. Earthscan, London, 236 p.

HOCQUET J.-C., 1994 - Production et commerce du sel à l’Age du Fer et à l’époque romaine dans l’Europe du Nord-Ouest. Revue du Nord. Archéologie, 76 (308), 9-20.

LESPEZ L., CLET-PELLERIN M., DAVIDSON R. & HERMIER G., 2010 - Middle to Late Holocene landscape changes and geoarchaeological implications in the marshes of the Dives estuary (NW France). Quaternary International, 216, 23-40.

MAGUER P., LANDREAU G. & MARTIGNOLE L., 2011 - West Salt Story : Les sauneries gauloises du littoral charentais. Archéopages, 31, 14-21.

MELLALIEU S.J., MASSÉ L., COQUILLAS D. ALFONSO S. & TASTET J.-P., 2000 - Holocene development of the east bank of the Gironde Estuary: geoarchaeological investigation of the Saint-Ciers-sur-Gironde marsh. In K. Pye & J.R.L. Allen (eds.), Coastal and estuarine environments: sedimentology, geomorphology and geoarchaeology. Special publication - Geological Society of London, 175. The Geological Society, London, 317-341.

MÉNANTEAU L., 1979 - Les Marismas du Guadalquivir : exemple de transformation d’un paysage alluvial au cours du Quaternaire récent. Thèse de 3e cycle, Université Paris-Sorbonne, Paris, 154 p.

MEURISSE-FORT M., PHILIPPE M., GOSSELIN G. & VAN VLIET-LANOË B., 2009 - Perspectives d’exploitation croisée des sources géomorphologiques et géoarchéologiques dans le Nord de la France et plus particulièrement dans le Val de Canche. Partie I : Nouvelles données environnementales, géomorphologiques et archéologiques. Revue du Nord. Hors-série, Collection Art et Archéologie, 14, 29-49.

MOBERG A., SONECHKIN D.M., HOLMGREN K., DATSENKO N.M., KARLÉN W., & LAURITZEN S.-E., 2005 - Highly variable Northern Hemisphere temperatures reconstructed from low- and high-resolution proxy data. Nature, 433 (7026), 613-617.

MILLER F., OSBAHR H., BOYD E., THOMMALA F., BHARWANI S., ZIERVOGEL G., WALKER B., BIRKMANN J., VAN DER LEEUW S., ROCKSTRÖM J., HINKEL J., DOWNING T., FOLKE C. & NELSON D., 2010 - Resilience and vulnerability: Complementary or conflicting concepts? Ecology and Society, 15 (3): 11. http://www.ecologyandsociety.org/vol15/iss3/art11/

MORHANGE C. & MARRINER N. 2009 - Paleo-Hazards in the Coastal Mediterranean: A Geoarchaeological Approach. In I.P. Martini & W. Chesworth (eds.), Landscapes and Societies: Selected Cases. Springer, Dordrecht & New York, 223-234.

MORHANGE C., MARRINER N., BONY G., CARAYON N., FLAUX C. & SHAH-HOSSEINI M., 2014 - Coastal Geoarchaeology and Neocatastrophism: a Dangerous Liaison? In S. Ladstätter, F. Pirson & T. Schmidts (eds.), Harbours and Harbour Cities in the Eastern Mediterranean. Byzas, 19. Ege Yayinlari, Istanbul, 705-726.

OSTROM E., JANSSEN M.A. & ANDERIES J.M., 2007 - Going beyond panaceas. Proceedings of the National Academy of Sciences of the United States of America, 104 (39), 15176-15178.

PAHL-WOSTL C., 2009 - A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes. Global Environmental Change, 19, 354-365.

PELLETIER P., 2011 - Effondrement ou détournement ? L’île et l’idéologie catastrophiste. Espace Géographique, 40, 88-90.

POIROT J. & GIRARDIN H., 2010 - L’attractivité des territoires : un concept multidimensionnel. Mondes en Développement, 149 (1), 27-41.

PONTEE N.I., TASTET J.-P. & MASSÉ L., 1998 - Morphosedimentary evidence of Holocene coastal changes near the mouth of the Gironde and on the Médoc Peninsula, SW France. Oceanologica Acta, 21 (2), 243-261.

PYE K. & ALLEN J.R.L. (eds.), 2000 - Coastal and estuarine environments: sedimentology, geomorphology and geoarchaeology. Special publication - Geological Society of London, 175. The Geological Society, London, 428 p.

QUENAULT B., 2014 - La résurgence/convergence du triptyque catastrophe-résilience-adaptation pour (re)penser la « fabrique urbaine » face aux risques climatiques. Développement Durable & Territoires, 5 (3). http://developpementdurable.revues.org/10683

RIPPON S. 2002 - Romney Marsh: Evolution of the Historic Landscape and its Wider Significance. In A. Long, S. Hipkin & H. Clarke (eds.), Coastal and Landscape Change through the Ages. Oxford University School of Archaeology Monograph, 56. Oxbow Books, Oxford, 84-100.

ROUSSOT-LARROQUE J., 1994 - L’âge du Fer en Aquitaine littorale : hommes et milieux naturels. Aquitania, 12, 13-25.

ROUSSOT-LARROQUE J., 2007 - Le temps qui passe et le temps qu’il fait : emprises et déprises agricoles en Médoc durant l’âge du Bronze. In H. Richard, M. Magny & C. Mordant (eds.), Environnements et cultures à l’Âge du Bronze en Europe occidentale. Actes du 129e Congrès national des sociétés historiques et scientifiques (Besançon, 2004). Documents préhistoriques (Paris), 21. Comité des Travaux Historiques et Scientifiques, Paris, 285-293.

SORREL P., DEBRET M., BILLEAUD I., JACCARD S.L., MCMANUS J.F. & TESSIER B., 2012 - Persistent non-solar forcing of Holocene storm dynamics in coastal sedimentary archives. Nature Geoscience, 5, 892-896.

STÉPHAN P. & GOSLIN J., 2014 - Évolution du niveau marin relatif à l’Holocène le long des côtes françaises de l’Atlantique et de la Manche : réactualisation des données par la méthode des « sea-level index points ». Quaternaire, 25 (4), 295-312.

STÉPHAN P., VERDIN F., ARNAUD-FASSETTA G., BERTRAND F., EYNAUD F., GARCÍA-ARTOLA A., BOSQ M., CULIOLI C., SUANEZ S., COUTELIER C., BERTRAN P. & COSTA S., 2019 - Holocene coastal changes along the Gironde estuary (SW France): new insights from the North Médoc peninsula beach/dune system. Quaternaire, 30, 47-75.

TURNER B.L., KASPERSON R.E., MATSON P.A., MCCARTHY J.J., CORELL R.W., CHRISTENSEN L., ECKLEY N., KASPERSON J.X., LUERS A, MARTELLO M.L., POLSKY C., PULSIPHER A. & SCHILLER A., 2003 - A framework for vulnerability analysis in sustainability science. Proceedings of the National Academy of Sciences of the United States of America, 100 (14), 8074-8079.

VAN DE NOORT R. 2004 - The Humber Wetlands: the archaeology of a dynamic landscape. Windgather, Bollington, 196 p.

VAN GEEL B., BUURMAN J. & WATERBOLK H.T., 1996 - Archaeological and palaeoecological indications of an abrupt climate change in the Netherlands, and evidence for climatological teleconnections around 2650 BP. Journal of Quaternary Science, 11, 451-460.

VERDIN F., 2015 - Habiter les marais estuariens à l’âge du Fer : quelques exemples en Médoc. Aquitania, 31, 85-106.

VERDIN F., ARD V., CARRÈRE I., EYNAUD F., HOFFMANN A., MANEN C., MARCHAND G. & SAINT-SEVER G., 2018 - Nouvelles recherches sur le site de la Lède du Gurp (Grayan-et-l’Hôpital, Gironde). In P. Marticorena, V. Ard, A. Hasler, J. Cauliez, C. Gilabert & I. Sénépart (dir.), « Entre deux mers » & actualité de la recherche. Actes des 12e Rencontres Méridionales de Préhistoire Récente : Bayonne (Pyrénées Atlantiques) – 27 septembre au 1er octobre 2016. Archives d’Ecologie Préhistorique, Toulouse, 209-220.

VERDIN F., EYNAUD F., STÉPHAN P., ARNAUD-FASSETTA G., BOSQ M., BERTRAND F., SUANEZ S., COUTELIER C., COMTE F., WAGNER S., BELINGARD C., ARD V., MANEN C., SAINT-SEVER G. & MARCHAND G., 2019 - Humans and their environment on the Médoc coastline from the Mesolithic to the Roman period. Quaternaire, 30, 77-95.

VINCENT K., 2007 - Uncertainty in adaptive capacity and the importance of scale. Global Environ. Change, 17, 12-24.

VISSET L. & BERNARD J., 2006 - Évolution du littoral et du paysage, de la presqu’île de Rhuys à la rivière d’Étel (Massif armoricain – France), du Néolithique au Moyen Âge. ArcheoSciences : Revue d’Archéométrie, 30, 143-156.

WATERS M.R., 1992 - Principles of Geoarcheology: A North American Perspective. University of Arizona Press, Tucson, 387 p.

WELLER O. & DESFOSSÉS Y., 2002 - Les ateliers sauniers de Sorrus (Pas-de-Calais) : un apport majeur aux techniques de production de sel et à leur évolution durant le second Âge du Fer. In O. Weller (ed.), Archéologie du sel : techniques et sociétés dans la Pré- et Protohistoire européenne. Internationale Archäologie. Arbeitsgemeinschaft, Symposium, Tagung, Kongress, 3. M. Leidorf, Rahden, 63-80.

WILLOWS R., REYNARD N., MEADOWCROFT I. & CONNEL R., 2003 - Climate adaptation: Risk, uncertainty and decision-making. UKCIP Technical Report. UKCIP, Oxford, 166 p.

YOHE G. & TOL R.S.J., 2002 - Indicators for social and economic coping capacity – moving toward a working definition of adaptive capacity. Global Environmental Change: Human and Policy Dimensions, 12 (1), 25-40.

Haut de page

Table des illustrations

Titre Fig. 1: Location maps of the study site.
Légende A/ Location of studied sites along the west coast of the Médoc peninsula (Sources: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, Swisstopo and the GIS User Community). B/ Detail of the historical map “Carte du premier quarré de la Generalle de Médoc et d’une partie de la Guienne et de la Saintonge”, surveyed and drawn by C. Masse in 1707 (Source: Institut National de l’Information Géographique et Forestière).
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-1.jpg
Fichier image/jpeg, 68k
Titre Fig. 2: Location map of archaeological sites and sedimentary outcrops along North-Amélie (A) and Google Earth vertical image providing a map‑like view of local coastal features (B).
Légende Description of units U1 to U5 (according to Stephan et al., 2019). Unit 1: base of the sedimentary sequence. Unit 1a/ 2 m-thick compact dark-green clay attributed to the “Argiles du Gurp” Pleistocene formation. Unit 1b/ Gravelly sandstone (colluvial material) locally referred to as “Grès de l’Amélie”. Unit 2: first stage of the Holocene estuarine sediment infilling of L’Amélie and La Glaneuse beaches. Unit 2a/ Medium to coarse sand deposit, dated to ca. 7290 cal. a BP and interpreted as the gradual infilling of a large tidal channel named the ‘Soulac channel’. Unit 2b/ Silty sand layer containing rootlets dated to 7420-7180 cal. a BP (early Neolithic). Unit 3: intertidal estuarine sediments interpreted as a first generation of estuarine deposits (late Neolithic to the early Iron Age). Unit 3a/ Sandy-silt at the base of the unit to a clayey-silt formed in an intertidal mudflat environment behind a progradational dune barrier. Unit 3b/ Dark organic-rich clayey-silt to silty sand corresponding to a late Neolithic high marsh environment. Unit 3c/ Humate-impregnated black sandstone (‘alios’). Unit 3d/ Thin layer of reworked lithoclasts (‘alios gravels’) with similar sedimentological characteristics to the unit 3c. Unit 4: clayey-silt deposit interpreted as a second generation of estuarine mudflats (ca. 2700 to 1250 cal. a BP). Unit 5: organic-rich clayey-silt, interpreted as a salt marsh deposit (ca. 2500 to 1070 cal. a BP). This unit was buried by aeolian sand supplies.
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-2.jpg
Fichier image/jpeg, 60k
Titre Fig. 3: Coastal morphology and features along North-Amélie beach.
Légende A/ Longitudinal view of the beach profile (north oriented) showing the Pleistocene sand terraces (unit 1b, locally referred to as “Grès de l’Amélie”) that dominate a thick compact dark-green clay attributed to an estuarine mudflat (unit 1a). A significant retreat of the overlooking dune front escarpment made possible the description of 28 sections (here between section 22 et 25) and the subdivision of the dune sequence into five aeolian sands units, separated by three layers of palaeosols or organic-rich sediment (dark lines) (photo: Y. Ceinturet, April 9, 20). B/ Cross-sectional view of the upper foreshore of AML-N-001 where the sandstone sands of the Amélie (unit 1b) are exposed at the foot of the dune. Oval archaeological structure (ST1) in the process of clearing its clay-sand filling horizon (unit 6) (photo: M. Seutin, March 22, 2015).
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-3.jpg
Fichier image/jpeg, 64k
Titre Fig. 4: Residual mound of La Lède du Gurp detached from the dune cliff after the complete uncovering of the underlying coastal marsh deposits by long-term erosion.
Légende A.  Aerial oblique back view northwest oriented showing the remnants of coastal marsh deposits after the recurrent winter storms of 2013-2014 (Photo: F. Didierjean, March 17, 2014). B. Front view towards the inland showing the Mid Holocene sedimentary filling sequence (from about 10 to 3 ka BP) of a circular depression partially connected to a former arm of the Gironde estuary. The background dune scarp exhibits thick aeolian sand and fossilized palaeosols deposits with very little evidence testifying for alternating periods (from 1275-1260 BC to the present-day) of aeolian activity and dune stability. Trace of resettlement after major environmental changes at the turn of the first millennium are rare (Photo: F. Bertrand, June 16, 2014).
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-4.jpg
Fichier image/jpeg, 68k
Titre Fig. 5: Vulnerability framework with emphasis placed on adaptive capacity and its role in influencing vulnerability through modulating exposure and sensitivity.
Légende Figure adapted from Turner et al. (2003) and Engle (2011). The components of vulnerability are linked to factors external to the studied system and interact at various spatial and temporal scales.
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-5.jpg
Fichier image/jpeg, 44k
Titre Fig. 6: Vulnerability, coping range and adaptation of coastal communities to climate variability and change along of the North Médoc peninsula over the last 12000 years.
Légende The figure, adapted from Willows et al. (2003), indicates the relationship between coping range, critical threshold, vulnerability and temperature considered as a key indicator of changes in the climate throughout the course of the Holocene. The coping range represents the tolerable climate (in both value and variability) that a system can endure at a given time given its vulnerability and resilience characteristics (Quenaut, 2014). The coping range boundaries lie about 0.2°C below the average value of reconstructions temperature proxies in the Northern hemisphere (Davis et al., 2003; Moberg et al., 2005).
URL http://journals.openedition.org/quaternaire/docannexe/image/11228/img-6.jpg
Fichier image/jpeg, 67k
Haut de page

Pour citer cet article

Référence électronique

Frédéric Bertrand, Florence Verdin, Frédérique Eynaud, Gilles Arnaud-Fassetta, Pierre Stéphan, Stéphane Costa et Serge Suanez, « Settlement potential and constraints on the lower Médoc coastline: results of the Litaq project and considerations on coastal palaeo‑risks in the protohistoric periods », Quaternaire [En ligne], vol. 30/1 | 2019, mis en ligne le 28 juin 2019, consulté le 21 août 2019. URL : http://journals.openedition.org/quaternaire/11228 ; DOI : 10.4000/quaternaire.11228

Haut de page

Auteurs

Frédéric Bertrand

Laboratoire PRODIG (UMR 8586 CNRS), Université Paris-Sorbonne, Institut de Géographie, 191 rue Saint-Jacques, FR-75005 PARIS. Email : frederic.bertrand@sorbonne‑universite.fr

Articles du même auteur

Florence Verdin

 Laboratoire AUSONIUS (UMR 5607 CNRS), Université Bordeaux Montaigne, Maison de l’archéologie, 8 Esplanade des Antilles, FR-33607 PESSAC. Email: florence.verdin@u‑bordeaux‑montaigne.fr

Articles du même auteur

Frédérique Eynaud

 Laboratoire EPOC (UMR 5805 CNRS), Université de Bordeaux, Allée Geoffroy Saint Hilaire, CS 50023, FR-33615 PESSAC. Email: frederique.eynaud@u‑bordeaux.fr

Articles du même auteur

Gilles Arnaud-Fassetta

Laboratoire PRODIG (UMR 8586 CNRS), Université Paris-Sorbonne, Institut de Géographie, 191 rue Saint‑Jacques, FR-75005 PARIS. Email : gilles.arnaud‑fassetta@univ‑paris‑diderot.fr

Pierre Stéphan

 Laboratoire LETG (UMR 6554 CNRS), Université de Bretagne occidentale, Institut Universitaire Européen de la Mer, Place Nicolas Copernic, 29280 PLOUZANÉ. Email : pierre.stephan@univ‑brest.fr

Articles du même auteur

Stéphane Costa

 Laboratoire LETG (UMR 6554 CNRS), Université de Bretagne occidentale, Institut Universitaire Européen de la Mer, Place Nicolas Copernic, 29280 PLOUZANÉ. Email : stephane.costa@unicaen.fr

Articles du même auteur

Serge Suanez

Laboratoire LETG (UMR 6554 CNRS), Université de Bretagne occidentale, Institut Universitaire Européen de la Mer, Place Nicolas Copernic, 29280 PLOUZANÉ. Email :serge.suanez@univ‑brest.fr

Articles du même auteur

Haut de page

Droits d’auteur

© Tous droits réservés

Haut de page
  • Logo Association française pour l’étude du Quaternaire
  • OpenEdition Journals