La villa des Chapelles, au lieu-dit le Grand Fief Chagnaud à Port-des-Barques (Charente-Maritime, Nouvelle-Aquitaine), est remarquablement située sur une langue de terre qui s’avance dans la mer, dominant d’un côté l’embouchure de la Charente, de l’autre le golfe de Brouage. Repéré depuis longtemps, le site a fait l’objet de sondages et de fouilles partielles entre 1957 et 1999. La position de la villa à proximité immédiate de l’Océan semble favorable au développement des activités en lien avec l’exploitation des ressources marines. Une approche intégrée incluant des données anciennes (documentation et mobilier archéologique) et nouvelles (prospections géophysiques) a été entreprise. Au terme de ce travail, une première synthèse relative à l’organisation architecturale, à la chronologie et aux productions a été proposée. Si l’activité viticole revêt une importance certaine, la pluriactivité entre terre et mer de la villa devait très probablement être inscrite dans une économie jouant sur la saisonnalité des ressources.
The Chapelles villa in the locality of Grand Fief Chagnaud at Port-des-Barques (Charente-Maritime, Nouvelle-Aquitaine) lies in a remarkable location on a spit of land overlooking the mouth of the River Charente on one side and the Gulf of Brougage on the other. In Roman times, this rural settlement lay in the lands of the Santones whose main city was Saintes. The site has long been known of and surveys and partial excavations made between 1957 and 1999. The plan of the villa appears to be almost complete (fig. 6) with its residential and working parts. Three buildings bounded a vast unconstructed space : to the north, a dwelling area ; to the west winemaking premises with vats ; to the east, undefined buildings. The whole can be dated to the late 1st century AD. The site does not seem to have been used for winegrowing in the 5th–6th century AD while building II was reoccupied from the early Middle Ages. Moreover, several probably medieval graves were laid out within the bounds of some of the buildings. The villa’s position beside the Atlantic seems to have promoted potential activities related to the exploitation of marine resources. An integrated approach including ancient data (archaeological records and material) and new data (geophysical surveys) has been undertaken. To do this it appeared necessary to review collections and take up the study of the material again. The objective of this inventory and the associated studies is to characterise the various farming enterprises and craft activities conducted on the villa during Antiquity. On completion of this review work, the value of an integrated approach including both ancient and new data no longer needs to be proved. The complementary sources yield a first summary of the architectural arrangement, the chronology and the activities practised on the villa. Extensive electrical and magnetic geophysical prospection has identified the footprint of the remains and supplemented the plan of the excavated structures. The linear plan of the pars urbana and the layout of the various production buildings are very much like the villae of Soubise and Royan, possibly attesting to an architectural scheme on a micro-regional scale. The ceramic study has confirmed the date of construction of the villa between AD 70 and 80. The rearrangement of the pool room and the addition of baths are evidence of stages of redevelopment whose chronology remains unknown in the current state of research. The presence of a midden, in the space of the monumental opening of the pars urbana , marks out one of these phases of redevelopment dated to the late 3rd century. Finally, the study of the pool fill has confirmed the date these structures were abandoned in the late 5th century. Although the winegrowing function has been established from the origin of work by the presence of winemaking vats and tools, other activities may have been carried out. It is very likely that cereal farming went on at Port-des-Barques, although most of the archaeological evidence remains indirect (presence of a barn). Only archaeological data (seeds, pollen) could confirm this hypothesis and provide a list of species grown. Nothing is known of the role of livestock farming and exploitation of the natural environment, for want of closer studies of the material. Archaeozoological study has produced novel hypotheses about forms of livestock farming with the presence of very young specimens among the members of the domestic meat suppliers (cattle, swine, sheep and goats) and fowl. The exploitation of the marine environment and wetlands is attested through the discovery of remains of shellfishes, crustaceans, fishes, seabirds and shorebirds. Finds of net shuttles and weights is evidence of fishing and possibly hunting. It seems that the intertidal zone was exploited at low tide for bird catching, shellfish collecting and fishing for species like sole. Lastly craft activities, and especially metal working, are also assumed from the presence of lead waste and furnaces identified from magnetic surveys. The table drawn up here serves as a review of available data and provides a foretaste to the resumption of field work that seems necessary to confirm the assumptions made. It would be worth checking the reconstructed plan of the villa against finer chronological items so as to better identify the various states and reworkings of this rural settlement. Despite widespread excavation, the vat area of the winemaking building has not yielded all its information about how this production was carried out. Moreover, the collating of archaeozoological and archaeobotanical data through suitable sample collection would help to characterise output, possible specialisations and the role of trade. In the same way, chemical analyses of the soil might be of great use in delimiting the areas where livestock were housed, whether open (enclosures) or closed (sheds). Lastly the spatial analysis of the distribution of material (ecofacts and artefacts) might prove decisive for defining the functions of spaces. It has to be recognised that this information can only be collected by excavation work. Alongside this, a geomorphological study backed by geophysical surveys seems essential for confirming certain hypotheses related to the exploitation of the shore and the question of a possible landing stage. Initial observations of the Charente estuary reveal a slightly different shape to that visible today. In early Roman times, it seems the Rochefort marsh was largely filled, even if channels maintaining lagoon systems might still have run through it. Such ecosystems might have provided ecological niches for certain animal species that could have been of some importance in farming in Antiquity. Similarly, such work might seek to better identify soil characteristics. Limestone rich areas are potentially suitable for cereal growing. The study of such data collected during field work would feed into a wider issue relative to the share of each activity and changes in its proportion. Moreover, the economic purpose (own-consumption or speculative activity) calls for particular attention. Accordingly, even if some particular activity may be important in the economics of a villa, other enterprises must have yielded additional resources. For instance, the Ausone estate in the Bazadais area (Gironde) amounted to 262 ha with 25 ha given over to vineyards. Multiple activities between land and sea at the Port-des-Barques villa would very likely have been part of an economy depending on the seasonal character of resources. This pattern is echoed in more recent periods. In medieval and modern times, certain coastal communities, particularly in Saintonge, did not by any means overlook agricultural activities.
