Résumé
Des études palynologiques et cartographiques récentes permettent d'affirmer que le golfe pliocène de Montaigu, dans le nord-ouest de la Vendée, fut en réalité l'embouchure d'un paléo-fleuve yprésien (Eocène inférieur). Des traces de ce fleuve semblent exister jusqu'en Brenne. Mais c'est surtout entre Poitiers et Bressuire, que les sédiments fluviatiles dessinent une traînée claire, large de 5 km en moyenne. Dans les Collines vendéennes, l'érosion a auit disparaître les sédiments fluviatiles. L'hypothèse selon laquelle le couloir de Saint-Mars-la-Réorthe pourrait représenter un vestige dégradé de la paléo-vallée est discutée. Dans le Bas-Bocage vendéen, les sédiments réapparaissent et. en aval de Montaigu, ils s'élargissent en un paléo-delta. Les sédiments fluviatiles dominants sont des cailloutis qui remanient des éléments provenant de la bordure sud-ouest du bassin de Paris : silex jurassiques, dragées de quartz et aussi quelques fossiles du Crétacé supérieur. Les sables qui accompagnent ces cailloutis acquièrent un façonnement marin dans l'embouchure. Les argiles sont riches en kaolinite. En une cinquantaine de localités, des sapropels ont livré une microflore yprésienne, rendant ainsi caduque l'âge pliocène qui était donné à cette formation jusqu'à ces dernières années. Des grès, plus ou moins assimilables aux "grès à Sabals" éocènes de l'Anjou, sont fréquents à la partie supérieure de ces sédiments. La disparition du fleuve, au Cuisien terminal ou au Lutétien inférieur est attribuée à la tectonique éocène, responsable du bombement des Collines vendéennes et du jeu vertical de certaines failles de direction sud-armoricaine, en particulier la faille de Pouzauges. Ces nouveaux reliefs ont divisé le bassin versant yprésien en deux bassins préfigurant le bassin versant de la Loire et celui des fleuves côtiers (Lay, Sèvre Niortaise, Charente). Les mouvements verticaux dus à cette tectonique ont pu être quantifiés dans la zone des Collines vendéennes, grâce aux déformations enregistrées par le profil longitudinal du fleuve. Leur intensité est de l'ordre de 150 m. Dans le nord-ouest de la Vendée, la révision de l'âge des sédiments fluviatiles, autrefois rapportés au Pliocène, désormais attribués à l'Yprésien, permet de reconsidérer l'âge des failles qui les affectent et d'attribuer une partie de leur jeu à l'Eocène. Par ailleurs, l'âge des sédiments piégés dans les principaux grabens montre que la tectonique tertiaire de la région fut principalement active de la fin de l'Yprésien au Stampien inclus. Enfin, la transition entre les domaines continental et marin de l'Yprésien, localisée actuellement à une altitude de 45 m, est bien inférieure au niveau moyen des mers yprésiennes. Ceci corrobore l'hypothèse d'une subsidence régionale au Néogène proposée par d'autres auteurs. Des révisions stratigraphiques sont proposées : une revue de l'ensemble des sédiments graveleux ("sables rouges") du Massif armoricain méridional, classiquement attribués au Pliocène, permet de montrer qu'une grande partie de ceux-ci est d'âge beaucoup plus ancien, yprésien ou crétacé supérieur.
Abstract
Recent palynologic (e.g. Ollivier- Pierre et al., 1985) and cartographic studies allow the assertion that the so-called "Pliocene gulf of Montaigu", in the north-west of the Vendée region (southern Armorican massif, western France), was in reality the mouth of an Ypresian (Lower Eocene) paleoriver which drained western France before the Loire river existed. The traces of this river appear to exist as far as Brenne, in central France (around La Roche-Posay in Fig. 2), where fluviatile sediments of a Cuisian age (upper Ypresian), lying at the base of continental Eocene deposits, outline an E-W trail (Donnadieu, 1976; Châteauneuf 1977; Rasplus, 1982). Between Poitiers and Bressuire, the fluviatile sediments outline a clear path averaging 5 km in width (Klein, 1961a; Coubès et al., 1985; Platel and Mourier, 1986). In the Vendée Hills (Fig. 1), erosion has caused the disappearance of these sediments, but the river may have left a surprising imprint: the Saint-Mars-la-Réorthe corridor, the origin of which is discussed. The edges of this hypothetical fossil valley are indicated by escarpments still well preserved (Fig. 3, 4, 5). This corridor is proved to be far older than the present rivers and it coincides, more or less, with a strip of schist within granite (Fig. 3) which may have guided the paleoriver. This structure cannot be satisfactorily explained by selective erosion between granite and schist as part of the edges does not coincide with the granite boundaries. Further west, on the western side of the Vendée Hills, the sediments reappear and, downstream from Montaigu, they enlarge into a paleodelta (Fig. 1). Detrital Ypresian sediments are known further NW, in western Brittany, and relic sediments of a similar facies have been observed as far as Morbihan (Mazères, 1931). South of the paleodelta, the shoreline was directed towards the Coutras region, near Bordeaux, where, at the same period, an other delta existed (Dubreuilh, 1989; Meunier et al., I989). The dominant fluviatile sediments are gravels which contain reworked clasts coming from the SW edge of the Paris Basin: Jurassic flint, quartz pebbles and some reworked fossils of the Upper Cretaceous, especially silicospongiae. These sediments are extremely similar on both sides of the Vendée Hills, indicating that they belonged to one and the same paleoriver, despite the gap of sediments in the Vendée Hills. In the paleodelta, the sands accompanying the gravels acquire marine characteristics and interbedded kaolinite-rich clays and sapropels appear. Stream channels are often observed (Fig. 7). Sandstones, more or less similar to the Eocene sandstones of the adjacent Anjou region ("Grès à Sabals"), are also frequently found in the upper part of these sediments. In about fifty different localities in the paleodelta, sapropels have yielded an Ypresian microflora (e.g. Ollivier-Pierre et al., 1985; this study: cf. Table 1), rendering null the Pliocene age which, until recently (Chevalier, 1987; Chevalier and Borne, 1989), had been assigned to this formation since Vasseur (1881). At Noirmoutier, the sandstones also contain a tropical macroflora of Cuisian age (Crié, 1881; Bonnet, 1905; Ters, 1978c; Borne and Vaudois-Miéja, 1986). The disappearance of this river, during the late Cuisian or early Lutetian period, is revealed by a change of sedimentation in the paleodelta where Lutetian calcareous sediments, without reworked elements from the Paris Basin, succeeded the Ypresian detrital fluviatile sediments (e.g. Borne, 1987). This disappearance is attributed to Eocene tectonism, also responsible for the uplift of the Vendée Hills and the vertical play of several faults, particularly the Pouzauges Fault (Fig. 5). These new reliefs divided the former Ypresian river basin into two new basins, prefiguring the Loire river basin, on the one hand, and the basin of coastal rivers (Lay, Sèvre Niortaise, Charente), on the other. The vertical movements related to these tectonic events have been quantified in the Vendée Hills, thanks to the deformation recorded by the longitudinal profile of the paleoriver (Fig. 8); their-measured intensity is in the order of 150 m. The reappraisal of the age of the fluviatile sediments in NW Vendée, previously placed in the Pliocene and henceforth attributed to the Ypresian, also allows a reconsideration of the age of the faults which distorted them; at least part of their play is Palaeogene. These arguments, together with the age of sediments trapped in grabens, lead to the conclusion that Cenozoic tectonism in the region was principally active from the end of the Ypresian to the Stampian inclusive, and could be related to the Pyrenean orogeny. Finally, the transition between the terrestrial and marine domains, currently located near Montaigu (Fig. 1) at an altitude of 45 m, is well below the mean sea level of the Ypresian seas (around 200 m: Haq et al., 1988). This corroborates the hypothesis of a regional subsidence during the Neogene proposed by Wyns (1991). A review of the various azoic gravelly sediments of the southern Armorican massif classically attributed to the Pliocene since Vasseur (1881), allows us to show that a large part of these are probably much older: in addition to the Ypresian fluviatile sediments, some are Upper Cretaceous marine sediments. Two main conclusions are proposed: a) The reappraisal of the age of various sediments and faults leads to the conclusion that the Cenozoic evolution of the Vendée region, which was thought to be recent, began much earlier. b) The Palaeogene "crisis", a consequence of the Pyrenean orogeny, is testified in western France by several tectonic disruptions and paleogeographical changes, especially the disappearance of the Ypresian paleoriver. This crisis and the subsequent evolution ended before the Miocene, when the creation of the Loire basin was achieved.
Dernière mise à jour le 28.07.2015