Découverte de mollusque d'eau douce stampiens (oligocène) et de traces de racines silicifiés au sommet des Grès de Fontainebleau près de Magny-les-Hameaux (Yvelines) ; implications paléoécologiques et conséquences sur lâge de la grésification

Abstract

The discovery of silicifed fresh water fossils in Fontainebleau Sandstones (Stampian, Oligocene) is a novelty. In the lower and middle part of the Fontainebleau Sands, concretions with marine shells were known for a long time, and also silicified roots in upper sandstones, and Taxodium trunks in the upper part of sands. Very thin layer of silicified limestone with limnea, planorbs and vegetal remains occur West of Versailles (fig. 5 and 6). Near Magny-les-Hameaux (fig. 1), an ancient quarry exhibits a good section of a sandstone body (fig. 2 and 3), 300 m long, 1-6 m thick, made of a complex association of more or less silicified masses. The Upper surface is accidented by pluridecimetric conical depressions, in a general botrioidal topography (pl.I, fig. 1). A great number of silicified shells of Lymnaea cornea BRONGN., associated with rare Planorbs and possible Hydrobia, are located in the uppermost 5 cm of the sandstone (pl. I, fig. 2, 3, 4). Root trace cross the 2 Upper meters of sandstones (fig. 4). These traces are interpreted as the silicification of primary calcareous rhizoconcretions (subcutanic features according with the pedological terminology). They represent vertical coalescent small roots (fascicles), frequently begining in the apex of a conical depression (fig. 5f and g), and crossing the mollusc fossiliferous layer. The sandstone shows many microfacies, from perfect quartzite (interpenetration grains without cement) to sandstone with chalcedony cement. Vegetal remains and shells are filled with sand grains (internal sediment) and chalcedony fans (pl. 1, fig. 5 to 10). The interpretation of this unusual biological association needs to consider the evolutions of the paleogeographies and the paleotopographies at the end of the Stampian. After the marine regression, a topography of parallel ridges (of controversed origin), is progressively submerged by the "Etampes Lake" waters (fig. 7), in which lived limnea and planorbs, and deposited calcareous sediments. The lake level suffered oscillations (and even complete exposures), and during low waters, and herbaceous vegetation was growing both on the quartz sand rigdes and the calcareous deposits; on the ridges, the roots were encrusted by calcareous coatings, related with hydromorphic processes. The duration of this phenomena has been short, because the Etampes Limestone covers all the Stampian irregularities (fig. 6). We note also that the present altitudes of the contact of Fontainebleau Sand-Sandstones/Etampes Limestone results from both the primary framework and Inter tectonical deformations (syn- clines/anticlines, fig. 6). Another problem is the age of the silicification. Silicified sandstones are mainly located near the top of the sand ridges (fig. 5 and 6), making one or several levels. Ancient hypothesis suggest silicifications at the top of aeolian ridges when the Etampes Lake occupies only the "interdunes" depressions (as on fig. 7). Others ancient hypothesis, recently adapted with geochemical argumentation, suppose that the silicification was Quaternary, contemporaneaous of the discontinuous modern valley incision (see section fig. 6, by seeping of phreatic waters. We do not agree with these later hypothesis, we believe at contrary that the silicification was early, because in Modern quartz sand dunes, all calcareous remains are quickly dissolved. The silicification can be caused by water tables related with the oscillations of the Etampes lake level, former, contemporaneous or later to the complete buring of the sandy rigdes. Geomorphic arguments emphazise the impossibility of a accumulation Quaternary age (reworking of Fontainebleau sandstone blocks in a Pliocene glacis formation).