Découverte d'une faune à Palaeotherium et Anoplotherium dans le bassin tertiaire de Domérat (Allier)

Abstract

The Montluçon-Domérat Basin is a relatively depressed area, about 15 km long from north to south, with a generally semi-circular western side and abutting against a well-developed N-S escarpment, the Cher fault, to the east. Geometric considerations (outcrop locations, borehole intersections) show the Cenozoic deposits to be 100 m or more thick in the deepest part of the basin, the Cher Trench. Most of the basin filling is composed of monotonous clayey-sandy deposits showing no stratification, but in places containing coarse blunted quartz gravel. These deposits end with calcareous encrustations (only seen in certain areas) intercalated in greenish clayey sediments or with micrite lenses; towards the original edges of the basin they are succeeded by red encrustations (fersiallitic palaeosols) that have fossilized not only the basement palaeoreliefs, but also Permian arkoses and Triassic sandstones. The facies similarities between these levels and the "terminal red crust" of the Brenne succession has led to the Montluçon-Domérat series being considered as equivalent to the main detrital body of the Brenne Basin (lying about 100 km away) and thus being similarly attributed a Ludian age. Finally, in the northern part of the Montluçon-Domérat Basin (Vallon, Urçay) these red encrustations are covered by veneers of flint or lacustrine limestone attributed to the Sannoisian. The Domérat fossil mammal locality described here is located in the upper part of the succession, i.e. in the limestone encrustations and micrite lenses, in an area where the general sedimentary pattern is interrupted by the Prémilhat granite massif. In the vicinity of the fossil locality the Tertiary succession begins, to the south, with associations of clay and sands with common facies variations and a crude stratification; silicification is seen at different levels. To the west, beneath the red-brown palaeosol, the detrital series contains calcareous encrustations. At a topographically lower level one finds beds and 10-m -size pockets of green clay. The fossil mammals come from one of these pockets, which has the shape of bell possibly due to compression of the sediments, and is capped by an unsorted quartz gravel which shows up the stratigraphy. The main fossil horizon contained within the clay forms a breccia that is about 10 cm thick and dips at 40-45° to the south (slightly less further to the east) at about 3-4 m depth. It is framed by other, smaller, fossil horizons. A sedimentological analysis of the deposits in the Domérat area reveals an eluvial environment to the east with alteration of the initial minerals and formation of kaolinite and individualization of iron oxides, and a site of illuvial accumulation to the west with kaolinic and smectitic clays (dominant smectites and secondary kaolinites) reflecting a confined environment. Reconstruction of the Late Eocene palaeogeography show a differentiated topography of palaeohills and palaeovalleys, and glacis fossilized by red beds. This toposequence is characteristic of tropical climates with strong seasonal contrast with a severe and extended dry season, a net hydric deficit, and (average?) temperatures in the region of 28°C. Mineralogical analysis of the green clays at the fossil locality shows a strong predominance of highly swelling interstratified illite-smectite and a particularly basic and reducing milieu. It was an area of shoals where pedogenesis favoured the development of vertisols, whereas the fersiallitic soils associated with the calcareous incrustations were formed on the emerged parts of the basin. The anarchic accumulation of bone fragments suggest that they were originally caught up in a mud flow and that the whole deposit was then affected by thixo tropic creep. The fauna is composed of six species of large ungulates, perissodactyls and artiodactyls. It is dominated quantitatively (95% of the samples) by a Palaeotherium magnum; some 100 teeth and many bones fragments of this species have been collected. The dimensions of the teeth, which are not very characteristic, and the morphological features of the premolars, which are ambigous owing toi their variability and to the juxtaposition of contradictory indications, prevent the attribution of a subspecific denomination to this form which lies somewhere between P. m. girondicum and P. m. magnum. Two other palaeotheriids are recor- ded. Plagiolophus minor has relatively large teeth and little coronary cement; characters that are considered as archaic for this lineage. A medium-size Palaeotherium is also recorded on the basis of a few teeth fragments that are not sufficiently characteristic for a bitter determination. The best represented artiodactyl is an Anoplotherium which, although difficult to determine at specific in the absence of post-cranian skeleton, should probably be assigned to A. commune. Finally, two teeth fragments have been found with a morphology that allows them to be attributed respectively to Xiphodon gracile and to Amphimeryx, possibly A. murinus, two Upper Late Eocene species. This very small fauna with no small mammals and the very, fragmentary state of the fossils indicate a selective transport of large fragments over a fairly large distance. We can surmise that mammals grouped together in the marshy environment (as indicated by the sedimentology) were caught up in a flood and bogged down, and that their remains were then transported to where they were discovered. The interest of this new locality is that for the first time we have a direct argument for dating a well-marked horizon in this detrital succession. All the species present are characteristic of levels MP 18 to MP 20 of the biostratigraphic scale of European Paleogene mammals. Moreover, Palaeotherium magnum is represented by a form intermediate between two subspecies, that of La Débruge (MP l8) and that of Montmartre (MP l9), whereas Plagiolophus in archaic in the minor lineage. The Domérat fauna can thus the attributed to an intermediate age between levels MP 18 and MP 19, certainly not later than MP 19, i.e. middle Ludiun. This dating confirms the parallelism which up to now was only assumed to exist between the detrital filling of the Montluçon-Domérat Basin and the main detrital body of the Brenne succession.