Sédimentation progradante au Cénozoïque inférieur en Aquitaine méridionale : un modèle

Abstract

The existing chrono-litho-stratigraphy of the Early Cenozoic succession (Paleocene to Eocene) of the southern Aquitaine Basin is outdated and no longer valid. Total revision of the data, especially geophysical seismic reflection data, has resulted in the proposal of a classic prograding model, which fits in with the existing one, east of the Garonne River. Background. Many authors, from the 19th Century (e.g. Delbos, 1854) up to recent times (e.g. BRGM, ELF-Re, ESSO-REP, SNPA, 1973), have contributed to our knowledge of the Aquitaine Basin. Initially based solely on outcrop and macrofauna study, the understanding of the basin progressed with the expansion of oil exploration; the combination of new geophysical methods (seismic reflection), deep wells and the correlative development of micropaleontological studies (planktonic and benthic foraminifera, nannofossils, palynology) has made it possible to identify the successive major periods of Mesozoic and Cenozoic history. Oil-company exploration had rapid success in the Mesozoic succession (the Lacq and Meillon gas fields in Jurassic and Barremian limestones; the Lagrave and Pecorade oil fields, etc., in Late Cretaceous sandstones and limestones). The Cenozoic rocks - the barren rocks -have been of lower priority in oil company studies (Fig. 1). Until recently very little oil company work was published. Following the synthesis of the Aquitaine Basin geology (op. cit., 1973 to the west of the Garonne Rive, most work in this area consisted of specialised studies by the academic sector; examples are the work on nummulites by Schaub (1981), on Eocene dinoflagellates from Chalosse by Michoux (1981), and on nannofossils and planktonic foraminifera by Seyve (1984), Mancion (1985) and Mathelin (1988). It is only very recently that various authors have attempted to define the lithostratigraphy and interpret the sequence stratigraphy based on a revised biostratigraphy (Mathelin and Sztràkos, 1993; Sztràkos, 1996; Sztràkos et al., 1997; Fondecave-Wallez et al., 1995; Peybernes et al., 1996, etc.). Conversely, research to the east of the Garonne River has advanced well over the last 30 or so years with the work of Villatte (1962), Tambareau (1972), Piaziat (1984), Crochet (1991), etc. Within this context, work on the forthcoming "Synthesis of the Pyrenees" led us to re-examine the Paleocene to Middle Eocene succession in the southern Aquitaine Basin to the west of the Garonne River. The present work aims to complete in space and extend in time the models recently proposed for the Paleocene to Early Eocene to the east af the Lannemezan Plateau (Tambareau et al., 1995). Review: complexity of the outdated nomenclao distinguish different lithostratigraphic units, or more precisely biolithostratigraphic units (cf. use of the term "layers" in the British sense of "beds", and apart from some rare exceptions (e.g. Bousquet et al., 1977; Platel, 1990; etc.), the lithology of the deposits is very similar and above ail repetitive: "molasse" with local conglomeratic bodies (Palassou conglomerate), "sand and sandstone" showing variations in cleanliness and quartz content, "carbonate" and "marl"; each of these lithologic units can be very thick and lithologically monotonous. However, interbedded limestonemarl, limestone-marl-sand, etc., also exist, in places over considerable thicknesses as shown by many oil wells. The repetition of these commonly 'simple' lithologies, which reflects a classic pro grading sedimentary type model from the 'middle' Ilerdian, has resulted in confusion concerning the contemporaneity of the various units. Certain lithofacies were considered as chonological units (e.g. the Palassou conglomerate), the presence of which for many years was interpreted as a signature o the Middle Eocene Pyrenean phase. This temporal unicity adopted for the base of the Palassou conglomerate is reflected in the stratigraphic assignment of the underlying units (nummulitic sandstone, etc.). Already demonstrated by the Geological Maps of the Aquitaine Basin (op. cit., 1973), and again by those of the forthcaming 'Synthesis of the Pyrenees' the history of the Aquitaine Basin during the Paleogene shows an evolution with two successive phases. From Paleocene to 'middle' Ilerdian times, the basin progressed eastwards, drowning the platform, resulting in a retrograding pattern within the different order transgressive-regressive cycles: cf. transgressions during the Danian, then the Selandian-Thanetian and finally the Ilerdian, which respectively reached the Mas d'Azil, the eastern Corbières and Béziers. Many examples of this evolution have been described on a more local scale, such as in the middle Early Ilerdian Mancioux limestone east of the Garonne River (cf. Bousquet et al., 1977). During the Middle Ilerdian, the pattern was reversed, giving way to a prograding evolution governed by basin filling within the transgressive-regressive cycles. Despite putting forward this "axial upstream to downstream progradation" as the principal sedimentation process, the authors of the Geology of the Aquitaine Basin (op. cit., 1973) practically restrict this mechanism in their reconstruction (cf. cross sections) to the end Early Eocene before deposition of the Palassou conglomerate. In the Middle Eocene, the conglomerate would thus reflect the peak phase of the Pyrenean orogeny, which caused emergence of the mountain belt. At the beginning of the 1970s, Tambareau (1972, 1976) had already demonstrated a Middle to Late Ilerdian age for the onset of Palassou conglomerate deposition, in continuitywith the older marine beds, to the east of the Garonne River. Analysys of the Palassou conglomérate by Crochet (1991) distinguished three units separated by two major angular uncanformities, respectively intra Lutetian and end-Bartonian in age. Progressively, the perfection and multiplicity of different dating methods has resulted in a demonstration of the heterochronology of the lithofacies and even of the bio-lithostratigraphic divisions; the examples below illustrate this well. By taking into consideration the planktonic foraminifera, nannoplankton and dinoflagellates and through improved determinations of the benthic foraminifera, alveolinids, nummulites, etc., it has been shown that, despite being geographically close, the typical beds with large nummulites at Fontaine de la Medaille (between Montfort en Chalosse and Gamarde) are clearly distinct from and younger than the "large nummulitic beds" at Jean Gazé ta the northwest of the Bastenne-Gaujacq dome (Kapellos and Schaub, 1973; Schaub, 1981). Another example is the "Xanthopsis" or "Trabay" marl between Horsarieu (Sourbe marl pit: NP12, P7) and the Louer bridge (NP11) (Kapellos and Schaub, 1973; Michoux, 1981). It must be noted that, in this area, the distinction between the "Xanthopsis beds" and the "discocylinid marl" with identical lithofacies and abundant Xanthopsis was only introduced to distinguish beds of different age. Even though the contemporaneity of distinct facies (consequence of the heterochronology of identical facies) had been demonstrated early on (e.g. Feinberg, 1964: lateral facies correlation between the "Trabay beds" and the "Lapêche beds" to the south of the Audignon Anticline), this concept is still only partially accepted, sometimes with difficulty. At basin scale, the general heterochronology is from east to west and can be demonstrated for example, along the North Pyrenean Front by the nummulitic sandstone. The sandstone is Middle Ilerdian in the Aude, Late Ilerdian along the Garonne and in Bigorre, Early Cuisian to the east of Bearn (Ossun), and end-Early Cuisian to early-Middle Cuisian south of Pau (Bosdaros). Much farther west, it is not until Urcuit near Bayonne (Schaub, 1981) that nummulitic sandstone of end-Cuisian is observed in an allochthonous position. The heterochronology of the "nummulitic sandstone" is even greater farther north (east of the Audignon ridge, Grenade/Adour 1 well) where the sandstone contains nannoplanktonic associations indicative of zones NP15 to NP16, i.e. Late Lutetian to Early Bartonian (Fig. 2). The heterochronalogy of the "nummulitic sandstone" or Grenade/Adour sandstone (Sztrakos 1996) and that of the molasse base (or Palassau conglomerate) implies the heterochronology of the interbedded sandy bodies ("inframolasse sands" or "Lussagnet sands". Other major sandy bodies are known in Aquitaine, such as the "Coudures sandstone", the lower and upper sands, etc., and even if these units locally cover extensive areas, their heterogeneity is visible in wells; they grade laterally and vertically into marl and/or carbonates. Camparison of the cross-sections of the platform area of the western Pyrenean foothills (well illustrated by Tambareau et al. [1995] and intersected by thee Saint-André 1 well), the Pont-Labau cross-section south of Pau (Seyve, 1984; Peybernes et al., 1996; Sztrakos et al., 1997) and the Gave d'Oloron cross-section at Oraas to the south af Orthez (Boltenhagen, 1966, Tambareau et al., 1994, and unpublished data of Aubry and Berggren) (Fig 3b), has made it possible to define a sedimentary model for tire Paleocene and Early Eocene. For the more recent successions, comparison can be made between the Grenade/Adour 1 well section and those of the Amou 1 and Pomps 1 wells (Fig 3a). The biostratigraphic attributions refer respectively to the zonation of Martini (1971) for nannoplankton (NP) and to that of Berggren and Miller (1988) modified by Berggren for planktonic foraminifera (P). Reinterpreted analytical data: a model. The generalised heterochronalogy of the facies, in addition to the spot biostratigraphic datings, means that it is difficult to use long-distance well-log correlations. However, notable unconformities of major geographic exent exist throughout the basin. The youngest for the considered period are tire two "angular unconformities" identified by Crochet (1991) within the Palassou conglomerate to the east of the Garonne River, and which can be traced far the west of the Garonne. They are exposed in the area to the west of Orthez where their age has been constrained, such as in the Urcuit area through the syntectonic deposition of Early Lutetian "breccia" (Schaub 1981), and in the Lagelouze quarry to the west of Belloc where sub-vertical end-Bartonian nummulitic limestone is overlain by Palassou conglomerate (unit III) dipping 25° to the north (Hourdebaigt, 1988). The many wells, and particularly the seismic profiles, demonstrate the generalisation of these unconformities in the western part of the basin. Similarly, the unconformity at the base of the Gan Marl, well known from the work of Gubler and Pomeyrol (1946) to the south of Pau, is easily recognised on the seismic profiles of the Tarbes-Arzacq Basin. Analysis of the older successions based on cross-sections (Pont-Labau, Oraas, etc.) and well data (Pontacq 1, Saint-André 1, etc.) enables correlation with the outcrops studied in detail to the east of the Garonne River (Tambareau et al., 1995). Two major unconformities can be traced over the entire basin (from the Mediterranean to the Gulf of Gascogne) where they are commonly overlain by detrital bodies (tire "lower sands" and "upper sands"). In the deep basin domain in the Eibar area, between Saint-Sébastien and Bilbao, these unconformities are overlain by very coarse breccia (Pujalte et al., 1994). Otirer unconformities, in places djfficult to trace in the basin domain, are recognised in the basin, such as at the Thanetian-Ilerdian boundary (between the Gloliveolina levis and Alveolina cucumiformis zones) or within the Bartonian beneath the dominantly transgressive Brassempouy limestone. To simplify matters, we shall call these major unconformities DO to D7 from the Cretaceous-Tertiary boundary (DO) to the top of the Middle Eocene (D7) (cf. table Fig. 3b). conclusion. The prograding pattern is reflected by the interpenetration of all the lithofacies, including the continental successions and their products (coarse sand). The examples presented incontestably show this interpenetration (Grenade/Adour 1 section: molasse above and below the Brassempouy limestone). In addition, clastic formations deposited under marine conditions exist at various levels ("lower sands" and Baliros sands in the Pontacq 1 well). Based on well data, seismic interpretation and available dating evidence (Fig. 4), a composite model has been established for the western Aquitaine area, from the Garonne River to the Atlantic coast, which fits in with the existing models farther east (Tambareau et al., 1995). Acknowledgements: The authors would like to express their thanks to Elf Aquitaine Exploration Production France and particularly to the Exploration Division for giving access to seismic and geological data that contributed to development of the present model. Thanks anks also to P. Pouchan and J.Rey for their meticulous reading of the text and comments that helped improve the final manuscript.