Faunes d'ammonites et interprétation des successions sédimentaires du Campanien-Maastrichtien sous-pyrénéen (Ariège, Haute-Garonne). Réfutation d'un modèle fondé sur des âges numériques hypothétiques

Abstract

From the Dreuilhe anticline in the east to the Dômes area of the "Petites Pyrénées" in the wesl, the Late Cretaceous (Campanian-Maastrichtian) successions have yielded faunas and floras at different levels in the sedimentary pile, and more particularly in the transition series between the fine-grained basin sediments (Plagne Marls, Saint-Martory Marls and Saint-Loup Marls, from east to west) and the, in places detrital, shelf carbonates (Labarre Sandstone, Nankin 1 and Nankin 2 limestones, Gensac Marly-Limestone, Larcan-Nizan Limestone, from east to west). These faunas and floras have been used (1) to demonstrate the diachronism of the Sub-Pyrenean foreland basin infill (Ricateau and Villemin, 1973; Ségura, 1979; Lepicard, 19S5; Bilotte, 1985; Kennedy et al., 1986; Martinet et al., 1990), and (2) as a chronostratigraphic framework for reconstructing the depositional sequences (Bilotte and Ségura, 1991; Bilotte, 1990, 1991). Recently, these data have been generally refuted and another model propo- sed (Deramond et al., 1993) that is based essentially on a method of dating through extrapolating the evolution rate of Rosita fornicata (Fondecave-Wallez, 1988). Comparison of the datings derived from the biostratigraphic method and those derived from the "grade-dating" method show sometimes considerable differences which appear responsible for "the identification of slope fan gravity deposits, sandy-bioclastic lowstand prograding wedges, carbonate transgressive systems tracts and ammonoid bearing condensed sections" (Deramond et al., 1993). After recalling a few fundamental facts, the present article emphasizes certain observations that contradict these assertions. The Sub-Pyrenean Campanian-Maastrichtian ammonite faunas do not come from condensation zones. Ammonites are, on the whole, rare in the Campanian-Maastrichtian successions. East of the Garonne they have been found in a suite of regressive sequences, 20-30 m thick, which mark a more or less rapid transition between the Plagne Marls and the Labarre Sandstone or Nankin 1 Limestoize (fig. 3A). From the sides of the Garonne valley west to the Dômes area of the "Petites Pyrénées", they are rare and sparse in the Saint-Martory-Marls and more common, but nevertheless dispersed, in the Saint-Loup Marls. A list of finds, including historical data, is as follows (fig. 1): - Dreuilhe anticline: two imprints of Hoplitoplacenticeras (H.) marroti (Coquand, 1859). - Richou-Montfa anticline: ten moulds, in different lithologies, from which five species have been identified (fig. 1) (Kennedy and Bilotte, 1994), including the first European record of Didymoceras stevensoni (Whitfield, 1877), an index fossil of the Upper Campanian Zone in the Western Interior (USA). - Plagne anticline: five different localities (fig. 1) have each yielded 1-2 examples of Baculites leopoliensis Nowak, 1908 (8 in all); one of these sites, Crabé, has also yielded a mould of Sphenodiscus ubaghsi de Grossouvre, 1894, but about 15 m below the level from which the Baculites leopoliensis was collected. - Hinge of the Plagne and Saint-Marcet anticlines: on side of the Garonne valley (fig. 1) the Saint-Martory Marls, in their lateral equivalent of the Nankin 1 Limestone, have yielded three ammonites - Pseudokossmaticeras tercerise (Seunes, 1891) and Hoploscaphites pumilis Stephenson, 1941, near Roquefort-sur-Garonne; a second H. pumilis, below Le Paillon. - Larcan and Dômes area: different levels of the Saint-Loup Marls and Gensac Marly Limestone (fig. 1) have yielded Hoploscaphites constrictus (Sowerby, 1817); locally, in the Saint-Loup quarry, this species is associated with Eubaculites carinatus (Morton, 1834) and, 30 m above, with H. constrictus crassus Lopuski, 1914, and Baculites vertebralis Lamark, 1801 (Kennedy et al., 1986). Chronostratigraphic constraints. The ammonite faunas are particularly interesting due to (1) the mixture of European and North American species, which include zone index species, and (2) thc absolute datings, on bentonites, attributed to certain North American marker horizons (Obradovich, 1988 and in press; MacArthur et al., 1994, Table 3). On the basis of the North American data combined with the chronostratigraphic values accorded to certain European Ammonite Zones, the Sub-Pyrenean faunas that have been collected fall within the following chronostratigraphic context (fig. 2): - Hoplitoplacenticeras (H.) marroti: collected from the top of the Plagne Marls in the Dreuilhe anticline, this species indexes the lower part of the late Campanian in Europe and is generally placed around 77-78 Ma. Didymoceras stevensoni: collected from the top of the Plagne Marls in the Montfa anticline. this species indexes a zone falling between the Jenneyi Zone (74.8 ± 0.5 Ma) and the Nebrascence Zone (75.9 ± 0.7 Ma) of the Western Interior (USA). It has been attributed an age of about 75.3 Ma (MacArthur et al., 1994, Table 3). - Baculites leopoliensis: originating from the top of the Plagne Marls in the Plagne anticline, this taxon is commonly associated in Aquitaine (Hancock and Kennedy, 1993) with Pseudokossmaticeras tercense and Nostoceras (N.) hyatti, its upper range limit would be equivalent to that of Hoploscaphites pumilis. - Hoploscaphites pumilis: collected from the Saint-Martory Marls at Roquefort and Le Paillon (fig. 1), this species belongs to the Jenseni Zone of the Western Interior (USA) (= zone with Nostoceras hyatti in Northern Europe; Kennedy and Cobban, 1993) the top of which is bracketed by radiometric determinations that enable its age to be estimated at around 71.3 Ma (Kennedy et al., 1992a). - Hoploscaphites constrictus: this strictly Maastrichtian species is relatively common in the Saint-Loup Marls and Gensac Marly-Limestone. Within the belemnite zoning of the Late Cretaceous in Central Europe, it ranges from the Lanceolata Zone to the Casimirovensis Zone (71.3 to 65.4 Ma); its association with Eubaculites carinatus indicates at least the upper part of the Junior Zone (Klinger and Kennedy, 1933); its association with H. crassus and Baculites vertebralis indicates the Casimirovensis Zone (Kennedy et al., 1986; Kennedy, I993), i.e. ranging within the Late Maastrichtian between 68 and 65.4 Ma. Comparison of the biostratigraphic and "Grade-dating" data. Between the Dreuilhe anticline in the east and the Plagne anticline in the west, four ammonite faunas have been collected either from the transition between the Plagne Marls and the Labarre Sandtone, then the Nankin 1 Limestone, or from the lateral equivalent of the Nankin 1 Limestone (fig. 1, 3). The faunas - H. (H.) marroti, D. stevensoni, B. leopoliensis/H. pumilis - show that, in the Sub-Pyrenean basin, progradation of the littoral terrigenous s.ystems began in the east, around 77-78 Ma, with the Labarre Sandstone and ended in the south of the Garonne valley, around 71 Ma, with the Nankin 1 Limestone (fig. 3A). The more carbonate series - Nankin 2 Limestone, Larcan-Nizan Limestone - complete the system, west o, the Garonne valley, between 71.3 and; 65.4 Ma. These dates do not agree with those provided by "grade-datings (fig. 3B), particularly those attributed to the Labarre Sandstone and Plagne Marls boundary which is uniformly situated around 71 Ma. "Grade-datings" and the depositional sequences. In the tectono-sedimentary evolution model of the Sub-Pyrenean basin proposed by J. Deramond et al. (1993), the sedimentary progression is not defined in relation to the geometry of the deposits nor to their sedimentological features, but was deduced from correlations with the table of eustatic cycles compiled by B. U. Haq et al. (1987) using data provided by "grade-dating". The arguments that have been developed above raise doubts as to the validity of these ages and reveal the artificial character of the resultant third-order depositional sequences. Conclusion Sub-Pyrenean Campanian-Maastrichtian ammonite faunas are rare, but in none of their localities do they form a condensation zone (F1, F2, F3 in Deramond et al., 1993). The faunas are contained within a radiochronological bracket which is accompanied by constraints that are difficult to circumvent. The differences of 1 to 6 Ma that one notes between the estimated ages and those that are derived more or less directly from palaeontological data raise doubts on the reliability of the "grade-dating" method. Based on numerical values alone, through correlation with the table of eustatic cycles drawn up by Huq et al. (1987), the third order depositional sequences in the model proposed by J. Derumond et al. (1993) are open to immense criticism.