Regional remagnetization, fluids migration and mineralization on the Cevennes border

Citation
H. Rouvier et al., Regional remagnetization, fluids migration and mineralization on the Cevennes border, B SOC GEOL, 172(4), 2001, pp. 503-516
Citations number
84
Categorie Soggetti
Earth Sciences
Journal title
BULLETIN DE LA SOCIETE GEOLOGIQUE DE FRANCE
ISSN journal
00379409 → ACNP
Volume
172
Issue
4
Year of publication
2001
Pages
503 - 516
Database
ISI
SICI code
0037-9409(2001)172:4<503:RRFMAM>2.0.ZU;2-S
Abstract
Widespread remagnetization has been identified in fold belts and forelands inboard of mountain ranges and has usually been interpreted as resulting fr om fluid migration related to orogenesis in these mountain ranges. The geoc hemical properties of these fluids should be compatible with the formation or the transformation of ferrimagnetic minerals, thus allowing acquisition of remanent magnetization during fluid migration. Carbonate hosted lead-zin c barite and fluorite) mineralization of the Mississippi Valley-type (MVT) are also generally considered to have formed during the migration of enormo us volumes of fluids and are commonly located in foreland fold belts or the ir forelands. This suggests a similar origin for widespread carbonate remag netization and MVT mineralization. The paleomagnetic dating of MVT deposits has been successfully applied in MVT districts, mainly of North America. T hus, it was used for the MVT deposits hosted by Mesozoic carbonate rocks of the Cevennes region of southern France. In view of the structurally complexity in the region, and because the resul ts presented here are intended to provide an initial reference direction on which further results can be based, only sites belonging to the most stabl e parts of this border are considered here. Most samples were collected fro m three surface sites in the area of the Largentiere mine and from independ ently oriented cores of the two deep boreholes (Balazuc and Morte-Merie) fr om the Geologie Profonde de la France program. Several samples were also co llected from three sites around the area of the Saint Felix-de-Pallieres mi ne. All these samples come from different stratigraphic levels : Carbonifer ous, Permian, Triassic, Liassic and Middle Jurassic. Geological setting (fig. 1). - The Cevennes area is a N020 striking horst c omprised of Proterozoic to Silurian formations and that is surrounded by tw o Mesozoic sedimentary basins. To the west and south of the Cevennes horst is the small Causses basin that contains in the center of the basin, about 2 000 in of Triassic to Upper Jurassic formations that are mainly carbonate s. To the east, the Southeast French basin contains 8 000 to 9 000 in of Tr iassic to Cretaceous formations, mainly shales and carbonates. The major fa ults that strike N020-040 and N090-135 cut the rocks of the Cevennes area a nd define the boundaries of the Cevennes horst. These faults formed mostly during the Upper Carboniferous or Permian and have been reactivated several times. From the Triassic to the end of the Middle Jurassic, the faults tha t mainly strike N020-040, were active during extension that resulted in the formation of the Cevennes horst. The Cevennes horst formed as a large asym metrical N020 striking ridge that resulted from the westward tilting of a b asement fault block. During the Cretaceous and perhaps part of the Eocene, the N020-040 faults were again reactivated which produced the large elevate d area, " Haut fond occitan ", located between the Aquitaine and southeast basins. The N090-135 faults were again reactivated and displaced rocks of t he horst and graben in the elevated area. Later, but before the Oligocene, the " Haut fond occitan " and its surrounding basins were subjected to the " Pyrenean " compressions. In the first stage of uncertain age (from the Up per Cretaceous to the Upper Eocene), the N090-135 striking faults were reac tivated as reverse faults. During the Upper Eocene, the N020-040 striking f aults were active as sinistral strike-slip faults. The ore deposits in the Cevennes (2.25 Mt Pb-Zn, about 5 Mt pyrite, 2 Mt fl uorite and few Mt barite) are hosted by various sandstone or carbonate form ations from the basement to the early Cretaceous. The deposits are mainly a ssociated with N020-040 faults on the eastern border of the Cevennes and wi th N090-135 on the western border of the Causses. Most of the deposits were affected by the Pyrenean compression. Paleomagnetic analysis. - The main m agnetic carrier in the carbonates of the Cevennes area should be magnetite, sometimes with rare pyrrhotite [lbouanga, in Kechra, 1997]. In red beds fr om the Balazuc borehole, Curie curves, obtained using CS2-KLY2 (Agico), poi nts out the presence of both hematite and magnetite. Because of the presence of sulfides in samples from some sites, the demagne tization process (heating and alternating field) differed from site to site and was chosen depending on the analysis of the pilot-specimens. A low tem perature magnetic component (A), as well as a characteristic remanent magne tization (ChRM) can often be isolated during progressive demagnetization. T he A component, usually removed during the first few thermal demagnetizatio n steps, is likely a recent magnetic overprint. The ChRM was obtained at hi gher fields, up to 40 mT, and sometimes temperatures up to 650 degreesC. Fo r the data from the Balazuc and Morte-Merie boreholes, the ChRM can be spli t into two different populations that vary according to lithology. In the T riassic red beds, ChRM with a mean inclination of the order of 38 to 48 deg rees was obtained (ChRM B). In all other formations (dominated by carbonate s) in the borehole and in all of the surface sites, the ChRM has a steeper inclination (ChRM C). ChRM B had both normal and reversed directions pointing out at least five e pisodes of normal polarity and four reversed. However the McFadden and McEl hinny [1990] reversal test is not positive, and ChRM B results of the super imposition of the primary magnetization with another magnetization componen t (likely the ChRM C). Age and origin of the ChRM C. - The samples retainin g the ChRM C are mostly from horizontal to very gently dipping beds. Howeve r, in two surface sites and in the deeper part of the Balazuc borehole, the dip is greater, reaching 34 degrees at Saint Martin. Progressive unfolding of the whole data set at once shows that the magnetization is "synfolding" and acquired at a mean unfolding of about 40-80 %. The mean direction for the remagnetization using small circles method [Surmont et al., 1990; Shipu nov, 1997] is very well defined as N=15 sites (244 samples), D = 3.3 degree s, l=55.6 degrees, k=2 155 and alpha (95) = 0.8 degrees (fig. 22 tab. 1). Comparison (fig. 3) of the paleomagnetic pole for the ChRM C (165.5 degrees E; 81.6 degreesN, K=1416; A(95)=1 degrees) with the APWP for stable Europe [Besse et Courtillot, 2001] shows that the pole of ChRM C is close to the e arly-middle Eocene (50-40 Ma) and the 100 Ma reference poles, but because t he magnetization is synfolding and the main folding began only at the end o f the Upper Cretaceous, the 100 Ma age has to be rejected. In the Balazuc borehole, the remagnetization was isolated using thermal dem agnetization at temperatures as high as 550 degreesC. So it cannot be a the rmo-remanent magnetization acquired at temperature as low as 130 degreesC o r even 210 degreesC (maximum temperature reached respectively at 1600m in t he borehole and locally along Uzer fault because of hydrothermal fluids) an d thus it must have a chemical origin. Clauer et al. [1997] showed that the age of illitization is different depending on the depth in the borehole. T he most recent age is 121 Ma in the lower units. Since paleomagnetic and st ructural relationships restrict the age of remagnetization to less than 60 Ma, it cannot be related to illitization [Katz et al., 1998] on the Cevenne s border. The remagnetization was therefore produced by migration of chemically activ e fluids. The fact that ChRM C is only of normal polarity seems to indicate a short duration (less than 1 m.g.) for the acquisition of the remagnetiza tion, i.e. of fluids migration. However, due to the chemical origin of the magnetization, the total duration of acquisition of the magnetization could have been somewhat longer. The ChRM C is therefore interpreted as being an early-middle Eocene remagne tization event that is related to widespread fluid migration. The peak phas e of metamorphism in the Pyrenees mountain range is late Cretaceous. In the North Pyrenean zone, erosion of the relief due to this phase starts before or during the Paleocene. The early-middle Eocene age of the remagnetizatio n corresponds to the end of, or possibly after, the post-metamorphic uplift in the mountain range of the Pyrenees. Thus, the paleomagnetic age propose d best corresponds to the time of fluid migration related to the topographi c recharge mechanism of Garven and Freeze [1984] and from the south (theref ore not from the Alps across the southeast basin). Relations with mineralization. - Remagnetization occurred here mainly in th e same units as the mineralization. Assuming a similar age for all the MVT deposits [see Sinclair et al., 1993], the MVT mineralization is younger tha n the early Cretaceous but predates the compressive deformation (Upper Eoce ne) in the Cevennes [Macquar, 1973]. Both the remagnetization and mineraliz ation therefore occurred within the same window of time. An early-middle Eo cene age for the mineralization has to be discussed in the light of some ot her geological data. A Liassic age has been obtained for green and violet f luorite within veins cutting the basement [Joseph ct al., 1973] and for ura nium deposits in the Permian Lodeve basin [Lancelot et al., 1984]. However, fluorites of the Cevennes Mesozoic cover (including in post-Liassic units) have a similar rare earth element spectrum different from that of the Lias sic green and violet fluorites of the basement, and thus cannot be of Liass ic age. Rhetian to Aalenian ages have been obtained with illites of the hos t rocks in three mineralized sites, in particular at Les Malines [Marini, 1 986]. In this site, half of the ores being in Bathonian units, these ages c annot be related to the mineralization. Ending, data about ages from matura tion of the organic matter appear to have very large uncertainty, and do no t seem incompatible with a window of age Upper Jurassic to Upper Eocene for the mineralization.