Characterization of wrench tectonics from dating of syn- to post-magmatism in the north-western French Massif Central

Abstract

This work establishes the relative timing of pluton emplacement and regional deformation from new dating and structural data. (1) Monazite and (2) zircon dating show Tournaisian ages for the Guéret granites [Aulon granite 352 ± 5 Ma (1), 351 ± 5 Ma (2) and Villatange tonalite 353 ± 6 Ma (1)] and Viseo-Namurian ages for the north Millevaches granites [Chavanat granite 336 ± 4 Ma (1), Goutelle granite 336 ± 3 Ma (1), Royère granite 323 ± 2 Ma (1) and 328 ± 6 Ma (2), Courcelles granite 318 ± 3 Ma (1)]. The Guéret and Millevaches granites are separated by the N110 Arrènes–la Courtine Shear Zone (ACSZ), composed from West to East by the Arrènes Fault (AF), the North Millevaches Shear Zone (NMSZ) and the la Courtine Shear Zone (CSZ), respectively. Tournaisian Guéret granites experienced a non-coaxial dextral shearing (NMSZ) recorded by the Villatange granite while the Aulon granite (Guéret granite) cuts across this dextral shear zone which thus stopped shearing during Tournaisian time. Visean to Namurian Millevaches granites experienced a coaxial deformation. Therefore, low displacements along the NMSZ and the CSZ occurred at the emplacement time of Chavanat and Pontarion-Royère granites (336–323 Ma). The structural analyses of Goutelle granite emphasizes a deformation related to the dextral Creuse Fault System (CFS) oriented N150–N160. From 360 to 300 Ma, the Z strain axis is always horizontal inferring a wrench setting for these granite emplacements. During this tectonic evolution, the Argentat zone acted as a minor normal fault and is related with a local Middle Visean (340–335 Ma) syn-orogenic extension on the western border of the Millevaches massif.

Appendices

Appendix 1: Analytical procedure

EPMA dating

The electron probe microanalyzer (EPMA) is becoming increasingly popular for determining Th–U–Pb_tot ages on monazite because it provides a true in situ high spatial resolution method and because it is a non-destructive method. The monazite mounted in resin and polished to obtain cross-sections of the grain was analysed using a Cameca SX 50 electron microprobe. The analytical procedure for monazite was detailed in Cocherie et al. (1998, 2005a). The interference of YLγ on PbMα is subtracted offline by applying a coefficient of interference to the value of Y. The different interference corrections have been validated by dating several monazite samples using this method and conventional isotopic methods (Cocherie et al. 1998). An acceleration voltage of 20 kV was considered while a beam current of 200 nA is now applied. According to this procedure, the calculated detection limits (2σ) are 110 ppm for Pb, 105 ppm for U and 130 ppm for Th, whereupon the absolute error is taken as 110, 105 and 130 ppm, respectively. A systematic relative error of 2% is considered for Th (whose concentration is generally above 6,500 ppm) and also for U concentrations above 5,250 ppm in order to avoid an unrealistic low error for U-enriched grains (Cocherie and Legendre 2006). For monazite the standards were galena (PbS) for Pb, uraninite (UO₂) for U, thorite (ThO₂) for Th, endmember synthetic phosphates (XPO₄) for each rare-earth element (REE) and Y, apatite for P and andradite for Si and Ca.

If two or more homogeneous age domains are separated by a gap that is lower than the analytical error on each individual spot analysis age, these can be identified by suitable isochron diagrams (Cocherie and Albarède 2001; Cocherie et al. 1998; Suzuki and Adachi 1991). A recent study shows how to provide precise ages of ± 5 to 10 Ma (2σ) using the most suitable isochron diagram according to the geochemistry of the studied grains (Cocherie et al. 2005a).

With an obvious need for a program to simplify individual age and isochron mean-age calculations, we created EPMA dating, a Microsoft Excel add-in program for determining U–Th–Pb_tot. ages from EPMA measurements (Pommier et al. 2002). All the parameters needed to calculate mean and intercept ages are computed, ready to be plotted using the ISOPLOT program (Ludwig 1999) in order to obtain statistics from suitable diagrams. Finally, EPMA dating produces (1) the age and error from the slope of the Pb versus Th* diagram, (2) the U–Th–Pb age at the centroid of the best-fit line and (3) the Th–Pb age (intercept with Th/Pb axis) and the U–Pb age (intercept with U/Pb axis) from the Th/Pb versus U/Pb diagram. All calculations were done at 2σ level. A special care was taken on the MSWD (mean squared weighted deviation) which must be below 1 + 2/(2/f)^0.5 (f: degree of freedom = number of analyses − number of dated events) in order to validate age calculation (Wendt and Carl 1991) for a single age population.

The three starting assumptions are: (1) common Pb is negligible as compared to the amount of thorogenic and uranogenic lead; (2) no radiogenic Pb loss has occurred since system closure; (3) a single age is involved at the size level of each individual spot analysis. After comparison with conventional isotopic U–Pb age determinations, it is now accepted that EPMA resolution allows to avoid inclusions and altered domains that could potentially contain common Pb. Systematic BSE study was performed to investigate monazite micro-texture for all mounted grains.

SHRIMP dating

Zircons were dated using the ion microprobe (SHRIMP II) of the Australian National University, Canberra, according to the procedure described by Williams (1998). The areas analysed (∼25 μm) were selected after studying images of the grains obtained by cathodoluminescence and by transmitted light microphotography. The determination of the ²³⁸U/²⁰⁶Pb ratio necessitated the external calibration of the measurements with the aid of a particularly homogeneous standard zircon of known composition: the Duluth Gabbro (Paces and Miller 1993). In general, areas very rich in U (> 2,000–3,000 ppm) were not selected in order to avoid, first, moving away from the area of validity of the U–Pb calibration line and, secondly, the risk of losses of radiogenic Pb, in relation to the metamictisation. For the relatively recent zircons (< 1,000 Ma), the imprecision of the ²⁰⁶Pb/²⁰⁴Pb ratio becomes critical; one then uses the Concordia diagram of Tera and Wasserburg (1972), modified by Compston et al. (1992), in which one plots the ²⁰⁷Pb/²⁰⁶Pb and ²³⁸U/²⁰⁶Pb ratios, not corrected for common Pb. In the absence of common Pb, the analyses of areas not affected by thermal events subsequent to the crystallisation of the zircon or by inherited cores are spread along this Concordia (Williams 1998). Although the variable quantities of common Pb adversely affect the values of the two ratios, these kinds of points form a straight line passing through the composition of common Pb (²⁰⁷Pb/²⁰⁶Pb) at the estimated age of the system given, in a first approximation, by the average ²³⁸U/²⁰⁶Pb ages stemming from the concordant analyses. The extrapolation of this line on the Concordia defined the sought age. This is what is called a correction of common Pb by the ²⁰⁷Pb method and not by the ²⁰⁴Pb method, as in the case of the conventional diagram. Using this correction method, one calculates the ²⁰⁶Pb*/²³⁸U (Pb* = radiogenic Pb) ratios for each point.

Whatever analytical approach was used, all the calculations were made at 2σ (95% confidence limit) using the ISOPLOT program (version 2) of Ludwig (1999). On the other hand, in the case of the data obtained using the SHRIMP, the uncertainties are given at 1σ in the corresponding table and, in the same way, the error ellipses are given at 1σ in order to make the figures easier to read.

Appendix 2: Sample location

Sample	X Lambert 2	Y Lambert 2	Locality	Rock type
CJ 51	554.95	2,116.76	Villatange	Granodiorite
LB 92	552.21	2,119.63	Aulon	Leucomonzogranite
LA 48	571.38	2,105.92	Chavanat	Two-mica granite
CJ 41	582.42	2,101.76	Goutelle	Leucogranite
CJ 47	569.45	2,102.36	Royère	Garnet, cordierite granite
CJ 53	576.38	2,105.20	Courcelles	Leucogranite