Résumé
The Montagne Noire (MN) migmatite dome in the French Massif Central has been interpreted as a double-dome produced by convergent flow of partially molten crust that ascended in a steep high-strain zone in the center of the dome and then flowed into two flanking subdomes. The driving force for upward flow of deep crust was formation of a pull-apart. Dome rocks are deeply sourced and contain a widespread gneiss protolith with zircon ages of approximately 520 and approximately 450 Ma. Monazite ages that represent crystallization of partially molten crust are approximately 315-300 Ma. Two eclogites in the dome contain unaltered omphacite + garnet: one in the dome-core in the high-strain zone, and the other at the dome margin. Zircons in these eclogites vary in extent to which they record different events: zircon in the dome-core eclogite contains a widespread record of crystallization in the presence of garnet at approximately 314 Ma (LA-ICPMS) and approximately 310 Ma (SHRIMP), with a sparse record of crystallization in the presence of plagioclase at approximately 360 Ma and an even more fragmentary record of crystallization at approximately 450 Ma. In contrast, zircon in dome-margin eclogite is dominated by cores that record the approximately 450 Ma event, reveal no trace of the 360 Ma event, and have very thin rims that record the same age of high-P metamorphism as dome-core eclogite. Mineral assemblage, bulk composition, and degree of retrogression are similar in the eclogites. A difference is the peak P-T condition: dome-core eclogite records approximately 1.4 GPa, 725 degrees C, and dome-margin eclogite records slightly lower P-T at approximately 1.2 GPa, 680 approximately C. The two eclogites also differ in their omphacite crystallographic fabric: core eclogite is characterized by flattening (transpression) whereas margin eclogite records constriction (transtension). These structural and geochronological data are compared to 3D numerical models of crustal flow in double domes. In models, material is exhumed first from the base of the transtensional domain and is rapidly transferred to the dome margin; this may explain the fragmentary record of zircon neocrystallization at high-P in the MN dome-margin eclogite. In contrast, material in the models that ends up in the core of the dome undergoes transpression before exhumation and spends more time at high P and T, promoting more extensive zircon resetting during dome evolution, consistent with the MN dome-core eclogite.