The Bossòst dome is a 35 km long and 15 km wide structural and metamorphic dome, located in the central Axial Zone, bound by the Northern Pyrenean Fault and by the Aran Valley synclinorium. The E-W-trending Bossòst Fault, part of orogen-parallel mylonite zones of latest Variscan or Alpine age (Carreras & Cirés 1986, McCaig & Miller 1986), divides the Bossòst dome into a northern half dome with an exhumed core of leucocratic muscovite-hornblende granites and two-mica granites, and a southern doubly plunging antiform with smaller, elongated, E-W-aligned bodies of granitic intrusions. The contact between granite and the mantling metasedimentary rocks is intrusive, generally crosscutting a pre-existing schistosity. Absence of internal deformation and ductile fabrics indicates that Bossòst granites are of Late Carboniferous age, similar to the Bassiès (312 Ma) and Mont Louis-Andorra (305 Ma) plutons (Gleizes et al. 1997).
Mantling the granites are mica-quartz schists and minor micaceous quartzites and quartzites of Cambro-Ordovician age (Zwart, 1963; Garcia-Sansegundo, 1996), which grade into phyllites towards the margins of the dome. They are overlain by Ordovician marble, carbonaceous Silurian slate and Devonian slate and marble (de Sitter & Zwart, 1960).
The main foliation within the Bossòst dome is a flat-lying schistosity in the core zone and a steeper foliation along the marginal areas and in the overlying post-Ordovician rocks. While the main schistosity in the northern section forms a half dome, south of the Bossòst fault it outlines a doubly plunging ESE-trending antiform. Mineral lineations are obliquely oriented to the antiformal fold axis and trend NW-SE (Mezger & Passchier, 2003). The tightly folded Devonian cover units have moved southward over a décollement located at the base of the Silurian black slates (Zwart, 1962; Matte & Xu Zhi, 1988; Garcia-Sansegundo, 1996).
Mezger & Passchier (2003) have shown that two distinct metamorphic events occurred after formation of the main S1 foliation (Figure 3). Medium pressure-medium temperature regional M1 metamorphism was followed by low pressure-high temperature M2 contact metamorphism, related to granitic intrusion. The two metamorphic events are divided by a period of strong non-coaxial D2a deformation in an extensional setting and distinguished by polymetamorphic assemblages. M2 is coeval with coaxial deformation within the contact aureole and doming during continuing emplacement of the granitic core. The final domal shape resulted from later NNE-SSW compression that formed local to regional folds.
M1 is characterized by successive growth of biotite, spessartine garnet and staurolite. Staurolite inclusion trails are straight, and random orientation of crystallographic length axis indicate that it nucleated in a period of deformational quiescence. Staurolite porphyroblasts experienced rotation with a uniform top-to-the-SE sense of shear, within a 1.5 km thick zone in the eastern part of the southern antiform. This is best preserved in staurolite-garnet schist, whose bulk rock composition did not favour growth of staurolite-consuming cordierite. Almandine garnets nucleated after staurolite. (Mezger & Passchier, 2003; Mezger et al., 2004).
M2 is indicated by overgrowth of staurolite by andalusite and cordierite. Centimetre-sized poikiloblastic andalusite porphyroblasts commonly possess inclusion trails have a convex geometry, indicating growth during coaxial deformation, flattening. Cordierite appears proximal to granitic intrusions in various parageneses with staurolite, garnet, andalusite and sillimanite, all of which are commonly found as inclusions within centimetre sized cordierite. Staurolite and andalusite inclusions in cordierite show strongly corroded rims, suggesting consumption of both phases to form cordierite. The presence of coexisting cordierite, andalusite and staurolite, and evidence for corrosion of andalusite suggest that the andalusite-producing reaction had not been completed, when cordierite-producing reaction began to consume andalusite. The presence of cordierite- and aluminosilicate-bearing assemblages within the proximity (< 1 km) to granitic intrusions suggests that M2 is related to contact metamorphism. In the northern part of the Bossòst dome a contact aureole is prominently developed and had almost completely annealed the M1 paragenesis, where only staurolite relics remain. Cordierite shows with few exceptions no indication of synkinematic growth. Fibrolitic sillimanite occurs close to granitic intrusions, commonly observed grow epitaxial on biotite and to form millimetre-long lenses, but rarely observed replacing andalusite.
M1 regional metamorphic conditions are best preserved in garnet-staurolite-biotite assemblages. Pressure conditions of this assemblage are poorly constrained by traditional geobarometry. By using computer software, THERMOCALC (v. 3.2.1) of Powell et al. (1998), P-T pseudosections and AFM diagrams can be calculated with known bulk rock chemistry. In this case pressures of 5.5 kbar and 580°C were calculated for M1 (Mezger et al. 2004). Pseudosection also show, that in the garnet-staurolite-andalusite-cordierite schist staurolite is not stable within the cordierite-andalusite-biotite stability field. Calculated pressures of 2 kbar and temperatures of 525°C indicate decompression and cooling during M2. Actual temperatures during peak-contact metamorphism were probably generally higher in the sillimanite-bearing assemblages closer to the granite intrusions.
The presence of a 1,5 km thick shear zone with hanging wall down sense of shear preserved in earlier medium pressure assemblages, the strong non-coaxial deformation that separates earlier medium pressure from a later low pressure metamorphism suggests that the D2a shear zone facilitated uplift of the core of the Bossòst dome in a regional extensional setting, similar to that of a metamorphic core complex. The difference of 3,5 kbar equals approximately 10-11 km of exhumation. The time frame for activity of the shear zone can only be guessed due to the lack of geochronological data, but can be constrained if a mid-Westphalian (c. 310 Ma) age for the formation of S1 (Matte & Xu Zhi 1988) and an approximate intrusion age of the granite of 300 Ma is assumed. Mezger & Passchier (2003) conclude that the Bossòst dome can be interpreted as the result of two compressional events separated by a temporally and probably spatially restricted dome-subparallel extension due to strain partitioning caused by rheological heterogeneities in an overall bulk compressional setting. Rheological heterogeneities could be found in orthogneiss cores of the structural-metamorphic domes. Such a core is not obvious in the Bossòst dome, but foliated diorite crosscut by granite has been observed in the core of the northern part of the dome. Geochronological dating of this diorite is planned by the author.