Introduction
Groix island is located in the southern Armorican domain which was structurated during Hercynian events. This part of south Brittany is called “the south Armorican grinded zone” because of the intense shearing processes which had shaped the area during Hercynian times. However, contrary to the coast which exposes high temperature metamorphic rocks associated to leucogranites occurrences, Groix island only presents low temperature metamorphic rocks with various pressure (high to medium).
Main field observations were made on the eastern part of the island and have permitted to distinguish metapelites with various paragenesis and meta-basics rocks with glaucophanites or eclogites occurences. A green schist facies has also be seen in the southern part of the island. Shearing markers with a S/N stretching direction were seen all around the island.
The field observations conduced to perform the nature of the protoliths and strain markers in order to replace the hypothetical evolution of the Groix island within Hercynian events.

Field observations
The Porz Melite area was the first zone of study. In this zone was observed intercalated metabasics horizons (centimetric to metric) in meta-sedimentary rocks (fine schistosity) (Figure 1). Meta-sedimentary are mainly micaschists with a Garnet-Phengite-Chloritoide high pressure mineralogy but some are micaschists to gneisses with an Albite-Chlorite-Quartz low pressure paragenesis. Metabasics are also various: glaucophanites high pressure with glaucophane-garnet-phengite paragenesis; low pressure with epidote-chlorite-albite mineralogy. Pebbles shows a high pressure eclogite facies with omphacite-garnet-glaucophane but it wasn’t observed in outcrops all around the islans study. The metabasics observed often look like sedimentary bed deposits (Figure 2).
Ghosts of lawsonites were seen in blue schist horizons and are infilled with albite or phengite. The contact between micaschists and amphibolites is various: folded levels of metabasics in a micaschist matrix with a sub-horizontal hinge; “pinch and swell” structures of lens amphibolites in ductile micaschists, with an horizontal stretching direction and a greater axis N-S for the amphibolites beds (Figure 3). C and S (shearing and schistosity) structures in the micaschists are related (Figure 4); they show small folds and a top-to-north shearing direction.
The second area of study was the “Pointe des Chats” area, wich is at the SE of the island. This place is mainly composed of blue schists (glaucophanites) occurences. The association of glaucophane (centimetric) with garnet and phengite is well-marked and glaucophanes lineations were observed (N/S stretching direction). Towars the west, metamorphics changed from glaucophanites to prasinites rocks (Figure 5). Prasinites belong to the green schists facies and the classical paragenesis is chlorite-albite-magnetite, wich formed small millimetric to centimetric beds. Thus, the two parageneses were observed in close contact.
The Groix field trip ended in the small valley of Kerigant, in the South. In this place, sheat folds were seen with sheared veins of quartz in the micaschists (Figure 6). Their stretching direction is N-S, the same as rods of quartz seen higher on the cliff.

Interpretations

1) Nature of protoliths
The various parageneses observed conduce to question the original protolith of these metamorphic rocks. High pressure micaschists with a grt-cld-pg mineralogy shows a fine schistosity and can be interpreted as metapelitics rocks. The low pressure facies with chlorite and albite can result from secondary phases.These rocks seem to be in a retrograde pathway from high pressure cld-grt-riched micaschists to low pressure ab-qz-chl micaschists (or gneisses). If these rocks were fine detritical sediments in the past, the sediment source is continental. So the protoliths deposits might occur in a distal marine sequency or in a marginal basin.
The occurrences of blues schists or eclogites with the dominant glaucophane mineral are witnesses of the high pressure metabasic sequence of probably basaltic protoliths. Their bedding structures suggest that the protolith was not only magmatic rocks but also volcano-sedimentary rocks like volcanic sandstones (tuffites). The intercalation of metabasic in the metapelites rocks suggests that volcanic eruptions occurred near the place of sedimentation of the pelitics rocks. Thus, basic or volcano-sedimentary levels can be melted or intercalated with sedimentary beds. The occurrence of epidote and chlorite in these amphibolites rocks mark a secondary phase and a retrograde pathway similar than the metapelitic rocks.
Prasinite rocks are closely linked with the glaucophanites and the predominance of chlorite suggests a low pressure-low temperature metamorphism (green schist facies) and a basaltic protolith.

2) Interpretation of the strain markers
The common features of both tectonic structures is that strain is mainly due to shearing processes in a sub-horizontal plane with a N/S direction. Indeed, amphibolites lens or sheats folds show a N/S stretching direction. Mineral lineations of glaucophane show the same direction. The C and S structures seen at Porz-Melite clearly show a northern vergence for the shear. However, there is also a minor shortening component visible in many places. Folds of quartz levels in micaschists are visible in folds with sub-horizontal hinge or in the sheat folds. Lenses of amphibolites in a ductile micaschists also show a vertical shortening.
So sub-horizontal shearing strains are related to South to North displacements and are associated with a minor vertical shortening.

Discussion-Conclusion
The Groix island is as a whole formed by metamorphic rocks: the main part are meta-sedimentary rocks (micaschists ang paragneisses) intercalated with a rest of meta-volcanics (glaucophanites, eclogites and prasinites). These rocks are the remnants of a high pressure - low temperature metamorphism (blue schists facies), which occurred on continental sediments and arc volcanites at Devonian times (around 360 Ma). Then, parts of these rocks were recrystallised in green schist facies (prasinites).
Shearing processes structures such as C and S or sheats folds show stretching in a S-N direction. They can be interpreted to give the strain direction from South to North, associated with the high pressure metamorphism.
The geological regional context show that Groix is situated southward of the roots of the south armorican magmatic-metamorphic belt (leucogranites and amphibolites) which is dated at upper Carboniferous, around 300-310 Ma, age of the Hercynian structuring in this area. So, Groix island presents a pre-Hercynian metamorphic belt. In order to explain the paleo-association of volcanites with sedimentary pelitics rocks near the continental coast, it is possible to imagine two possibilities for the Groix formation.
First, the high pressure metamorphism with blue schists facies and shearing processes can indicate the northward subduction of an oceanic crust under the Armorican plate at Devonian times. In that case, parts of the oceanic crust could go down in the subduction trench with sedimentary rocks of an accretion prism. At depth, these rocks suffered a high pressure metamorphism with intense shearing processes. Then a rapid exhumation preserves them from high temperature.
The second hypothesis is that all the original material was in a back-arc basin context. Here, the beginning of the continental collision between Iberia and Armorica caused the closure of the basin and the "subduction" under Armorica (as the Karakorum subduction under the Asian plate before the collision with the indian plate). Then, huge parts of this basin material suffered various medium to high pressure metamorphism with intense shearing and could be exxhumed towards the South as in the Himalayas. But there are no remnants of a batholith with ages older than the Hercynian granites.
The origin of this subduction still remains unknown.