Geodynamic setting of volcano-plutonic rocks in so-called “paleo-accretionary prisms”: Fore-arc activity or post-collisional magmatism? The Shimanto belt as a case study

Abstract

The Cretaceous-Tertiary Shimanto belt of southwestern Japan is usually considered to be a simple accretionary prism (Taira et al., 1982; Ogawa, 1985), whereas, based on structural and sedimentological evidences, Charvet and Fabbri (1987) and Charvet et al. (1990) proposed an alternative model in which a collision with an unknown microblock followed the formation of the accretionary prism, and induced the Early Miocene main tectonism. One characteristic of the Shimanto belt is the occurrence of well developed magmatic bodies of Middle Miocene age (14±1 Ma) which can be used for testing the model. They show numerous peculiarities. (1) They consist predominantly of acidic volcano-plutonic complexes. Mafic sills or lavas are present in a lesser amount. (2) With respect to the supposed position of the subduction trench during Miocene time and to the northern location of the Setouchi volcanics, considered as a volcanic front ca 13 Ma, this magmatism is a near-trench magmatism. (3) Various tectono-magmatic affinities are present. The basaltic sills or lavas have mostly T-to E-MORB type affinities. An alkaline complex is present at Ashizuri cape. The most abundant magmatism consists of calc-alkaline volcano-plutonic bodies, typically peraluminous in the southern part of Shimanto and metaluminous near the northern border of the belt. (4) Metamorphic enclaves inter alia are often included in the calc-alkaline complexes. Their highest P- T conditions range between 0.7–0.8 GPa and 630–860°C. A simple subduction model is not able to explain all these features. The first necessity is to find a mechanism which develops a high thermal anomaly allowing the formation of magmas close to the trench. According to thermal modelling, the subduction of an oceanic plate, even young, does not provide enough heat. Secondly, in well known mature accretionary prisms (e.g. the Barbados one), the thickest part of the complex does not exceed 15 km. In such conditions, it is difficult to obtain pressures ranging from 0.7 to 0.8 GPa. This is much more difficult if we consider that some granitic rocks were formed in the youngest part of the so-called prism, in the thinnest portion. Finally, a simple subduction model does not explain the various magmatic affinities. Geochemical data show that the magmatic source of the Ashizuri alkaline complex is likely to be an enriched mantle comparable to OIB sources. On the other hand, the calc-alkaline rocks may have been derived by mixing between mantellic components and crustal ones, or in some cases by pure anatectic processes. The upwelling of a hot asthenospheric mantle, from which the Ashizuri suite could have been derived, may represent a heating source strong enough to make crustal materials melt. The proposed collision model could explain (1) the heating source, (2) the various magmatic affinities and also (3) the 0.7–0.8 GPa pressure invoking crustal thickening induced by the collisional event.