International workshop on the tectonics of Eastern Turkey and the Northern Arabian Plate in Erzurum, Turkey, Turkey, 1 - 04 September 2002, pp.16-17
The Eastern Anatolia region contains one of the most complete and best-exposed sequence of collision-related volcanic rocks in an active collision terrain. Radiometric dating results indicate that volcanism started during the Middle Miocene (11-13 Ma) and continued until Quaternary time. The volcanic products cover almost two third of the region, exceeding 1 km in thickness in places. This volume, in fact, should reflect only a fraction of the melt generated in Eastern Anatolian collision zone, since a greater proportion is presumed to have emplaced deep in the crust as plutonic intrusions. This indicates that huge volumes of melt were generated over a relatively short period of time (~12 My) beneath Eastern Anatolia. Although still debated, there is now a general consensus that collision started at around Late Eocene (~40 Ma), predating the regional block uplift taken place in the Middle Miocene (~12-15 Ma), by about 25 My. Initiation of collision-related volcanism (CRV) was coincident with rapid uplift that created a fundamental palaeogeographic, sedimentologic and tectonic change all over the region. This remarkable relationship may imply that what caused the uplift should also have controlled the magma generation at depth. Therefore, collision-related volcanic units appear to have the potential of recording some important data about the lithospheric evolution of the region. In fact, in most places, they are the only material that contain the geological record, since they mask older formations over a very large area.
Volcanological and Geochemical studies on the CRV to date have revealed a number of important issues as follows: (1) The CRV is dominated by pyroclastic products produced by explosive eruptions, implying that magmas were rich in H2O. (2) They span the full compositional range from basalts to high-silica rhyolites. (3) The volcanic products have trace element and isotope ratios, indicating an “enriched lithospheric mantle source”, that contains no subduction component (SC) in the south but displays a distinct SC in the north around Erzurum. This component is thought to have been inherited from a previous subduction event (i.e. Pontide subduction). (4) The existence of a SC in the N in the absence of a subducting slab is also consistent with lithospheric origin rather than asthenospheric origin, because SC cannot be retained in the asthenosphere for a long time due to convection. (5) The lavas contain variable crustal contamination (i.e. r<0.4) via assimilation combined with fractional crystallisation (AFC) process but “none of them are crustal melts”. (6) Volcanic activity is not contemporaneous in the north around the Erzurum-Kars Plateau, being earlier in the West than in the East. This may be related to the eastward migration of pull-apart basins created by strike-slip fault systems which provided fractures that enable magma to reach the surface.
Temporal relationship between regional uplift and magma genesis coupled with petrologic findings about source characteristics and the magma evolution of the CRV indicate that “catastrophic delamination of the sub-continental lithospheric mantle” (first proposed and thoroughly discussed by Pearce et al., 1990 for the region and further refined by Keskin et al., 1998) appears to be the most viable mechanism for the Eastern Anatolian collision zone. This hypothesis holds that the base of the sub-continental lithosphere becomes detached along the thermal boundary layer because of its greater density relative to the asthenosphere. Not only this causes the hotter asthenosphere to rise and conductively heat and hence melt lithospheric mantle, but also does it create a sudden block uplift which is exactly the case for Eastern Anatolia. There exist inconsistencies in all other alternative hypotheses: (1) decompression melting of asthenosphere on a regional scale is inconsistent with the presence of a thick crust and geochemistry of the CRV; (2) hot spot activity is inconsistent with the nature and timing of regional uplift; and (3) renewed subduction is not supported by seismic evidence.