Petrogenesis of the post-collisional Eocene volcanic rocks from the Central Sakarya Zone (Northwestern Anatolia, Turkey): Implications for source characteristics, magma evolution, and tectonic setting

Abstract

Eocene volcanic rocks from the Central Sakarya Zone in the north of the İzmir-Ankara suture zone (IASZ) are predominantly intermediate-acidic lava flows and pyroclastics that crop out with a W-E orientation. The volcanic rocks include Bozaniç lavas, agglomerates, İğdir lavas, and Kapıkaya tuffs. The Bozaniç lavas contain plagioclase, hornblende, clinopyroxene, and biotite, whereas the İğdir lavas consist of plagioclase, hornblende, biotite, and quartz. According to the total alkali-silica (TAS) diagram, the Bozaniç lavas are mainly composed of andesite and dacite, with one sample of trachyandesite, whereas the İğdir lavas mainly contain dacite and minor andesite. Bozaniç and İğdir lavas exhibit moderate-to high-K calc-alkaline character. Major oxide and trace element variations suggest the effects of fractional crystallization in the evolution of the volcanic rocks. N-type midocean ridge basalt (MORB) normalized trace element patterns of the lavas exhibit enrichment in large ion lithophile elements (LILEs; K, Rb, Ba, Th) and depletion in high field strength elements (HFSEs; Zr, Ti, Y). In addition, the chondritenormalized rare earth element (REE) plots of the rocks show moderately enriched and nearly concave-shaped patterns (La N /Yb N = 5.4-17.6 for the Bozaniç lavas and 6.5-13.1 for the İğdir lavas), suggesting clinopyroxene (Cpx) and hornblende dominated fractionation. Negative Eu anomalies in the acidic lavas reveal plagioclase fractionation. Some trace element ratios of the lavas demonstrate a subduction signature and crustal contamination in the generation of the parent magma(s). Multi-dimensional tectonic discrimination diagrams suggest that the studied volcanic rocks have developed in a collisional setting.

Collision Chronology Along the İzmir‐Ankara‐Erzincan Suture Zone: Insights From the Sarıcakaya Basin, Western Anatolia

“…The two basement units in the Sakarya Zone are (1) Central Sakarya Basement, composed of Precambrian to Paleozoic crystalline basement, metamorphosed continental units called the Söğüt Metamorphics, and Carboniferous Söğüt granodiorites (also termed Söğüt magmatics, Central Sakarya granite, and Sarıcakaya granitoid) (Kibici et al, ; Okay, ; P. Ustaömer et al, ) and (2) the Permian‐Triassic subduction‐accretion Karakaya Complex composed of metamorphic rocks, ophiolitic mélange, deformed flysch, and limestone (Okay, ; Okay & Göncüoglu, ; T. Ustaömer et al, ). Basement rock units are exposed in south directed thrust sheets and overlain by a thick sequence of unmetamorphosed Jurassic‐Miocene sedimentary and volcanic rocks, which are well exposed in the Sarıcakaya and Central Sakarya Basins (Kasapoğlu et al, ; Ocakoğlu et al, ; Yildiz et al, ). North of the İAESZ and Karakaya Complex is the Sarıcakaya Basin (SB), a triangular sedimentary basin (~90 × 0 × 10 km) delimited to the north by the Söğüt Thrust (also termed Nallıhan Thrust), a south directed, basement‐involved thrust fault.…”

“…In most places in the SB, the thick Paleogene sedimentary sequence unconformably overlies the Karakaya Complex (Figure ; Gedik & Aksay, ). The basal Paleogene unit is dated from Early Paleocene to Middle Eocene and given various names: the Mihalgazi Formation, including the Çamaklı Member (Yildiz et al, ), Kabalar Formation, and Kızılçay Member/Group (Gedik & Aksay, ; Kasapoğlu et al, ). Uğuz () briefly describes and interprets this unit as debris flow, alluvial fan, flood plain, and lacustrine limestone facies.…”

“…The tectonic setting and geodynamic implications of postcollisional magmatism is debated. There are three belts of postcollisional magmatism in western Anatolia: (1) 54‐ to 48‐Ma granodiorite plutons and associated HP/LT metamorphism in the Tavşanlı Zone (Figure ) (Altunkaynak , ; Dilek & Altunkaynak , ; Harris et al, ; Okay & Satir , ; Özdamar et al, ), (2) 53‐ to 47‐Ma basaltic to rhyolitic Nallıhan volcanic rocks in the SB and eastern CSB (Figure ; Harris et al, ; Kasapoğlu et al, ; Yildiz et al, ), and (3) the Kizderbent belt of 53‐ to 38‐Ma volcano‐plutonic rocks ~80‐km north of the İAESZ (Altunkaynak , ; Ersoy, Akal, et al, ; Ersoy, Palmer, et al, ). Because the Kizderbent volcanic rocks exhibit different geochemical properties (Ersoy, Palmer, et al, ) and could belong to a different subduction system (Okay & Satir , ), we only consider the Nallıhan and Tavşanlı igneous rocks.…”

“…Whole rock major and trace elements and Sr–Nd–Pb isotopic compositions of the Nallıhan rocks indicate some fractional crystallization and crustal contamination, and mixture modeling supports increased asthenospheric (Ersoy, Palmer, et al, ) or metasomatized mantle lithospheric sources (Altunkaynak , ). The geochemistry, in addition to the linear geometry of the volcano‐plutonic belt, the late Paleocene‐early Eocene transition from marine to terrestrial deposition (e.g., Okay et al, ), and a possible relict slab interpreted from seismic tomography (e.g., Portner et al, ; van Hinsbergen et al, ) have led many authors to conclude that slab breakoff or delamination processes generated the volcanism (Altunkaynak, ; Altunkaynak et al, ; Ersoy, Akal, et al, ; Kasapoğlu et al, ; Yildiz et al, ). However, P. Ustaömer et al () outlined the inconsistencies in the slab breakoff hypothesis, most notably that magmatism from slab breakoff or delamination is emplaced into thickened and elevated crust, yet there was Early‐Middle Eocene marine deposition in the three foreland basins.…”

“…In western Anatolia, the timing of Pontide‐ATB collision is poorly resolved, sometime in the Late Cretaceous, Paleocene, or Early Eocene (e.g., Pourteau et al, ). Eocene slab breakoff is inferred from geochemical data (e.g., Altunkaynak, ; Dilek & Altunkaynak, ; Ersoy, Akal, et al, ; Ersoy, Palmer, et al, ; Harris et al, ; Kasapoğlu et al, ; Yildiz et al, ) but is either not supported or unresolved in mantle tomography (e.g., Portner et al, ; van Hinsbergen et al, ) or topographic expression (e.g., Okay et al, ; P. Ustaömer et al, ). Models consistent with collision or slab breakoff can be tested by examining the stratigraphic record against the expected crustal response.…”