The Strandja Massif and the İstanbul Zone were once parts of the same palaeotectonic unit: new data from Triassic detrital zircons

ABSTRACT Spatially continuous rock assemblages that share similar environmental evolution or structural features can be classified as a single tectonic unit. This approach enables to link dispersed units or massifs with each other and sometimes can be subjective, depending on the classification criteria. The relationship and the nature of the contact between the Strandja Massif and the İstanbul Zone have been controversial due to the Cainozoic cover. Amalgamation of these units was claimed as early as the Aptian-Albian. Lower Triassic sedimentary rocks, which are overlain by the Carboniferous flysch with a N-verging thrust fault are exposed NW of the İstanbul Zone. This study reveals the spatial relationship between the Strandja Massif and the İstanbul Zone deduced from the U-Pb dating and Lu-Hf isotopes of the detrital zircons from these Lower Triassic clastics. Our results show that the early Triassic basin was fed from a provenance that included arc-related Upper Carboniferous-Lower Permian magmatic rocks which is much more likely to be the Strandja Massif than the İstanbul Zone. The second outcome of this study is that a unit that previously assigned to Palaeozoic turned out to be Triassic, which brings the Strandja Massif farther to the east, into the northern İstanbul Zone.


Introduction
The Rhodope-Pontide Fragment (Şengör & Yilmaz, 1981) is the northernmost part of the Alpides in Anatolia. It stretches from the eastern Carpathians in the west (Burchfiel, 1980) to the Caucasus in the east (Şengör & Yilmaz, 1981), embracing the Moesian Platform form the west and extending along the southern Black Sea coast. It is separated from the southern tectonic units in Turkey by the northern branch of the Neo-Tethyan suture (the İzmir-Ankara-Erzincan suture, Ketin, 1966;Şengör & Yilmaz, 1981). According to Şengör and Yilmaz (1981) it essentially consists of four tectonic units that amalgamated before the late Cretaceous, from west to east; the Rhodope Massif, the Strandja Massif, the İstanbul Zone and the Eastern Pontides. The main difference in its definition is that Ketin includes the Sakarya Continent into his Pontides whereas Şengör and Yilmaz (1981) separates the Sakarya Continent from the İstanbul Zone by the Intra-Pontide suture. Okay, Şengör, and Görür (1994) considered the Eastern Pontides as the eastern continuation of the Sakarya Zone, an interpretation not here followed. Within the Rhodope-Pontide Fragment the Strandja Massif and the İstanbul Zone form the northern part of Ketin's (1966) Western Pontides.
The basement of the İstanbul Zone consists of Neoproterozoic (here Pan-African) metamorphic rocks (Chen et al., 2002). This basement is overlain by a continuous, well-developed Atlantic-type sedimentary sequence extending from the Lower Ordovician to the Upper Devonian. The basin, which was tectonically stable from the Ordovician to the end of the Devonian, became tectonically active and turned into a site of turbiditic flysch deposition during the Early Carboniferous. The Carboniferous flysch has long been interpreted to mark the progress of a collision (Görür et al., 1997;Şengör & Özgül, 2010). The Palaeozoic sequence is unconformably overlain by late Permian and younger sedimentary strata (Özgül, 2012). The geology of the İstanbul Zone resembles that of the Hercynides with its abundant Lower Carboniferous flysch deposits passing into Upper Carboniferous molasses carrying with a sharp unconformity upon its sediments.
The relationship between the İstanbul Zone and the Strandja Massif and their contact have long been a matter of discussion. Early publications suspected an overthrust contact between these units (Şengör, 1984; Şengör et al., 1984) and interpreted amalgamation through a suture zone (Yilmaz, 1977) whereas Okay et al. (1994) considered a right lateral strike-slip fault between them and brought the İstanbul Zone during the opening of the Black Sea from the north, where the Moesian platform is exposed. The contact between them concealed under the sediments of the Trace basin somewhere in the west of İstanbul, in the Çatalca region. The nature of the contact thus remains uncertain.
Northwest of Bosphorus, in a very limited area, a Triassic sequence crops out (Figure 1; Kaya & Lys, 1979-1980. The base of this sequence consists of Lower Triassic massive basaltic lava flows that are overlain by Upper Triassic (Scythian) white to light grey, thicklybedded Çiftalan Sandstone, which was investigated in this study. The overlying late Scythian-Anisian Köseler Limestone passes into Bakırlıkıran Formation, represented by sandstone and shale.
The Triassic units exposed in the northwestern part of İstanbul are thrust by Carboniferous flysch of the İstanbul Zone (Gedik, Timur, Duru, & Pehlivan, 2005;Kaya & Lys, 1979-1980Özgül, 2011). Age of thrusting remains unclear: it could be medial or late Mesozoic related to the final closure of the Palaeo-Tethys here.
In this study we examined the ages of the detrital zircons obtained from the Triassic rocks in northwest İstanbul and correlated them with the known age data reported from the İstanbul Zone and the Strandja Massif. According to our provenance analyses of the Triassic rocks we came to a conclusion that they were fed from the Strandja Massif rather than the İstanbul Zone. Thus, the Triassic rocks can be attributed to the Strandja Massif and the buried contact between these two units in the NW part of İstanbul must be thrust fault as it is in the exposed area ( Figure 3). The second result of the present study is the necessity of reexamination of some clastic rocks of the İstanbul Zone attributed to Palaeozoic from which there have so far been no palaeontological data. We here show that one such  allegedly Silurian unit turned out to be at the oldest post-Carboniferous!

Analytical techniques
Standard mineral separation techniques, which include grinding, sieving, Frantz isodynamic separator and heavy liquids were used for zircon extraction from rock samples. Extracted zircons were handpicked under binocular microscope and mounted. Due to low number of the extracted zircon grains all of the 161 zircons between 63-180 µm are used for dating. Cathodoluminescence (CL) images of the polished zircons were taken in Hacettepe University by a Zeiss Evo 50 EP scanning electron microscope.
Uranium, thorium and lead isotope analyzes were carried out by laser ablationinductively coupled plasmamass spectrometry (LA-ICP-MS) at the Goethe University of Frankfurt (GUF), using a slightly modified method, as previously described in Zeh (2006, 2009)) and Zeh and Gerdes (2012). A ThermoScientific Element 2 sector field ICP-MS was coupled to a Resolution S-155 (Resonetics) 193 nm ArF Excimer laser (CompexPro 102, Coherent) equipped with two-volume ablation cell (Laurin Technic, Australia). Laser spots of 20-30 µm were fired with 5.5 Hz at a fluence of about 1.5 J cm −2 . Raw data were corrected offline for background signal, common Pb, laser induced elemental fractionation, instrumental mass discrimination, and time-dependent elemental fractionation of Pb/U using an in-house MS Excel © spreadsheet program (Gerdes & Zeh, 2006, 2009. Laser-induced elemental fractionation and instrumental mass discrimination were corrected by normalization to the reference zircon GJ-1 (0.0982 ± 0.0003; ID-TIMS GUF value). Repeated analyses of the reference zircon Plesovice, BB-16 and 91,500 (Santos et al., 2017;Sláma et al., 2008;Wiedenbeck et al., 1995) during the same analytical session yielded an accuracy of better 1% and a reproducibility of <2% (2 SD). All uncertainties are reported at the 2sigma level. The analytical results are presented in Table 1 (supplementary material). Concordia diagrams are prepared by using ISOPLOT spreadsheet (Ludwig, 2003) and AgeDisplay (Sircombe, 2004) were used to prepare the probability density distribution (PDD) and the histograms.
Hafnium isotope measurements carried by using Thermo-Finningan Neptune multi collector ICP-MS coupled to the same laser system as for U-Pb analysis following the method described in Zeh (2006, 2009)). Spots of 40 µm in diameter were drilled with a repetition rate of 5.5 Hz and an energy density of 5 J/cm 2 during 50s of data acquisition. All data were adjusted relative to the JMC475 of 176 Hf/ 177 Hf ratio = 0.282160 and quoted uncertainties are quadratic additions of the within run precision of each analysis and the reproducibility of the JMC475 (2SD = 0.0028%, n = 8). Accuracy of the Hf isotope measurements were verified by repeated analyses of reference zircon GJ-1 and Temora, which yielded a 176 Hf/ 177 Hf of 0.282008 ± 0.000019 (2 SD, n = 62) and 0.282685 ± 0.000022 (n = 14), respectively. This is in perfect agreement with the LA-MC-ICPMS longterm average of GJ-1 (0.282010 ± 0.000025; n > 800) and Temora (0.282483 ± 0.000023, n > 250) reference zircon at GUF.

Description of samples
Two samples were collected from the western side of the Bosphorusşo; the sample 107-b was taken from near the Black Sea coast, close to Gümüşdere village and the second one, the sample 139, was taken from NW coast of the Bosphorus, near Sarıyer village ( Figure 3). The sample 107-b was collected from below the thrust fault boundary of the Lower Triassic Çiftalan sandstone (Kaya & Lys, 1979-1980 with Carboniferous flysch of the Trakya Formation ( Figure 3). The sample is white to light grey sandstone (subarkose) with predominantly quartz, lesser plagioclase, alkali feldspar and minor lithic grains. Sum of 132 zircon grains were extracted from this sample.
The sample 139 is yellow to light brown, greenishgrey, fine-grained sandstone and siltstone alternation of so-called Upper Ordovician-Lower Silurian Gözdağ Member of the Yayalar Formation (Özgül, 2012). Only 30 zircons could be extracted from this sample. Figure 4 shows cathodoluminescence (CL) images of the selected zircons representing the peak ages in the probability distribution diagrams illustrated in Figure 5. Zircon grains from the sample 107-b are colourless, transparent and often idiomorphic. Their sizes range from 70 and 150 µm. Grains have aspect ratios between 1:1.5 and 1:4. The dominant structures in the grains are well developed oscillatory magmatic zoning, which gets sometime blurred. Other grains show resorption and CL bright overgrowth rims (e.g. grains A425, A419 and A437) implying an overprinting metamorphic event.

Zircon external and internal structures
Zircon grains obtained from the sample 139 are also colourless, transparent and mostly idiomorphic but their sizes are smaller (between 70 µm and 100 µm). Many grains are slightly rounded and their aspect ratios range from 1:1.2 and 1:2 While the majority of the crystals show normal oscillatory zoning some grains display in addition sector zoning (e.g. A12, A39, A29), either. Compared to sample 107-b zircon grains, the oscillatory zoning is wider (e.g. A13, 28 and 40).
Both samples are poor in terms of inheritance. Core structure is rare and roundness is not well expressed, which requires short transport and less reworking. Dating spots have been defined in places where zircons have oscillatory magmatic zoning (Figure 4). During the identification of the laser spots, we try to avoid placing them into mixed domains. Therefore, we believe that all ages obtained represent real geological events. Especially magmatic character of the late Carboniferous ages spots is obvious in the CL images of the zircons from sample 107-b (Figure 4).

U-Pb ages and Lu-Hf isotopes of the clastics
The Triassic is an important time interval for both the İstanbul Zone and the Strandja Massif. In the Strandja Massif 2 different Triassic successions defined; Triassic volcanic and sedimentary rocks (Chatalov, 1990(Chatalov, , 1991 and metaclastics with minor volcanics (Çağlayan & Yurtsever, 1998;Gočev, 1976Gočev, , 1979Natal'in et al., 2016Natal'in et al., , 2012Okay et al., 2001). Furthermore, the İstanbul Zone also displays two different yet correlatable Triassic successions; one in the west, to the NW of İstanbul and the second one in the east, in the Kocaeli region. First, we would like to discuss the age of the deposition of the rocks (Triassic and Silurian?) studied. The age of the Çiftalan Sandstone, exposed in the İstanbul area, previously reported as Scythian (Kaya & Lys, 1979-1980 based on the paleontological data from overlying Anisian limestone. The youngest detrital zircon age obtained from Çiftalan sandstone in this study is 167 ± 3 Ma (Middle Jurassic), but internal structure of this zircon is complex and the age has low concordance value (91%). Therefore, we consider the second youngest age 240 ± 3 Ma (Anisian) as more reliable to establish the maximum depositional age for this unit. Sample 139 was collected from a rock unit, which had been assigned to the Upper Ordovician-Lower Silurian Gözdağ Member of the Yayalar Formation (Özgül, 2012 and references herein). However, 8 of the 30 zircon ages gave late Carboniferous and younger ages and the youngest    (Okay et al., 2011;Ustaömer et al., 2011;Yılmaz Şahin et al., 2010) and the pre-Cambrian orthgneisses and the Palaeozoic basement schists of the Strandja Massif (Natal'in et al., 2016;Şahin et al., 2014;Sunal et al., 2006Sunal et al., , 2008 (Figure 8). It should be mentioned that we have only 25 ages from the sample 139 and it is difficult to be sure about the satisfactory comparison between it and the sample 107-b. Furthermore, ages obtained from these two samples confirm that although there are similarities in their Carboniferous and Permian ages, their εHf(t) values differ considerably revealing distinct origins (see Figure 7).
From the provenance perspective, the first similarity between the unit studied here and the Strandja Massif is the presence of a large number of Cambrian ages which represents the Pan-African events. No The Cambrian ages were described from the basement of the İstanbul Zone. There are only ages from detrital zircon grains (6%) (Okay et al., 2011;Ustaömer et al., 2011). However, Cambrian ages (544 Ma) were reported both from the orthogneisses and the detrital zircons of the Strandja Massif (Natal'in et al., 2016;Sunal et al., 2008). The Ediacaran ages of the İstanbul Zone has a higher percentage than in the Triassic rocks studied here, which shows that the Ediacaran was not the main source for the Triassic rocks. Besides, Ediacaran rocks were described both from the Strandja Massif (Natal'in et al., 2016) and from the basement of the İstanbul Zone (Chen et al., 2002).
Detrital zircon ages and their distribution show that Triassic clastics represented by the sample 107b could not have been fed from the İstanbul Zone because of the presence of Upper Carboniferous representing Hercynian orogenic events and Permian ages. In contrast, it perfectly matches with the Strandja Massif when this interval plus distribution of the ages older than early Carboniferous are considered. Figure 9 shows the percentage of age distributions in specific geological time periods for ages older than the late Carboniferous for the sample 107-b, the Strandja Massif and the İstanbul Zone. Even if the late Carboniferous and Permian ages are excluded from the Triassic clastics of the Çiftalan sandstone, pie charts clearly show close similarity with the Strandja Massif rather than with the İstanbul Zone. Almost continuous age distribution from the Ordovician to the early Permian require a source area or areas with synchronous arc magmatism, which was already proposed for the Strandja Massif by Natal'in et al. (2012). However, there is no direct indication of magmatism between the Ordovician and the late Carboniferous. Only some detrital ages from the Lower Carboniferous unit reveal that there are early Carboniferous zircons supplied to the basin (Okay et al., 2011).
Zircon internal structures for both of the samples indicate that source area of the zircons are magmatic. Zircon crystals are prismatic and reveal good magmatic oscillatory zoning. In addition, high Th/U ratios support a magmatic origin for the zircons. Presence of resorption structures in the rims of the dated Ordovician zircon grains and their absence in the Carboniferous ones indicate that they came from a source that was metamorphosed sometime between the Ordovician and the Carboniferous.
In Şengör et al. (1984), the İstanbul Zone is considered as a northward thrusting nappe over the Strandja Massif. Kaya andLys (1979-1980) indicate that Carboniferous flysch of İstanbul Zone is thrust over the Triassic rocks and our personal observations also confirm this relationship. Figure 10 illustrates thrusting of the İstanbul Zone onto the Strandja Massif as suggested by Şengör et al. (1984).
Sandstone and siltstone alternation of the alleged Silurian west of Sarıyer is definitely not a part of the İstanbul Palaeozoic sequence. Both the sample 139 and the sample 107-b represent of Upper Carboniferous and Permian rocks. However, as pointed out above, the εHf(t) values of the two samples differ from each other. εHf(t) values of the sample 107-b mainly grouped around the CHUR line in the diagram (Figure 7) indicating that zircons transformed from rocks that are products of a continental margin arc magmatism where mature and juvenile crustal material mixed. In contrast, zircons from the sample 139 mainly show positive εHf(t) values for Permian and Ediacaran intervals. This data show that zircons collected from the sample 139 must be a product of a juvenile crust. Model ages of these two samples also show differences. The sample 139 has younger model ages such as 0,65 Ga, whereas the youngest model age in the sample 107-b is 0,92 Ga. Most probably, source areas of these two units were different.
(2) Analyzed detrital zircon ages show that provenance of the early Triassic units in the western part of the İstanbul Zone is a continuation of the Strandja Massif.
(3) In this region boundary between the İstanbul Zone and the Strandja Massif is tectonic. The İstanbul Zone was thrust over the Strandja Massif sometime after the Triassic. The age of thrusting remains unclear, but was most likely medial to late Mesozoic. (4) Before the Triassic the İstanbul Zone was located to the south of the Strandja Massif before its movement to north (present geographical positions).  We sincerely thank to Timur Ustaömer for constructive comments on U-Pb dating and Lu-Hf analyses. SCÜ thanks to Zehra Deveci Aral for her support on the evaluation of the data, Richard Albert Ropert and Linda Marko for there helps during isotope analyses and data processing and collaboration at GUF.

Disclosure statement
No potential conflict of interest was reported by the authors.