Devonian magmatism in the western Sakarya Zone, Karacabey region, NW Turkey

Abstract

The Karacabey Pluton is a large magmatic body in the northwestern Turkey overthrust by the probable Triassic metamorphic rocks of the Lower Karakaya Complex. Both the metamorphic rocks and the Karacabey Pluton are unconformably overlain by a Lower Jurassic and younger sedimentary sequence. The Karacabey Pluton was regarded as a Carboniferous intrusion based on the previous K-Ar biotite geochronological data. Here, we provide new geological, geochemical and geochronological data from the Karacabey Pluton. Zircon U-Pb results from two samples yielded ages of 393.8 +/−2.7 to 395.9 +/−4.09 Ma, suggesting that the granitoids intruded in the crust throughout the Early to Middle Devonian. The Karacabey Pluton consists mainly of biotite and locally hornblende bearing granitoid with lesser amounts of S-type leucocratic granodiorite, all of which are cut by pegmatitic bodies. It belongs to the high-K calc-alkaline series with distinct Nb and Ta anomalies in multi-element spider diagram. Sr and Nd isotopes’ initial values are 0.709–0.712 and 0.511–0.512, respectively. εNd(i) values range between −7.8 and −9.4. The isotopic characteristics of the rocks indicate lower crustal sources of both metapelitic and metaigneous origin. Geochemical features of the rocks suggest that they developed in an arc-related environment, along with the other Devonian granitoids described from the Biga Peninsula in northwest Turkey. The granitoid shows a low-temperature alteration/metamorphism marked by recrystallization of quartz, sericitization of the feldspar and formation of late chlorite, epidote and muscovite. Possibly because of these, the Ar–Ar biotite ages are scattered with a possible concentration at around Permo–Carboniferous boundary. Zircon (U-Th)/He geochronology suggests that after the granitoid was reburied during Early Jurassic to Early Cretaceous sedimentation, there was renewed uplift and erosion during the Late Cretaceous (Turonian), which is possibly related to the closure of the Intra-Pontide Ocean in the north.

Keywords:

• Karacabey Pluton
• Sakarya Zone
• geochemistry
• zircon U-Pb
• biotite Ar–Ar and zircon (U-Th)/He ages

1. Introduction

The Sakarya Zone of the Pontides is an E-W trending belt bounded by Istanbul Zone in the north and by the Izmir-Ankara Suture in the south (Figures 1 and 2). It consists of a complex basement overlain by Jurassic and younger sedimentary sequence. The basement rocks of the Sakarya Zone can be divided into two separate parts: 1. Paleozoic plutonic and metamorphic rocks possibly related to the Variscan orogeny and 2. Triassic accretionary complexes with blueschists and eclogites, called as the Karakaya Complex (Okay & Göncüoğlu, 2004; Okay & Tüysüz, 1999; Okay, Satır & Siebel, 2006).

Figure 1. Distribution of the Paleo-Tethyan accretionary complex and Variscan basements in the Pontides, N Turkey (modified from Okay & Göncüoğlu, 2004). Distribution of the Variscan magmatism is adopted from; 1- Okay et al., 1996 and Okay et al. (2006), 2- Aysal et al. (2012), 3- this study, 4- personal unpublished data, 5- Ustaömer et al. (2012), Nzegge et al. (2006), 6- Topuz et al. (2010); Dokuz (2011), 7- Ustaömer et al. (2012b).

Figure 2. The geological map of the Biga Peninsula (slightly modified from Altunkaynak et al. (2012) and the references therein).

Most of the plutonic rocks in the basement of the Sakarya Zone are of Carboniferous age (350–320 Ma, Figure 1). These can traced from the Kazdağ Massif in the Biga peninsula (308–329 Ma, Okay et al., 1996, 2006), through west of Ankara (~290 Ma, Okay et al., 2006, 319–327 Ma; Ustaömer, Ustaömer & Robertson, 2012) and Central Pontides (275–295 Ma, Nzegge, 2008; Nzegge, Satır, Siebel & Taubald, 2006) to the Eastern Pontides (308–325 Ma, Dokuz, 2011; Kaygusuz, Arslan, Siebel, Sipahi & İlbeyli, 2012; Topuz et al., 2010; Ustaömer, Robertson, Ustaömer, Gerdes & Peytcheva, 2012). These granitoids are often associated with high-temperature metamorphic rocks. The best studied of the Carboniferous metamorphic complexes are those of the Pulur region, characterized by partial melting and sillimanite–cordierite–garnet mineral paragenesis, which have yielded 331–327 Ma monazite ages (Topuz, Altherr, Satır, Schwarz & Sadıklar, 2001). In contrast to the widespread Carboniferous granitoids, Devonian granitoids are rare and are apparently restricted to the western part of the Sakarya Zone. They were first reported from the southern part of the Biga Peninsula (Figure 1). This Çamlık Granitoid gave zircon Pb-Pb ages of 399–397 Ma (Okay et al., 1996, 2006). Aysal et al. (2012) showed that some of the other granitoids in the northern part of the Biga Peninsula are also of Devonian age (401–389 Ma) and provided geochemical data showing that they are related to subduction or delamination and/or slab-breakoff models. Here, we show that the Devonian granitoids extend farther east toward Bursa. We provide new geological, isotopic and geochemical data from the Devonian Karacabey Pluton from west of Bursa (Figures 1 and 2), which was previously considered of Permian age on the basis of K-Ar biotite data (Delaloye & Bingöl, 2000). We also chart its subsequent thermal history through new Ar–Ar biotite and zircon U-Th/He ages.

2. Geological setting

The region between Karacabey and the Sea of Marmara is constituted of a thick series of metamorphic rocks of the Lower Karakaya Complex. The metamorphic rocks are dominated by metabasites (more than 80% of the sequence) with lesser amounts of phyllite and marble and a few lenses of serpentinite. The metamorphism is in greenschist facies with actinolite/hornblende + albite + epidote + chlorite assemblage in the metabasites. The metabasites include at least one tectonic lens of Triassic eclogite dated at 205 ± 3 Ma (Okay & Monié, 1997). In the south, the metamorphic rocks lie tectonically above a granitoid, the Karacabey Pluton, which makes a 6 km long and 2.5 km wide body. The Karacabey Pluton is unconformably overlain by Lower Jurassic sandstone and conglomerates, which pass up into a thick sequence of Middle-Upper Jurassic to Lower Cretaceous marine limestones (Figure 2, Altıner, 1991; Ergül et al., 1986). The limestones are conformably overlain by a thin horizon of red mudstones, which includes clasts derived from underlying carbonates, which passes up into a thick sequence of turbidites. The sedimentary section ends with thin-bedded cherts (Figure 3).

Figure 3. Geology of the Karacabey region (modified from Ergül et al., 1986). Note the uneven trace of the contact between Permo-Triassic metamorphic rocks and the Karacabey Pluton, which is interpreted by us as a thrust.

North of the main body of the Karacabey Pluton, a small stock of granodiorite porphyry intrudes the metamorphic rocks of the Lower Karakaya Complex. This shallow-level intrusion, which was considered as part of the Karacabey Pluton in previous studies (Ergül et al., 1986), is of Early Eocene age (personal unpublished data).

3. Lithology and petrography of the Karacabey Pluton

The country rocks, in which the Karacabey Pluton intruded, are only observed north of the village of Kıranlar (Figure 3). They consist of intercalated gneisses and schists; the gneisses are light gray, banded and consist mainly of 0.5–1.0 cm long porphyroclasts of feldspar and quartz. The schists are gray, foliated and fine grained (<1 mm). The mineral assemblage in the gneisses and schists is quartz, K-feldspar, plagioclase, biotite, chlorite and minor muscovite. The metamorphic rocks are cut by pegmatitic veins that most probably related to the magmatics of the Karacabey Pluton.

The Karacabey Pluton includes a range of different rock types. The common rock type is biotite-rich granodiorite consisting of quartz (~23%), plagioclase (~42%), K-feldspar (~23%), biotite (3–10%), ±hornblende (<5%) and minor ± sphene, apatite, zircon and opaque minerals. Secondary minerals are epidote, chlorite and muscovite. Biotites are largely replaced by chlorite, and secondary muscovite has developed along the veins or/and cracks. Plagioclase is replaced by sericite. Hornblende and sphene are only observed in the western part of the complex (samples KG 07 and 17) (Figure 3). Granodiorites are mostly equigranular and medium grained (up to 5 mm), but sometimes are fine grained (~1 mm) along 2–5 m thick zones. The granitoid is massive and does not show any foliation.

Other rock types in the Karacabey Pluton are muscovite- and garnet-bearing leucocratic granite and pegmatitic veins. Muscovite- and garnet-bearing leucocratic granite is only exposed north of Kıranlar Village (Figure 3). It contains quartz, K-feldspar, plagioclase, muscovite, garnet and minor opaque minerals. Pegmatitic veins cut both the biotite-rich granodiorite and the leucocratic granite indicating that they are the last phase of the magmatism in the region. The muscovite- and garnet-bearing leucocratic granite is not analyzed in this study because of its small exposure and scarce outcrops in the field. Contact relationship with other rock types is not also obvious.

A low-grade alteration and metamorphism in the Karacabey Pluton are shown by recrystallized quartz, strained biotite flakes and secondary chlorite, epidote and muscovite (Figure 4). However, no development of cleavage and lineation is observed in the Karacabey Pluton. A gneissic texture is observed in a few places, but it is not clear whether it is metamorphic or formed during the intrusion. A late cataclastic fabric is common in the Karacabey Pluton, which is cut by late veins filled by quartz and muscovite.

Figure 4. Cross-polarized thin-section photographs of; a- Hornblende bearing granodiorites (KG 07 and 17) and b- Biotite rich granodiorites (KG 27 and 28).

3. Analytical techniques

3.1. Whole-rock analyses

The whole-rock powders were split from 1–5 kg of crushed rocks. Whole-rock analyses were performed at Acme Analytical Laboratories Ltd. in Vancouver, Canada. Two hundred milligrams of rock powder was mixed with 1.5 g of LiBO₂ flux in a graphite crucible. Subsequently, the crucible was placed in an oven and heated to 1050 °C for 15 min. The molten samples were dissolved in 5% HNO₃ (ACS-grade nitric acid diluted in demineralized water). International reference samples and reagent blanks were added to the sample sequence. For analyses of major elements and the trace elements Ba, Nb, Ni, Sr, Sc, Y and Zr, sample solutions were aspirated into an ICP emission spectrograph (Jarrel Ash AtomComb 975). For the determination of other trace elements including rare earth elements, the solutions were aspirated into an ICP mass spectrometer (PerkinElmer Elan 6000). Accuracy for major elements is better than 2% and for trace elements better than 10% relative.

3.2. Sr-Nd isotope analyses

Isotope ratio measurements were performed on the Finnigan MAT 262 TIMS located at the Isotope Geochemistry Group of the University of Tübingen (Germany). Sample powders were weighted into Teflon beakers and mixed with adequate amounts of tracer solutions enriched in ⁸⁷Rb-⁸⁴Sr and ¹⁴⁹Sm-¹⁵⁰Nd, respectively. Digestion was achieved via HF (48%) – HNO₃ (65%) acid attack by placing the closed beakers on a hot plate at 140 °C. Digested samples were dried and transformed to chlorides in 6 N HCl. The samples were then redissolved in 2.5 N HCl for chromatographic separation. Rb, Sr and light rare earth elements were purified by conventional ion exchange chromatography using quartz glass columns filled with BioRad AG 50 W-X12 (200–400 mesh). Subsequent separation of Sm and Nd was achieved using a reverse ion chromatographic procedure using quartz glass columns filled with HDEHP resin (Richard, Shimizu & Allegre, 1976). Sr separates were loaded with a Ta-activator on W single filaments and isotope ratio measurements were performed in dynamic mode. Rb was loaded on Re double filaments. Analytical mass fractionation was corrected using a ⁸⁸Sr/⁸⁶Sr ratio of 8.375209 and exponential law. Long-term reproducibility for NBS SRM 987 (n = 30) is 0.710248 ± 11 for ⁸⁷Sr/⁸⁶Sr-ratio. Total procedural blanks (chemistry and loading) were <450 pg for Sr and <15 pg for Rb. Sm and Nd were loaded as phosphate on Re double filaments and measured in static multiple collector mode. Analytical mass fractionation was corrected with ¹⁴⁶Nd/¹⁴⁴Nd of 0.7219 using exponential law. Repeated measurements of the La Jolla Nd standard (n = 12) gave a ¹⁴³Nd/¹⁴⁴Nd ratio of 0.511838 ± 13. Procedural blanks were <90 pg for Nd and <20 pg for Sm.

3.3. Zircon U-Pb LA-ICP-MS technique

Zircons were extracted from rock samples by standard mineral separation techniques, crushing, sieving, Frantz isodynamic magnetic separator, heavy liquids, and were finally handpicked under a binocular microscope. Then, a fraction with grain sizes 63–200 μm was classified according to crystal properties (i.e. euhedral morphology, lack of overgrowth and visible inclusions). The grains were mounted in epoxy, polished and imaged using cathodoluminescence (CL) on a Zeiss Evo 50 EP scanning electron microscope at the Hacettepe University in Ankara.

Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) was conducted at the University of California, Santa Barbara following Kylander-Clark, Hacker and Cottle (2013). The laser spot size, repetition rate and ablation rate were 30 μm, 4 Hz and ~0.1 μm/pulse, respectively. U-Th-Pb isotopic ratios and their uncertainties were calculated using Iolite (Paton et al., 2010), which corrects for downhole fractionation and machine drift using a primary reference material. The primary reference material was 91500 zircon; SL-1, GJ-1 and Plešovice zircon were used as a secondary reference materials and yielded ages within 2% of their accepted values. As such, we conservatively included 2% error in the final age calculation of our unknowns.

Final age calculations performed using computer program Isoplot 3.0 (Ludwig, 2003). TuffZirc way is defined for the age calculations. TuffZirc method is a statistical way to eliminate mixed ages that mainly caused by inherited cores or metamorphic zircon overgrowths (for details see program manual Isoplot 3.0, Ludwig, 2003).

3.4. Biotite Ar–Ar dating method

⁴⁰Ar/³⁹Ar dating was performed at the Open University at Milton Keynes, UK. Biotite separate was obtained by conventional techniques including crushing, sieving, magnetic and heavy liquid separation, hand picking and ultrasonic cleaning. All samples were irradiated at McMaster University in Canada. Irradiation flux was monitored using the GA1550 biotite standard with an age of 98.79 ± 0.54 Ma (Renne et al., 1998). Sample J values were calculated by linear interpolation between two bracketing standards; a standard was included between every 8–10 samples in the irradiation tube. J values are included in the full dataset presented in the Auxiliary material. Results were corrected for blanks measured either side of every two mineral analysis (average blanks are included in the Auxiliary material), ³⁷Ar decay and neutron-induced interference reactions. The correction factors used as follows: (³⁹Ar/³⁷Ar) Ca = 0.00065, (³⁶Ar/³⁷Ar)Ca = 0.000265 and (⁴⁰Ar/³⁹Ar)K = 0.0085 were based on analyses of Ca and K salts. Analyses were also corrected for mass spectrometer discrimination (283).

Samples were loaded into an ultra-high vacuum laser port and placed under a heat lamp for 8 h to reduce atmospheric blank levels. Total fusion of single grains was achieved using a continuous wave SPI infrared (IR) laser (1050–1250 nm) coupled to an automated gas handling vacuum system and admitted into an MAP 215–50 noble gas mass spectrometer.

3.5. Zircon (U-Th)/He dating method

Single-crystal aliquots were dated, usually 3 aliquots per sample. The crystals were selected carefully; only fissure-free specimens were used, with well-defined completely convex external morphology. Euhedral crystals were preferred. The shape parameters were determined and archived by multiple digital microphotographs. In case of zircon crystals besides length and widths, the proportion of the length of the prismatic and pyramidal zones was also considered.

The crystals were wrapped in platinum capsules of ca. 1 × 1 mm size. The Pt capsules were heated up by an infra-red laser (7 min for zircon). The extracted gas was purified using a SAES Ti-Zr getter at 450 °C. The chemically inert noble gases and a minor amount of other rest gases were then expanded into a Hiden triple-filter quadrupole mass spectrometer equipped with a positive ion counting detector. Beyond the detection of helium, the partial pressures of some rest gases were continuously monitored (H₂, CH₄, H₂O, N₂, Ar, CO₂). He blanks were estimated using the same procedure on empty Pt tubes (max. 0.0003 and 0.0008 ncc ⁴He; cold and hot blanks, respectively). Crystals were checked for degassing of He by sequential reheating and He measurement. The residual gas is usually around 1 to 2% after the first extraction in case of zircon. The analysis procedure was operated by HeLID automation software through a K8000/Poirot interface board (developed by I. Dunkl, Göttingen).

Following degassing, samples were retrieved from the gas extraction line, spiked with calibrated ²³⁰Th and ²³³U solutions. Zircon crystals were dissolved in Teflon bombs using a mixture of double distilled 48% HF and 65% HNO₃ at 220 °C during 5 days. Each sample batch was prepared with a series of procedural blanks and spiked normals to check the purity and calibration of the reagents and spikes. Spiked solutions were analyzed as 1.4 or 2 ml of ~0.5 ppb U-Th solutions by isotope dilution on a PerkinElmer Elan DRC II ICP-MS with an APEX micro-flow nebulizer. Procedural U and Th blanks by this method are usually very stable in a measurement session and below 1.5 pg. Sm, Pt and Ca were determined by external calibration. The oxide formation rate and the PtAr–U interference were always monitored, but the effects of these isobaric argides were negligible relatively to the signal of actinides.

The He signal was processed and evaluated by the factory-made software of the mass spectrometer (MASsoft, HIDEN). During standard and sample measurements, 240 readings of the mass spectrometer were recorded. The data of the ICP-MS measurements were processed by an in-house freeware software (PEPITA, url: www.sediment.uni-goettingen.de/staff/dunkl/software). Usually 40–70 readings were considered, and the individual outliers (spikes) of the ²³³U/²³⁸U and ²³⁰Th/²³³Th ratios were tested and rejected according to the 2 SD criterion.

The ejection correction factors (F_t) were determined for the single crystals by a modified algorithm of Farley, Wolf and Silver (1996) using an in-house spread sheet. Note that alternative methods result in different F_t values. The (U-Th-Sm)/He ages were calculated by the Taylor expansion method (Des Patterson, CSIRO, Sidney). The Sm content was also considered for age calculation.

4. Results

4.1. Geochronology

Geochronological studies were carried by three different methods to reveal histories of emplacement, metamorphism and exhumation of the rocks. Two samples are dated using zircon U-Pb LA-ICP-MS method (KG 07 and 7000). From one of these samples (7000), biotites were separated and dated using Ar–Ar laser probe method and zircons from the sample KG 07 were dated using (U-Th)/He method.

4.1.1. Zircon U-Pb LA-ICP-MS dating

Two samples of the Karacabey Pluton were dated using zircon U-Pb LA-ICPMS method (Figure 3). Figure 5 shows CL images of the zircons extracted from the sample 7000. All of the zircons have internal oscillatory magmatic zoning. In addition, almost every zircon grain has inherited cores. Low CL intensity observed in the zircons is an indication high U amount (Poller, Huth, Hoppe & Williams, 2001). Their U amount ranges between 300 and 1100 ppm (Table 1).

Figure 5. CL images of the zircons extracted from the sample 7000. All grains have oscillatory magmatic zoning and almost inherited old cores. For inherited ages see Table

Table 1. U-Pb zircon geochronologic data of the samples.

				Isotopic ratios							Apparent ages (Ma)
Analysis	U	Th	U/Th	^207Pb	±	^206Pb	±	error	^208Pb	±	^206Pb	±	^207Pb	±	^208Pb	±
	(ppm)	(ppm)		^235U	error	^238U	error	corr.	^232U	error	^238U	(Ma)	^235U	(Ma)	^232Th	(Ma)		Concordance
Sample KG-07
1	698	159	4.384	0.47540	0.00440	0.06284	0.00052	0.9	0.02053	0.00019	393	3	395	3	411	4		99.5%
2	483	126	3.841	0.47510	0.00530	0.06317	0.00062	0.9	0.02019	0.00022	395	4	395	4	404	4		100.1%
3	521	435	1.1933	5.07100	0.04700	0.31280	0.00290	1.0	0.09063	0.00086	1754	14	1833	8	1753	16		95.7%
4	1001	278	3.633	0.48880	0.00480	0.06353	0.00056	0.9	0.02040	0.00021	397	3	404	3	408	4		98.2%
5	566	152	3.724	0.49070	0.00480	0.06469	0.00054	0.9	0.02133	0.00020	404	3	405	3	427	4		99.7%
6	885	264	3.357	0.47800	0.00420	0.06276	0.00045	0.9	0.02005	0.00019	392	3	397	3	401	4		98.8%
7	667	137	4.882	0.46700	0.00430	0.06188	0.00048	0.9	0.01967	0.00021	387	3	389	3	394	4		99.5%
8	749	272	2.749	0.50540	0.00680	0.06297	0.00062	0.7	0.02120	0.00031	394	4	414	4	424	6		95.0%
9	621	151	4.111	0.49140	0.00510	0.06430	0.00069	0.9	0.02075	0.00025	402	4	406	4	415	5		99.0%
10	504	179	2.813	0.75810	0.00810	0.09234	0.00099	0.9	0.02840	0.00032	569	6	573	5	566	6		99.4%
11	638	144	4.404	0.47910	0.00470	0.06254	0.00050	0.9	0.02032	0.00020	391	3	397	3	407	4		98.4%
12	494	169	2.902	0.48880	0.00700	0.06348	0.00069	0.8	0.02010	0.00024	397	4	404	5	402	5		98.2%
13	831	188	4.415	0.47750	0.00510	0.06300	0.00062	0.9	0.01967	0.00023	394	4	397	4	394	5		99.3%
14	985	381	2.617	0.47600	0.00540	0.05847	0.00055	0.6	0.01565	0.00016	366	3	395	4	314	3		92.7%
15	573	167	3.421	0.60160	0.00620	0.06453	0.00059	0.9	0.02780	0.00032	403	4	478	4	554	6		84.3%
16	528	121	4.346	0.48000	0.00610	0.06338	0.00068	0.9	0.02042	0.00024	396	4	398	4	409	5		99.5%
17	696	235	2.937	0.47600	0.00440	0.06252	0.00058	0.9	0.01968	0.00021	391	4	395	3	394	4		98.9%
18	483	137	3.526	0.47510	0.00510	0.06235	0.00061	0.9	0.01987	0.00022	390	4	395	4	398	4		98.8%
19	471	143	3.497	0.50030	0.00730	0.06316	0.00078	0.9	0.02148	0.00040	395	5	412	5	430	8		95.8%
20	380	94	4.045	0.49300	0.00860	0.06298	0.00065	0.6	0.02167	0.00047	394	4	407	6	433	9		96.8%
21	302	80	3.786	0.50180	0.00730	0.06390	0.00065	0.8	0.02158	0.00039	399	4	413	5	432	8		96.7%
Sample 7000
1	434	138	3.15	0.48260	0.00490	0.06366	0.00059	0.9	0.01987	0.00020	398	4	400	3	398	4		99.5%
2	840	307	2.73	0.47850	0.00470	0.06319	0.00064	1.0	0.01968	0.00022	395	4	397	3	394	4		99.5%
3	1095	1260	1.07	0.80590	0.00760	0.08990	0.00120	0.7	0.01450	0.00140	555	7	600	4	290	29		92.5%
4	270	72	3.77	0.48090	0.00480	0.06334	0.00062	0.8	0.01981	0.00022	396	4	399	3	397	4		99.3%
5	292	82	3.56	0.48440	0.00590	0.06401	0.00066	0.9	0.02054	0.00027	400	4	401	4	411	5		99.7%
6	471	155	3.03	0.49010	0.00600	0.06465	0.00077	0.9	0.02064	0.00026	404	5	405	4	413	5		99.8%
7	379	105	3.62	0.48210	0.00590	0.06334	0.00070	0.9	0.02000	0.00025	396	4	399	4	400	5		99.1%
8	411	120	3.42	0.47370	0.00520	0.06320	0.00058	0.9	0.02039	0.00024	395	4	394	4	408	5		100.4%
9	194	61	3.17	0.71710	0.00910	0.08876	0.00091	0.9	0.02937	0.00037	548	5	549	5	585	7		99.8%
10	807	357	2.26	0.47530	0.00390	0.06284	0.00054	0.9	0.01931	0.00018	393	3	395	3	387	4		99.5%
11	705	288	2.45	0.48640	0.00440	0.06423	0.00054	0.9	0.02002	0.00018	401	3	402	3	401	4		99.8%
12	578	224	2.57	0.48660	0.00540	0.06418	0.00066	0.9	0.02041	0.00024	401	4	402	4	408	5		99.7%
13	281	91	3.08	5.13800	0.05800	0.31550	0.00340	1.0	0.09568	0.00093	1767	17	1843	10	1847	17		95.9%
14	312	80	3.86	0.48440	0.00520	0.06324	0.00065	0.9	0.01997	0.00024	395	4	401	4	400	5		98.6%
15	616	324	1.90	0.48140	0.00460	0.06246	0.00059	0.9	0.01600	0.00018	391	4	399	3	321	4		97.9%
16	532	142	3.73	0.61700	0.01100	0.07330	0.00100	0.9	0.01644	0.00030	456	6	488	7	330	6		93.5%

1.

Twenty-one spots from the sample KG 07 and 17 spots from the sample 7000 have been analyzed by laser ablation (Table 1). Nineteen spots from the sample KG 07 yielded ²⁰⁶Pb/²³⁸U ages between 387 and 404 Ma. Most of the ages are concordant (between 95 and 101%). Only two spots gave older inherited zircon ages (569 and 1754 Ma, respectively). The TuffZirc age calculated (Ludwig, 2003) is 393.75 + 2.35/−2.65 Ma (Figure 6). Twelve spots out of 16 in the sample 7000 gave ²⁰⁶Pb/²³⁸U ages ranging from 391 to 404 Ma (Table 1). Except for two spots (spots 3 and 16, Table 1), ages are concordant (between 95% and 101%). The TuffZirc age calculated (Ludwig, 2003) is 395.9 + 4.0/−0.9 Ma (Figure 6). Four spots gave older inherited ages (456, 548, 555 and 1767, respectively). Except for the Ordovician age from sample 7000, the ages from the two samples are very similar and indicate a Middle Devonian (ca. 395 Ma) intrusion age. These ages are also similar to those obtained in the western Sakarya Zone (399–397 Ma in Okay et al., 1996, 2006; 401–389 Ma in Aysal et al., 2012).

Figure 6. Tera-Wasserburg diagrams of the dated samples (KG 07 and 7000). TuffZirc ages calculated (Isoplot 3.0, Ludwig, 2003) for the samples are given as inset diagrams. For single spot ages see Table 1.

4.1.2. Biotite ⁴⁰Ar/³⁹Ar dating

Twelve biotite grains from the sample 7000 were dated using Ar–Ar laser beam method (see Analytical Techniques). The resultant Ar–Ar ages show a wide scatter (Table 2) possibly because of alteration and range between 223 and 304 Ma. Probability distribution diagram of the ages is given in Figure 7(a) and the mean age calculation in Figure 7(b). Because of the wide scatter of the ages, result of the mean age calculation is not meaningful. Furthermore, Roberts, Kelley and Dahl (2001)’s study has shown that different degrees of alteration in biotites give generally anomalous results. They dated fresh, partly altered and altered biotite grains using both conventional stepwise and laser Ar–Ar methods. According to their results, the amount of ³⁶Ar increases with increasing degree of alteration. Real ages obtained from the fresh samples show smaller ³⁶Ar/³⁹Ar ratios and ³⁶Ar/³⁹Ar ratios of biotites that have different degree of alteration deviates from real age value. In Figure 7(c), we plotted ³⁶Ar/³⁹Ar ratios of the biotites against their calculated ages. Eight ratios show two different trends. Four of the values lie outside these two trends and are accepted as outliers (Figure 7(c)). The lowest ³⁶Ar/³⁹Ar ratios in two trends correspond ~285 and 291 Ma, respectively. This Early Permian age can possibly be interpreted as the age of low-grade metamorphism or alteration. However, the data can be further evaluated by assuming that fresh biotite would have the lowest ³⁶Ar/³⁹Ar ratio, which would also represent the “true” metamorphic age; in this case, the trend of altered ³⁶Ar/³⁹Ar ratios should join in a point where ³⁶Ar/³⁹Ar ratio is the minimum (ratio of theoretical fresh biotite). In our data set, trends of ³⁶Ar/³⁹Ar ratios crosscut in an age value of ~307 Ma (late Carboniferous). However, it is not possible to judge which approach is more correct, and it can only be said that the Ar–Ar data indicate low-grade metamorphism and/or alteration close to the Permian–Carboniferous boundary.

Table 2. ⁴⁰Ar/³⁹Ar age results of UV laser microprobe analysis of individual biotite grains.

Grains	39/40	+/-	36/40	+/-	37/39	+/-	38/39	+/-	^39Ar	+/-	^40Ar*/^39Ar	+/-	Age (Ma)	Error (Ma)
7000–1	0.04180	0.00059	0.00072	0.00009	−0.02951	−0.04302	0.01454	0.00297	0.00804	0.00010	18.85	0.67	223.1	7.5
7000–2	0.03054	0.00016	0.00073	0.00005	0.12250	0.02305	0.01595	0.00159	0.01502	0.00007	25.70	0.49	297.9	5.5
7000–3	0.03256	0.00024	0.00073	0.00003	0.06830	0.01533	0.01603	0.00106	0.02260	0.00016	24.07	0.36	280.4	4.1
7000–4	0.03354	0.00015	0.00074	0.00002	0.10127	0.00778	0.01592	0.00096	0.04459	0.00019	23.33	0.19	272.4	2.4
7000–5	0.03352	0.00021	0.00067	0.00002	0.03724	0.01293	0.01597	0.00080	0.04346	0.00027	23.89	0.22	278.4	2.7
7000–6	0.03246	0.00016	0.00063	0.00002	0.10987	0.02449	0.01420	0.00113	0.02297	0.00011	25.08	0.23	291.2	2.8
7000–7	0.03318	0.00014	0.00064	0.00001	0.05020	0.01119	0.01665	0.00069	0.05029	0.00021	24.47	0.14	284.6	2.0
7000–8	0.02927	0.00017	0.00078	0.00003	0.06778	0.02354	0.01577	0.00108	0.02392	0.00014	26.32	0.32	304.5	3.7
7000–9	0.03150	0.00011	0.00078	0.00002	0.07243	0.01189	0.01636	0.00093	0.04741	0.00016	24.42	0.17	284.1	2.2
7000–10	0.03256	0.00066	0.00090	0.00003	0.11491	0.01405	0.01566	0.00068	0.04041	0.00059	22.55	0.59	263.9	6.5
7000–11	0.03364	0.00023	0.00082	0.00003	0.08468	0.02480	0.01792	0.00114	0.02275	0.00016	22.51	0.34	263.5	3.9
7000–12	0.03419	0.00010	0.00089	0.00003	0.06643	0.01002	0.01704	0.00113	0.05632	0.00016	21.57	0.23	253.1	2.8
Average blanks													J=0.006984
40Ar	+/-	39Ar	+/-	38Ar	+/-	37Ar	+/-	36Ar	+/-				Error=3.49E-05
0.0057	0.0024	1.3E-05	1.5E-06	1.6E-05	8E-06	2.1E-05	6.4E-06	0.000111	2E-05

Figure 7. Ar–Ar age plots of the individual biotite grains; (a) relative probability distribution of the ages, (b) mean age calculation of the ages (Isoplot 3.0, Ludwig, 2003), (c) ³⁶Ar/³⁹Ar vs. age distribution of the ages (adopted from Roberts et al., 2001). See text for details.

A similar scatter of K-Ar biotite ages was obtained by Delaloye and Bingöl (2000). In their study three different biotites separate yielded ages between 199 and 298 Ma. This shows that Ar-based dating methods do not give satisfactory results for the intrusion age of the Karacabey Pluton.

4.1.3. Zircon (U-Th)/He dating

Three zircon grains from the sample 7000 have been analyzed using (U-Th)/He dating method. Results are given in Table 3. (U-Th)/He ages range between 83 and 106 Ma (Albian–Santonian). Unweighted average of three zircon ages is 93.0 ± 6.9 Ma (Table 3, Turonian). The zircon (U-Th)/He thermochronometer has a closure temperature of 170–190 °C and correlated with K-feldspar ⁴⁰Ar/³⁹Ar cooling temperature (Reiners, Spell, Nicolescu & Zanetti, 2004).

Table 3. Zircon (U-Th)/He age results of the sample GK 07.

	He		^238U			^232Th				Sm			Ejection	Uncorr.	Ft-Corr.		Sample unweighted aver. ± 2 s.e.
	vol.	1s	mass	1s	conc.	mass	1s	conc.	Th/U	mass	1s	conc.	correct.	He-age	He-age	2s
Aliq.	[ncc]	[%]	[ng]	[%]	[ppm]	[ng]	[%]	[ppm]	ratio	[ng]	[%]	[ppm]	(Ft)	[Ma]	[Ma]	[Ma]	[Ma]	[Ma]
1	98.865	1.5	10.29019	1.8	n.d.	2.755546	2.4	n.d.	0.27	0.49	10.8	n.d.	0.831	74.4	89.5	6.0
2	89.523	1.5	9.706849	1.8	n.d.	3.022634	2.4	n.d.	0.31	0.64	12.8	n.d.	0.850	70.7	83.2	5.3
3	106.638	1.5	9.146853	1.8	n.d.	3.087565	2.4	n.d.	0.34	0.47	11.5	n.d.	0.834	88.7	106.4	7.1	93.0	6.9

Amount of helium is given in nano-cubic-cm in standard temperature and pressure.

Amount of radioactive elements are given in nanograms.

Ft: alpha-ejection correction (according to Farley, 2002).

Error on average age is 1s, as (SD)/(n)^1/2; where SD = standard deviation of the age replicates and n = number of age determinations.

4.2. Major and trace element geochemistry

In this study, four samples from the Karacabey Pluton were analyzed for their geochemistry (Table 4). All the samples are from the biotite-rich granodiorite. Two samples are collected from the western part of the pluton and two from the east (Figure 3). The main difference between western and eastern samples is the presence of hornblende in western ones (Figure 4). The eastern samples have the mineral assemblage of quartz, plagioclase, K-feldspar, biotite and minor apatite, zircon and opaque minerals, whereas those from the east have hornblende and titanite in addition to the minerals above.

Table 4. Geochemical analyzes of granodiorite of the Karacabey Pluton.

Sample	KG28	KG 27	KG17	KG07–2
SiO2	64.8	64.7	63.5	63.6
TiO2	0.51	0.49	0.52	0.52
Al2O3	15.41	15.83	15.44	15.34
Fe2O3^tot	4.64	4.61	5.23	5.37
MnO	0.10	0.10	0.10	0.10
MgO	3.52	3.58	4.03	4.08
CaO	2.58	2.22	3.64	3.71
Na2O	2.81	3	3.06	3.04
K2O	3.36	3.16	2.95	2.99
P2O5	0.13	0.15	0.13	0.13
LOI	2.24	2.55	1.62	1.52
Total	100.2	100.4	100.3	100.5
Ni	35.1	44	35.6	37
Co	8.8	18.2	15.3	13.6
V	71	87	84	85
Cu	13.7	30.4	25.8	29.1
Zn	62	65	60	59
Cs	8.8	6.7	9.4	8.8
Rb	116.5	121.2	110.9	113.5
Ba	523	520	542	509
U	1.3	2.9	1.5	1
Th	11.1	9.7	25	23.9
Pb	10.7	35.2	14.2	9.8
Sr	244.8	258.6	279.9	255.8
Nb	8.5	9.9	9.7	10.7
Ta	1.0	0.8	1.0	1.0
Zr	121.2	115.2	135.3	134.7
Hf	3.6	3.5	3.4	4.1
Y	17.5	20.3	35.2	33.7
Ga	17.8	18.1	18	17.7
La	24.9	23.8	58.5	46.1
Ce	50.2	44.2	124.7	109.6
Pr	5.55	5.55	13.42	11.3
Nd	18.6	19	49.5	40.7
Sm	4.13	4.11	8.07	8.06
Eu	0.91	1.02	1.08	1.00
Gd	3.35	3.84	6.69	6.9
Tb	0.52	0.55	1.03	1.09
Dy	3.05	3.75	6.12	6.95
Ho	0.65	0.64	1.28	1.37
Er	1.93	2.12	3.63	3.88
Tm	0.27	0.3	0.52	0.56
Yb	1.75	2.16	3.18	3.7
Lu	0.27	0.3	0.5	0.51
K2O/Na2O	1.20	1.05	0.96	0.98
Mg#	60.05	60.61	60.42	60.08
ASI	1.19	1.28	1.04	1.02
(La/Yb)cn	9.59	7.43	12.40	12.46
Eu/Eu*	0.75	0.79	0.45	0.41

Mg# = 100 × (MgO/(MgO + FeO^tot)) in molar poroportions; ASI = aluminum saturation index = molar Al₂O₃/(CaO + Na₂O + K₂O); (La/Yb)cn = chondrite-normalized La/Yb ratio; oxides are given in wt.%. trace elements in μg/g.

Modal compositions of the samples fall into the granodiorite field when plotted on SiO₂ vs. total alkali diagrams (Figure 8, Middlemost, 1985). In AFM classification (Irvine & Baragar, 1971), they plotted in the field of calc-alkaline series. On the SiO₂ vs. K₂O (Peccerillo & Taylor, 1976) diagram, all samples plotted in the field of high-K calc-alkaline series (Figure 9). The samples collected from the western part of the pluton (KG 07 and 17) fall in the peraluminous I-type field whereas those from the east plot in the S-type area of Shand (1943) (Figure 10).

Figure 8. Geochemical nomenclature of the samples of the Karacabey Pluton in total alkali vs. SiO₂ diagram (Middlemost, 1985). They all fall into the granodiorite fields.

Figure 9. The AFM and K₂O-SiO₂ diagrams of Irvine and Baragar (1971) and Peccerillo and Taylor (1976), respectivelly. All samples belong to the high-K, calc-alkaline series.

Figure 10. The distribution of the samples in A/CNK-A/NK diagram of Shand (1943). All samples are peraluminous in character but western samples are I-type whereas eastern samples are S-type.

SiO₂ amounts of the samples vary between 63 and 65%. TiO₂, Fe₂O₃, MgO, CaO and Na₂O amounts of the western samples are higher than the eastern ones. Only K₂O amount of the eastern samples is higher than the western samples. The Al₂O₃ amount and Mg# do not change between the samples (Table 1).

In SiO₂ vs. trace element diagrams, the differences between these two regions are more evident (Figure 11). Especially, Y, Zr, La and Ce amounts differ significantly. Y, Zr, La and Ce are incompatible and immobile elements, and their variations should be dependent on the different magmatic origins of the rocks. However, elements such as Rb and Ba are compatible and mobile elements, and they show only changes according to their SiO₂ amounts (Figure 11). The Y, Zr, La and Ce amounts of the western samples (KG 07 and 17) are higher than the eastern samples (KG 27 and 28).

Figure 11. Harker diagrams of the samples analyzed. Note that only elements Y, La, Ce and Zr reveal real difference when plotted against their SiO₂ amounts. Other elements such as Rb, Sr and Ba show only distribution according to their SiO₂ amounts.

In the chondirite normalized spider diagram, La, Ce, Nd, Tb, Y, Tm and Yb ratios of the eastern samples are different from the western samples (Figure 12(a), Sun & Mcdonough, 1989). The western samples KG 07 and 17 have higher values. However, other elements have similar pattern for all samples in both regions. All samples exhibit negative anomalies of Ta and Nb relative to their adjacent LILE (large-ionlithophile elements: Sr, K, Rb, Ba and Th) and LREE (light rare earth elements: La to Nd). These geochemical features are characteristic for arc-related magmas (e.g. Pearce & Peate, 1995). In the chondrite normalized REE pattern (Figure 12(b), Nakamura, 1974), the western samples are much enriched than the eastern samples. The La_n/Yb_n values are 7.5 and 9.6 for the eastern samples, and 12.4 and 12.5 for western samples (Table 4). The clear difference between the samples is the western samples have strong negative Eu anomaly, whereas the eastern samples do not. The negative Eu anomalies (Eu_n/(Sm_n × Gd_n)^0.5) are 0.75 and 0.79 for the eastern samples and 0.41 and 0.45 for western ones. The strong negative anomaly in the eastern samples reveals that they have fractionation of plagioclase during their magma chamber evolution. Although samples have some differences between the western and the eastern samples, it should be taken into account that the later events such as metamorphism and alteration, as it was revealed by the scatter of Ar–Ar ages, can cause such differences in the geochemical characteristics of the samples. This is clearly seen in their alumina saturation characteristics. Samples are distributed from I- and S-type characteristics. Field observations show that exposures of both rock types are transitional and U-Pb zircon ages obtained from both rock types are identical.

Figure 12. Multi-element spider and REE distribution of the samples from the Karacabey Granitoid. Normalization values are; (a) primitive mantle from Sun and Mcdonough (1989) and (b) chondirite from Nakamura (1974). Note that only western samples (KG 07 and 17) have negative Eu anomaly. Shaded area represents distribution of the samples analyzed in Aysal et al. (2012).

4.3. Sr-Nd isotope geochemistry

All of the samples have similar Sr initial isotopic ratios between 0.709735 and 0.711698 (Table 5). Their Nd initial ratios are also similar to each other (0.51165–0.51173). Initial epsilon Nd values of the western samples (−9.4 and −8.5) are slightly higher than the eastern samples (−8.2 and −7.8). In the Sr_(i) vs. epsilon Nd_(i) diagram, all samples plotted into the field between lower continental crust of metaigneous and metapelitic sources (Figure 13). A similar isotopic distribution was also obtained from Devonian granitoids in other parts of the western Sakarya Zone (see Figure 14, Aysal et al., 2012).

Table 5. Sr and Nd isotopic data of the samples of the Karacabey Pluton.

Sample	^87Rb/^86Sr	^87Sr/^86Sr	^87/86Sr (i)	^147Sm/^144Nd	^143Nd/^144Nd	^143/^144Nd(i)	εNd(i)	TDM (Ga)
KG07	1.4162	0.718590(14)	0.710766	0.12213	0.511966(10)	0.511656	−9.4	1.96
KG17	1.2246	0.718463(19)	0.711698	0.09053	0.511934(10)	0.511704	−8.5	1.50
KG27	1.3398	0.717923(12)	0.710521	0.12511	0.512038(09)	0.511720	−8.2	1.90
KG28	1.5805	0.718467(18)	0.709735	0.11876	0.512039(09)	0.511737	−7.8	1.78

Uncertainties for the ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios are 2σ errors in the last two digits (inparantheses).

Initial values are calculated for obtained zircon age of 393 Ma. εNd(i) values are calculated relative to CHUR with present day values of ¹⁴³Nd/¹⁴⁴Nd = 0.512638 and ¹⁴⁷Sm/¹⁴⁴Nd = 0.1967 (Jacobsen & Wasserburg,1980).

Nd modelages (TDM) are calculated with a depleted-mantle reservoir and present-day values of ¹⁴³Nd/¹⁴⁴Nd = 0.513151 and ¹⁴⁷Sm/¹⁴⁴Nd = 0.219 (e.g. Liew & Hofmann, 1988).

Figure 13. ƐNd(₃₉₃) vs. ⁸⁷Sr/⁸⁶Sr(₃₉₃) diagram of the analyzed granodiorites. MORB (Mid-ocean ridge basalts), OIB (Ocean island basalts), BSE (Bulk silicate earth) and DMM (Depleted MORB mantle) fields are after Zindler and Hart (1986); LM (Lithospheric mantle) is from Yang, Wu, Chung, Wilde, and Chu (2004) and Yang, Wu, Chung, Wilde, and Chu (2006); LCC (1) (lower continental crust of meta-igneous source) and LCC (2) (lower continental crust of meta-pelitic source) are from Villaseca, Downes, Pin, and Barbero (1999). Pale-blue area represents distribution of the samples analyzed in Aysal et al. (2012).

Figure 14. Tectonic discrimination of the samples in the Pearce et al. (1984) diagrams. Pale-blue area represents distribution of the samples analyzed in Aysal et al. (2012).

5. Discussion

5.1. The Variscan magmatism in the Sakarya Zone

Early-Middle Devonian granitoids in the western Sakarya Zone have been first reported by Okay et al. (1996, 2006). More recently Aysal et al. (2012) showed a wider distribution of the Early-Middle Devonian granitoids in the Biga Peninsula in the western Sakarya Zone. No other evidence of Devonian magmatism is known in the Pontides or in Turkey in general, except detrital zircons of Devonian age, reported from the Yusufeli Area (Ustaömer, Robertson, et al., 2012) and from the Late Carboniferous metasediments in the Strandja Massif (Sunal, Natal’in, Satır & Toraman, 2006).

In contrast to the scarcity of the Devonian magmatism, Carboniferous and Permian granitoids crop out widely in the Sakarya Zone. In the central Sakarya zone, to the north of Eskisehir, granitoids of 319–327 Ma old have been published by Ustaömer, Ustaömer, et al. (2012). In the central Pontides, Nzegge et al. (2006) dated several granitoids between 303 and 275 Ma. In the eastern Sakarya Zone, 318–329-Ma-old granitoids from Gümüşhane and surrounding plutons have been dated by Topuz et al. (2010) and Dokuz (2011). Recently, Ustaömer, Robertson, et al. (2012) published 356–325-Ma-old magmatics in the Yusufeli area (the eastern Sakarya zone).

The presence of the garnet-bearing leucocratic granitoids in the region (not analyzed in this study) and inherited cores in the zircons in the dated samples indicate contribution of Precambrian continental crust to generation of Middle Devonian magmas. Furthermore, TDM (age of the depleted mantle) values (1.5–1.96 Ga, Table 5) also suggest involvement of the older crust to the Middle Devonian magma generation. Isotopic data show compositions close to the lower crustal values (Figure 9). However, high Mg# of the samples (~60) is also indicative of a basic magma derived from a mantle source. Karacabey Pluton has distinct Nb and Ta anomalies, which is commonly take as a subduction sign (Baier, Audétat & Keppler, 2008). Samples from the granodiorite also plot in the fields of volcanic-arc magmatics on the tectonic discrimination diagrams of Pearce, Haris and Trindle (1984) (Figure 14). Besides, isotopic characteristics of the samples are very close to the enriched mantel source (EMII in Zindler & Hart, 1986) that represented by low Nd but generally high and variable Sr isotopic ratios. Zindler and Hart (1986) stated different involvements of the mechanisms to generation of EMII but also stated that they always play roles in convergent plate boundaries. Aysal et al (2012) also described Early-Middle Devonian magmatics with similar petrography and geochemistry from farther west; they interpreted their genesis to delamination and/or slab-breakoff but also discussed subduction possibility. Even they show differences in their petrography and some geochemical characteristics, later events are able to change geochemistry of the rocks, especially alteration. Thus, we interpret our geochemical data obtained from the Middle Devonian granodiorites as subduction related.

5.2. The Variscan metamorphism in the Sakarya Zone

The Variscan high to low degree metamorphism in the Sakarya Zone is described from different regions and with different ages (Figure 1). The oldest metamorphic event is a HP/LT metamorphism in the Eastern Pontides of Middle Devonian age (Topuz, Altherr, Schwarz, Zack & Sunal, 2009). The age of the granulite facies metamorphism in the Pulur region, in the Eastern Pontides, is determined between 331 and 323 Ma (Topuz & Altherr, 2004; Topuz et al., 2001, 2004a; Topuz, Altherr, Schwarz, Dokuz & Meyer, 2007). The cooling of the granulite facies metamorphic rocks to 300 C^o lasted to c. 310 Ma (Topuz et al., 2004a). The timing low-grade metamorphism in the Pulur region is much younger than high grade metamorphism (~260 Ma, Topuz, Altherr, Satır, & Schwarz, 2004b). Beside this direct data, rims of the detrital zircons from pre-Early Jurassic metaclastic rocks reveal two different intervals for metamorphism in the eastern Sakarya Zone (Ustaömer, Robertson, et al., 2012). First event dated is about 334 Ma, and the younger one is about 314 Ma. Ustaömer, Robertson, et al. (2012) claim that these two metamorphic events were separated by a hiatus (320–325 Ma). In the western Sakarya Zone, in Kazdağları region, the age of high grade metamorphism is given as c. 319 Ma (Okay et al., 1996, 2006).

The K-Ar ages from the western Sakarya zone show a wide scatter, and generally fall within the Permian; those from the Karacabey Pluton range from 199 to 298 Ma (K-Ar method, Delaloye & Bingöl, 2000) somewhat similar to our Ar–Ar biotite results (223–304 Ma). In Söğüt region, K-Ar biotite and Pb-Pb zircon evaporation ages are ~290 Ma (Çoğulu & Krummenacher, 1967; Çoğulu, Delaloye & Chessex, 1965), and one Ar–Ar amphibole age is 272 ± 2 Ma (Okay, Monod & Monié, 2002). However, the granitoid ages in these regions are known to be Carboniferous, hence younger ages from the granitoids should represent later metamorphic (heating?) events. The timing of low-grade metamorphism in the western Sakarya Zone may correspond to the Permian low-grade metamorphism reported from the Eastern Pontides (Topuz et al., 2004a).

5.3. Thermal History of the region

Thermal history of the Karacabey region is given in Figure 15. The main constraints are the ages obtained from minerals with different isotopic closure temperatures (this study), and the apatite fission-track ages published by Cavazza, Federici, Okay, and Zattin (2012) from the nearby areas and the stratigraphy of the region (indicated as A, B, C and D in Figure 15). Time-temperature pathways of the Karacabey region are indicated as numbers in Figure 15.

Figure 15. Thermal history of the Karacabey region calculated using computer program HeFTy (Ketcham, 2005). Constraints are (A) Early-Middle Jurassic regional unconformity, (B) Late Jurassic- Early Cretaceous rifting and carbonate deposition, (C) Santonian regional unconformity, (D) Eocene deposition and relevant burial. Oligocene apatite FT age (c. 33 Ma) are from Cavazza et al. (2012). For references related to other constraints see text. Thermal pathways (numbers) are explained in the text in detail. Green lines represent acceptable paths and pink lines good pats. Thick black line is the best-fit model path and blue thick line is the weighted mean path.

The first pathway (number 1 in Figure 15) is constructed between the zircon U-Pb and biotite Ar-Ar ages obtained in this study. The uncertainty here is the lack of data between Middle Devonian and end of Permian. The second pathway (number 2) is constrained between the biotite Ar-Ar ages and the Early Jurassic unconformity (indicated as A in Figure 15). The Karacabey Pluton and the metamorphic rocks of the Lower Karakaya Complex were juxtaposed by thrusting before the Early Jurassic, as shown by a common Lower Jurassic sedimentary cover. The third pathway represents the reburial of the Karacabey Pluton under the Jurassic-Cretaceous sedimentary cover. During this event, the Karacabey Pluton must have been subjected to the temperatures of over 200 °C because zircons from the granitoid have been reset for (U/Th)-He system, suggesting burial to depths of over 5–6 km’s. The fourth pathway is between Early Jurassic deposition and the Zr (U/Th)-He ages, indicating cooling to temperatures of 170–190 °C (Reiners et al., 2004). After resetting Zr (U/Th)-He ages during post-Jurassic burial, blocking of the Zr (U/Th)-He ages occurred during the Late Cretaceous (Turonian) possibly related to uplift and erosion. Akbayram, Okay and Satır (2013) have claimed that main closure of the Intra-Pontide suture is at Early Cretaceous (latest age was given as c. 112 Ma), which might have induced regional uplift as also shown by the stratigraphy of the Cretaceous sequences in the Bursa region (Özgörüş, Okay, & Özcan, 2009).

The fifth pathway drawn between zircon (U/Th)-He age obtained in this study and the exhumation due to closure of the Neo-Tethys Ocean. After closure of the Neo-Tethys Ocean, during almost whole Paleocene interval, regional uplift and subsequent erosion occurred in the Sakarya Zone (Okay & Tüysüz, 1999; Okay et al., 1996; Şengör & Yılmaz, 1981). The pathway 6 represents Eocene deposition that unconformably overlies all units beneath (see Cavazza et al., 2012 and the references therein) and concerning burial and reheating. Reheating is deduced from opening of the apatite FT ages during Eocene interval. The apatite FT ages at surrounding regions of the Karacabey region were given as c. 33 Ma (Cavazza et al., 2012). This indicates that just after deposition of Eocene volcanic and clastics, region started to exhume during early Oligocene.

5.4. Regional implications

In the European Variscides opening of the Rheic Ocean (Proto-Tethys) started in the Late Cambrian-Early Ordovician (e.g. von Raumer & Stampfli, 2008; Schät, Reischmann, Tait, Bachtadse, Bahlburg & Martin 2002). The initiation of subduction has been assigned to Silurian-Early Devonian (Dokuz et al., 2011; Kröner & Romer, 2013; Matte, 2001; Reischmann & Anthes, 1996; Stampfli, 2000; von Raumer & Stampfli, 2008; von Raumer, Stampfli & Bussy, 2003) and almost in every region where the Variscan orogeny is defined, granitoids belonging to this interval have been reported. In the European Variscides, main arc-related magmatism is constrained between Devonian and Early Carboniferous (410–350 Ma, Kröner & Romer, 2013; Neubauer & Handler 2000) and magmatism between 325 and 310 Ma is regarded as post-collisional (Siebel, Chen, & Satır, 2003). Especially in the northern Carpathians, Bohemian Massif and Massif Central Early-Middle Devonian granitoids are widely described (see Kröner and Romer (2013) for the review of Variscan orogeny in Europe). However, Matte (2001) has claimed that during the Variscan orogeny, the Himalayan-type orogeny occurred in Europe in Early Carboniferous interval (~345 Ma) but at the same time the Andean-type orogeny was still active in the east (most probably in the Pontides). In the Sakarya Zone, even if we consider the older metamorphism, it is still younger than the European ones. The intense magmatic activity in the Sakarya Zone occurred between 330 and 318 Ma (except one age gave 354 Ma, Ustaömer, Robertson, et al., 2012). Even if there is no age data published between 390 and 354 Ma, starting from Early Devonian, most probably subduction was active during the whole Paleozoic Era in the Sakarya Zone and its neighboring regions. One of the evidence of this early subduction is Middle Devonian blueschist metamorphism documented in the Eastern Sakarya Zone (Topuz et al., 2009). In the eastern Sakarya Zone, Dokuz et al. (2011) proposed a dual subduction model for the closure of the Rheic Ocean. Their model constructed according to special distribution of the mantle and metamorphic rocks in the region. Their dual subduction model infers collision of Andean-type continental margin in the north while the southward directed SSZ in the south (Dokuz et al., 2011).

Furthermore, in Sredna Gora region (Bulgaria), eclogite bearing amphibolites yielded 398.4 ± 5.2 Ma (Ar–Ar dating on hornblende, Cortesogno, Gaggero, Haydoutov & Buzzi, 2005). In the southern Carpathians, the timing of HP metamorphism is younger. Medaris, Ducea, Ghent and Iancu (2003) have reported that age of prograde eclogite metamorphism as between 358 and 342 Ma. Taking into account all data outlined above and our new results, similar to European history of the Rheic Ocean, subduction of the Rheic Ocean started in the Sakarya Zone in Early Devonian and most probably lasted during the whole Paleozoic Era and closure of the Rheic Ocean and subsequent orogeny occurred during Late Carboniferous and most probably Early Permian interval.

6. Conclusions

The Karacabey Pluton in the western Sakarya Zone, west of Bursa is a large magmatic body, previously considered as of Permian age, and in tectonic contact with the probably Triassic metamorphic rocks of the Lower Karakaya Complex. Here, it is shown to be an Early Devonian intrusion with 395 Ma U-Pb zircon ages. The geochemistry of the granitoid suggests an arc-related magmatic setting. Along with the other Devonian granitoids in the western part of the Sakarya Zone (Aysal et al., 2012; Okay et al., 1996, 2006), the Karacabey Pluton is considered to have been formed in a subduction setting.

The Karacabey Pluton had a long complex thermal history. New Ar–Ar biotite ages suggest thermal reheating during the Permian–Carboniferous boundary. It was exhumed during the Early Jurassic, when sandstones and conglomerates were deposited on the Karacabey Pluton and on the Triassic metamorphic rocks of the Lower Karakaya Complex; the clastic rocks were succeeded by the deposition of the Upper Jurassic-Lower Cretaceous carbonates. The uplift occurred during the Late Cretaceous (Turonian) as deduced from zircon (U-Th)/He dating. This event is probably related to the Closure of the Intra-Pontide Ocean in the north of the Sakarya Zone (Akbayram et al., 2013).

Acknowledgments

This study was supported by TÜBITAK (The Scientific and Technical Research Council of Turkey, Project No: CAYDAG 111Y314). I wish to thank W. Siebel and E. Reitter for help in isotopic measurements. I warmly thank M.K. Erturaç for assist in the field studies. I sincerely thank A.I. Okay who proposed this study and provided some of the samples. A.I. Okay is also thanked for reading early draft of the manuscript and making numerous changes and constructive comments. I thank S. Sherlock for Ar–Ar dating. I wish to thank A.R.C. Kylander-Clark for U-Pb dating of the rocks. A special thank is due to I. Dunkl for (U-Th)/He dating of zircons. E. Aydar and E. Çubukçu are thanked for helping CL imaging in Hacettepe University. Detailed comments and editorial handlings of E. Bozkurt, constructive reviews by E. Koralay and O. Karslı are greatly appreciated.