Geological and palaeoseismological evidence for late Pleistocene−Holocene activity on the Manisa fault zone, Western Anatolia

In West Anatolia near the cities of İzmir and Manisa, the historical occurrence of large earthquakes suggests the presence of important seismogenic faults. However, these faults have yet to be investigated in detail. The Manisa Fault Zone (MFZ) is an active large-scale normal fault system in this area, and thus field observations and palaeoseismological studies of this zone are important for predicting future earthquakes.

Turkish Journal of Earth Sciences (Turkish J Earth Sci.),Ç.Vol 20, 2011, pp Copyright ©TÜBİTAK ƯZKAYMAK ET449–474 AL doi:10.3906/yer-0906-18 First published online 08 November 2010 Geological and Palaeoseismological Evidence for Late Pleistocene−Holocene Activity on the Manisa Fault Zone, Western Anatolia 1 ÇAĞLAR ƯZKAYMAK , HASAN SƯZBİLİR , BORA UZEL & H SERDAR AKYÜZ Dokuz Eylül University, Engineering Faculty, Department of Geological Engineering, Tınaztepe Campus, Buca, TR−35160 İzmir, Turkey (E-mail: caglar.ozkaymak@deu.edu.tr) İstanbul Technical University, Faculty of Mines, Department of Geological Engineering, Maslak, TR−34469 İstanbul, Turkey Received 23 July 2009; revised typescripts receipt 15 March 2010, 24 August 2010 & 22 October 2010; accepted 08 November 2010 Abstract: In West Anatolia near the cities of İzmir and Manisa, the historical occurrence of large earthquakes suggests the presence of important seismogenic faults However, these faults have yet to be investigated in detail The Manisa Fault Zone (MFZ) is an active large-scale normal fault system in this area, and thus field observations and palaeoseismological studies of this zone are important for predicting future earthquakes Hence we sought to document geological and palaeoseismological evidence for Holocene activity on the MFZ We performed trenching to determine the magnitude and timing of past surface-faulting events using detailed fault-trace mapping, measurements of Upper Pleistocene− Lower Holocene sediments, and radiocarbon dating By comparing the trench data with palaeoearthquake records, we find evidence for three palaeoearthquakes which correspond to 926 AD, 1595 or 1664 AD, with the most recent event in 1845 AD We also find this in the central and western sectors of the MFZ, which together with the eastern sector comprise the three major seismogenic zones The Pliocene−Quaternary vertical offset at fault scarps is far less than that in the western sector, suggesting that activities of these sectors are highly independent Evaluation of field observations suggests that the MFZ has been the source of multiple Late Pleistocene and Holocene surface-rupturing earthquakes Our results constitute the first palaeoseismic evidence on the causative faults of historical earthquakes that affected Manisa, and point to their underlying tectonic mechanisms Key Words: Manisa Fault Zone, palaeoseismology, late Pleistocene, Holocene, Gediz Graben, Western Anatolia Manisa Fay Zonunun Geỗ PleyistosenHolosen Aktivitesine Ait Jeolojik ve Paleosismolojik Veriler, Batı Anadolu Özet: Batı Anadolu’da İzmir ve Manisa şehri yakınlarında tarihsel dönemlerde büyük depremlerin meydana gelmiş olması, bu bölgede önemli sismojenik fayların varlığına işaret etmektedir Ancak, bu faylar şimdiye kadar ayrıntılı bir şekilde araştırılmamışlardır Bu bửlgede bulunan Manisa Fay Zonu (MFZ) bỹyỹk ửlỗekli aktif normal fay sistemidir ve bu zon ỹzerinde gerỗekletirilen arazi gửzlemleri ile paleosismolojik ỗalmalar, gelecekteki depremlerin tahmin edilebilmesi aỗsndan ửnemlidir Bu ỗalmada, MFZ’nun Holosen aktivitesine ait jeolojik ve paleosismolojik veriler sunulmuştur Tarihsel dửnem yỹzey faylanmalarnn zamann ve bỹyỹklỹklerini ortaya ỗkarmak amacyla, ayrntl fay izi haritalamas, geỗ Pleyistosenerken Holosen yal sedimanlarn incelenmesi, radyokarbon yalandrma yửntemleri kullanlarak hendek ỗalmalar gerỗekletirilmitir Hendek verileriyle eski deprem katalou bilgileri karlatrldnda ỹỗ depreme ait izler saptanmtr; bunlar srasyla, 926, 1595 veya 1664 ve 1845 depremlerine karşılık gelmektedir Ayrıca, bat, orta ve dou bửlgelerden oluan ỹỗ ana sismojenik zon tanımlanmıştır Fay sarplığının Pliyosen−Kuvaterner zamanındaki düşey atım miktarının batı bölümde daha az olması bu bölümlerin bağımsız olarak hareket ettiklerini göstermektedir Arazi ỗalmalar, MFZnun geỗ Pleyistosen ve Holosende bửlgede meydana gelen ve yüzey kırığı oluşturan depremlerin kaynağı olduğunu göstermektedir Elde edilen sonuỗlar, Manisa ehrini etkileyen tarihsel depremlere neden olan faylar ve bu fayların tektonik mekanizmaları üzerine elde edilen ilk paleosismolojik verileri oluturmaktadr Anahtar Sửzcỹkler: Manisa Fay Zonu, Paleosismoloji, Geỗ Pleyistosen, Holosen, Gediz Grabeni, Batı Anadolu 449 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA Introduction Western Anatolia represents a good example of post-collisional extensional tectonics dominated by approximately E–W-trending active normal faults (with maximum lengths typically in the range of 15–25 km), as well as NE–SW- and NW–SE-trending active strike-slip faults (Figure 1a; Dewey & Şengör 1979; Şengör & Yılmaz 1981; Jackson & McKenzie 1988; Eyidoğan & Jackson 1985; Şengör et al 1985; Şengör 1987; Seyitoğlu & Scott 1991; Sözbilir 2000, 2005; Bozkurt 2001; England 2003; Koỗyiit & ệzacar 2003; Lenk et al 2003; Kaymakỗ 2006; Sửzbilir et al 2006, 2007, 2008, 2009; Ưzkaymak & Sưzbilir 2008; Uzel & Sưzbilir 2008) In the western part of West Anatolia, near the cities of İzmir and Manisa, the historical occurrence of large earthquakes suggests the presence of important seismogenic faults, although none have yet been investigated in detail One such active large-scale normal fault system is the Manisa Fault Zone (MFZ), which exhibits prominent Quaternary fault scarps and significant morphologic variations (Figure 1b; Bozkurt & Sưzbilir 2006; Ưzkaymak & Sưzbilir 2008) Therefore, the MFZ itself is a likely source of future earthquakes in the region, and thus field observations of Holocene activity and palaeoseismological studies of this fault zone are important for predicting future earthquakes (Hakyemez et al 1999; Emre et al 2005; Bozkurt & Sửzbilir 2006; ầiftỗi & Bozkurt 2007, 2008, 2009; Ưzkaymak & Sưzbilir 2008) This manuscript, which contains palaeoseismological data for the MFZ, aims for the first time to document geological and palaeoseismological evidence for Holocene activity Trenching was carried out across the westernmost segments to determine the magnitude and timing of past surface-faulting The report includes a detailed map of the studied area at a scale of 1/5000, the logging of two trench walls across the Paşadeğirmeni fault zone, measurements of a stratigraphic section of the upper Pleistocene–lower Holocene sediments, and radiocarbon dating Seismotectonic Setting The Aegean and surrounding area is considered to be one of the most seismically active regions of the world 450 where N–S-trending contraction is overprinted by N–S-trending extension from at least Pliocene times (Şengör et al 1985; Seyitoğlu & Scott 1991; Taymaz et al 1991; Pavlides 1996; Papazachos & Papazachou 1997; Altunel 1998, 1999; Koỗyiit et al 1999; Akyỹz & Altunel 2001; Bozkurt 2001; Caputo et al 2004; Pavlides & Caputo 2004; Caputo & Helly 2005, 2008; Akyol et al 2006; Özkaymak et al 2008, 2009) The studied area, between longitudes of approximately 27° to 28° and latitudes of approximately 38.3° to 38.7° north, is deformed by two active fault systems: strikeslip and dip-slip faults (Figures 1a & 2) The former are characteristically NE–SW-trending dextral and NW–SE-trending sinistral strike-slip faults, while the latter dip-slip normal faults are mainly E–W oriented (Figure 2a) These active tectonic structures work together and are responsible for most of the earthquakes that occurred in both historical and instrumental periods in the region (Taymaz et al 1991; Emre et al 2005; Akyol et al 2006; Zhu et al 2006; Aktar et al 2007; Sözbilir et al 2008, 2009; Tan et al 2008; Uzel et al 2011) Although the İzmir-Manisa region has not suffered destructive earthquakes between 1902 and 2010, several publications containing information about large historical earthquakes that damaged cities in the region (Ergin et al 1967; Soysal et al 1981; Ambraseys 1988; Guidoboni et al 1994; Ambraseys & Finkel 1995; Papazachos & Papazachou 1997; Ambraseys & Jackson 1998; Tan et al 2008) Table summarizes the descriptions of large historical earthquakes that affected the region (Figure 2a) and specifically damaged the city of Manisa One of the best-documented historical earthquakes in West Anatolia is the event in 17 AD According to Ambraseys (1988), 16 ancient cities, most of them located within the Manisa Basin, collapsed and were damaged by the event (Table 1) Soysal et al (1981) and Guidoboni et al (1994) noted that 13 cities were completely demolished by this event Some researchers affirm that the 17 AD earthquake occurred along the Gediz Graben (Ambraseys & Jackson 1998), but other studies give the location as the Muradiye district (western part of Manisa, Figure 2a) with an intensity of IX (Soysal et al 1981; Guidoboni et al 1994; Tan et al 2008) According to Guidoboni et al (1994), it was the Ç ƯZKAYMAK ET AL o N G ör de s o in MFZ Ged iz Figure 1b Figure 1a 38 o 200 km AEGEAN İzmir G b en Gediz detachment fault Figure O FZ 36° N İF Se le Ba nd s i Manisa Anatolia raben nderes G e Kỹỗỹk M Sack Gulf S EE AA Büyük Menderes Menderses detachment fault EF 25 km Büyük Menderes Graben b o 27 30 30’ Man Muradiye Ge d MANİSA i sa N Bas in km iz Ri ve r KeF aF F Ta M GFZ F PA Figure as in Ba si n B 39 The Balkans 28 en rab yG iB 30 rỗa ak 42° 22°E 27o Midilli Island De m irc a MFZ F Ka pre-Neogene basement rocks Quaternary alluvial fan Karaburun Platform and slope deposits Pleistocene Sakarya Zone continental clastics pre-Neogene Quaternary alluvium Neogene volcanoMenderes Massif basement rocks sedimentary rocks Quaternary alluvial fan Karaburun Platform major thrust and andand/or slope deposits normal suture zones Pleistocene oblique fault Sakarya Zone continental strike-slip fault clastics river channel Neogene volcanoMenderes Massif detachment fault rocks sedimentary and shear zone 30’ 38o 30 Quaternary alluvium Karaoğlanlı normal and/or oblique fault sinistral strike-slip faulting (earlier motion) dextral strike-slip faulting (the latest motion) detachment fault and shear zone Figure (a) Outline geological map of western Turkey showing major tectonostratigraphic units and location of the study area (compiled from Okay & Siyako 1993; Bozkurt & Park 1994; Bozkurt 2001, 2004; Sưzbilir 2001, 2002, 2005; Collins & Robertson 2003; Ưzer & Sözbilir 2003; Bozkurt & Sözbilir 2004; Işık et al 2004; Ưzkaymak & Sưzbilir 2006, 2007, 2008) Abbreviations EF, İF, MFZ, and OFZ, refer to the Efes Fault, İzmir Fault, Manisa Fault Zone, and Orhanlı Fault Zone, respectively Inset shows the location of Figure Bold dotted lines indicate the location of the İzmir-Balıkesir Transfer Zone; white and black arrows show the reactivation of the transfer zone as sinistral and dextral strike-slip faults, respectively (b) Simplified geological map showing the curvature of the Manisa Fault Zone (compiled from Bozkurt & Sưzbilir 2006; Ưzkaymak & Sưzbilir 2008) See Figure 1a for location of the map Abbreviations PFZ, GFZ, MaF, TaF, KeF, and MFZ, refer to the Paşadeğirmeni Fault Zone, Gürle Fault Zone, Manastr Fault, Talburun Fault, Keỗilikửy Fault, and Manisa Fault Zone, respectively 451 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA a L1 38.7 1250 17(4,8) Manisa Ka 44(2,8) İzmi r İzmi r Bay OF Z 926(6,8) Tu rgutlu 17(6) 750 1595(6,8) Kemalpaşa 1664(1,2) İF 1000 1845(2,8) MF Z F Menemen KF L2 500 250 N 17(2,3) 10 km 38.3 27 27.5 b epicenter focal depth (km) 10 53 28 elevation (m) strike-slip faults inferred faults hypocenter 0 normal faults 10 instrumental earthquakes (M: 5.4 - 2.6) 43 historical earthquakes distance from the fault (km) Figure (a) Seismotectonic map of the İzmir-Manisa region, showing the epicentres of both instrumental and historical earthquakes Instrumental earthquakes exceeding M 5.4 are reported from 1902 to 2010 Magnitude is shown by the size of the red-filled circles Epicentres of historical earthquakes that affected the region and specifically damaged the city of Manisa are shown by blue stars Dates and coordinate references (given in brackets) of the historical earthquakes are given above or to the right of the symbol (see Table 1, for details) The map also shows fault-plane solutions of two earthquakes (L1: 28.01.1994, Mb: 5.2 and L2: 16.12.1977, Mb: 5.3) Abbreviations MFZ, KaF, İF, KF, and OFZ, refer to the Manisa Fault Zone, Karaỗay Fault, zmir Fault, Kemalpaa Fault, and Orhanlı Fault Zone, respectively (b) A cross section across the Manisa Fault Zone showing the hypocentre and epicentre of the Manisa earthquake on 28.01.1994 The location of focal depth and the epicentre of the earthquake are taken from Taymaz et al (2004) and Tan et al (2008) Note that the dip of the surface-slip vector (53°) was measured in the field, while the dip of the fault at the hypocentre (43°) was obtained from the fault-plane solution The hypocentre of the earthquake is marked by a red-filled star largest and most damaging earthquake ever known in the Manisa Basin An intensity distribution map of the 17 AD earthquake (Guidoboni et al 1994) shows a relatively large amount of damage to 13 ancient cities; the authors mention that especially in Magnesia (Manisa) and Sardeis (Sart), located almost at the centre of the intensity map, there were wide and deep surface ruptures Twenty-seven years after this 452 event, in 44 AD, the ancient cities of Magnesia and Ephesus (Efes) were shaken by an earthquake with an intensity of VIII (Ergin et al 1967; Soysal et al 1981, Table 1) Ambraseys & Jackson (1998) refer to an earthquake in Manisa in August of 926, although there is no detailed information about this event Ambraseys & Finkel (1995) and Ambraseys & Jackson (1998) reported an earthquake that caused Ç ƯZKAYMAK ET AL Table List of recorded historical earthquakes in the Manisa region References: (1) Ergin et al 1967; (2) Soysal et al 1981; (3) Ambraseys 1988; (4) Guidoboni et al 1994; (5) Papazachos & Papazachou 1997; (6) Ambraseys & Jackson 1998; (7) Ambraseys & Finkel 1995; (8) Tan et al 2008 I– intensity M– magnitude N Date 17 (1, 2, 3) Coordinate Lat (N)–Long (E) Locality affected 38.40–27.50 (2, 3) Manisa (Magnesia), Muradiye, Sart (Sardes) (2); Magnesia, or 38.5–27.8 (6) or Sardes, Temnos, Myrina, Ephesus, Appolonia, Hyrcanis, 38.6168–27.3992 VS Mostheni, Aegae, Hierocaesaria, Euthena, Ulloron, (4, 8) Philadelphia, Tmolus, Cyme, Thyatira (3); Gediz River (6) Magnesia, Ephesus (2) 44 (1, 2) 38.50–27.40 (2) August 926 (6) 38.50–27.50 (6) 38.50–27.90 (6) Manisa (6) Manisa, Urganlı, Sart, Ahmetli, Gedik, Bostancı, Hamza I M IX (2) 7.4 (4, 8) VIII (2) 1, 2, ? 6, 7, June (7) 1664 (1) 38.41–27.20 (1, 2) İzmir (2), Manisa (7) VII (1,2) June 23 1845 (1, 2, 5) 38.60–27.50 (2) Manisa (1, 2, 5) VIII (2) The instrumental earthquake data of Figure 2a, whose magnitudes range from 2.6–5.4 in the rectangular area specified by the coordinates, are acquired from sources documented by Tan et al (2008) and KOERI (2010) According to these reports, the most recent earthquakes occurred on 28 January 1994 near Manisa with a magnitude of 5.2 at a depth of 10 km (Figure 2b) and on 16 December 1977 near İzmir with a magnitude of 5.3 at a depth of 24 km (Tan et al 2008) Solutions of their focal mechanism indicate the existence of normal faulting with a minor right-lateral slip component (Figure 2a) The Manisa earthquake triggered 13 aftershocks within two months after the main shock, all recorded by a local seismic network (KOERI) The spatial distribution 6, 6, September 22 1595 (6, 7) surface ruptures between the cities of Manisa and Ahmetli on 22 September 1595 Another earthquake in İzmir on June 1664 with an intensity of VII has been reported (Ergin et al 1967; Soysal et al 1981; Ambraseys & Finkel 1995) According to Ambraseys and Jackson (1998), this earthquake occurred near İzmir, probably in Manisa The last historical and destructive earthquake, which damaged the city of Manisa, occurred on 23 June 1845 (Ergin et al 1967; Soysal et al 1981) The location of this earthquake is given as the city centre of Manisa, and its intensity was VIII (Soysal et al 1981) According to Papazachos & Papazachou (1997), the MFZ is thought to have produced historically significant earthquakes such as the M= 6.7 event in 1845, and fault activity was also manifest during an M= 5.2 earthquake in 1994 1, 2, 3, 4, ? Çavuş, Azizlü villages (7) Ahmetli (6) Reference 1, 2, 6.7 (5) 1, 2, 5, of the main shock and aftershocks with magnitudes between 3.5 and 4.0 shows that these shocks form a cluster approximately 10 km away from the MFZ (Figure 2a) As Figure 2b shows, this event appears to have a focal depth of up to 10 km which indicates that recent earthquakes in the Manisa Basin originate near the base of the brittle upper crust The Manisa Fault Zone: Fault Geometry and Segment Characteristics The fault zone, a 35-km-long northeastward arched active corrugated fault system with distinct alongstrike bends, trends NW–SE for some distance in the southeast, then bends into an approximately E–W direction in the north (Figure 1b, Bozkurt & Sözbilir 2006) It cuts and displaces Miocene lacustrine carbonates in the footwall and Upper Pleistocene to Holocene deposits in the hanging wall The eastern sector of the well-defined fault trace has been mapped in detail for more than 15 km (Bozkurt & Sözbilir 2006) and can be followed westwards up to a possible total length of about 35 km (Özkaymak & Sözbilir 2008) According to geological and geomorphological investigations, the MFZ has been geometrically and kinematically characterised as a typical Aegean-type active fault similar to the Tyrnavos Fault (Thessaly, Central Greece) (Caputo et al 2004) The MFZ was divided into separate segments arranged en échelon Individual array length varies from one to several kilometres During the early 453 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA early fault geometries prior to linkage 4,5 3k km H G post-linkage geometry 67 Manisa m 1,5 F km km 10 km 65 O 65 10 km O 76 O 53 km D C KoR ÇaR B GuR N 50 KıR O km km SıR KaR km KocR A a O 15 E km O ER b km 1500 N 1000 500 0m altitude re Manisa Figu alluvial fan normal fault dextral strike-slip fault c sinistral strike-slip fault Figure Fault geometry and segment characteristics of the Manisa Fault Zone (a) Early fault geometry prior to linkage, (b) post-linkage geometry of the fault zone, (c) a 3D block diagram showing present-day configuration of the study area Note that the large alluvial fans frequently take advantage of transfer fault zones and areas between en échelon faults Note also that early fault segments are km long on average prior to linkage; but after linkage and breaching of the fault segments, the fault zone resulted in three main segments of about 10 to 15 km long Abbreviations GuR, KaR, KoR, ÇaR, KıR, SıR, KocR and ER refer to the Gỹrleỗay River, Karaỗay River, Kocadere River, ầayba River, Krtk River, Srtlangửỗỹ River, Kocakzl River and Eref River respectively linkage stage, faults were not connected but remained as isolated fault segments separated by a distorted ramp (soft linkage) or transfer fault (Figure 3a) As the faults grew, the fault segments became connected along the strike to form a continuous zigzag-shaped fault trace (hard linkage) Thus, kilometre-scale segmentation occurs as the displacement changes from one fault to another and is accommodated either by a relay ramp or transfer fault (Figure 3b, c) Similar segmentation has been described in Greece (Roberts & Jackson 1991) 454 Our studies suggest that the courses of some northeast-flowing rivers, which correspond to the NE–SW-striking faults, may mark segment terminations One of these is the Kırtık River, which enters the graben km southeast of Manisa (Figure 3b) where the MFZ bends to align E–W To the west, the Karaỗay River, km west of Manisa, enters the graben where there is an offset in the fault of about 0.5 km The Gỹrleỗay River enters the graben at the end of the Manastır Fault (Figures & 4) These rivers carry the largest sediment loads and thereby Ç ƯZKAYMAK ET AL dominate lateral alluvial fan deposition in the area Thus, the largest fans with the coarsest sediments develop at the ends of fault segments These NE– SW-trending strike-slip faults (i.e the Gürle Fault Zone (GFZ) and the Karaỗay Fault (KaF)) are clearly recognised by morphologically deep valleys in the west and east of the study area (Figures 1b & 3) The Karaỗay Fault comprises several well-exposed fault surfaces with a relief of to 10 m and displays well-preserved slickensides Structural observations on these slickensides show that the Karaỗay Fault is a polyphase structure Evidence for reactivation is similarly established on the slip surfaces of NW– SE-striking normal faults (see Ưzkaymak & Sưzbilir 2008 for a detailed description of NE–SW-trending strike-slip faults) The GFZ is about 1.5 km wide, and consists of parallel-subparallel bifurcated fault segments which juxtapose Neogene volcanic rocks and the Bornova Flysch Zone This fault is cut by the NW–SE-striking Paşadeğirmeni Fault Zone (PFZ) which can be traced northwest of Paşadeğirmeni Hill as a single fault about km long: to the SE it consists of two fault branches about km long In the southeastern part of the PFZ, we mapped three en échelon faults, one of which is the Manastır Fault This fault, which is 4.5 km long, defines the southwestern boundary of the Manisa Basin Its trace is marked by large well-preserved scarps with maximum heights of 140 m, a series of screes, landslides, and triangular facets Along the range-front, the faceted spurs show at least two generations of facets (Figure 3c) This suggests that the front has experienced at least two uplift periods separated by tectonic quiescence The bottom of the front shows evidence of very recent movement, such as a well-preserved and continuous scarp The scarps have all the main characteristics of the fault scarps described in different erosion models (e.g., Wallace 1977; Nash 1980; Mayer 1984) This type of morphotectonic structure has also been observed on the eastern sector of the active Tyrnavos Fault (Thessaly, central Greece, Caputo 1993; Caputo et al 2004) Northeast of the Manastır Fault, there is a 3-kmlong fault (i.e the Taşlıburun Fault) at a left-step configuration (Figure 1b) These two segments are connected by a N–S-trending fault Another segment (i.e the Keỗilikửy Fault), 1.5 km long with resistant, striated, and corrugated fault planes developed within Mesozoic carbonates of the Bornova Flysch Zone, was mapped northwest of the Taşlıburun Fault The main fault planes are characterised by millimetre-scale frictional-wear striae, and metrescale corrugations are typically underlain by several centimetre- to metre-thick fault gouges preserved within the footwall It is possible that the three leftstepping en échelon segments merge into a single fault plane at seismogenic depths All segments are clearly defined by linear escarpments, across which variable reliefs exist (20–400 m) It is notable that the southern segments occur at higher elevations (Figure 3c) They display a persistent zigzag trace with a total length of up to 10 km (Ưzkaymak & Sưzbilir 2008) Based on empirical relationships (Wells & Coppersmith 1994; Pavlides & Caputo 2004), a 10-km-long fault is capable of generating an earthquake with a magnitude of 6.5 The length of this segment is similar to the thickness of the seismogenic layer in the region (e.g., 10 km, Akyol et al 2006) We mapped several northeast dipping faults between the Manastır Fault and the Paşadeğirmeni Fault Zone with a similar strike (N60°W) They are composed of a series of parallel fault strands, each with a much lower relief of typically m, spread over a region km wide All cut and deform the upper Pleistocene–lower Holocene Emlakdere Formation, and thus have been active at least until the Late Holocene (Figure 4b) Fault-1 is closest to the Manastır Fault It cuts and back-tilts the strata of the Emlakdere Formation The vertical offset associated with rotation of a downthrown block around the horizontal axis implies the activation of Fault-1 Fault-2 is indicated by a discontinuous morphological scarp 1.5 km long running parallel to the trace of the Manastır Fault Fault-3, which can also be traced as a morphological scarp, sinistrally offsets the N–S-trending and northflowing stream along its eastern portion These may be shallow synthetic normal faults caused by refraction of the Manastır Fault Although they may move simultaneously with the Manastır Fault, they merge with it at depth and may even then produce negligible seismic moment The colluvial sediments of the Emlakdere Formation are likely to have been deposited at dips of about 5–10° northwards but now dip at angles of 455 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA about 50–60° southwards (Figures & 5a–c) Backtilting of the down-thrown block towards the fault scarp was observed along the quarry road, suggesting the listric geometry of the Manastır Fault Listric faults, flattening at shallow depth, are common features in extensional tectonic environments They may produce repeated surface displacements during strong earthquakes, thus generating typical tectonicgravitational landforms such as secondary faultbounded tilted blocks (Dramis & Blumetti 2005) Stratigraphy and Facies Analysis of the Quaternary Deposits While Neogene tectonic evolution of the West Anatolian extensional province has been well studied (Emre & Sửzbilir 1997; Koỗyiit et al 1999; Bozkurt 2000; Sửzbilir 2001, 2002; Bozkurt & Sửzbilir 2004; ầiftỗi & Bozkurt 2008, 2009, 2010), little information is available on its Quaternary evolution (Hakyemez et al 1999; Bozkurt & Sưzbilir 2006; Ưzkaymak & Sözbilir 2008) The indicators of Quaternary deformation are mainly located in the studied area Structural analysis of normal faults offsetting Quaternary sediments was based on the study of striated fault planes and offsets of the upper Pleistocene–lower Holocene Emlakdere Formation (Ưzkaymak & Sưzbilir 2008) These types of structures are scarce and concentrated mainly in the western end of the MFZ, between the villages of Gürle and Emlakdere (Figure 4) The region in which the study area is located contains four unconformity-bounded units: the upper Cretaceous–Palaeogene Bornova Flysch Zone, a Miocene volcano-sedimentary unit, the upper Pleistocene–lower Holocene age Emlakdere Formation, and upper Holocene modern graben fill (Figure 4) The Emlakdere Formation Back-tilted colluvial and alluvial sediments in the western part of the MFZ were mapped and named the Emlakdere Formation by Özkaymak & Sözbilir (2008) and dated as late Pleistocene–early Holocene in the present study The Emlakdere Formation comprises unsorted crudely stratified gravel and cobble-pebble conglomerate alternating with several palaeosol layers (Figures 5a, b & 6) The unit is 456 overlain with an angular unconformity by upper Holocene colluvial/alluvial fans (Figure 5c) In an attempt to obtain a representative typesection of overall upper Pleistocene–lower Holocene debris-flow activity within the area, two sections were measured along the quarry road (Figure 6) They proved to cut approximately the full thickness of the Emlakdere succession Individual sediment layers were characterised in terms of matrix and clast characteristics, grain-size, grading, sediment structures, and the nature of contacts between adjacent layers (Harms et al 1975; Collinson & Thomson 1982; Blikra & Nemec 1998; Sletten & Blikra 2007) In each profile, we identified the main units and morphological features; we also measured soil colour using the Munsell® colour book (Munsell Color Company 1994) The characteristics and interpretation of the main sedimentary facies are outlined below and illustrated in two measured sediment logs (Figures & 7) The observed sedimentary facies can be differentiated into four major groups: rock fall, debris fall, debris flow, and palaeosol The diagnostic sedimentological characteristics of the main sediment types are shown in Figure Debris Flow– This facies consists of tabular beds with large floating clasts Large clasts are mainly aligned downflow Most of the beds show inverse grading and are characterised by a matrix-rich to clast-supported sandy-muddy matrix including boulders up to 212 cm in diameter (Figures & 8a) Some beds are lenticular with imbricate or more complex stacking These deposits are interpreted as proximal to distal facies of high- to low-viscosity debris flows (cf Blikra & Nemec 1998) Debris Fall– This facies is characterised by immature to mature debris including subangular to subrounded clasts (Figures & 8b) Massive to upward-fining and typically clast-supported units are common Deposits are often infilled with sandy mud, although some openwork structure is also visible Rock Fall– This consists of highly immature debris, mainly angular clasts in a pebbly to sandymuddy matrix (Figures & 8c) Massive to normallygraded and clast-supported units showing openwork structures are common (Figure 8d) Clasts in such units are up to macross Some rock-fall units Ç ƯZKAYMAK ET AL a PF Z 28 28 29 60 C N GFZ G ỹrle ỗ a y r i v er 42 75 000 50 27 000 Gürl e Village 74 153 Paşadeğirmeni Hill T2 Mugirtepe Hil l 70 T1 S1 61 18 S2 157 Taşlıburun Ridge 158 52 12 10 F Ta 18 33 B MaF Emlakdere Village 73 Manastır Dağ ı H igh Sivritepe Hill 758 A Yassımeşe Ridge 72 662 Taşlıbayır Hill Aỗarlktepe Hill 768 723 b SW late Holocene recent alluvium late Pleistocene-early Holocene 750 m 500 m NE SSW colluvial / alluvial fan Emlakdere Formation Neogene volcanic rocks Neogene volcano-sedimentary rocks late Cretaceous–Paleocene A 250 m basement rocks NNE MaF Mugirtepe Hill PFZ strike-slip fault normal fault scarp normal fault inferred fault line of cross section measured stratigraphic section (S1 and S2) 250 m C Figure (a) Detailed geological map of the study area showing NW–SE-trending active faults and location of trench sites on the southern branch of the PFZ (see Figure 1b for map location) Abbreviations PFZ, MaF, TaF, GFZ, T1, T2, S1, and S2 refer to the Paşadeğirmeni Fault Zone, Manastır Fault, Taşlıburun Fault, and Gürle Fault Zone, Trench-1, Trench-2, stratigraphic section 1, and stratigraphic section 2, respectively (b) Geological cross section showing stratigraphic and structural relationships of the units Note that the offset of Quaternary deposits by several instances of synthetic Holocene faulting in the hanging wall of the Manastır Fault is the most direct evidence for their activity 457 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA S N PS3 PS6 PS5 PS2 PS4 a S PS12 PS10 N PS6 b SE c NW dip of the layers unconformity Figure Field views showing (a) lower, (b) middle, and (c) upper part of the Emlakdere Formation Note the steeply dipping lower part and gently dipping upper part of the bedding exhibit a flow-parallel orientation a(p), indicating sliding fabric Bed thickness is typically 70–200 cm Contacts between units are generally sharp but apparently conformable, indicating little or no 458 erosion of underlying units The colour of the matrix varies from light-yellowish brown (10YR 6/4 in Munsell colour value) to reddish yellow (7.5YR 6/6), similar to the underlying organic soils PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA DEPOSITIONAL SEDIMENTARY FEATURES PROCESSES debrisflow rockfall/debrisfall fresh rock debris TYPE/GEOMETRY varied runout OF DEPOSITS AVALANCHES relatively broad resedimented lobes gravel highly elongate tongue-shaped lobes (upslope fining) levees upslope fining spill-over lobes tree-dimensional view scattered clasts lentic vertical cross-section lobate or 'patchy' accumulations of debris; scattered large 'outrunners' ular b high-viscosity debrisflow upward eds coarse low-viscosity/watery debrisflow parallel imbricate beds ning upward fining lenticular beds with openwork tabular infilled imbricate or more beds large by 'tail' complex stacking 'floating' clasts immature to matrix-rich to clasthighly immature debris clast-supported, debris; supported mainly angular clasts mature bouldery to cobby subangular to sandy/muddy matrix 'heads' and clast to subrounded common 'coarse tail' matrix-supported, pebbly clasts inverse grading and upslope 'tails', outsized cobbles or common normal grading boulder to sand size grade, boulders clast-supported and commonly openwork, common carbonate-rich slope wash with pebbly to sandy-muddy infill at the material derived from weathering top Deposits often infilled with sandy limestone bedrock mud and redeposited soil material upward fining openwork TEXTURE AND STRUCTURE CLAST FABRIC DEBRIS SOURCE boulders and large cobbles often shown 'rolling' fabric,a(t) or a(t)b(i) many large clasts upslope show 'sliding' fabric a(p).But a disordery 'adjustment' fabric predominates; 'shear' fabric a(p) often typifies the avalanche’s overriding tail, when evolved in to grainflow weathered bedrock older debrisflow sediments large clast mainly common 'rolling' fabric aligned downflow, a(t) in the frontal and top part of the a(p) or a(p)a(i), debrisflow head; but showing a(t) common 'shear' orientation along fabric a(p) or a(p)a(i) the lobe front in the flow’s tail upper-slope colluvium Figure Diagnostic sedimentological features of the main sedimentary facies observed in the measured stratigraphic sections of the Upper Pleistocene–Lower Holocene Emlakdere Formation (compiled from Blikra & Nemec 1998) 460 Ç ÖZKAYMAK ET AL 212 cm a b c dip of the layer PS3 S d N Figure 8f e f Figure Field photographs of the Emlakdere Formation, showing colluvial facies and depositional processes (a) Debris-flow deposits Immature cobbles and boulders are randomly distributed within the sandy muddy matrix, the largest one 212 cm long (b) Resedimented debris-fall deposits consist of relatively mature subrounded to rounded gravels (c) Close-up view of the rock-fall deposit composed of boulders and large cobbles Clasts are angular, very immature, and show random orientation (d) Closeup view of the clast-supported lenticular bed within debris-flow deposits Note that the non-stratified openwork gravels show multimodal grain-size distributions (e) View of the steeply dipping section of the Emlakdere Formation showing debris-flow units alternating with a thick palaeosol (f) Close-up view of PS3 palaeosol level given in Figure 8e Note that the upper levels of the palaeosol have vertical/subvertical cracks filled by carbonate-rich sediments 461 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA Palaeosol– Fourteen buried soils were found downslope of the Manastır Fault plane Organic layers in the sediment successions range in thickness from a few to 118 cm (Figure 5) Characteristically, greyishyellowish brown beds (10YR 5/2) mostly include randomly dispersed limestone clasts Some of the thick palaeosols include randomly dispersed cracks filled by clay- and carbonate-rich materials (Figure 8e, f) Bioturbation was not observed in these levels A total of samples were extracted for radiocarbon dating from different palaeosols interbedded with coarse-grained debris units (see Figures & for the locations of the samples) Radiocarbon ages were calibrated to calendar years using the programme OxCal 4.1.3 (Ramsey 2009) based on atmospheric data from Reimer et al (2009) Age ranges are rounded to the nearest decade The range of possible ages is shown only for the best level of confidence (2σ= 95.4%) Dating analyses were carried out in the laboratory at the University of Arizona (AA) and the laboratory for Ion Beam Physics (ETH) in Zurich Results of the radiocarbon dating are given in Table [see Hajdas et al (2004) for a detailed description of sample preparation] As a result, the bulk sediments from the Emlakdere Formation yielded radiocarbon ages between 19030±560 and 9199±67 cal yr BP, indicating a late Pleistocene–early Holocene age (Table 2) Samples from the lowest palaeosol level of the formation yielded a radiocarbon age of 19030±560 cal yr BP Younger ages of 16210±90 and 14270±360 cal yr BP were obtained on samples from the upper part of the formation (see Figure 6) In addition, in order to evaluate the ages of the uppermost Emlakdere sediments located in the hanging wall of fault-1, we collected a sample for radiocarbon dating from organic-rich sediments, which yielded an age of 9199±67 years BP (Table 2, Figure 9a, b) This result led us to two basic conclusions: (1) deposition of the Emlakdere Formation ended at the beginning of the Holocene and (2) fault-1 is post-Early Holocene in age Late Holocene Alluvial/Colluvial and Fluvial Units The recent graben floor is filled by lateral alluvial fans of diverse size and axial fluvial sediments (Figures 1, & 4) The lateral fans and the axial fluvial deposits interfinger and are typical of continental basin fill with axial through-drainage (Leeder & Gawthorpe 1987) In these areas, colluvial slope deposits often merge with those of adjacent alluvial fans, forming an alluvial/colluvial wedge that extends along the base of the valley wall Although alluvial fans represent sediments deposited from side canyons into the graben, the axial river is composed of sand/ clay and gravel, which is deposited in a meandering stream channel of the N–NE-flowing river We dated the upper part of the colluvial deposits as 988±37 cal years BP, indicating probable Late Holocene deposition Palaeoseismology The southwestern part of the Manisa Basin is covered by large predominantly upper Holocene alluvial fans Their surfaces are displaced by several normal faults trending semi-parallel to the strike of the main lineament One of these, the southern branch of the PFZ (i.e the Mugirtepe Fault) was selected for palaeoseismic studies because it showed the most prominent and well-preserved fault scarps displacing Table Radiocarbon age data from the western part of the MFZ Note that the 14C data obtained from the Emlakdere Formation suggest a Late Pleistocene–Early Holocene age of formation Sample number Unit Laboratory ID no Material type Radiocarbon age (BP) Two sigma range (AD or BC) (95.4 %) M0724 colluvial fan AA-78578 charcoal 988±37 990 AD – 1150 AD M0727 Emlakdere Fm AA-78579 bulk sediment 9199±67 8600 BC – 8290 BC MED09 Emlakdere Fm ETH-37114 bulk sediment 14270±360 16540 BC – 14810 BC MED08 Emlakdere Fm ETH-37113 bulk sediment 16210±90 17610 BC – 16990 BC MED03 Emlakdere Fm ETH-37110 bulk sediment 19030±560 22220 BC – 19510 BC 462 Ç ƯZKAYMAK ET AL 9,199 67 (BP) fault 140 m S Emlakdere Formation 135 m RS N ,6 11 a b Karadağ Formation 8m 5m Figure Field photograph (a) and a detailed sketch from (b) Emlakdere Formation against the fault Black-filled circles show the locations of samples for radiocarbon dating upper Holocene alluvial fan surfaces Trenches were opened across this fault because of its proximity to the bottom of the graben, which we expected to be filled by Late Holocene sediments and soils: hence its suitability for palaeoseismic trenching The Mugirtepe scarp has a NW–SE strike and a maximum vertical topographical separation of about m It separates Mesozoic limestone from Holocene sediments In order to date the most recent movements of the fault, two trenches were dug perpendicular to the scarp in the Holocene deposits of the hanging wall (trenches T1 & T2) The exact location of these trench sites was decided by the results of a groundpenetrating radar investigation showing several cumulative dislocations of the dielectric horizons The clear geomorphic expression of the fault scarp, as well as the presence of fault-like structures visible in the geophysical profiles intersecting Quaternary alluvial sediments, were the main criteria for site selection The microstratigraphy of the trenching area was determined by detailed mapping of the trench walls (1: 10 mapping) and C14 dating of the lithostratigraphic units Detailed microstratigraphic and structural analyses were performed for each trench on both walls Accordingly, in Figures 10 and 11, we represent only one side of each trench In total, 10 stratigraphic units were identified and are presented in detail in Table The individual trenches are described below Trench-1 (T1) Trench-1 is N–S-oriented, roughly orthogonal to the local trend of the fault trace It is nearly m wide and about 10 m long On the hanging wall, the oldest unit (unit 2) consists of dark-reddish-brown sandy clay (Figure 10a, b) The conformably overlying unit consists of moderate-reddish-brown silty sand and is cut and displaced by the dip-slip normal faults as is unit Unit is overlain by intercalations of lightyellowish-pink to strong-yellowish-brown gravelly sandy mud and pale blue-green gravelly fine-grained sand (unit 4) This unit also includes randomly distributed clay and lime nodules A light-reddishbrown organic soil horizon (Unit 5) overlies the older unit These units are cut and displaced by a normal fault forming a 50-cm-thick-sheared fault zone The minimum vertical displacement of the fault is m with respect to the upper contact of unit observed at the hanging-wall block A cone-shaped structure observed between unit and the fault zone is filled by sediments of units and On the hanging wall, unit consists of highly jointed reddish-yellow to pale yellowish-green silty sand Unit comprises pale yellowish-green well-sorted gravelly sand and has an erosive base into unit 5, with discontinuously preserved lenses of clast-supported cobbly gravel It shows a cone-shaped geometry that dies out towards the footwall block 463 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA 7 1 S a N RS 1,638 57 (BP) unit unit cf unit unit unit unit unit b striations s1 s1 c1 s2 striations s2 c c2 Figure 10 Field photographs and logged section of trench-1 (a) Oblique photomosaic and (b) log of trench-1, eastern wall (see Table for description of lithostratigraphic units) (c) Close-up views of trench-1 showing two slip surfaces with distinct rake angle Note that the C1 surface overprints the C2 surface, indicating reactivation of the fault White plots represent the stereographic projection of the planes 464 Ç ƯZKAYMAK ET AL a 5m 1m b SW 2m unit 10 5m NE unit 798 37 (BP) ch c RS 156 37 (BP) unit unit unit Figure 11 (a) Field view of the Mugirtepe scarp (the southern branch of the PFZ), with (b) photomosaic and (c) log of trench-2, western wall, excavated across the Mugirtepe scarp (see Table for description of lithostratigraphic units) White plots represent the stereographic projection of the fault plane 465 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA Table Description of lithostratigraphic units identified in the trenches, their estimated and absolute ages and interpretation of their environments Unit Interpretation Lithology Base Sedimentary structures and remarks color (Munsell code) RS recent soil silts, sand, mud and organic material gradual average slope of 7° in trench1 and 2° in trench2 greyish yellowish brown 10YR 5/2 (trench1) light yellowish brown 10YR 6/4 (trench2) highly immature debris gravel to sand size grade gradual mainly angular clast, partly clast supported average slope of 6° moderate yellowish brown 10YR 5/4 unit 10 colluvial wedge-2 unit palaeosol muddy silts and organic material sharp average slope of 5° moderate yellowish brown 10YR 5/4 to light yellowish brown 10YR 6/4 unit fluvial deposits gravely sand erosive mainly well sorted, loose, includes charcoal fragments pale yellowish green 10GY 7/2 unit fluvial deposits silty sand sharp highly jointed; cracks filled by carbonate-rich clay reddish yellow 7.5YR 6/6 pale yellowish green 10GY 7/2 (ch) colluvial wedge-1 highly immature gravels and boulders, sandy matrix sharp unsorted, angular clast partly clast supported pale yellow 5GY 8/6 cf crack fill muddy silt and sandy mud faulted v-shaped in plan view, crack-fill materials mainly derived from unit and strong yellowish brown 10YR 5/6 unit palaeosol muddy silt, and organic material gradual average slope of 9°, includes charcoal fragments light reddish brown 2.5YR 5/4 unit fluvial deposits gravely sandy mud (a) alternating with gravely finegrained sand (b) erosive consist of clay and lime nodules; limestone gravels characteristically parallel lamination (mm–cm scale) strong yellowish brown 10YR 5/6 (a); pale blue green 5BG 8/2 (b); light yellowish pink 10R 9/6 © unit fluvial deposits silty sand sharp well–consolidated moderate reddish brown 2.5YR 4/4 unit fluvial deposits sandy clay unit basement rock limestone rocks unit 156 ± 37 years (BP) 798 ± 37 years (BP) 1638 ± 57 years (BP) dark reddish brown 10R 3/4 massive bedrock carbonates The upper portion of the exposure in trench-1 is incomplete and truncated by the basal surface of an agricultural layer Normal faults dissect the sequence below the agricultural layer recording at least one event at this site In the central sector of the trench, a well-defined sheared zone can be observed, consisting of numerous shortly spaced shear-planes oriented parallel to the main contact surface We have obtained two different slip surfaces including distinct rake angles on the shear planes The computed results of these shear surfaces indicate oblique-slip normal faulting The C2 surface including the slip lines with an average rake angle of 292° are superimposed by a C1 surface including striations with an average rake angle of 320° (Figure 10c) Trench-2 (T2) Trench-2 trends N–S, almost orthogonal to the trend of the fault trace, represented by a scarp more than 466 Age medium light grey N6 Upper Cretaceous– Palaeogene m high (Figure 11a) The trench is nearly m wide and more than m long, and exposes colluvial, fluvial, and palaeosol units which overlie the bedrock (unit 1) In the footwall, the displacement of the NWtrending fault has generated a relatively fresh scarp 2.5 m high on Mesozoic limestones The sequence exposed in the hanging-wall block includes at least six sedimentary units (Figure 11b, c) In the downthrown block, the older soil is traced beneath the colluvial wedge-1 (unit 6), containing blocks of limestone derived from the upthrown block They are located at the base of the free face composed of Bornova Flysch Zone bedrock The toe of the colluvial wedge is composed of highly immature clasts with pale-yellow sandy matrix Note that the largest blocks rolled to the tip of the wedge Clasts in the lower colluvial wedge range between millimetres to centimetres in size and are generally subangular to angular The matrix is locally abundant close to the fault plane, and commonly the texture is clast- Ç ƯZKAYMAK ET AL supported at the tip of the wedge This facies is similar to the sorted debris facies of Nelson (1992) The lower colluvial unit is m thick and overlain by fluvial deposits (unit 7), which were affected by a subsidiary normal fault, unconformably overlain by a light- to moderate-yellowish-brown palaeosol Unit is restricted to a channel that parallels the fault and was probably deposited by a stream diverted into the fault trace after faulting Unit represents a palaeosol overlying the stream deposit of unit Unit 10 consists of highly immature colluvial gravel and coarse sand embedded in a midyellowish-brown muddy matrix (Figure 11a, b) The poorly-sorted, sand-rich gravel with cobbles and boulders of bedrock limestones, passes upward to fining-upward cobble and then to granular gravel horizons, and corresponds well with the ‘debriselement’ facies association (wedge-shaped deposits resulting primarily from degradation of the free face) of Nelson (1992) The soil, which represents the present-day pedogenic layer, is a greyish-yellowish and lightyellowish-brown layer about 30 cm thick, consisting of relatively organic-rich fine-grained material with scattered subangular carbonate clasts The soil developed on a post-faulting colluvial wedge merges downslope with the pre-faulting soil (unit 9) away from the fault (Figure 11a, b) The soils represent periods of landscape stability, whereas the colluvial deposits represent periods of landscape adjustment after a fault rupture event (Machette 1978; McCalpin 2009) Therefore, soil horizons on and under colluvial wedges were used to date rupture events on the fault (see McCalpin 2009 for further reading) Materials for radiocarbon dating including detrital charcoal and palaeosols were collected in the trenches, providing age constraints for the stratigraphy and history of active faulting The age of the succession was constrained by means of absolute 14C Samples were sent to ETH laboratory (Zurich, Switzerland) and the Arizona laboratory (Phoenix, Arizona) for absolute dating Table summarizes the radiometric (R) 14C analyses of sampled detrital charcoal and soil/ deposit bulk The detrital charcoal sample was subjected to a standard pretreatment consisting of acid/alkali/acid washes The relative 1σ (68%) and 2σ (95%) areas under the probability distribution are shown on the right side of Table In the text and in the trenchlogs, the 2σ calendar age is reported for all samples Detrital charcoal from a pre-event palaeosol (unit 9) yielded a radiocarbon age of 156 14C cal yr BP and a calendar age of 1670 AD We collected a bulk sample from the unit where it exhibited the darkest colour Radiocarbon dating of the sample yielded 798 yr BP Bulk sediment from the palaeosol (unit 5) yielded radiocarbon ages of 1638 BP The corresponding calendar age is 260 AD Interpretation of Palaeoearthquakes The displaced upper Holocene sediments exposed in the trenches are the youngest faulted deposits at this site, and provide evidence for the most recent surface faulting at the western end of the MFZ Unfortunately, differences in sedimentary environments, the truncation of the upper part of the sequences, and uncertainties in the available ages prevent a clear correlation of the events between the two trench sites In addition, evidence for the youngest two events was not preserved at trench In trench 1, as post-event erosion has removed the scarp, as well as geological units on the upthrown side of the fault and the colluvial wedge associated with the fault, we could neither correlate geological units across the fault to directly measure the net slip, nor estimate displacement using colluvial-wedge size relations However, we could measure at least m of Table Samples collected from palaeoseismological trenches and and the results of radiocarbon dating Sample number Trench no Unit Laboratory ID no Material type Radiocarbon age (BP) Two sigma range (AD or BC) (95.4%) M0734 Unit AA-78583 charcoal 156±37 1670 AD – 1950 AD M0735 Unit AA-78577 bulk sediment 798±37 1170 AD – 1280 AD M0729 Unit AA-78581 bulk sediment 1638±57 260 AD – 550 AD 467 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA vertical displacement across the fault by projecting the upper contact of unit from the down-dropped side of the fault to the upthrown side Considering the vertical displacement on the fault, we expected that the trench might contain evidence for two or three events Evidence for event is provided by a series of normal faults diffused in a 50-cm-wide sheared zone that deformed the sequence up to unit However, the presence of a single crack fill and only one faulted palaeosol (unit 5) indicates that only one older earthquake is recorded in the trench Thus, the exposure in trenchwall records one coseismic deformation event that occurred after 1638 yr BP This is the oldest documented geological record of a palaeoearthquake in the MFZ In the hanging wall of trench 2, two post-event, scarp-derived wedge-shaped colluvial deposits are present on the downthrown side of the Mugirtepe Fault Similar colluvial deposits, which result from the erosion of nearby steep fault scarps, are the characteristic stratigraphic signatures of normal-slip surface faulting (McCalpin 1996) Several studies deal with the existence of syntectonic sedimentation in the form of colluvial-wedge deposits (e.g., Machette 1978; Nelson 1992; McCalpin 1996, 2009) Colluvium, however, is not the only type of sediment that occurs in the wedge-shaped space created on the hanging-wall block of the Mugirtepe Fault as a result of fault rupture Fluvial sediments were also deposited along the fault scarp created by a normal fault (Nelson 1992) The studied wedge-shaped space includes colluvial deposits produced by shedding of material into the depression adjacent to the footwall block of the fault and fault-parallel fluvial deposits The colluvial wedge closest to the fault can be in either depositional or fault contact with footwall deposits (McCalpin 2009) The upper colluvial wedge is in depositional contact with the fault face, suggesting that the colluvium shed from a single-event fault scarp However, the older scarp-derived colluvium is in fault contact with the upthrown block This suggests that at least two discrete displacement events occurred: the first generated the older colluvium and the second faulted it In addition to recent soil, two palaeosols, representing a period of non-deposition, are developed on bedrock and Holocene fluvial deposits 468 in the area The stratigraphically lowest palaeosol is relatively thin and weakly developed It formed on the recrystallised limestone bedrock (unit 1) The second palaeosol (unit 9), which is relatively thick, formed on fine-grained overbank deposits containing channel-fill coarse-grained sandstones Evidence for event is provided by a series of antithetic and synthetic normal faults that cut the sequence up to unit Thus, event cut the sequence up to the base of unit 9, which can be considered to be the exposed surface when this event occurred Samples from units and 9, which pre- and post-date event 2, respectively, constrain this event to between 798 and 156 BP The palaeosol (unit 9) on the downthrown side of the fault (southern branches of PFZ) was buried by a scarp-derived colluvial-wedge, after the most recent surface-faulting earthquake on the fault A calendarcalibrated age of organics from the palaeosol (unit 9) beneath the colluvial wedge places the time of soil burial at 156 cal yr BP Given that the upper colluvialwedge material was derived from the fault scarp produced by the most recent surface faulting at the site, the earthquake occurred just after that time With respect to the scarp degradation model, the maximum thickness of scarp-derived colluvium is limited to half of the free-face height from which it washed (McCalpin 2009) Thus, a first approximation of the initial fault-scarp height is twice the maximum colluvial thickness exposed in the trench In addition, considering the height of the scarp (2.5 m), we infer that the trench contains evidence for two or more events Finally, structural and stratigraphic relationships at both trench sites show the occurrence of at least three individual faulting events However, according to the historical catalogue, six seismic events caused damage in the city of Manisa and its surrounding areas (Table 1) The epicentres of the first and last events were situated near the centre of Manisa and each event had an intensity of VIII However, the seismic events (17 and 1595 AD) were located in the eastern sector near Turgutlu and the Kemalpaşa district According to the palaeoseismological data presented above, evidence for the first two historical earthquakes cannot be documented in the trench logs However, event can be associated with 926 AD, event with 1595 or 1664 AD, and event with 1845 AD (Figure 12) Ç ƯZKAYMAK ET AL OxCal v4.1.3 Bronk Ramsey (2009); r:5 Atmospheric data from Reimer et al (2009) HIST ORICAL EQ EVENT M0734 EVENT M0735 EVENT M0729 Figure 12 Probability distribution of calibrated 14C ages obtained from sequential radiocarbon dates collected from the trench walls Discussion and Conclusion The West Anatolian extensional province is an 800-kmwide region of continental extension consisting of mostly E–W-striking Quaternary normal faults that separate the footwalls of longitudinal horsts from shallow (1–4 km) basins with a typical relief of 100 m The region is one of the most seismically active in the Aegean extensional province The present rate of N–S extension across this part of the province is in the range of 10–20 mm/a (Taymaz et al 1991) The MFZ separates Mount Spildağı to the south from the Manisa Basin to the north To the west, the fault appears to transfer much of its displacement on to the NE-trending faults, whereas the slip on the fault apparently dies out to the southeast Along most of its trace, the MFZ juxtaposes pre-Pliocene bedrock in the footwall with upper Pleistocene to Holocene colluvial/alluvial fans, river deposits, and interfan colluvial aprons in the hanging wall The footwall of the MFZ exhibits some of the best examples of faceted spurs in the West Anatolian extensional province (Bozkurt & Sözbilir 2006) Many localities along the mountain fronts display triangular facets, a geomorphic feature characteristic of normal fault activity Faults showing geomorphic evidence 469 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA for recent activity such as prominent range-front escarpments, V-shaped valleys, and triangular facets are commonly accepted as evidence of active faults (Ganas et al 2004) Along the range-front, the faceted spurs show at least two generations of facets This suggests that the range-front has experienced at least two uplift periods separated by tectonic quiescence The characteristics of the MFZ may provide a key to understanding how new secondary faults develop in the hanging-wall position of a larger fault The migration of normal faults into their hanging walls has been identified from field evidence in the southern margin of Gediz Graben (Paton 1992; Dart et al 1995; Bozkurt & Sửzbilir 2004; ầiftỗi & Bozkurt 2008, 2009) There is little comparable evidence for the relative ages of the main fault and of the secondary hanging-wall faults The clearest evidence for migration of fault activity in the study area is that the uplifted footwall of the basinward fault contains back-tilted strata of the Emlakdere Formation The hanging-wall block of the Manastır Fault contains a network of NW-trending normal faults that postdate the upper Pleistocene–lower Holocene Emlakdere Formation which is uplifted and tilted in their footwalls In this example, the major faults dip northeast and the syn-rift sediments dip southwest, possibly implying that the secondary basinward fault is younger than the main Manastır Fault, although this is clearly speculative in the absence of subsurface data This study also shows evidence for basinward migration of active normal faults Both of the trenches were excavated on one of the branches of the Paşadeğirmeni Fault Zone, which is the longest, straightest basinward fault affecting the Quaternary deposits, as well as the pre-Neogene basement rocks Therefore, it should be the most active fault among those in Figure We found evidence of repeated surface ruptures in the past 1000 years, the most recent dated to a period fitting with the catastrophic earthquake of 1845 AD (M= 6.7), which caused damage to the city of Manisa Our data definitively show late Pleistocene– Holocene activity of the MFZ The most striking evidence of activity along the MFZ was observed on the Manastır Fault and its synthetic minor faults mapped on the hanging wall, which are responsible 470 for the displacement of upper Pleistocene–lower Holocene deposits Offset upper Pleistocene–lower Holocene sedimentary deposits, close to the study area, indicate that most normal faults have ruptured the surface at least once during the Quaternary period The latest snapshot of activity is recorded by seismicity in narrow belts near Alaşehir in the east and the Sığacık Gulf in the west (Figure 1) In both of these belts, earthquake-triggered surface deformations have been documented (e.g., Arpat & Bingöl 1969; Sözbilir et al 2009) Earthquakes in the Alaşehir area generally occur on north-striking normal faults (Arpat & Bingöl 1969; Eyidoğan & Jackson 1985), whereas those in the Sığacık Gulf occur predominantly on NW-striking left-lateral and NE-striking right-lateral strike-slip faults (Aktar et al 2007; Sözbilir et al 2009) Similarly, the Dinar earthquake (M= 6.1) on October 1995 resulted in a 10-km-long surface rupture which was associated with faulting in the hanging wall of the NW-trending normal fault (Eyidoğan & Barka 1996; Altunel et al 1999) The Menderes earthquake on 20 September 1899 was also associated with normal faulting, resulting in a vertical displacement of as much as m (Altunel 1999) To the northeast, in the Manisa Basin, actual earthquakes show predominantly and nearly pure normal motion Our studies suggest that the eastern branch of the MFZ represents a single 15-km-long seismogenic structure associated with earthquakes of about M= 7, whereas the central and western branches represent the expression of an independent 10-km-long seismogenic zone capable of producing earthquakes of M= 6.5 This result is in accordance with the large majority of active faults in the Aegean region, which is characterized by moderate earthquakes (Pavlides & Caputo 2004; Caputo & Helly 2008) We attempted to quantify the slip rate along the Manastır Fault by looking for suitable stratigraphic markers We used the cumulative displacement across the fault, stratigraphically measured from the separation of the correlative basal unconformity of upper Miocene lacustrine limestones (5 Ma; Bozkurt & Sözbilir 2006), which may be the oldest beds of the basin deposited prior to the current extensional regime Our mapping showed uplifted lacustrine beds at a height of about 580 m By correlating the base of the lacustrine units on either side of the fault Ç ƯZKAYMAK ET AL we estimate a vertical offset (throw) of about 600 m This yields a mean slip rate of about 0.1 mm/yr for the Manastır Fault However, based on the position of the upper Miocene lacustrine limestone, Bozkurt and Sözbilir (2006) calculated a vertical Plio–Quaternary offset of 1500 m across the central sector of the MFZ Assuming a maximum time interval of million years for the deformation, a minimum slip rate of 0.3 mm/yr can be computed for this part of the fault zone The slip rate differences between the western and central sectors of the MFZ may thus suggest that they have reactivated independently of one another These deformation values are similar to the slip rates calculated for active normal faults in Greece (Caputo et al 2004; Pavlides & Caputo 2004) In this study, although we find evidence for three historical earthquakes, the data mentioned above is not sufficient to estimate the recurrence interval of earthquakes shaking the city of Manisa However, palaeoseismological studies in Greece have shown that the average recurrence interval of similar active faults is commonly longer than 500 years and usually some thousands of years (Pavlides & Caputo 2004) Further work should be carried out in order to determine: (1) the seismic behaviour of each fault segment of the MFZ, (2) the exact dating and amount of slip of the last movement of these faults, (3) the dating of landslides and of material embedded in the extensional fissure fills, and (4) the cosmogenic dating of the free fault face along the main and secondary faults By obtaining these data, evidence could be provided for the seismic hazard of Manisa Acknowledgements This work is a part of PhD Thesis undertaken by Çağlar Özkaymak at the Institute of Natural and Applied Sciences in Dokuz Eylül University This research was supported by the Dokuz Eylül University Research Foundation (Project number: DEU-BAP-2006.KB.FEN.008) Comments and remarks by Ömer Emre and four excellent reviewers improved the paper and are gratefully acknowledged We thank Özgür Karaoğlu for his help in measuring stratigraphic sections We are also grateful to Erhan Altunel and C ầalar Yalỗner for their help with the fieldwork carried out for trench site selection The radiocarbon dating was performed at the University of Arizona and at the ETH/PSI AMS facility in Zurich We thank Irka Hajdas and Mitzi de Martino for their helpful discussions and contributions on radiocarbon dating Cengiz Zabcı kindly ran the radiocarbon dates using the OxCal calibration programme The paper was edited by the International Science Editing, the English language editing service (Editage) offered by Wiley-Blackwell Mỹjgan Koỗ from the American Cultural Center also helped with the English References Aktar, M., Karabulut, H., Özalaybey, S & Childs, D 2007 A conjugate strike-slip fault system within the extensional tectonics of Western Turkey Geophysical Journal International 171, 1363–1375 Altunel, E., Barka, A.A & Akyüz, H.S 1999 Paleoseismicity of the Dinar Fault, SW Turkey TerraNova 11, 297–302 Akyol, N., Zhu, L., Mitchell, B.J., Sözbİlİr, H & Kekovali, K 2006 Crustal structure and local seismicity in western Anatolia Geophysical Journal International 166, 1259–1269 Ambraseys, N.N & Finkel, C.F 1995 The Seismicity of Turkey and Adjacent Areas: A Historical Review, 1500–1800 Eren Publishing & Book Trade, İstanbul Akyüz, H.S & Altunel, E 2001 Geological and archaeological evidence for post-Roman earthquake surface faulting at Cibyra, SW Turkey Geodynamica Acta 14, 95–101 Ambraseys, N.N & Jackson, J.A 1998 Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region Geophysical Journal International 133, 390–406 Altunel, E 1998 Evidence for damaging historical earthguakes at Priene, western Turkey Turkish Journal of Earth Sciences 7, 25–35 Arpat, E & Bİngöl, E 1969 The rift system of the western Turkey: thoughts on its development Mineral Research and Exploration Institute (MTA) of Turkey Bulletin 73, 1–9 Altunel, E 1999 Geologic and geomorphologic observations in relation to 20th september 1899 Menderes earthguake, western Turkey Journal of the Geological Society, London 156, 241–246 Blikra, L.H & Nemec, W 1998 Postglacial colluvium in western Norway: depositional processes, facies and palaeoclimatic record Sedimentology 45, 909–959 Ambraseys, N.N 1988 Engineering seismology Earthquake Engineering and Structural Dynamics 17, 1–105 471 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA Bozkurt, E 2000 Timing of extension on the Büyük Menderes Graben, western Turkey, and its tectonic implications In: Bozkurt, E., Winchester, J.A & Piper, J.D.A (eds), Tectonics and Magmatism in Turkey and the Surrounding Area Geological Society, London, Special Publications 173, 385–403 Bozkurt, E 2001 Neotectonics of Turkey: a synthesis Geodinamica Acta 14, 3–30 Bozkurt, E 2004 Granitoid rocks of the southern Menderes Massif (southwestern Turkey): field evidence for Tertiary magmatism in an extensional shear zone International Journal of Earth Sciences 93, 52–71 Bozkurt, E & Park, R.G 1994 Southern Menderes Massif: an incipient metamorphic core complex in western Anatolia, Turkey Journal of the Geological Society, London 151, 213–216 Bozkurt, E & Sözbİlİr, H 2004 Tectonic evolution of the Gediz Graben; field evidence for an episodic, two-stage extension in western Turkey Geological Magazine 141, 63–79 Bozkurt, E & Sözbİlİr, H 2006 Evolution of the large-scale active Manisa Fault, southwest Turkey: implications on fault development and regional tectonics Geodynamica Acta 19, 427–453 Caputo, R 1993 Morphogenic earthquakes: a proposal Bulletin INQUA-NC 16, 24 Caputo, R & Helly, B 2005 The Holocene activity of the Rodia Fault, Central Greece Journal of Geodynamics 40, 153–169 Caputo, R & Helly, B 2008 The use of distinct disciplines to investigate past earthquakes Tectonophysics 453, 7–19 Caputo, R., Helly, B., Pavlides, S & Papadopoulos, G 2004 Palaeoseismological investigation of the Tyrnavos Fault, Central Greece A contribution to the seismic hazard assessment of Thessaly Tectonophysics 394, 1–20 ầftỗ, N.B & Bozkurt, E 2007 Anomalous stress field and active breaching at relay ramps: a field example from Gediz Graben, SW Turkey Geological Magazine 144, 687699 ầftỗ, N.B & Bozkurt, E 2008 Folding of the Gediz Graben fill, SW Turkey: extensional and/or contractional origin? Geodinamica Acta 21, 145167 ầftỗ, N.B & Bozkurt, E 2009 Evolution of the Miocene sedimentary fill of the Gediz Graben, SW Turkey Sedimentary Geology 216, 4979 ầftỗ, N.B & Bozkurt, E 2010 Structural evolution of the Gediz Graben, SW Turkey: temporal and spatial variation of the graben basin Basin Research 22, 846–873 Collins, A.S & Robertson, A.H.F 2003 Kinematic evidence for Late Mesozoic–Miocene emplacement of the Lycian Allochthon over the Western Anatolia Belt, SW Turkey Geological Journal 38, 1–16 Collinson, J.D & Thompson D.B 1982 Sedimentary Structures Allen & Unwin, London Dart, C., Cohen, H., Akyüz, H.S & Barka, A.A 1995 Basinward migration of rift-border faults: implications for facies distributions and preservation potential Geology 23, 69–72 472 Dewey, J.F & Şengör, A.M.C 1979 Aegean and surrounding regions: complex multiplate and continuum tectonics in a convergent zone Geological Society of America Bulletin 90, 89–92 Dramis, F & Blumetti, A.M 2005 Some considerations about seismic geomorphology and paleoseismology Tectonophysics 408, 177–191 Emre, Ö., Ere, Ö., Özalp, S., Doğan, A., Ưzaksoy, V., Yildirim, C & Gưktaş, F 2005 İzmir Yakın Çevresinin Diri Fayları ve Deprem Potansiyelleri [Earthquake Potential and Active Faults of İzmir and Surroundings] Mineral Research and Exploration Institute (MTA) of Turkey, Report no 10754, Ankara [in Turkish, unpublished] Emre, T & Sözbİlİr, H 1997 Field evidence for metamorphic core complex, detachment faulting and accommodation faults in the Gediz and Büyük Menderes Grabens (Western Turkey) In: Pİşkİn, ệ., Ergỹn, M., Savaỗn, M.Y & Tarcan, G (eds), Proceedings, International Earth Sciences Colloquium on the Aegean Region, İzmir, 73–94 England, P 2003 The alignment of earthquake T-axes with the principal axes of geodetic strain in the Aegean region Turkish Journal of Earth Sciences 12, 4754 Ergn, K., Gỹỗlỹ, U & Uz, Z 1967 A Catalogue of Earthquakes for Turkey and Surrounding Area (11 A.D.–1964) Technical University of İstanbul, Publications, no 24 [in Turkish] Eyİdoğan, H & Barka, A.A 1996 The October 1995 Dinar earthquake, SW Turkey Terra Nova 8, 479–485 Eyİdoğan, H & Jackson, J.A 1985 A seismological study of normal faulting in the Demirci, Alaşehir and Gediz earthquake of 1969–70 in western Turkey: implications for the nature and geometry of deformation in the continental crust Geophysical Journal of the Royal Astronomical Society 81, 569–607 Ganas, A., Pavlides, S.B., Sboras, S., Valkaniotis, S., Papaioannou, S., Alexandris, G.A., Plessa, A & Papadopoulos, G.A 2004 Active fault geometry and kinematics in Parnitha Mountain, Attica, Greece Journal of Structural Geology 26, 2103–2118 Guidoboni, E., Comastri, A & triana, G 1994 Catalogue of Ancient Earthquakes in the Mediterranean Area up to the10th Century Istituto Nazionale di Geofisica, ISBN: 88-85213-06-5 Hajdas, I., Bonanı, G., Thut, J., Leone, G., Pfennınger, R & Maden, C 2004 A report on sample preparation at the ETH/ PSI AMS facility in Zurich Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 223–224, 267–271 Hakyemez, H.Y., Erkal, T & Göktaş, F 1999 Late Quaternary evolution of the Gediz and Büyük Menderes grabens, western Anatolia, Turkey Quaternary Science Reviews 18, 549–554 Harms, J.C., Southard, J.B., Spearing, D.R & Walker, R.G 1975 Depositional Environments as Interpreted from Primary Sedimentary Structures and Stratification Sequences SEPM Short Course No Lecture Notes Society of Economic Palaeontologists and Mineralogists, Tulsa, OK Ç ƯZKAYMAK ET AL Işik, V., Tekelİ, O & Seyİtoğlu, G 2004 The 40Ar/39Ar age of extensional ductile deformation and granitoid intrusion in the northern Menderes core complex: implications for the initiation of extensional tectonics in western Turkey Journal of Asian Earth Sciences 23, 555–566 Jackson, J & Mckenzie, D 1988 The relationship between plate motions and seismic moment tensors and rates of active deformation in the Mediterranean and Middle East Geophysical Journal 93, 4573 Kaymakỗi, N 2006 Kinematic development and paleostress analysis of Denizli basin (west Turkey): implications of spatial variation of relative paleostress magnitudes and orientations Journal of Asian Earth Sciences 27, 207222 Koỗyt, A & Özacar, A 2003 Extensional neotectonic regime through the NE edge of outer Isparta Angle, SW Turkey: new field and seismic data Turkish Journal of Earth Sciences 12, 67–90 Koỗyt, A., Yusufolu, H & Bozkurt, E 1999 Evidence from the Gediz Graben for episodic two stage extension in western Turkey Journal of the Geological Society, London 156, 605–616 KOERI 2010 Boaziỗi University Kandilli Observatory and Earthquake Research Institute web site http://www.koeri boun.edu.tr/scripts/lst5.asp Leeder, M.R & Gawthorpe, R.L 1987 Sedimentary models for extensional tilt block/half-graben basins In: Coward, M.P., Dewey, J.F & Hancock, P.L (eds), Continental Extensional Tectonics Geological Society, London, Special Publications 28, 139–152 Lenk, O., Türkezer, A., Ergintav, S., Kurt, A.I & Belgen, A 2003 Monitoring the kinematics of Anatolia using permenant GPS network stations Turkish Journal of Earth Sciences 12, 55–66 Machette, M.N 1978 Dating Quaternary faults in the southwestern United States by using buried calcareous paleosols Journal of Research U.S Geological Survey 6, 369–381 Mayer, L 1984 Dating Quaternary fault scarps formed in alluvium using morphologic parameters Quaternary Research 22, 300– 313 McCalpin, J 1996 Paleoseismology Academic Press, San Diego, California McCalpin, J 2009 Paleoseismology, Second Edition Academic Press, an imprint of Elsevier Özer, S & Sözbİlİr, H 2003 Presence and tectonic significance of Cretaceous rudist species in the so-called Permo– Carboniferous Göktepe Formation, central Menderes metamorphic massif, western Turkey International Journal of Earth Sciences 92, 397–404 Ưzkaymak, Ç & Sưzbİlİr, H 2006 Aktif Manisa fayının batı bölümünün jeolojik özellikleri; deprem üreten faylarda reaktivasyon [Geological properties of western part of the active Manisa Fault; fault reactivation on active faults] 10th Meeting of Active Tectonics Research Group, Abstracts, 68–70 Ưzkaymak, Ç & Sưzbİlİr, H 2007 Aktif Manisa Fayında reaktivasyona işaret eden stratigrafik ve jeomorfolojik kriterler [Stratigraphic and geomorphologic criterias indicating reactivation on active Manisa Fault] Çukurova University 30th Year Geology Symposium, Abstracts, 177–178 Ưzkaymak, Ç & Sưzbİlİr, H 2008 Stratigraphic and structural evidence for fault reactivation: the active Manisa Fault Zone, Western Anatolia Turkish Journal of Earth Sciences 17, 615– 635 Ưzkaymak, Ç & Sưzbİlİr, H., Uzel, B., Akyỹz, H.S., Altunel, E., Yalỗiner, C.ầ., Merỗ, N & Atlihan, M.A 2008 Manisa Fay Zonu’ndaki Holosen aktivitesine ait veriler ve fay zonunun bat segmentinde yaplan paleosismolojik ỗalmalar [Evidence for Holocene activity on the Manisa Fault Zone and paleoseismic studies along the western part of the fault zone] 12th Workshop of Active Tectonics Research Group, Abstracts, 35–36 Ưzkaymak, Ç., Sưzbİlİr, H., Uzel, B & Akyüz, H.S 2009 Palaeoseismological evidence for Holocene activity on the Manisa Fault Zone, Western Anatolia EGU General Assembly 2009 Geophysical Research, Abstracts, 11-13407 Papazachos, B & Papazachou, C 1997 The Earthquakes of Greece Technical books Edition, Thessaloniki Paton, S 1992 Active normal faulting, drainage patterns and sedimentation in southwestern Turkey Journal of the Geological Society, London 149, 1031–1044 Pavlides, S 1996 First palaeoseismological results from Greece Annali Geofis 34, 545–555 Munsell Color Company 1994 Munsell Soil Color Charts, Revised Edition Macbeth Division of Kollmorgen, New Windsor, NY Pavlides, S & Caputo, R 2004 Magnitude versus faults’ surface parameters: quantitative relationships from the Aegean Tectonophysics 380, 159–188 Nash, D.B 1980 Morphological dating of degraded normal fault scarps Journal of Geology 88, 353–360 Ramsey, B.C 2009 Dealing with outliers and offsets in radiocarbon dating Radiocarbon 51, 1023–1045 Nelson, A.R 1992 Lithofacies analysis of colluvial sediments – an aid in interpreting the recent history of Quaternary normal faults in the Basin and Range province, western United States Journal of Sedimentary Petrology 62, 607–621 Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, B.C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J & Weyhenmeyer, C.E 2009 IntCal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP Radiocarbon 51, 1111–1150 Okay, A.I & Sİyako, M 1993 The revised location of the İzmirAnkara Suture in the region between Balıkesir and İzmir In: Turgut, S (ed), Tectonics and Hydrocarbon Potential of Anatolia and Surrounding Regions Ozan Sungurlu Symposium Proceedings, Ankara, 333–355 [in Turkish] 473 PLEISTOCENE−HOLOCENE ACTIVITY OF THE MANISA FAULT ZONE, W ANATOLIA Roberts, S & Jackson, J.A 1991 Active normal faulting in central Greece: an overview In: Roberts, A.M., Yielding, G & Freeman, B (eds), The Geometry of Normal Faults Geological Society, London, Special Publications 56, 125–142 Seyİtoğlu, G & Scott, B.C 1991 Late Cenozoic crustal extension and basin formation in west Turkey Geological Magazine 128, 155–166 Sletten, K & Blikra, L.H 2007 Holocene colluvial (debris-flow and water-flow) processes in eastern Norway: stratigraphy, chronology and palaeoenvironmental implications Journal of Quaternary Science 22, 619–635 Soysal, H., Spaholu, S., Kolỗak, D & Altinok, Y 1981 Historical Earthquake Catalogue of Turkey and Surrounding Area (2100 B.C.–1900 A.D.) Technical Report, TÜBİTAK, no TBAG-341 Sözbİlİr, H 2000 Batı Anadolu’dan örneklerle aktif faylar ve potansiyel aktif faylar [Active faults and potentially active faults: examples from western Anatolia] Symposium on Earthquake Potential of Western Anatolia, Abstracts, 133–142 Sözbİlİr, H 2001 Extensional tectonics and the geometry of related macroscopic structures: field evidence from the Gediz detachment, western Turkey Turkish Journal of Earth Sciences 10, 51–67 Sözbİlİr, H 2002 Geometry and origin of folding in the Neogene sediments of the Gediz Graben, western Anatolia, Turkey Geodinamica Acta 15, 277–288 Sözbİlİr, H 2005 Oligo–Miocene extension in the Lycian orogen: evidence from the Lycian molasse basin, SW Turkey Geodinamica Acta 18, 257–284 Sözbİlİr, H., Koỗer, T., Demrta, R., Uzel, B., Ersoy, Y., Akgỹn, M & ệzkaymak, ầ 2006 Seferihisar-Yelki Fay zonunda paleosismolojik ỗalmalar: dorultu atımlı fay zonunda eğim atımlı aktif normal faylanmaya ait veriler [ Paleoseismological studies on Seferihisar-Yelki Fault Zone: evidence for active normal faulting on strike-slip fault zone] 10th Meeting of Active Tectonics Research Group, Abstracts, 76–77 Sưzbİlİr, H., Sümer, Ư., Uzel, B., Ưzkaymak, Ç., Ersoy, Y., Erkül, F., İncİ, U & Helvac, C 2007 Bat Anadoluda zmir-Balkesir Transfer Zonu iỗinde kalan aktif doğrultu-atımlı faylarla sınırlı havzaların jeolojik özellikleri [Geological features of strikeslip basins located within the İzmir-Balıkesir Transfer Zone, western Anatolia] 11th Workshop of Active Tectonics Research Group Meeting, Abstracts, p 42 Sưzbİlİr, H., Uzel, B., Sümer, Ư., İncİ, U., Ersoy, E.Y., Koỗer, T., Demrta, R & ệzkaymak, ầ 2008 D-B Uzanml zmir fay ile KD-uzanml Seferihisar Faynn birlikte ỗaltna dair veriler: zmir Kửrfezini oluturan aktif faylarda kinematik ve paleosismolojik ỗalmalar, Batı Anadolu, Türkiye [Evidence for a kinematically linked E–W- trending İzmir Fault and NEtrending Seferihisar Fault: kinematic and paleoseismogical studies carried out on active faults forming the İzmir Bay, Western Anatolia] Geological Bulletin of Turkey 51, 91–114 474 Sözbİlİr, H., Sümer, Ö., Uzel, B., Ersoy, Y., Erkül, F., İncİ, U., Helvacı, C & Ưzkaymak, Ç 2009 17–20 Ekim 2005-Sığacık Kưrfezi (İzmir) depremlerinin sismik jeomorfolojisi ve bölgedeki gerilme alanları ile ilişkisi, Batı Anadolu [The seismic geomorphology of the Sığacık Gulf (İzmir) earthquakes of October 17 to 20, 2005 and their relationships with the stress field of their Western Anatolian region] Geological Bulletin of Turkey 52, 217–238 Şengör, A.M.C 1987 Cross-faults and differential stretching of hangingwalls in regions of low-angle normal faulting: examples from western Turkey In: Coward, M.P., Dewey, J.F & Hancock, P.L (eds), Continental Extensional Tectonics Geological Society, London, Special Publications 28, 575–589 Şengör, A.M.C & Yilmaz, Y 1981, Tethyan evolution of Turkey: a plate tectonic approach Tectonophysics 75, 181–241 Şengör, A.M.C., Görür, N & Şaroğlu, F 1985 Strike-slip faulting and related basin formation in zones of tectonic escape: Turkey as a case study In: Biddle, K & Christie-Blick, N (eds), Strike-slip Deformation, Basin Formation and Sedimentation Society of Economic Paleontologists and Mineralogists, Special Publications 37, 227–264 Tan, O., Tapirdamaz, M.C & Yörük, A 2008 The earthquakes catalogues for Turkey Turkish Journal of Earth Sciences 17, 405–418 Taymaz, T., Jackson, J & McKenzie, D 1991 Active tectonics of the North and Central Aegean Sea Geophysical Journal International 106, 433–490 Taymaz, T., Tan, O & Yolsal, S 2004 Seismotectonics of western Turkey: a synthesis of source parameters and rupture histories of recent earthquakes AGU Fall Meeting, Session T14, San Francisco-California, EOS Transactions 85, 47 Uzel, B & Sözbİlİr, H 2008 A first record of strike-slip basin in western Anatolia and its tectonic implication: The Cumaovası basin as an example Turkish Journal of Earth Sciences 17, 559– 591 Uzel, B., Sözbİlİr, H & Ư zkaymak, Ç 2011 Neotectonic evolution of an actively growing superimposed Basin in Western Anatolia: the inner Bay of İzmir, Turkey Turkish Journal of Earth Sciences [in press] doi: 10.3906/yer-0910-11 Wallace, R.E 1977 Profiles and ages of young fault scarps, northcentral Nevada Geological Society of America Bulletin 88, 1267–1281 Wells D.L & Coppersmith K.J 1994 New empirical relationship among magnitude, rupture length, rupture width, rupture area, and surface displacement Bulletin of the Seismological Society of America 84, 974–1002 Zhu, L., Akyol, N., Mitchell, B.J & Sözbİlİr, H 2006 Seismotectonics of western Turkey from high resolutions and moment tensor determinations Geophysical Research Letters 33, L07316 ... to the Paşadeğirmeni Fault Zone, Gürle Fault Zone, Manastır Fault, Taşlıburun Fault, Keỗilikửy Fault, and Manisa Fault Zone, respectively 451 PLEISTOCENEHOLOCENE ACTIVITY OF THE MANISA FAULT ZONE,. .. Abbreviations MFZ, KaF, İF, KF, and OFZ, refer to the Manisa Fault Zone, Karaỗay Fault, İzmir Fault, Kemalpaşa Fault, and Orhanlı Fault Zone, respectively (b) A cross section across the Manisa Fault Zone... ages of the main fault and of the secondary hanging-wall faults The clearest evidence for migration of fault activity in the study area is that the uplifted footwall of the basinward fault contains