Landscape formation on Cyprus is controlled by processes directly linked to the uplift and unroofing of the Troodos ophiolite complex since mainly early Miocene times. Understanding the island's geology and dating individual tectonic events will help in differentiating between tectonically controlled uplift and eustatic sea level or climatic changes. In order to improve the timing of these events, a magnetostratigraphic study was carried out on two terraces in the Mesaoria Basin of central Cyprus. At Vlokkariá, southwest of Nicosia, an artificial cliff exposes the sedimentary Apalós Formation of early Pleistocene age. A nearby section situated on a terrace, probably postdating the Vlokkariá section, was sampled at Kókkinos. Remanence-carrying minerals are end-member magnetite and maghemite formed by exsolution and oxidation from ophiolitic titanomagnetite. Hematite as a late alteration product is also present. Alternating field demagnetized paleomagnetic samples yield predominantly reversed polarities interpreted to have been acquired as early detrital remanent magnetization during the Matuyama chron. Two zones of normal polarity within the Apalós Formation are interpreted to correlate to the Olduvai and Jaramillo subchrons. Pre-Apalós marine sediments at the bottom of the Vlokkariá section show transitional polarity behavior and therefore might correlate with the onset of the Reunion event. Results from the Kókkinos section are exclusively of normal polarity, which has been acquired during the Brunhes chron. Sedimentation rates derived from these magnetostratigraphic results are in the order of 3–6 cm/kyr with a marked increase to 50 cm/kyr just below the Matuyama-Jaramillo polarity transition, which is interpreted to reflect increased uplift of the source area and supports the hypothesis of pulsed uplift of the Troodos ophiolite complex.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 The island of Cyprus is situated in the frontal range of a fore-arc region associated to a northward dipping subduction zone between Africa to the south and Anatolia to the north. The island itself can be divided into four tectonostratigraphic elements (Figure 1), which from west to east are the Mamonia and Moni complexes, the Troodos Massif, the Mesaoria Basin and the Kyrenia Range [Robertson, 2000, and references therein]. The Mesaoria Basin, subject of this study, is situated between the Cretaceous Troodos ophiolite complex to the south and the Kyrenia Range to the north and can be described as an asymmetric halfgraben with maximum subsidence along the northern margin. Its tectonic evolution, however, is still not well understood in detail [Baroz and Bizon, 1974; McCallum and Robertson, 1990; Weiler, 1969]. It is either linked to collision in southeast Turkey and overthrusting along the southern margin of the Anatolian orogen [Calon et al., 2005] or to crustal thinning in a fore-arc setting during the Oligocene to Miocene [Robertson and Woodcock, 1986] and to subduction rollback in the south [Robertson et al., 1996]. It seems clear, however, that the timing of tectonic events in the eastern Mediterranean is controlled by changes in the motion of Africa with respect to Turkey [Dewey et al., 1973; Dolson et al., 2001; Harrison et al., 2004; Morris, 1996], around 9 Myr (Miocene, compressional or transpressional tectonism) and 5 Myr ago (Pliocene, directional change of the African Plate motion, see Harrison et al. [2004, and references therein]). Since then, the deformational setting seems to have been controlled by differential convergence of the Eurasian and African plates and resulting westward escape of the Anatolian microplate [Dolson et al., 2001; Sengör et al., 1985; Westaway, 1994]. Detailed knowledge of the uplift history of Cyprus will potentially lead to better insight into the mode and timing of the tectonic processes in the eastern Mediterranean.
2. Uplift History of Cyprus
 The onset of northward subduction of the African Plate under the Turkish Plate along the Cyprean Arc during the early Miocene in combination with serpentinite diapirism uplift of the island of Cyprus controls the geological and geomorphological development of the island during the Neogene [Moores and Vine, 1971; Kempler and Ben-Avraham, 1987; Kempler and Garfunkel, 1994]. Consequently, the development of alluvial fans and marine terraces is closely associated with uplift and therefore, if stratigraphically correlated across the island and if dated, will yield further insight into the tectonic evolution along an active plate boundary [Robertson, 1977].
 After an initial uplift during Miocene times [Robertson, 1977; Eaton and Robertson, 1993; Schirmer, 2000], a second and stronger phase of Troodos uplift has been identified based on the formation of a well developed fan delta (Kakkaristra Formation) along the northern edge of Troodos during the late Pliocene [McCallum and Robertson, 1995a; Poole and Robertson, 1998]. The oldest terrace of Plio-Pleistocene age lies at roughly 500 m above sea level (asl) in the southwest of the island [Poole and Robertson, 1991]. A further (younger) terrace at 350–360 m asl is stratigraphically correlated to a terrace on the southern flank of the Kyrenia Range and the Athalassa and Kakkaristra formations of the Mesaoria Plain [Poole and Robertson, 1991, and references therein]. An age of 2.2 to 1.9 Myr has been assigned for the younger terraces by McCallum and Robertson [1995b]. However, the main phase of uplift has been derived from the deposition of fluvial fanglomerates, unconformably overlying the sand and silts of the Apalós Formation [Poole and Robertson, 1991, 1998]. Although poorly defined, the Pleistocene age of the Apalós Formation [Harrison et al., 2004] defines a lower time constraint for this event [Poole and Robertson, 1991]. The transition from the underlying marine formations (Kakkaristra, Nicosia and Athalassa) into the fluviatile Apalós Formation was gradational and occurred approximately at the Pleistocene-Pliocene boundary [Newell et al., 2004; Poole and Robertson, 1991]. Very young uplift events can be traced by marine terraces in coastal southern Cyprus at 8 to 11 m asl and less than 3 m asl and U/Th ages have been dated to 185–204 ± 8 kyr and 116–134 ± 10 kyr [Poole and Robertson, 1991], respectively.
 Two Plio-Pleistocene sections in the Mesaoria Plain have been sampled for magnetostratigraphic dating. The Apalós Formation, subject of this study, is well exposed in a man-made cliff at Vlokkariá (latitude: 35.0895°N, longitude: 33.2496°E) some 12 km southwest of downtown Nicosia and 343 m above sea level. The cliff can be traced along 130 m and reaches a maximum height of 20 m. During fieldwork, the outcrop was extended to a stratigraphic depth of 44.5 m along the hillslope by a 146.5 m long and 2.5–3 m deep trench. The sediments of the Apalós Formation show clearly cyclic deposition, which is diagnostic for the fluviatile series [Schirmer, 1983]. Each cycle starts with dark gravel beds consisting of dark ophiolite derived rudites with medium-grained sand of dark olive color (channel deposits). The gravels are overlain by sandy silty deposits fining upward (floodplain deposits) that are topped by floodplain soils. These three constituents form a fluviatile series or fluviatile unit. Although all three members are not always present, 24 fluviatile units (Figure 2) have been identified in the Vlokkariá section [Schirmer et al., 2010]. The Apalós fluvial units (starting above 1.6 m in Figure 2) are underlain by a greenish gray fine-grained, silty shale with abundant secondary calcrete, which forms the upper part of the marine Nicosia Formation (NF in Figure 2) of Pliocene age.
 In addition to the Vlokkariá section a further site was studied at Kókkinos. Here, the next younger terrace following geomorphologically below the Vlokkariá site yielded a 5 m thick section in a flat hill with the toponym Kókkinos 324 m above sea level (latitude: 35.0821°N; longitude: 33.2341°E). Below 4 m of boulder gravel of Troodos ophiolitic provenance there was 0.7 m silty yellow brown flood loam with a thick cambisol on top underlain by a lower gravel. Lower gravel, flood loam and cambisol form one unit, the topping 4 m thick gravel forms a next one. Again they lie in superposition as it is the case in the Apalós Formation at Vlokkariá.
 A total of 445 samples was collected from the fine-grained parts of the Apalós and the uppermost part of the underlying marine Nicosia Formation at Vlokkariá using two portable coring devices with different diameter [Weber, 2005]. Both consist of nonmagnetic steel tubes, which were gently pressed or hammered into the sediments at an average vertical distance between samples of about 8 cm. The tubes were oriented with an inclinometer and a magnetic compass. Using a silicon piston the sample material was pressed out of the tubes into cylindrical perspex containers with a volume of 3.51 cm3 or 8.55 cm3 depending on the coring tube used. Unfortunately, the containers inhibit thermal demagnetization of the natural remanent magnetization (NRM). Hence the bulk of the specimens was exposed to stepwise alternating field demagnetization (AF) up to peak fields of 90 mT. AF demagnetization and all measurements of remanence were carried out with a 2G degaussing system and a 2G cryogenic magnetometer in a magnetically shielded room at the paleomagnetic laboratory of the Department of Earth and Environment Sciences at the Ludwig-Maximilians-University Munich. The measured paleomagnetic data are displayed on orthogonal vector plots [Zijderveld, 1967] and stereographic projections (cf. Figures 3 and 4). Linear or curved demagnetization trajectories were identified by eye and analyzed using principal component analysis (PCA) [Kirschvink, 1980].
 NRM intensities of all samples measured are ranging between 1.55 mA/m and 1850 mA/m. AF demagnetization at peak fields of about 15 mT reveals the presence of a small secondary component with unsystematic directions in some of the specimens studied (Figure 4). Further demagnetization at increasingly higher fields is successful in isolating a characteristic remanent magnetization (ChRM), which is directed either to the north and down (Dec: 358.3°, Inc: +45.5°, α95: 2.9°, k: 37.6, N: 65) or to the south and up (Dec: 176.3°, Inc: −49.6°, α95: 1.9°, k: 33.3, N: 175, cf. Figure 3). A positive reversal test [McFadden and Lowes, 1981] with classification “A” suggests the primary character of this magnetization. Intermediate directions, for example close to reversal boundaries, and directions of poor quality have been excluded for determining mean directions and the reversal test.
 The resulting latitudes of the individually determined virtual geomagnetic poles (VGP) have been compared with published polarity timescales [Lourens et al., 2004, and references therein] in order to estimate the age of the Vlokkariá section (Figure 2). Most of the section is reversely magnetized and is attributed to the main part of the Matuyama polarity zone. Normal polarity zones in between are thought to represent the Olduvai (1.945–1.778 Myr) and Jaramillo (1.072–0.988 Myr) subchrons, respectively. If this interpretation is correct, the Cobb Mountain event of roughly 10.000 years of duration has not been identified. This might either be due to sampling gaps or to a hiatus in sedimentation. We note, however, that in the middle part of the profile, a section of 250 cm was not sampled due to the presence of massive gravel, thus supporting our speculation that the lack of the Cobb Mountain might be due to sampling gaps. The lower Jaramillo transition from reversed to normal (M→J) apparently has been recorded in great detail over a length of roughly 4.0 m. The transitional ChRM directions are well defined during AF demagnetization like those in the remainder of the section (cf. Figure 4) and the resulting VGPs cluster to the east of South America (Figure 5). The result is comparable to M→J data of Mazaud et al.  observed in North Atlantic sediment u channel cores. Their data show a similar cluster of VGPs over the southern tip of South America, which supports our interpretation that we have identified the M→J transition. The high VGP path resolution in conjunction with a positive reversal test at Vlokkariá is taken as evidence for early detrital remanent magnetization alignment.
 Alternatively, one might argue that we have missed the Jaramillo subchron and that the normal polarity interval identified in the upper part of the section represents the Cobb Mountain subchron. The spatial distribution of VGPs during the Cobb Mountain reversed-to-normal transition [Hsu et al., 1990; Yang et al., 2001; Clement, 1992; Clement and Martinson, 1992; Chauvin et al., 1990; Leonhardt et al., 2009] defines a path across the central Pacific Ocean and does not resemble the transitional path presented here. This strengthens our interpretation that the Jaramillo subchron has been correctly identified.
 Paleomagnetic directions of samples from the Kókkinos section show exclusively normal polarity, which is thought to have been acquired during the Brunhes chron. However, there remains some uncertainty whether they could represent a younger Pleistocene terrace or a ridge of the Pliocene Kephales Member [Newell et al., 2004] of the upper Nicosia Formation [Schirmer et al., 2010]. VGP latitudes of the Kókkinos section have been plotted in the uppermost part of the polarity pattern in Figure 2. A stratigraphic gap between the Vlokkariá and the Kókkinos section is likely as indicated in Figure 2.
4. Magnetic Mineralogy
 To get higher concentrations of magnetic particles, sediment samples have been mixed with distilled water and stirred up in an ultrasonic bath for mineral separation. By dipping a magnetic finger repeatedly into this suspension magnetic particles accumulate at its tip and thus can be extracted. The magnetic extracts were analyzed using scanning electron microscopy (SEM; JSM-5900 LV, JEOL). Backscattered electron images of the extract surfaces in which all particles with atoms of higher atomic number appear brighter than particles with lighter atoms allow the identification of relatively heavy magnetic particles in a mixture with silicates that are still present in the extract (Figure 6). The diameter of the bulk of the magnetic particles in the samples was found to be between about 1 and 25 μm.
 Energy-dispersive X-ray spectroscopy (EDX) shows that the analyzed grains can be divided in two different types of magnetic particles: pure iron-oxides (Figure 7a) and iron-titanium-oxides (Figure 7b) (ilmenite, titano-magnetite, titano-maghemite). These results are in accord with a volcanic source such as the Troodos ophiolitic complex. In order to characterize the mineral inventory further, magnetic high field experiments were carried out on 76 representative specimens using a variable field translation balance (VFTB) [Krása et al., 2007].
Figure 8a summarizes the standardized isothermal remanent magnetization (IRM) acquisition curves of 56 specimens from Vlokkariá. The IRM reaches 90% of the maximum magnetization at an external field strength of 300 mT documenting predominance of ferromagnetic (titano-)magnetite and (titano-)maghemite. Since saturation occurs only in fields of about 600 mT, the sediments at Vlokkariá must contain particles of high-coercivity minerals like hematite or goethite as well.
 S ratios [Bloemendal et al., 1992] ranging between 0.93 and 0.96 support this evidence. They indicate again that the main part of the magnetization is carried by low-coercivity minerals such as (titano-)magnetite or (titano-)maghemite with additional contributions from high-coercivity minerals. Since the saturation magnetization of hematite is weaker than that of magnetite by roughly a factor of 250, we estimate the content of hematite (and goethite) to be higher by a factor of 10 than that of magnetite.
 Hysteresis loops of selected specimens show generally the same shape and are predominated by ferrimagnetic minerals like (titano-)magnetite or maghemite. A slight narrowing in the middle part of the loop (wasp-waisted shape) might stem from interaction of minerals with different coercivities [Roberts et al., 1995; Tauxe et al., 1996]. Again this can be interpreted as a substantial contribution of hematite (and/or goethite). Saturation magnetization, saturation remanence and coercive force vary significantly across the Vlokkariá profile (Ms: 0.1–1.6 Am2/kg; Mrs: 0.017–0.33 Am2/kg; Hc: 6–25 mT) reflecting different concentrations preferably of (titano-)magnetite. These are likely to be controlled by variability in lithology and sediment source.
 In order to get further information about the type of magnetic minerals in the samples, thermomagnetic curves have been measured. The strength of the external field has been chosen at the point where the closing of the hysteresis loop could be observed in the foregoing hysteresis measurements in order to minimize the influence of paramagnetic minerals in the specimens. All measured thermomagnetic curves have a similar shape (Figure 8c). Two Curie temperatures could be identified in each specimen upon heating. The first one is indicated by a slight inflexion between 380°C and 400°C, the second around 580°C is more prominent and diagnostic of pure magnetite.
 The magnetic phase responsible for the characteristic drop in saturation magnetization between 380°C and 400°C is not readily identified. The cooling curve shows clearly reduced magnetization compared to the heating curve and the decrease appears to be due to chemical alteration of magnetic minerals during heating. Repeated M(T) runs with successively increasing maximum temperatures (Figure 8d) show that the drop between 380°C and 400°C does not result from an actual Curie temperature transition, but arises from magnetomineralogical alteration between 350°C and 400°C, most likely the inversion of maghemite to stable but weakly magnetic hematite [Krása et al., 2007]. This interpretation is reinforced by intensive reddening of the samples during the heating experiments. The relatively low amplitude of the applied field and the small saturation magnetization of hematite prohibit its clear identification except for one specimen, where hematite with a Curie temperature of about 675°C was unambiguously identified.
 If a specimen consists of a mixture of different magnetic minerals with variable grain size or concentration, the rock magnetic parameters measured using the VFTB merely represent mean values. Therefore, several methods of unmixing have been developed in order to analyze the magnetic composition [e.g., Egli, 2003; Kruiver et al., 2001]. In this work, single rock magnetic components have been characterized using detailed IRM acquisition curves measured at 0.05 mT intervals up to 0.5 T to be analyzed with the method of Egli [2003, 2004].
 Detailed IRM acquisition curves of eight specimens have been determined and analyzed with the computer programs CODICA (COercivity DIstribution Analyzer) and GECA (Generalized Coercivity Analyzer) [Egli, 2003, 2004]. Figure 9a shows a typical coercivity distribution that is analyzed using a model with three different coercivity components (A, B, C; Figure 9b). In trying to identify these components, the foregone VFTB investigations and the scanning electron microscope observations have been included in the interpretation. Component A of lowest coercivity is assigned to (titano-)magnetite. Intermediate coercivity component B probably represents (titano-)maghemite while the magnetic relatively hard component C is allocated to (titano-)hematite.
 A Day plot [Day et al., 1977] of the Mrs/Ms versus the Hcr/Hc ratios places the Vlokkariá samples in a narrow field within the pseudo-single-domain (PSD) region of pure magnetite confirming a complex but homogeneous magnetomineralogy as it was observed in the IRM acquisition curves, hysteresis loops and thermomagnetic curves (Figure 8b). Mrs/Ms ratios between 0.18 and 0.25 and Hcr/Hc ratios between 2.75 and 4.8, however, are not in line with theoretical mixing curves of pure magnetite. The data cluster between the theoretical SD-MD and SD-SP mixing curves [Dunlop, 2002a, 2002b]. This leads to the assumption that the specimens contain also superparamagnetic (SP) particles. Alternatively titanomagnetite of intermediate composition, although not observed in the thermomagnetic curves, may shift the data to higher Hcr/Hc values.
5. Discussion and Conclusions
 Titanomagnetite from the ophiolites of the Troodos Massif is supposed to be the ferromagnetic source mineral of the sediments at the Vlokkariá locality [Borradaile et al., 2010]. Titanomagnetite of intermediate composition (TM60 with x = 0.6 in the solid solution series Fe3-xTixO4 with 0 < x < 1) formed originally in the pillow basalts, sheeted dykes and gabbros. Depending on the cooling, hydrothermal or surface alteration history, the TM60 usually exsolved or oxidized to end-member magnetite and titanium-rich oxides such as ulvöspinel, ilmenite, titanite, etc. [Shau et al., 2000]. When the massif was elevated above sea level and eroded, the ophiolites were exhumed. The ferromagnetic phases were eventually exposed to alteration and slow oxidation to (titano-)maghemite. Finally ferromagnetic minerals in different oxidation states were deposited in the fluviatile sediments of the Apalós Formation. Further in situ weathering or soil formation under variable paleoclimatic conditions caused the formation of hematite. This is also suggested by the generally strong red coloration of the sediments at Vlokkariá.
 Our rock magnetic studies demonstrate the main carriers of remanent magnetization to be end-member magnetite and end-member maghemite in variable concentrations and grain sizes (Figures 8 and 9). Irreversible thermomagnetic curves, Curie temperatures and hysteresis properties point to the presence of magnetite between the SD-MD and SP-SD mixing lines and maghemite. No clear evidence of the original TM60 extensively occurring in the Troodos ophiolites [Borradaile et al., 2010] has been detected in the Apalós sediments. High-temperature exsolution with micrometer-sized ilmenite lamellae in magnetite is well documented (Figure 6). Superparamagnetic magnetite or maghemite appears to be of some importance although secondary (viscous) NRM components are often negligible during AF demagnetization (Figure 4). In addition, significant amounts of hematite could be identified, especially in the NRM experiments. Continued oxidation and alteration of the ferromagnetic mineral phases during and/or after sedimentation has produced higher Hcr/Hc values (Figure 8b) than those observed by Borradaile et al.  in the original Troodos ophiolites.
 The magnetostratigraphy of the Vlokkariá section is predominated by reversed polarity, which is interpreted to have been imprinted during the Matuyama polarity zone. If this interpretation is valid, the two zones of normal polarity identified, represent the Olduvai and Jaramillo subchrons, respectively with ages ranging between 1.945 and 1.778 Myr and 1.072 and 0.988 Myr, respectively [Lourens et al., 2004, and references therein]. Pre-Apalós marine sediments (Nicosia Formation) in the lowermost part of the Vlokkariá section show transitional polarity behavior and therefore might correlate with the onset of the Reunion event (2.148 Myr). Based on this interpretation, we arrive at a sedimentation rate of >2.7 cm/kyr for the time span represented by the lowermost 10 m of the section up to the top of the Olduvai subchron. This is quite similar to the sedimentation rate of 3.5 cm/kyr calculated for the part of the section between Olduvai and Jaramillo (up to profile distance 33 m in Figure 2). We note, however, that the lower Jaramillo transition from reversed to normal polarity (M→J) has been recorded in the section across a long interval of 400 cm, suggesting increased sedimentation rates just below the onset of the Jaramillo event. If the transition period from reversed to normal polarity for the lower Jaramillo is taken to be in the order of 8 kyr as suggested by Clement , sedimentation rates of about 50 cm/kyr result. Because of this relatively high rate of sedimentation a cluster of transitional VGPs situated to the east of South America (Figure 5) has been recorded during the lower Jaramillo polarity change. During the Jaramillo event itself, sedimentation rates with 6.25 cm/kyr fall back to about twice the pre-Jaramillo values. As the rate of sedimentation is directly linked to the amount of uplift, these results are consistent with the work of McCallum and Robertson , who propose that uplift of the Troodos Massif was not uniform but has been pulsing. Although, climatic and/or eustatic events cannot be completely ruled out to have influenced sedimentation rates, Poole and Robertson  and Spezzaferri and Tamburini  have demonstrated that these did not play a major role in terrace evolution on the island of Cyprus during the Pliocene and early Pleistocene.
McCallum and Robertson  conclude that uplift slowed down toward the end of the Pliocene leading to a decreased sedimentation rate in the Mesaoria Basin. This agrees with the average rates of about 3.5 cm/kyr derived in this study. They also deduce that major compression was followed by drastic uplift in the Pleistocene, which might correspond to the strongly increased sedimentation rate we observe during the transition from the Matuyama chron to the lower Jaramillo subchron around 1.1 million years ago.
 Thanks go to Michael Wack, Lars Cofflet, and Alexander Ikinger for their enthusiastic support during field work. Thorough reviews by Joshua Feinberg (Associate Editor), Alastair H. F. Robertson, and two anonymous reviewers significantly improved the quality of the paper. This research was supported by a grant of the German Research Foundation (DFG) to VB and WS (Ba1210/8–1).