Late extensional shear zones and associated recumbent folds in the Alpujarride subduction complex, Betic Cordillera, southern Spain

This work is licensed under Creative Commons Attribution-ShareAlike 4.0 International.

The western (Ionian Sea) margin of Greece is one of the most seismically active parts of the Mediterranean, with intense shallow depth seismicity and magnitudes reaching up to 7.4 (e.g. Papazachos, 1990; Louvari et al., 1999). The area is a zone of post-Pliocene contraction (Sorel, 1989), reflected in numerous inverse faults/thrusts, folds and evaporite doming and intrusions (British Petroleum Co., 1971; Underhill, 1988).

Crustal deformation of Ionian Sea has been studied through GPS measurements (e.g. Kahle et al., 1995; Peter et al., 1998; Cocard et al., 1999; Hollenstein et al., 2006; Serpelloni et al., 2007, 2013; Anzidei et al., 2014). According to Hollenstein et al. (2006), the north part of the Ionian Sea is characterized by vertical velocities of subsidence of the order of -2 to -2.5mm/yr, while the south part by -3.5 to -4.5mm/yr. Furthermore, the authors report an interseismic subsidence of -4mm/yr for the Ionian Islands.

In particular, the Cephalonia Transform Fault (CTF), separating the Apulia or Adriatic (Africa) and Aegean (Eurasia) lithospheric plates, which is clearly visible from the bathymetric maps of the area, developed close to the western side of Leukas and Cephalonia islands (Fig. 1) (Valkaniotis et al., 2014). Its slip rate varies between 7 and 30mm/yr based on GPS measurements (Anzidei et al., 1996; Hollenstein et al., 2006).

A prolongation of the CTF is found parallel to the western coast of Leukas Island and constitutes a very active seismic fault segment (Leukas segment; Louvari et al., 1999). According to Louvari et al. (1999) the Leukas segment is characterized by a dextral strike-slip motion combined with a small thrust component. Leukas Island is characterized by intense seismicity. The instrumental records show that strong earthquakes with surface-wave magnitude (Ms) up to 6.5 have taken place frequently in the area, associated with the strike-slip fault segment to the west of the island striking NNE-SSW and the thrust structure offshore NW Leukas (Fokaefs and Papadopoulos, 2004). Since 1612, at least 20 earthquake events have been reported for Leukas along with descriptions of the damages triggered by the earthquakes (Papathanassiou et al., 2005). According to these authors the list of historical events suggests that earthquakes appear in couples (twin or cluster events) and the re-occurrence period ranges between 2 months and 5 years.

FIGURE 1
Tectonic setting of the study area (after Valkaniotis et al., 2014) and seismicity since 550B.C

In this paper, we focused on the investigation of tidal notches in Leukas and Meganisi Islands. We discuss four submerged shorelines at Meganisi (Ionian Islands, Greece) and provide some chronological estimations that allowed us to identify a sequence of coseismic subsidences that occurred during the last few millennia. We have also estimated the average coseismic subsidence displacement and the average return time of these vertical displacements. Furthermore, we examine the tectonic behavior of the Island of Leukas, after the uplift that affected the northernmost part of the island around 4 to 5ka BP, and correlate the coseismic subsidences observed in Meganisi during the latest Holocene with a corresponding sequence of coseismic movements as suggested by submerged tidal notches in Leukas Island.

Field observations, mapping and measurement in both Meganisi and Leukas islands took place in 2012. All coasts were systematically surveyed in detail, using a boat in order to access all sites and establish the continuity of observations. Figure 2 depicts the sites where we located evidence of relative sea level changes, which are discussed below.

FIGURE 2
Location of tidal notches in Leukas and Meganisi islands, discussed in the paper.

Former sea-level positions were deduced from sea- level indicators, such as emerged and submerged tidal notches. Notch geometry i.e.height, inward depth and vertex depth from present mean sea level were measured according to Pirazzoli (1986) and Evelpidou and Pirazzoli (2014). The rocky shoreline was investigated through snorkeling and all indications suggesting relative sea level changes were recorded. We report here only the measurements of tidal notches that were continuous. Considering that the factors affecting the erosion of the tidal zone are various (e.g. Trenhaile, 2015), only the sites where the tidal notch profiles provided clear information were taken into account (e.g. with no irregularities on the rock, wave-protected sites, continuity). For each site, measurements were taken along extended outcrops of continuous notches. The final values correspond to the average of 10 measurements for each morphological characteristic of the notches (notch height, vertex depth, inward depth). In particular, the vertex depth measurements in relation to the present sea level were accomplished during favorable weather conditions. Depth measurements were subsequently corrected for tide and atmospheric pressure during the fieldwork period. In addition, the error estimations for the depth measurements were applied by taking into consideration the wave conditions and the measuring method. Atmospheric pressure and tidal records at the time of measurements were provided by the Hellenic Navy Hydrographical Service (HNHS) at the station of Leukas. The tidal range in the investigated area is 0.17 m and the mean sea level (MSL) is situated at 0.510m based on 21 years of statistical data (1990–2011) of HNHS.

An interpretation of the relative sea level change was made based on the profiles of the notches. Figure 3 depicts the different types of profiles identified in the study area. Given the difficulties in obtaining a chronology for the submerged erosion features, either by direct dating (all submerged fossils contemporary with the time of deposition or having grown on the notch profile are rapidly destroyed by bioerosion), or by comparison with already dated levels of assumed global significance, we have limited our study to the following:

• The Late Holocene, i.e. about the last 5–6ka during which global sea level is generally considered to have remained almost stable;

• an area of the eastern Mediterranean where the tectonic influence on relative sea level change can be identified and isolated (e.g. Nixon et al., 2009; Evelpidou et al. 2012a, 2013a, b). Accurate underwater measurements of appropriate sea-level indicators (i.e. tidal notches) were used to identify changes in relative sea-level and to distinguish the speed (slow or rapid) of observed vertical displacements (e.g. Anzidei et al., 2011);

• an estimation of the time required for tidal notch formation based on local rates of intertidal erosion (a range between maximum and minimum possible values) (Furlani et al., 2010, 2014; Evelpidou and Pirazzoli, 2016b). This technique is useful in areas lacking archaeological remains or stratigraphical data from coastal boreholes;

• a determination of sequences of vertical movements, in the case of several tidal notches existing in the same area, and quantification of coseismic subsidences and estimation of their return time.

FIGURE 3
Different types of tidal notch profiles identified in the study area (MSL is mean sea level, HF is the height of the notch floor and HR is the height of the notch roof).

Vött et al. (2007) interpreted the northern border of Leukas as representing remnants of a former mid-Holocene strandline made of beachrock that was subsequently partly dislocated by tsunami. Near Cape Gurapetra some in situslightly elevated beachrock was reported by Pirazzoli et al. (1994) from which two shell samples, collected at an estimated palaeo - MSL of +1m, provided radiocarbon ages between 4.2 and 5.3ka BP, thus possibly dating the formation of the elevated notch of Aghios Ioannis. This uplift was probably originated the seismo-tectonic activity related to the Cephalonia Transform Fault and the Leukas Fault (Louvari et al., 1999). The seismicity, following the western coast of Leukas Island, shows the continuation of the Cephalonia Transform Fault to the north and defines the Leukas segment. The maximum magnitude for this fault segment, according to its length, is 6.8 (Papazachos and Papazachou, 1997). Historical reports of earthquakes that caused damage to Leukas Island cited in Papazachos and Papazachou (1997) clearly indicate that this is the maximum magnitude ever observed in the island (Table V). Vött et al. (2007, p. 52) assume that the beachrocks dated by Pirazzoli et al. (1994) “are not in situ but probably relocated and uplifted by tsunamigenic influence”. The size of beachrock (it extends almost continuously parallel to the coast for more than 2km) and the condition of the beachrock slabs (undisturbed, in-situ slabs, maintaining their morphological characteristics, such as their inclination parallel to the coast) makes difficult to adopt this assumption. This mid-Holocene uplift seems to have been local and limited to the northernmost border of Leukas Island. Tectonic movements during the next centuries, as shown below, consist of a sequence of coseismic subsidences, deduced from submerged tidal notches, which affected not only most of Leukas Island, but also the nearby Meganisi Island.

A comparison between the two islands may now be attempted (Table VI). If the total observed subsidence (about 140cm) and the range of possible return times of coseismic subsidence (especially for a medium bioerosion rate) are similar in the two islands, several differences appear:

• number of coseismic events: 5 in Leukas against 4 in Meganisi;

• absence of modern notches in Meganisi;

• and the type of certain tidal notch profiles, e.g. for the -100±6cm shoreline, which is of b’-type in Leukas, while it is of d’-type in Meganisi.

The aforementioned suggest that although the area was affected by a similar tectonic strain, certain coseismic events were only recorded in one of the two islands and, furthermore, some events may have affected only part of one island.

Five coseismic subsidences have been identified from submerged notches in Leukas Island in the last few millennia. Their cumulative effect was a total subsidence of 140±6cm in a period of 1.1 to 5.7kyr. Four submerged fossil shorelines were found in Meganisi Island. The cumulative effect was a total subsidence of 135±6cm during a period lasting between 0.9 and 3 kyr. The average coseismic subsidence was between 29±6 and 34±6cm with a return time between 0.2 and 1kyr respectively.

Although the observed subsidences are similar in Leukas Island and the nearby Meganisi Island, several differences appear indicating that certain coseismic events have a local effect and in some cases have affected only part of the study area.

The authors thank the Mayor of Leukas, Kostas Aravanis, who coordinated all efforts from the part of the municipality ensuring boat facilities for the researchers, Deputy Mayor, Ioannis Kartanos, who was close to the research team, and Mr. Livitsanos and the Hotel association for accommodation facilities. We wish to thank also Mr. Stathis Conidaris who was close to the research team throughout the whole fieldwork in Leukas.

Authors wish to thank also the Mayor of Meganisi, Mr. Zavitsanos Stathis, and the Deputy Mayor, Ms. Kavvada Elli who coordinated all efforts from the part of the municipality ensuring boat facilities for the researchers and Mr. Malamas Andreas who was close to the research team throughout the whole fieldwork.

The authors also thank two anonymous reviewers whose constructive comments and suggestions improved the final version of this paper.