The Geotectonic State of the Greek Area: Vulcanism, Intermediate Earthquakes and Plate Tectonics

The regional and temporal wandering of the Greek geosynclines from NE to SW is the result of the action of the convection currents. The African plate (Gondwana) has always played the role of the foreland, which has produced the impression that this plate underthrusts below the Aegean Sea.

The intermediate earthquakes are concentrated in the geosynclinal Eastern Hellenic zone (Subpelagonian), characterized by the occurrence of ophiolites and the presence of large longitudinal tensional fractures. The andesitic volcanoes occur in the inner part of this zone, partly metamorphosed, and are accompanied by the deepest of the shocks.

The Greek volcanoes seem to be connected with transversal fractures and accompanied by intermediate earthquakes. Intermediate earthquakes are present, independent of the occurrence of volcanoes, in the geosynclinal zones, where deep tensional fractures appear (northern part of the Eastern Hellenic zone, Olonos - Pindos zone, Arcadian nappe zone). The andesitic lavas of the Aegean are derived from orogenetic magma. The observed differences of the K₂O content may be considered as a local phenomenon of differentiation of the magma and assimilation of metamorphic igneous rocks.

The presence of ophiolites inside of the Hellenic arc in places where andesites occur indicates that both these rocks are connected with the eugeosyncline of the Eastern Hellenic zone. They are derived from a basaltic magma introduced into the crust through the fissures in the substratum of the geosyncline and are the cause of the presence of the positive gravity and magnetic anomalies, and the high heat flow values in the Aegean Sea.

The so called subduction of the African plate below the Aegean area is apparent. In reality, this plate does not move northward, but it has retreated (extinguished) in the opposite direction. The question here is of a gradual destruction of the African plate from NE to SW, due to the geosynclinization of its parts being the cause of the convection currents acting in this direction. The occurrence of intermediate earthquakes in the convex side of the Greek arc indicates that the destruction of the African plate and the genesis of new geosynclines still continue.

The existence of andesitic volcanoes and the occurrence of intermediate earthquakes in the Aegean Sea, inside of the Greek orogenetic arc led us (Kiskyras 1960) to the conclusion that the Aegean Sea resembles from the geotectonic point of view the inner side of the Eastern Indian arc, the Japan Sea and the Antilles arc, and more, that in the Aegean Sea there may be present positive gravity anomalies. Further, it was supposed that below the Aegean area convection currents should move, due to the temperature differences between the substratum under the Greek orogen (continent) and the bottom at the corresponding depth under the deep Ionian Sea (Gondwana). These convection currents are moving below the Greek orogen nearly from E to W and turning down in the front of the Gondwana.

In the author's opinion (Kiskyras 1960) such local convection currents are present in the alpine geosyncline area, bounded by the cratonic bottom of the deep sea. A similar convection current cycle, but in the opposite direction, was suggested for the Tyrrhenian Sea. In both these cases the currents move toward the deep Ionian Sea (Gondwana) which plays the role of the foreland (See Fig.1). All the orogenetic processes are developed in the same direction as the convection currents.

Later on, Caputo et al. (1970), McKenzie (1970), Derwey & Bird (1970 - 1971), Ninkovich & Hays (1971), Papazachos & Komninakis (1971), Galanopoulos (1972) and many others tried to explain the volcanism and the occurrence of intermediate earthquakes, according to the theory of plate-tectonics, as the result of the subduction of the African plate northward underneath the Aegean plate. We shall discuss below all the arguments, geological, petrochemical, and geophysical (seismological, magnetical, gravitational etc.) supporting the concept of the subduction of the African plate below the Aegean area, and furthermore we shall investigate systematically the geotectonic phenomena in the Aegean area in order to come to the most probable conclusion.

SEISMOLOGICAL ARGUMENTS

Caputo et al. (1970) suppose that the African plate is wedged under the European plate with a slope of approximately 58° in the Tyrrhenian region and a slope of 35° below the Aegean area. Papazachos & Komninakis (1971) suggest that the African plate underthrusts the Eurasian lithospheric plate in the Aegean island arc at a mean angle of about 30°. The length of this plate (Benioff zone) under the Aegean arc is 250 km. and its thickness 90 km. In a new paper (1973) Papazachos notes that the mean dip of the Benioff zone in the Aegean is about 35°, while its length is about 280 km. The thickness of the Aegean plate is approximately 65 km., while according to Makris (1973) the Moho's depth below the Aegean Sea is only 26 - 28 km. According to Galanopoulos (1973), using a list of earthquakes covering the period 1841 - 1959, the dip of the African plate under the Hellenic island arc is much more gentle, nearly 20°, and the strike of the plate NW 55°.

Besides the two big plates, the European and the African, McKenzie (1970) believes in the existence of two small rapidly moving plates. One of these, the Aegean plate, is moving towards the southwest relative to the European plate, producing extension and strike slip in the boundary between them and also a deep trough (Sea of Marmar, northern Aegean Sea etc.) associated with large basic inqusions. This feature is, according to McKenzie, comparable to that of the East African Rift, but the occurrence of some shocks of intermediate focal depth on the northern boundary of the so called Aegean plate makes this suggestion weak. (Galanopoulos 1972). The two-plates model of McKenzie must also be reconsidered from a seismological point of view (Papazachos 1973).

The concept of the existence of a Benioff zone in the Aegean area, dipping northwards is founded on the presence of intermediate shocks and the appearance of andesitic volcanoes in this region. Nevertheless, a careful examination of the spatial distribution of the intermediate foci in the Eastern Mediterranean does not favour this concept. If a Benioff zone were under the Aegean Sea, the average depth of the intermediate foci should increase with the distance from the Greek arc. But the delineation of the epicenters of the intermediate earthquakes of this area does not permit us to assume such a relation between the depth of the seismic foci and their place. Sometimes there is a clear increase in the focal depth, but from west to east (Kiskyras 1960; Galanopoulos 1973b).

As van Bemmelen (1971) notices, the section across Crete with the depths of the earthquake foci published by Galanopoulos (1968, fig. 5 on p. 187) does not show an inclining Benioff zone. It should be mentioned here that Galanopoulos (1973b) suggests later that the intermediate earthquakes beneath the Aegean Sea fall into a pattern that does not fit a simple model of a continuous inclined seismic zone dipping from the trench beneath the volcanoes. For that reason, in order to explain the anomalous distribution of the hypocentres of the intermediate earthquakes in the area of Greece, he supposes a disruption of the African plate in two segments along the Cretan furrow. On the basis of seismological data we should, according to theory of tectonics, conclude that the floor of the Ionian basin plunges simultaneously in a NNW direction underneath the Tyrrhenian area and in a NE to N direction underneath the Aegean area. But it is mechanically impossible (Rizema 1970). It is better to suppose that the Ionian basin plays only a passive role. The Calabrian arc overrides this basin in an ESE direction and the Hellenides - Cretan arc in a SW to S direction (Rizema 1970).

That is in agreement with the opinion (Kiskyras 1960) that the convection currents are moving below the Tyrrhenian Sea towards the deep Ionian Sea, as they do below the Greek area. The area of the deep Ionian Sea (Gondwana) is the foreland of both the orogens, the Greek and the Calabrian. The fact that the Greek orogen is thrust over the Ionian area produces the impression that the Ionian area (African plate) is underthrusting the Greek area.

As Fig. 2 shows, the foci of the Greek intermediate shocks are preferably concentrated in the area of the Eastern Hellenic zone, i.e. the main Greek eugeosyncline (Kiskyras 1972). This zone is characterized by the occurrence of igneous rocks, especially of ophiolites. These rocks are from the deep water outpouring of magma from the upper mantle (Brunn 1976). The andesitic volcanoes occur in the inner part of this zone partly metamorphosed, and are accompanied by the deepest of the intermediate earthquakes. These'shocks are connected with longitudinal tensional fractures (Kiskyras 1959, 1964 and 1972). Intermediate earthquakes occur also in the northern part of the Eastern Hellenic zone where volcanoes are not present.

Fig. 2 shows also that the Greek volcanoes are connected with transversal fractures. The intermediate earthquakes occurring in - the Pelagonian zone are associated with these fractures. There are five groups of transversal fractures accompanied by volcanoes and intermediate shocks:

1. The fractures along the North-Anatolian Line from the Aegean Sea across the Greek Mainland;
2. the second group of volcanoes and intermediate shocks is connected with the SW - NE fractures, cutting off the Cyclades from the Attica area;
3. the third group is along the islands of Milos and Antiparos;
4. the fourth group occurs along the islands Santorin - Antiparos, i.e. in a SE - NW direction; and
5. the fifth group with a SE  -NW direction is parallel to the large fracture of the Eastern Aegean.

Intermediate earthquakes occur also in the Olonos - Pindos zone. Precise focal determinations of earthquakes by using seismological stations installed in the Peloponnese have shown (Leydecker 1975) that the intermediate earthquakes are present east of 22^o E and east of 23^o E. This implies that intermediate earthquakes do occur in the eastern part of the Glonos - Pindos zone along the longitudinal tensional fractures of the Messenian Gulf, and also in the inner part of the Arcadian nappe zone, along the fractures of the Argolikos Gulf. The lack of volcanoes in both these zones may be attributed to the big thickness of their crust. In the area between 24^o E and 26^o E longitude, and north of 37^o N there are neither volcanoes nor intermediate earthquakes present. That may be explained by the lack of large fractures in this area. Generally the zones inside the Eastern Hellenic zone are poor in intermediate earthquakes; on the contrary, the zones outside of it are rich in intermediate earthquakes. The fact that intermediate shocks with a depth of 100 km or more occur in the area of Crete does not favour the concept of the subduction of the African plate below the Aegean plate.

From the above we can conclude that the andesitic volcanoes and the intermediate earthquakes in the Aegean area are closely connected with large tensional fractures, but intermediate earthquakes may occur independently of the occurrence of andesitic volcanoes; the presence of these requires a small thickness of the crust. In both cases the question has to do with conditions habitual in the geosynclinal area. With the occurrence of andesitic volcanoes and intermediate earthquakes in the Aegean Sea, there are not sufficient arguments to support the concept of the existence of a Benioff zone dipping northward below the Aegean area. Futhermore, the continuation of the Cyreneica sedimentary series till the Hellenic arc does not justify the underthrusting of the African plate below the Aegean Sea (Morelli et al. 1973).

Galanopoulos (1975), in view of the result of Morelli et al and taking into account the fact that intermediate earthquakes appear along the so -called Mediterranean Ridge in the convex side of the Hellenic arc, suggests the hypothesis of a new subduction zone along the Hellenic foredeep (Ionian and Levantine foredeeps) and a southward shift of the subduction process. That is supported by the opinion (Kiskyras 1960 and 1962) that the region west of West Greece and south of Crete constitutes a basin of sedimentation with the features of a narrow oblong geosyncline in evolution, where future Greek mountains will emerge.

But the question here is of an overriding by the Hellenic - Cretan system over the Ionian basin (piece of the African plate) and not of a subduction of the African plate.

New seismic data reveal according to Papazachos (1976), the existence of a Benioff zone in the northern Aegean - Marmara region. In this case the foci of the intermediate earthquakes are also concentrated in the area of geosynclinal zones i.e. the Axios zone and the Eastern Chalkidiki - Lesbos zone (eastern part of the Serbomacedonian zone after Mercier). The volcances and the intermediate earthquakes are connected with large tensional fractures in the inner part of these zones as also with the big fractures of the northern Aegean Sea.

The high heat flow measured in the northern Aegean, two very high values associated with intense magnetic anomalies (Morelli et al. 1975) may be attributed to geosynclinal volcanism and not to a rift, similar to that of the Red Sea.

Here also we have to do with an overthrusting of the geosynclinal zone to SW.

PETROCHEMICAL ARGUMENTS

The volcanic rocks of Isthmus, Methana, Aegina, Milos, Santorin and Nisyros, with a small potash silica ratio, are characterised as volcanic lava lying 120 - 150 km above the Benioff zone, while the Kos and Antiparos rocks with a higher potash content as volcanoes corresponded to a Benioff zone of 150 - 200 km depth (Ninkovich & Hays 1971). According to the observation of Dickinson & Hatherton (1967) in the Pacific, where the ratio K₂O/SiO₂ of volcanic rocks increases with the increasing depth of the Benioff zone below the volcanoes, a dipping of the Benioff zone northwards below the Aegean region is suggested.

It may be of interest to notice here that it has been shown (Nielson and Stoiber 1973) that the ratio of potash to silica content of the volcanic rocks is different for different volcanic arcs in the circum-Pacific. Besides that, the andesites of the Pacific are probably formed by fractional crystallization of the basaltic magma, while the andesitic rocks of the Aegean area are formed from synorogenetic magma (Kiskyras 1964) i.e. from magma with assimilated materials of the geosyncline (andesitic rocks rich in alumina with pyroclasts and xenoliths).

The assumption of the existence of a Benioff zone below the Aegean region does not explain the presence of both the alkaline and the calc-alkaline-lavas in the Islands Kos and Patmos.

Besides, the lavas with a high K₂O /SiO₂ ratio are preferably present in the volcanoes of Patmos, Kos, Pserimos, Lesbos, Imbros and Samothrace, lying in the area of the eastern Aegean Sea.

The observed differences in the K₂O content of the lavas may be attributed to the enrichment in potassium during the differentiation of the magma. The K₂O content increases with the removal of the elements Ca, Mg and Fe from the magma, bound to its early solidified ingredients (see Fig. 3a and 3b). Potassium with an ionic radius r = 1,38 also suppresses the sodium in the new lavas, for example in the rhyolitic rocks of Samos, Patmos, Kos and Antiparos (Kiskyras 1964). The lavas have a higher alcali content than their inclusions.

The lavas of Mediterranean type of the Aegean Sea are rich in SiO₂. The alkaline lavas of the Aegean Sea are probably derived from a magma with assimilated metamorphic igneous rocks (Kiskyras 1964).

The volcanic arc of Sousaki - Corinth - Aegina, Methana, Poros, Milos, Santorin and Nisyros argues against the presence of a Benioff zone, dipping northward with an angle of 30°. The mentioned volcanic arc corresponds to an arc of a small circle the pole of which lies near the island of Chios. If the volcancoes of this arc were connected with a fault plane it would have an inclination equal to its radius (polar distance) measured in degrees. In this case the polar distance (Santorin - Chios) is 2° 30 '. On the basis of that we can calculate the maximum depth of the magma chamber penetrated by this fault as equal to 12 km. It is of interest to notice here that the magma chamber of the Santorin explosion 1925 - 1926 had, according to F.v. Wolff (1938), a depth about one kilometer.

GEOLOGICAL ARGUMENTS

The orogeny, volcanism, seismicity and other geological and geophysical events in the Greek region are closely connected with the regional and temporal wandering of the geosynclines from NE to SW.

If the andesitic volcanoes of the Aegean area were associated with the subduction of the African plate dipping northwards below the Aegean plate the volcanoes of Methana, Poros, Milos, Santorin and Nisyros should be older than the volcanoes of Antiparos, Patmos and Samos lying northwards of the mentioned volcanoes. In the same way, the last volcanoes should be older than the volcanoes of Evoia (Euboea), St. Eustratios and Lemnos, lying further northwards of the first ones. But it does not happen; on the contrary, the volcanoes of the southern Aegean Sea are younger (Quaternary) than those of the northern Aegean Sea (Tertiary). Besides that, active Volcanism in Greece only appears in the southern part of the Aegean (Methana, Santorin and Nisyros).

It is very important to mention that the andesitic rocks of Sousaki (Corinth), Methana and Poros are present at the same places where ophiolites occur (Kiskyras 1964). The location of the ophiolites inside the Hellenic arc indicates that both these rocks are connected with the eugeosyncline. Therefore we don't consider these ophiolites to be slices of mantle emplaced cold into orogenetic belts and thrusting over the foreland.

The genesis of a geosyncline is due to a continuous down-buckling and attenuation of its substratum. That is the result partly of the weight of the surface sediments upon the substratum of the basin, and partly of the tensional movements due to the action of the convection currents upon the same substratum (Kiskyras 1964). These currents, moving from the Aegean Sea towards the cold substratum of the Ionian Sea (Gondwana), produced a continuous heating of the faced parts of the Gondwana and therefore a fusion or at least a plastic yield of it. After that, the geosyncline between the Aegean Sea and the Gondwana could be expanded SW into the plastic Gondwana in order to take the shape of an arc. The high positive values (+ 320 mgal) of the gravity anomalies in the Ionian Sea is an indication that the bottom of the deep Ionian Sea is an oceanic plate (Gondwana). The same conclusion is also suggested by the presence of low heat-flow values in the Ionian Sea (R. Haenel 1972).

The concept that the African plate underthrusts the Aegean area does not explain the building of the Hellenic arc with the convex side to south, i.e. in the opposite direction to the subduction (Schwan 1976). Besides that, the presence of strong compressional forces in Western and Southern Greece, and that of extensional forces in the Aegean area, i.e. inside the Hellenic arc, are not in agreement with the concept of plate tectonics.

When the deformation of the substratum of the eugeosyncline, due to the action of the convection currents and to the weight of the new sediments, was carried far enough, ruptures took place in the thin crust and blocks of it subsided into the upper mantle. After that a large mass of the underlying basaltic magma was introduced into the crust, with as a result the uplifting of the magma zone in this area. The derivatives of this basaltic magma are the cause of the strong positive magnetic and gravity anomalies in the Eastern Hellenic zone (Argolis and Aegean area). The fact that in the area between Crete and Africa the strong positive magnetic and gravity anomalies of the Aegean Sea do not continue, indicates that the crust there is non oceanic, and the so-called "Mediterranean Ridge" is not a ridge (Morelli et al. 1975).

Another result of the plastic yield of the northern parts of the African plate (Gondwana) due to the action of the convection currents was their subsidence and the building of new geosynclines SW of the main geosyncline. Fig. 4 (taken from Kiskyras 1972 p. 99) shows the wandering of the Greek geosynclines from NE to SW and the thickening of the Hellenides in the same direction. Recent geophysical investigations (Makris 1973) revealed that the crustal thickness of the Hellenides are of the order of 42 - 48 km in Western Greece, and of 32 to 34 km in Eastern Greece, whereas the Aegean area has a much thinner crust, about 26 to 30 km. This thickening of the Hellenides can't be explained by strong compressional forces acting on the Hellenides from WSW to ENE as a consequence of the opening of the Atlantic. It is the result of sedimentation and overthrusting, so that the external zones are thicker than the internal ones.

Summarizing, we can say that the Gondwana (African plate) leap from the cratonic to the geosynclinal state is because of convection currents. The consequence of that is the gradual destruction of the African plate from NE to SW and the building of new geosynclines in the opposite direction. The sediments of the new geosynclines are thrust over the area of the Gondwana, which produces the impression that the African plate underthrusts the Aegean area.

---------------------------------------