The Geological Evolution of the Aegean Region

It offers in a unique way the possibility to study the interrelation between the evolution of an Alpine orogenic belt and the development of a marginal basin in it. The intimate relation of young continental orogenies and marginal basins seems to be a general rule. Gravity instability caused by nonhydrostatic forces in a laterally nonhomogeneous crust is regarded as the initiator of disintegration in the marginal basin. Upwelling mantle material caused by lithospere subduction reinforces the back arc expansion. This is shown by model experiment.

INTRODUCTION

The Aegean region is a small but geodynamic ally extremely active part of the Alpine - Himalayan orogenic system. Its seismicity is the highest in the European - Mediterranean region. Seen from the perspective of a satellite it seems that the mountain chains of the Dinarides - Hellenides are spreading apart and submerging in the floods of the Aegean Sea before reappearing in the Tauridic mountain belt of Anatolia. The traces of the foundered land masses are marked by islands, scattered or grouped into chains. This rather superficial picture already contains quite correctly some essential features of the geologic history of the Aegean World.

In this paper an attempt is made to summarize and evaluate our present knowledge of geological and geophysical information with the aim of a better understanding of the processes which have finally led to the present situation.

Fortunately, in recent years, stimulated by the International Geodynamics Project, comprehensive systematic research has been carried out in many fields of geosciences in this area.

Physiographically the Aegean Sea is a marginal basin bordering or even penetrating the European continent, and bounded toward the eastern Mediterranean Sea by the Hellenic Island Arc. A first question arises: Is this structural similarity to Pacific, Indonesian, or Caribbean island arc systems only superficial, or are these processes of essentially the same kind? There exists a broad spectrum of widely differing opinions. A systematic comparison may help us to find the answer.

The most comprehensively investigated part of the Hellenic Arc is the Peloponnesos region, which may, therefore, serve as an example of the arc structure.

Taking into account recent results from marine geophysics (Weigel 1974, Finetti 1976, Jongsma et al. 1977), neotectonics (Kelletat et al. 1976), precision hypocenter determinations and focal mechanism studies of earthquakes (Leydecker et al. 1977, Ritsema 1974), explosion deep seismic sounding (Makris 1977), and gravity field survey (Makris et al. 1973) it may be justifiable to draw a generalized West - East cross-section through the Ionian Sea, Hellenic Trench, Peloponnesos into the Aegean Sea as shown in Figure 1 (Berckhemer & Kovalczyk 1977).

Thick lines indicate seismic refractors or reflectors, chiefly crust-mantle boundaries. Low angle underthrusting of the Ionian crust below the Peloponnesos and / or equivalent thrusting of the Peloponnesos over the Ionian crust is evidenced by a gently dipping refractor, the seismic stress field, a thrust fan of sediments in the trench region, and the upward bending and exposure of the lower Hellenic nappes near the Ionian coast. The antithetic block rotation within the body of the Peloponnesos might be the consequence of under-thrusting in the West.

Toward the Aegean Sea the tectonic style changes into predominantly normal faulting with extension of the crust. A "Benioff Zone" of earthquake hypocenters, branching off where the crust reaches its greatest thickness and dipping down under 40^o toward the central Aegean to a depth of 160 -180 km, is clearly visible. The tension axes of some intermediate deep shocks point along the dip of the Benioff Zone indicating, in terms of plate tectonics, a lithospheric slab pulling downward under its own weight. Where the foci reach a depth of 100 - 150 km Quarternary andesitic volcanism occurs at the surface. The K₂O/SiO₂ - ratio of these volcanic rocks compares well with those from other island arc regions where a Benioff Zone lies at a depth of 120 - 150 km (Ninkovich & Hays 1973). Although one has to be careful in generalizing over the situation of the Peloponnesos, the following features, typical for marginal sea-island arc structures, are found in the Aegean region: a deep sea trench with a pronounced negative free air gravity anomaly af - 130 to - 250 mgal (e. g. Fleischer 1964); a non-volcanic frontal arc (Hellenic Island Arc) stretching from Kephalinia via Peloponnesos, Crete, to Rhodes; high seismicity with foci in normal and intermediate depth; a Benioff Zone of amphitheatrical shape, dipping down toward the central Aegean, is at least indicated in a general sense (Papazachos & Comninakis, this volume); an inner island arc with Quarternary andesitic volcanism (Aegina, Methana, Milos, Santorini, Nisyros, Antiparos, Kos) above hypocenters in 100 - 150 km depth at a distance of about 200 km from the trench; tensional tectonics in the back-arc region (see Figure 2); high heat flow up to 2.7 HFU in the central Aegean Sea, (Jongsma 1974) and low heatflow in front of the arc; high absorption of seismic waves in the upper mantle behind the island arc (Papazachos & Comninakis 1971).

There are, on the other hand, differences compared with many other island arcs: Small dimensions (the radius of curvature of the frontal arc being only 300 km); no typical oceanic crust (thickness exceeding 20 km including some 10 km of Neogene sediments) in the eastern Mediterranean, and a huge sedimentary chain (East Mediterranean Ridge) in front of the Hellenic Trench, which, in its eastern part, doubles itself into the Pliny and Strabo Trenches; a thick, predominantly sialic (continental) crust of some 30 km thickness below Crete and the Cyclades and of 20 km below the Cretan Sea (Figure 3), and a comparatively low sub-moho velocity of v_p = 7.7 - 7.9 km/s (Makris 1977). There was, with the possible exception of the Saros Trench in the northern Aegean, no true sea floor spreading. Some scattered intermediately deep shocks occurred outside the Benioff Zone, in particular in the northern Aegean (Papazachos & Comninakis, this volume, and Galanopoulos 1972). The non-volcanic, sedimentary frontal arc is clearly dominating while the volcanic arc is only poorly developed.

Even the atypical features are, at least in tendency, known from a few other island arcs. Differences seem to be more of a gradual than of a principal kind.

According to Karig's classification (Karig 1971), the Aegean region can be considered a young, small, inactive (with regard to sea floor spreading) marginal basin with high heatflow in a somewhat hampered state of development. If this is essentially true, the same physical and chemical processes should be responsible for its evolution and no special mechanism is required. On the other hand, conclusions drawn from observations in the Aegean might be of use for a better understanding of other marginal seas.

As a unique case, Alpine orogenic belts can be traced from the Hellenides through the island arc up to the Taurides, following the multitude of islands in the Aegean Sea (Figure 4). The apparently intimate connection of an Alpidic orogeny and the successive island arc development in one and the same region is a fact which needs to be emphasized. The frontal arc and the back-arc islands were fully involved in the Alpine orogeny of the Dinarides - Hellenides - Taurides realm. This might be briefly sketched below.

In the beginning of the Alpine orogeny the situation was characterized by a broad, stable shelf platform of the African Continent, some rather mobile depressions like the Gavrovo, Tripolitza and Pindos troughs and, separated by a crystalline anticline, the Vardar Basin which represented the central Mesozoic Tethyan ocean. Together with the shelf region in the North, the width of the whole marine area might have amounted to some 1500 - 2000 km. According to the process of rifting in the Atlantic Ocean (Dewey et al. 1973) lateral and convergent movements occured which led, in the course of the closure of the Tethys and of the continental collision, to the Alpine orogeny. Following Jacobshagen (1977), in the Hellenides five tectogenetic phases can be distinguished, which migrated from the internal (northern) to the external (southern) zones and were in action successively, or overlapping in time and space, along the belt. Deformations and metamorphism in the circum-Rhodope belt and in the internal Vardar zone took place in the Jurassic. During the Eohellenic Phase (Upper Jurassic/Lower Cretaceous) crustal shortening caused overthrusting Of Ophiolitic nappes from the external Vardar zone onto the Pelagonian platform. In the Upper Cretaceous a similar phase of crustal shortening and subduction affected the southern and eastern Aegean and seems to be directly connected with the over thrusting of Ophiolitic nappes in the Lycian Taurus. In the Eocene a phase of folding traversed obliquely existing orogenic trends in the Aegean.

Beginning in the Oligocene and culminating in the Lower Miocene (Burdigalian), an in tense tectogenetic phase of compression with underthrusting, folding and crustal thickening affected the central Aegean area. With increasing temperature the Median Crystalline Massif was uprising, causing the Hellenic nappes to slide onto the external foreland. The former proximity and the continuous contact between the central Aegean and the frontal island arc up to the Pliocene is impressively demonstrated by piles of allochthonous nappes transported far outward from the internal zones, and resting now on the Mesozoic carbonate base of the Cretan crustal block (Wachendorf et al. 1975), Figure 2.

In the rear of the outward migrating zone of compression, already in the Upper Miocene/Lower Pliocene, distractive tectonics commenced with intense normal faulting, differential block movements, general subsidence and simultaneous extension. The Aegean Archipelago proper developed in the Upper Pliocene. The block tectonics were reactivated when, some 3 m.y ago, the island arc volcanism began. 

It is clearly seen on a compilation of neotectonic structures in the Aegean region (Figure 3) that normal faulting and graben-like structures with the tendency of expanding the Aegean Basin are abundant in particular in the southern part (Cretan Sea), and in the northern part (Saros Trench), and penetrating deeply into western Anatolia. Miocene volcanism is widespread in northwestern Anatolia, and volcanic intrusions within the Aegean Sea are marked by magnetic anomalies. Present day extensional tectonics in most of the Aegean and western Anatolian crust, and a frontal zone of compression stress, are evident from earthquake focus analysis (Papazachos & Comninakis, this volume).

An already existing orogenic ridge disintegrated and was destroyed. Why just here and not somewhere else in the Alpine belt? A simple and obvious answer is that island arc formation requires a continental margin and only this part of the Alpine chain is now bordering at least a pseudo - oceanic crust, perhaps the remainder of a Mesozoic Tethys basin (Monod et al. 1974). This answer is still unsatisfactory for at least two reasons. A crucial question is why the crust bordering the Aegean region in the South was once able to stand or even to provide the compressional forces for the orogenic phases in the Aegean Alpidic belts, but was, later on, overridden and subducted. A belt of Mesozoic carbonate series which bottoms the external Dinarides, Hellenides, the Cretan Arc, and the Lycian Taurus (Laubscher & Bernoulli 1977) indicates the former existence of a very extended, relatively rigid and stable shelf platform of Africa, like the present Pre -Apulian and Adriatic Platform. In the rear of this platform a Mesozoic Tethian basin (the present Eastern Mediterranean) might have existed (Monod et al. 1974) which, with the outward migration of the Alpidic orogeny, approached the Aegean region and finally allowed the island arc to develop. This was the particular situation in the Aegean region, but it seems not to be an absolutely unique one. Where ever an island arc approaches a continent (Andaman Sea, Bering Sea, eastern Caribbean) the non volcanic frontal arc merges continuously into a continental mountain system of Tertiary age and the back arc basin contains areas of relatively thick, epicontinental crust (Figure 5).

We still don't understand quite clearly what caused the Aegean orogen to spread apart, to subside, and to form the island arcs and the marginal sea. By analogy we assume the same mechanism as in other comparable areas. Toward the end of the high orogenic phase in the Aegean area probably a young mountain range of considerable relief existed (see above) which was bordered by a Mesozoic Tethian ocean. It can be shown by a simple derivation (Berckhemer 1977), and this has already been discussed in detail by Artyushkov (1973), that a laterally nonhomogeneous crust of variable thickness, even in isostatic equilibrium with its viscous mantle below, exerts horizontal forces acting from areas of high relief to those of low relief. One third only of the surplus energy of the thick crust is stored in the relief itself, and two thirds in the root of the mountain. It has been shown (Berckhemer 1977) that this potential energy could easily account for the total seismic energy released in the Aegean region during the past five million years, provided the present activity consitutes a representative average for this whole period.

It has also been found that the mean horizontal stress is about equal to the load of the surface relief. For example, a mountain range of 3 km elevation relative to the adjacent seafloor produces a mean horizontal stress of about 800 bar which has to be taken up by counter forces provided by the adjacent crustal blocks, or by the internal strength of the crust. Otherwise the thick crust will flow apart with the tendency to equalize the crustal thickness. Low strength, however, is expected inside a young mountain body, because it is still under devated temperature and the viscosity or any kind of creep strongly depends on temperature. 

This gravity controlled disintegration is regarded by the author as a possible cause for the initiation of the marginal sea development. First the eastern Mediterranean crust was passively overridden from the North and the frontal arc was created. Subsequently, under the weight of the overload, the oceanic lithosphere became actively pulling downward due to increase of its density caused by high pressure phase transformation of minerals such as Gabbro - Eclogite.

Subduction began. The rate and direction of subduction was governed by the relative motion of Africa, Europe and Arabia, and by the expansion of the Aegean sea. This expansion, in turn, is reinforced by upwelling mantle material behind the arc, which is simply a consequence of conservation of volume when the mantle gives way before the intruding oceanic lithosphere. This has been demonstrated by the author in model experiments. Honey was used as the model substance to simulate the viscous mantle and a slab made from plexiglas strips served as the model of the oceanic lithosphere. Phase transformation to higher density was simulated by loading the edge of the slab with steel balls when it submerged into the viscous fluid. The free motion of the slab under its own weight and the countercurrent in front of the regressing edge, which corresponds to the spreading of the back arc basin, is seen in Figure 6.

Somewhat similar results, but from different reasoning, were obtained theoretically by Sleep & Toksöz (1971) and by Andrews & Sleep (1974), (see also Papazachos & Comninakis, this volume). The active volcanism in the zone of subduction, to which also the island of Thera owes its existence, is not shown by the model experiment, but the usual assumption of remelting of the wet, subducted crust with mobilisation and contamination of mantle material seems acceptable to the author.

The active back-arc expansion may also account for the strong curvature of the Hellenic Island Arc which cannot be understood simply by unidirectional subduction of an oceanic plate. The mantle diapir postulated in a general sense by Van Bemmelen (1975) and in specific form for the Aegean by Makris (1976) and by Wachendorf et al. (1975) as an ad hoc agent for the evolution of the Aegean Sea, or the convection current acting against the African plate, as postulated by Kiskyras (this volume), might perhaps be considered particular aspects of the total process.

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

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