Thera as the Tectonic Focus of the South Aegean: Archaeological Evidence from the Aegean Margin

The central Aegean islands may thus be decoupled form each other and move tectonically quite independently.

INTRODUCTION

Although there are regional variations in estimates of the net eustatic sea level change in the last 4000 years (Scholl & Stuiver 1967; Flemming 1969; Mörner 1969; Pirazzoli 1976; Kraft et al. 1977) there is no suggestion that the world sea level in 1500 B.C. was more than 1 - 2 m different from present sea level.

The present author is of the opinion that, for the Mediterranean region, the relative change of sea level which is uniform throughout the area has been less than 1.0 m, and probably less than 0.5 m, during the last 4000 years (Flemming 1969, 1972, 1977, 1978 in press, and Flamming et al. 1973). Thus changes in the vertical position relative to the sea for the island of Thera, and changes in coastal topography, must be due to tectonics, local volcanism, and coastal erosion or deposition.

Vertical tectonic movements and small eustatic changes can be detected at many archaeological sites on the Aegean coast. In practice only the relative change can be observed for each single site, and the tectonic and eustatic factors are separated statistically for groups of sites (Flemming 1972). I do not have feild evidence for relative vertical movements on Thera, nor for most of the islands of the Cyclades and Dodecanese. In the present brief review paper I will summarize the evidence available from the Aegean margin, and from this purpose some hypotheses about the nature of vertical movements which might be expected at Thera itself.

FIELD OBSERVATIONS

Two hundred and two sites were visited between 1967 and 1977 on the coasts of the Peloponnese, Kythera, Crete, Karpathos, Rhodes, south-west and south Turkey, and Cyprus. Surveys were carried out using snorkel swimmers and divers in those cases where the archaeological remains extended under the water. Observations of relative changes of sea level were made using the methods described by Flemming (1969), and Blackman (1973). Full bibliographic references were assembled for each site, together with maps and charts of previous surveys, and these are published more fully elsewhere.

The field measurements were treated statistically using computer methods to contour blocks showing systematic tectonic trends, and separating out a eustatic factor. The observations show that the margins of the Peloponnese are subsiding at rates of 1 - 2 m per millenium, with the centre of the Peloponnese doming upwards, and the ends of the Messenia and Argolis peninsulae being the most depressed. From Monemvasia to Neapolis there is a zone of reduced submergence, which extends through Kythera, and leads into a region of uplift at Antikythera. The channel between Antikethera and Crete is probably a graben structure, of which the floor is subsiding, while the two flanking islands are uplifted, but at different rates. Western Crete is tilting upwards at the south-west corner, with Elaphonisi rising at a rate of nearly 5.0 m per millenium about a hinge-line from Rethymnon to Matala. Central Crete is irregularly crumpled, with some depression along the northern coast. East of Hierapetra isthmus eastern Crete is depressed at the north where Cap Sidhero is submerged by about 2.0 m, and uplifted in the south where the island of Kuphonisi is raised by several meters. There is a discontinuity between Crete and Kassos. Karpathos is subsiding at about 1.0 m per millenium, whilst maintaining horizontallity, and not tilting. There are further discontinuities between Karpathos and Rhodes, and between Rhodes and Turkey. Rhodes is folding in an antiform manner, with an axis of uplift extending approximately westwards from Kalithea: at Kalithea itself the uplift is about 1.0 m per millenium, with one limb sloping northwards towards Rhodes city, and the other sloping southwards towards Prasonisi. Echo sounding show that the margins of the Rhodes channel are broken by step features which may be due to slumps, or the faults of the graben structure.

Southern Turkey from Fetiye to Gelidonya is subsiding at 2.0 m per millenium, but the zone north-west from Marmaris, through Cnidos to Kusadasi and Ephesus, is almost totally stable in relation to present sea level. Within this stable zone there are a few sites: Loryma, Bodrum, and Myndos, which show subsidence of the order of 0.3 - 0.5 m per millenium. The Izmir - Cesme peninsula shows increasing subsidence from east to west, with rates of submergence of the order of 1.5 m per millenium at Erythrae, Ildir.

COMMENTARY

There is scattered literary evidence from various reports to show that earth movements could be detected in the Cyclades islands from the archaeological remains. However, most of these observations are not backed up by accurate measurements or geological descriptions. Thus we have to try and interpret the possible nature of earth movements in the centre of the Aegean from the data which we have from the margins.

Various authors (McKenzie 1970, 1972; Galanopoulos 1973) have interpreted the tectonics of the Aegean area in terms of plates, including the so-called Aegean plate. It is however extremely difficult to define the rigid extent of this plate, and equally difficult to define its active boundaries. The data summarized in the previous section are used by Flemming (1978 in press) to support the hypotheses that the margin of the south Aegean is composed of a large number of small lithospheric blocks with lateral dimensions of the order of 50 - 100 km, which are decoupled from each other, and which may be tilting as rigid blocks, or may be stable, or slightly distorted. This hypotheses is not compatible with the proposition that the Hellenic arc is a conventional island arc and lithospheric subduction zone, nor with the proposition that there is a rigid Aegean plate defined by sharp boundaries.

It is proposed (Flemming 1978 in press) that the south Aegean is a complex region within which the crust has been thinned by mantle upwelling, leading to radial surficial extension, thrusting the islands of the Hellenic arc and the Peloponnese outwards, overriding the floor of the Mediterranean. The Mediterranean lithosphere is partially subducted under the Hellenic arc, but, because of the small radius of curvature, the subducted slab must break into numerous separated fingers (Frank 1968). Further speculation suggests the following general model for the south Aegean. Because the descending slab is broken into discrete fingers it is easier for mantle material to rise. The fingers actually diverge toward the centre of the Aegean, so that the area is likely to have very low rigidity (Hedervari 1973), and it is possible that many of the central Aegean islands will be tectonically decoupled from each other, liable to move independently vertically or laterally, with overall radial dispersion. Thera is at the focus of this movement, with radial lines of dormant volcanoes, seismic activity, and magnetic anomalies, extending from it.

CONCLUSIONS

Statistical tests designed to detect a common eustatic component in the field data for the vertical displacement of south west Turkey resulted in a maximum estimate of 0.3 m (Flemming 1972). Vertical tectonic movements of the south Aegean area are of the order of 2.0 - 5.0 m per millenium in the more active blocks, which may have lateral demensions of many tens of km.

Thera, almost by definition, is in a highly active zone, and by analogy with other volcanic regions such as the Bay of Naples, Indonesia or New Zealand, vertical movements could be of the order of 5 - 10 m per millenium. However, the present data indicate that there is no way of extrapolating from one island to the next, since the islands of the Aegean are usually decoupled from each other. It is equally probable that Thera may have moved 10 m per millenium either up or down.

Until incontrovertible evidence is available from the island itself, the magnitude and direction of the tectonic movements can not be estimated. However the tectonic movement is liable to be large, while the eustatic component is probably less than 1.0 m over several thousand years, so that the solution of the problem lies almost entirely in the realm of local tectonics.

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