New Evidence of Late Holocene Sea-level Changes in the Aegean and Santorini: Tectonic, Eustatic and Volcanological Effects

The main characteristic of the observed relative sea-level changes is that they have a clear, differential character, and this reflects tectonic movements. However, the possibility of eustatic effects cannot be rejected, while in at least one case sea-level fluctuations possibly reflecting an interaction of volcanic and hydrothermal activity have been identified.

With specific reference to Santorini, there is ample evidence of a sea-level rise, which in association with marine erosion causes severe changes in the Classical landscape, at least, but the available data are too scarce to permit the assigning of this movement to volcanic activity.

INTRODUCTION

The study of the changing coastal topography in the Aegean and the reconstruction of old coastlines, especially of the prehistoric period, is a relatively new, but very important field of archaeological research. Detailed, rather local studies and palaeogeographic reconstructions have been made in several coastal locations in Greece and in Asia Minor, for example in the Argive plain (Finke 1988), in Messenia (Kraft et al. 1975), in the Thermopylae plain (Kraft et al. 1987) and in Troy (Kraft et al. 1980), while larger scale surveys have been completed by Flemming (1973; 1978a) and Pirazzoli et al. (1982) or are still in progress (for example Stiros 1988; Stiros and Papageorgiou 1988a, b).

For Thera, especially, the reconstruction of the pre-eruption coastal topography is more important and of higher priority than for any other ancient site, for the Minoan eruption was associated with an impressive alteration of the landscape: collapse of the caldera, disintegration of the remaining part of the island with blocks moving possibly independently, and finally their burial under a thick layer of volcanic products.

Although this was recognized some time ago and has been discussed in previous congresses on Thera and the Aegean World (Rapp and Kraft 1978; Flemming 1978b), until now no detailed studies for the reconstruction of the pre-eruption landscape have been made, and no assessment, for example, of Doumas's hypothesis about the location of the harbour of Akrotiri (Doumas 1983, 55) can yet be provided. However, in 1985, a multidisciplinary project supported by the Institute of Geology and Mineral Exploration (IGME) started in order to study systematically the problem of the relative sea-level changes and the large-scale palaeogeography of the Aegean mainland and island coasts (see Stiros 1988), including those of Thera; this last point is the subject of the present paper.

THE PROJECT

Our research is mainly focused on the central and northern part of the Aegean and on the island coasts where no systematic surveys have ever been made, in contrast with the near-arc areas, where sea-level fluctuations have been studied in detail by N. Flemming and P. Pirazzoli (Flemming 1973; 1978; Flemming and Pirazzoli 1981; Pirazzoli 1986a, b).

The aim of our research is to identify and study traces of ancient coastlines, emerged or submerged, based on archaeological data, most of them collected from library work and field geomorphological observations. The main objectives of this Project are: i) identification of recent tectonic motions from differential sea-level changes, and ii) correlation of these tectonic motions with longer-scale motions deduced from geological data. For archaeological research: i) identification of all the ancient coastal sites and compilation of a data-base that will help in their study and preservation; and ii) contribution to the study of the changing coastal topography and the palaeogeography of ancient coastal sites.

METHODOLOGY AND DATA

The data basically used in our study are of three kinds: first, archaeological marks of an old coastline, i.e. ancient constructions, which today appear to be in abnormal or normal position relative to their original use, and therefore indicate a relative sea-level change or stability. Such marks are harbour installations (breakwaters and quays, for example) which today are above the present sea-level, or land constructions (houses, temples, basilicas, etc.) and other coastal installations (fish tanks, Roman piscinae, coastal quarries, etc.) which today are completely or partly submerged; second, geological-geomorphological marks, including marine notches (wave rock-cuttings) and surf-benches, the typical cross sections of which are shown in Fig. 1a and 1b respectively. These are formed on coastal solid rocks, mainly by sea wave action over a few thousand years. Although their evolution is still a matter of debate among scientists (for example Higgins 1979; Pirazzoli 1986a), it is widely accepted that the roof of a typical limestone notch forms at about 30 to 40 cm above high tide. Consequently, if notches with roofs below or above this level are observed, a relative sea-level rise or fall, respectively, can be deduced; third, combination of archaeological and geomorphological observations: such examples are exposed marine fossils on ancient constructions or ancient pottery fragments cemented offshore (Stiros 1988; Stiros and Papageorgiou 1988a).

FIRST RESULTS OF THE PROJECT

The first results of this Project, obtained from the data collected so far, are summarized in the following: first, subsidence prevails almost in the whole of the central and northern Aegean, but still uplifted coasts, occasionally more than 100 km long have been identified (for example north Peloponnesos, central-east Euboea); as far as the area around the arc is concerned, subsidence and uplift seem to alternate (see Fig. 2); second, the observed relative sea-level changes have a clear, differential character, since adjacent areas of uneven subsidence or areas of subsidence next to uplift have been identified; third, tectonic motions are responsible for these differential sea-level changes, and possibly for the large-scale subsidence in the Aegean. However, the available data do not permit acceptance or negation of the hypothesis of important eustatic effects, since the latter may have been overprinted by tectonic motions; fourth, sea-level fluctuations of a few metres at the wavelength of a few hundred or thousand years are not confined to Crete and other islands of the arc (Pirazzoli 1986b), but have also been observed elsewhere, for example, in the vicinity of Thermopylae (T. Niemi and T.J. Van Andel pers. comm.). It is likely those motions reflect volcanic and hydrothermal activity, associated with a nearby Quaternary volcano.

SEA-LEVEL CHANGES IN THERA

Especially concerning the coasts of Thera and of surrounding islets, our research revealed that at certain sites shown in Fig. 3 there exist some signs of old coastlines. More explicitly:

       Cape Exomytis:      In the coastal area between Cape Exomytis, at the southernmost point of the island, and Vlychada, some hundred metres south-west of St. George's chapel, two huge, submerged moles were reported (Sperling 1973, 18) and marked on the Admiralty Chart prepared by T. Graves in 1848. Lehmann-Hartleben (1923) suggested that they belong to the ancient city of Eleusis, mentioned by Ptolemy (III, 14, 23), the location of which is still a matter of debate (Sperling 1973, 18, 31, Appendix 2). Our investigation, however, revealed that these 'ancient moles' are only natural rock formations.

In the same area, ancient graves washed by sea-water and buried under the sandy beach have been reported in the past (Scholes 1956). These graves are occasionally exposed and observed during storms, when the sand cover is removed, and are probably progressively destroyed by erosion. Their top is below sea-level, while the pottery indicates an Archaic age.

       Kamari:      Remains of the important coastal settlement of ancient Oia are found along the southern end of present-day Kamari. Evidence of a sea-level change in the area is provided by the following:

i) An ancient wall in situ, between the rock mass and the Irini tavern, below the black sand, periodically exposed, depending on prevailing winds.

ii) An early Byzantine basilica excavated by Ch. Sigalas, next to Kamari Hotel and very close to the present-day beach. The floor of this basilica is about 1 m above sea-level, and was covered by about 1.5 m thick sediment. Thus, even if the 15-20 m advance of the coastline due to subtraction of huge amounts of sand from the area for the rebuilding of Thera after the 1956 earthquakes is taken into consideration, this church is at a very low level and close to the open sea front, and consequently it is difficult to accept that originally it was built in a position not protected during storms. Thus, a considerable advance of the sea-front, possibly associated with a rise in its level, is inferred. It is interesting to note that a layer of sand exists between the present-day soil and the floor of the chapel (Fig.4). This probably reflects development of dunal environment after the abandonment of the site and not sea-level fluctuations.

iii) At St. George Thalassitis chapel, to the north of Kamari and close to the beach, during an excavation in 1978, well-preserved remains of a possibly Roman villa have been identified by Ch. Sigalas. The threshold of this building was only about 0.2 m above sea-level, indicating a considerable sea-level rise since that period.

iv) Rather poorly developed notches exist on the limestone mass marking the southern limit of Kamari beach; the top of these notches is below the limit of 30-40 cm, indicating a relative sea-level rise (see above).

       Perissa:      Some investigators (Sperling 1973, 31) suggest that the southern harbour of the ancient city of Thera may be identified in the small embayment of present-day Perissa, where there is some poor evidence of ancient inhabitation. Recently, medieval remains have been excavated below the church and other buildings near the shore, but although there is some indication that their floor was at a relatively low level, no reliable conclusions about sea-level changes are possible yet.

       Therasia, south coast:      Considerable remains of a prehistoric settlement have been excavated by Fouqué and others at 'Alafousos property' on the south coast (Sperling 1973, 39; Page 1970, 25). Sperling, based on this excavation, assumes that the Minoan coastline was further south, and a marine transgression was associated with the eruption. Today no signs of this excavation exist, so this information cannot be evaluated.

       Phaneromeni:      Submerged ancient (medieval?) remains are reported on the northern side of Therasia, near the chapel of Phaneromeni (Sperling 1973, 41), where surficial evidence of ancient inhabitation exists, but we failed to observe them.

       Other sites:      Indication of ancient occupation in several other near-coastal, present-day sites exists. For example, pottery sherds have been observed at the low hill west of Koloumbos, not away from the sea (see also Scholes 1956, 13), where changes in the coastal topography associated with the 1650 eruption of the nearby crater are likely.

On Christiana, an islet about 20 km south-west of Thera, a cemetery has been excavated by C. Doumas. This indicates some important ancient inhabitation in this area, which probably would have been associated with some coastal installations. Thus, detailed investigations in such areas may provide some useful information about ancient coastal topography.

DISCUSSION

The above data suggest that important changes in the coastal topography of Thera and the adjacent islets have taken place since Minoan and Early Byzantine times, and that a marine transgression, due to a sea-level rise and erosion is to be inferred. However, the available data are too scarce to permit conclusions about the amplitude, age and origin of these changes, as well as their possible correlation with volcanic effects.

As far as longer term motions are concerned, exposed Upper Quaternary marine strata near Phira have been reported (Fouqué 1879; Keraundren 1970), but again the available data are too poor to describe the corresponding sea-level changes.

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

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