Aegean Sea Level Changes in the Bronze Age

The conclusions from these local studies are not directly applicable over a wider geographical area, but similar field studies on Thera could help to answer questions about Thera's Bronze Age coastline.

In work related to his excavations at Akrotiri, Spyridon Marinatos (1971) observed "that the lowermost layer of the eruption (pumice about 4 m. in thickness) as well as part of the superimposed ashes, are now below sea-level along the south shore of Thera." Marinatos believed (oral communication) that the "Minoan" shoreline near Akrotiri was at least 800 m. seaward of its present position, and that this relative sea-level change was due to a sinking/tilting of the southern tip of the island contemporaneous with the eruption.

Relative sea-level changes for any local area are predominantly due to some mix of three distinct geological processes:

1. eustatic sea level rise or fall,
2. vertical tectonic movements of a regional land mass,
3. addition to or removal of sedimentary deposits from coastal areas.

Many examples of sediment infill which extended the land mass seaward during the late Holocene are detailed in Kraft et al. (1977). It is rarely possible to quantitatively determine the contribution of each process to relative sea-level change in any area. However, it is feasible to generate relative sea-level data for the local area where the tectonic and sedimentary processes are operating. Data from one area are not fully applicable to any other area. With this caveat in mind we will attempt to discuss Aegean sea level changes in the light of data developed from our detailed studies of Holocene coastal sediment deposits on mainland Greece. Although to date our studies have not included the Aegean islands, a common trend of eustatic - tectonic sea level change is emerging from the results obtained on the mainland. Thus this paper might provide data and a conceptual framework for those scholars attempting to reconstruct the paleogeography of Minoan Thera.

The study of coastal change in areas where sediments have been deposited must include an analysis of vertical sedimentary sequences in order to successfully produce a three-dimensional model. The record of Holocene sedimentation lies in the subsurface. Efforts to secure this record have centered on obtaining cores from a drilling program undertaken in the unconsolidated coastal sediments in a number of archaeologically important areas (Kraft et al. 1975; Kraft et al. 1977; Rapp et al. 1978). Figure 1 is a map of the Aegean Sea and the surrounding landmass showing major basins of sedimentary deposition in the Holocene and portions of such basins in the Pleistocene. Many of these basins are tectonically controlled grabens (downfaulted blocks).

Although figure 1 emphasizes alluvial deposition in valleys and embayments of the mainland surrounding the Aegean Sea, the same processes occurred on a smaller scale in the islands. For example, similar patterns of Holocene deposition are common on Crete and Lesbos. Volcanic deposits such as those on Thera represent an extension of the potential for geomorphic change. Where deposits extend the shoreline seaward, a marine regression occurs. Such relative change combined with absolute (eustatic) sea level change determines where shorelines and other coastal features were located in historic and prehistoric times.

Although no extensive attempts at paleogeographic construction for Thera have yet been made, H. Reck (1936) has proposed that central Thera was a volcanic dome of medium height surrounded by hills cut by fairly large valleys.

In the interior of the island he suggested there may have been a fertile plain and some fresh water.

Massive eruptions deplete the magma chamber and broad tracts surrounding a volcano may subside or collapse. This subsidence would be one form of vertical tectonic movement effecting a local relative sea level change. Marinatos (1971) suggested extensive subsidence and tilting. Mavors (1969 p. 235) draws a hypothetical 1500 B.C. beach line approximately 200 - 400 meters offshore through remains of buildings in three meters of water off the south coast of Thera. That there has been relative rise in sea level on the south coast of Thera since 1500 B.C. cannot be questioned. How much relative rise for what coastal areas of Thera and due to what mix of geological factors is still in question.

Figure 2 is a generalized topographic/bathymetric map of Thera. There are a few coastal sedimentary deposits associated with small streams draining the south coast that could be used for the type of investigation we have undertaken elsewhere in the eastern Mediterranean. So far time has not permitted a core drilling program. However, Marinatos' comment (1971) cited at the beginning of this paper is of major significance. Comparison of our relative sea-level curves for various embayments in the Peloponnese and Hafemann's (1960) projections shows reasonably close agreement since Bronze Age times. Our data from three embayments at the Bay of Navarino, the Gulf of Messenia, and the Gulf of Argos show variable sea level rise components from two to four meters since the Bronze Age. Hafemann's (1960) projections are of the same order. Since multiple study areas show two to four meters of relative sea-level rise it is reasonable to assume a similar eustatic component would apply to the coasts of Thera.

A common hypothesis concerning recent eustatic sea level changes in the Mediterranean assumes a relatively stable sea-level. Flemming (1969) states "a survey of the literature on archaeological evidence for eustatic change reveals that selections from the field data have been used to prove everything from eustatic constancy to oscillations of 2.5 feet every 600 years." Flemming concluded that "there has been no net eustatic change in the last 2,000 years; that all submerged sites are due to earth movements; and that tectonic movements in the basin are predominantly downward." Flemming's data here referred to the western Mediterranean. For the Aegean, Flemming (1972) found "no consistent eustatic trends" but allows that there has been some eustatic change in the last 3,000 years. Flemming (1973) presents a synthesis of all eastern Mediterranean data.

Another school of thought suggests that world sea-level rose fairly rapidly throughout the early and middle Holocene. At about 3700 radiocarbon years B. C. the rate of sea level rise began to decelerate rapidly. Many investigators believe that sea-level has continued to rise at a rate of approximately 1 to 1.5 meters per millennium since that time. Hafemann (1960) puts the sea-level rise for the eastern Mediterranean at 2.5 to 2.8 meters over the past 2,500 years. Our data would support a rise of approximately one meter per millennium. Pirazzoli (1976) reached similar conclusions.

Fairbridge (1960) and Morner (1969) contend that world sea level rose to its present position approximately 3,500 to 6,000 years B.P. and has fluctuated above and below this level ever since. Morrison (1976) has recently done a comprehensive analysis of Holocene radiocarbon dates for the western seaboard of Europe. From this analysis he determined a number of "eustatic" transgressions and regressions. Whether or not Morrison's detailed transgressive-regressive sequences are relative and therefore nontransferable or eustatic (absolute) and therefore applicable to coastal studies in the Mediterranean remains in doubt.

The interpretation of coastal sediment sequences is affected by whether one accepts a stable, a positive or an oscillating eustatic component from the Bronze Age to the present. However, the local stratigraphic record must remain the definitive data base from which one can extrapolate local relative sea-level change and develop paleogeographic reconstructions.

Research on the coastal areas of mainland Greece by Kraft et al. (1975, 1977) and Rapp et al. (1978) is summarized in figure 3. This figure shows simplified stratigraphic-environmental sections and paleogeographic maps for four coastal areas in the Peloponnese. Each of these areas displays similar transgressive and regressive phases with, however, considerable local variation due to sedimentation rates, previous morphology and possible tectonic movements.

The transgressive phase reached these areas about 5,000 B.P. and the regressive phase about 2,000 B.P. Regression apparently resulted from sediment infill with modification by gradual eustatic change and possible tectonic movement.

The embayments at Messenia, Laconia, and the Argolid share many common elements. Each is probably a large Late Neogene graben. Pliocene marine sedimentation occurred further inland and these deposits are now exposed in outcrops elevated above present sea level by tectonic movement. Each embayment underwent deeply incised erosion during the Quarternary low sea stands.

Deposits of earlier Quarternary sediment both flank and underlie the Holocene sediment which infilled the heads of the embayments. In part the morphology of these "low sea stand" embayments may be observed surrounding the plains of Messenia, Laconia, and the Argolid. With the waning of the Riss - Würm glaciation and resultant eustatic sea-level rise, the major transgression began towards the heads of the embayments. As the landward transgression continued marine sedimentation developed over a sequence of coastal environments. Greater than 90 meters of these Holocene floodplain, swamp, coastal, and shallow marine sediments have been encountered in the subsurface of the plains at the heads of the Gulfs of Messenia and Laconia (figure 3).

Initially, early Holocene transgression was rapid. By Bronze Age times this eustatic rise began to slow down. However, erosional shoreline modification continued until Hellenistic or Roman times. Since then, sedimentary extension of the coastal plain and coastal environments has been dominant resulting in large numbers of archaeological sites being abandoned inland as the shoreline receded.

Figure 3 illustrates four variants of the vertical stratigraphic sequences and their related resultant present-day maps of the embayment heads. Each is accompanied by a paleogeographic projection showing the coastal environmental setting from Bronze Age to Roman times.

Radiocarbon dates (summarized in part in Kraft et al. 1977 and in part from on-going studies at the Bay of Navarino) provide the basis for figure 4, relative sea level rise curves for areas investigated in the Peloponnese.

Abundant organic materials were obtained from drill cores at Navarino, and elsewhere in Messenia, Laconia and the Argolid. These materials include highly varied organics from coastal marshes and swamps as well as shallow water marine mollusca. Seventeen radiocarbon dates from our studies were combined with five radiocarbon dates obtained by Wright (1972) in the Bay of Navarino area. In addition, partially submerged archaeological features and other site information proved to be important in defining relative sea level positions. The curves shown on figure 4 vary slightly and should be regarded as preliminary.

However, they show a commonality of two to four meters of relative sea-level change since Bronze Age times. This, coupled with the interpretations of Hafemann (1960) and Pirazzoli (1976) strongly suggests that a common eustatic sea level rise element is present and must be considered in all coastal paleoenvironmental changes in the Aegean area.

Figure 5 presents schematically the situation relating a rise in eustatic sea level to sediment infill and tectonic subsidence. In the hypothetical situation illustrated, if there was no tectonic subsidence, sediment infill would form new land (above sea-level) beginning about 2500 B.C. With subsidence this would not have happened.

Figure 6 illustrates the hypothetical situation for the south coast of Thera near Akrotiri. Paleogeographic reconstruction of the Thera coastal regions for Minoan times will require field studies of the type our group has used. Perhaps these studies could be supplemented by the method of Flemming (1972) who subjects coastal archaeological data to a multiple regression analysis to determine vertical earth movements.

Marinatos' (1971) observation of volcanic deposition near Akrotiri in a position lower than present sea-level needs explanation. Tilting with possible local subsidence is one hypothesis. Flemming et al. (1973) suggest a broad regional subsidence for the southwestern Aegean. Without additional field evidence local subsidence cannot be regarded as a strong hypothesis. This is particularly true in view of the evidence of relative and absolute sea-level rise in adjacent mainland regions of the Peloponnese. Eustatic rise of 2 to 4 meters since Bronze Age times is certainly a strong possibility based on evidence presented in figure 4 and supported by the work of Hafemann (1960) and Pirazzoli (1976). Should our hypothesis of a continuing eustatic sea-level rise in the eastern Mediterranean Sea or relative sea-level rise due to tectonic forces be correct, then it might well apply to paleogeographic studies in the islands of the Aegean. Marinatos' projection of the "Minoan shoreline" near Akrotiri 800 meters seaward of its present position certainly appears viable. However, the land did not necessarily subside. Careful field study of surface and subsurface stratigraphic sequences may ultimately provide the answer.

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

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