Prehistoric Pumice Tephra on Aegean Islands

1. Minoan Santorini tephra has been identified as a distinct layer, 30 cm thick on the island of Kos. The distance is 190 km eastwards from Thera.
2. Prehistoric explosive activity has been demonstrated from a second source in the Aegean: pumice deposits cover prehistoric levels with obsidian artefacts and pottery on the island of Yali. Petrographic distinction between Yali and Santorini pumice is given.
3. Pumice material from six localities on the island of Anaphi has been examined. All deposits belong to Santorini activity older than the Minoan event. Therefore, pumice occurrences on Anaphi island cannot be used to estimate the strength of tsunamis possibly related to the Minoan eruption on Santorini.

INTRODUCTION

Following the First International Conference on the Volcano of Thera various efforts have been made to trace the tephra of the Minoan Santorini explosion to Crete and to establish its quantitative distribution over the Aegean. Threefold was the approach adopted:

1. Microscopical inspection of soils.
2. Extension of the deep-sea record.
3. Search for megascopially visible layers in Holocene sections.

Examples for the important progress made are the papers by Dorothy and Charles Vitaliano and by Watkins et al. (1978). The Vitalianos (1974, 1978) provided compelling evidence for tephra fallout over settlements in Eastern Crete and Milos. These results also contributed most important arguments to the discussion of a LM IA or IB age of the Minoan eruption.

Systematic deep-sea coring by the R/V Trident cruise 172 (Watkins et al. 1978) resulted in isopach maps of Minoan tephra in the sector of the Mediterranean Sea southeast of the volcano. These maps indicate that tephra actually fell over Eastern Crete but that the thickness probably did not exceed a few centimeters. Similar results are presented by McCoy, this volume.

Still wanted for our problems is a distinct and measurable tephra layer in a well defined LM I stratigraphical position. Until having this ideal, we have to accumulate single pieces of direct and indirect evidence.

During geological field trips over many of the Aegean Islands and Crete since the last congress, the author was on constant outlook for megascopically visible tephra layers. Only a few findings can be reported here.

TEPHRA LAYERS ON THE ISLAND OF KOS

In three places on the island of Kos in the eastern Aegean fine-grained tephra layers with the petrographical characteristics of Santorini pumice have been found. The key locality is in the coastal section of easternmost Kos, 500 meters north of Cape Fokas (Fig. 1).

The tephra is intercalated as a distinct layer in an alluvional fan of Holocene age. The minimum thickness is 30 cm. (In one place the thickness reaches even 60 cm, but may result from secondary accumulation).

Coarsest pumice grains are 1.5 mm in ø. Agranulometric Median Value(MD) of 0.06 mm is defined by the following grain-size analysis of a sample from 5 cm above the bottom of the layer:

The correlation of this tephra layer is possible on the basis of its refractive index (n = 1.508 ± 0.002) and its mineralogy with a hypersthene-plagioclase dominated phenocryst assemblage. These parameters are identical to the Minoan tephra of Santorini and no other Holocene tephra in the Aegean is known with the same characteristics. Watkins et al. (1978) mention an identical refractive index for the older Y-4 tephra in deep-sea sediments. However, a chemical distinction is possible and shown in Fig. 3. Microprobe data for tephra shards from Kos are compared in Table 1 with the original Minoan material from Thera. Both samples have been measured in the same preparation in order to have identical analytical conditions.

Table 1: Microprobe analysis of single glass shards from Minoan Santorini tephras. Averages are given for 3-8 grains. Variation is small. All measurements on one preparation. Analysis with TPD-microprobe and ORTEC energy-dispersive detector. Avoidence of Na-loss monitored. All analyses by courtesy of N.G. Ware, Canberra.

* cf. C. Doumas, in prep.

Another pumice deposit with similar refractive index (n = 1.506) has been found on the Island of Yali (Fig. 4). This material is however distinguished by its augite-olivine dominated phenocryst assemblage.

The tephra locality on Kos is 190 km eastnortheast of Thera (Fig. 2). This new point on land is in good agreement with the results of Watkins et al., who found at deep-sea core point 172 - 25 a thickness of 26 cm. This core point is about the same distance to Santorini. The possibility of a more easterly dispersion axis of the Minoan tephra than indicated by the deep-sea sampling was not ruled out by Watkins and co-workers and is now fully confirmed.

We see now that serious damage by tephra fall must have affected the whole western Anatolian coastal region. Hence, in the context of this conference the possible influences on people living contemporaneously in that area should also be discussed. Moreover, the Eastern Aegean and the Western Anatolian coastal area seem most promising for getting the Minoan tephra in clear stratigraphical position if cultural levels related to the LM I period are excavated.

PREHISTORIC ERUPTIONS ON lliE ISLET OF YALl

The small island of Yali is built up of obsidian domes in its northeastern part and of a > 160 m thick pumice deposit in its southwestern half. In a review of the multiple sources of pumice tephra in the Aegean (Keller 1971) a geologically rather young age of the Yali volcanism was deduced. This is now fully confirmed by a fission-track age for the Yali obsidian of 24.000 years B.P. (Wagner et al. 1976).

The tephra findings on the neighbouring island of Kos led us to revisit the island of Yali, in order to exclude this island reliably as a possible source volcano. Two visits in 1972 and 1973 showed on top of the main pumice deposit the superposition of 3 - 4 pumice beds, separated by paleosoils.

Two of the soils contained abundant neolithic obsidian artefacts* and a primitive pottery. A sketch profile of the sequence, together with petrographical characteristics, is given in Fig. 4.

*     The obsidian artefacts are well-shaped blades of the whitespotted typical Yali obsidian. One obsidian blade is typical Milos type and this provenance was checked by trace element analysis, extending the distribution of Milos obsidian even to the Yali source.

It is deduced from this situation (Fig. 4) that at least two pumice eruptions occured on Yali during neolithic and postneolithic times. The Minoan eruption of Santorini is therefore not the only prehistoric explosive event in the Aegean. However, already by using refractive indexes a petrographical distinction is possible. Also the mineral content is different. For further characterization, the chemical bulk composition is given in Table 2. The prehistoric deposits Yali 3 and 4 and the main pumice unit Yali 1 are chemically almost identical. Yali 2 is more basic, in accordance with its higher refractive index. A second, more silicic glass phase (Table 2, Nr. 2) was found as an admixture in Yali 2. This phase is similar chemically to the Nisyros caldera-phase pumice (Table 2, Nr. 5). For geological position see Keller 1971 p. 160.

Table 2: Chemical bulk compositions of prehistoric pumices on Yali (Yali 3 & 4), with comparisons to similar pumice occurrences of Yali and Nisyros.

Yali 1 - 4, Nisyros pumice from northern slope, for significance of "caldera phase" see Keller 1971, p. 160.

Analyses 5 and 6 represent respectively bulk composition and glass composition of the same pumice, as an example of a strong chemical difference, which is of great tephrochronological importance. From the petrological point of view the change of K₂O/Na₂O from < 1 (0.85) to > 1 (1.54) by this fractionation is of great interest. Analyses 1-5 by XRF, no. 6 by energy-dispersive microprobe analysis as in tab.1

* by difference

PUMICE DEPOSITS ON THE ISLAND OF ANAPHI

Anaphi, the island 25 km east of Santorini has received much attention since Marinos & Melidonis (1971) described several pumice deposits at altitudes up to 250 m above sea-level. The deposits had been interpreted as tsunami depositions serving in various calculations as a measure of the strength of Minoan tsunami waves (see also Yokoyama 1978).

Keller & Ninkovich (1972) examined samples from the topographically highest deposits of Vounia and Georganou (both 250 m above sea level). These deposits could be correlated petrographically with the Upper Quaternary "Unterer Bimshorizont" BU of Santorini. The correlation is based on petrography (inlcuding a refractive index of 1.514 compared with 1.508 - 1.509 of the Minoan pumice), on chemical composition and on trace elements.

Furthermore, radiocarbon dating of the CaCO₃ - cementation of the pumice deposit showed that cementation occurred already at least 18,000 years ago.

Several more pumice deposits from the localities Metochi, Chalepa, Sklavogeorgio and Brassa-Theologou have been sampled and analysed. All show the higher refractive index of 1.514, which is distinctive for older, pre-Minoan Santorini pumice units.

As a result, the occurrences of Santorini pumice on Anaphi are of no value to estimate the strength of the Minoan explosion nor of possibly related tsunamis.

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

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