Radiocarbon Dated Plant Remains from the Akrotiri Excavation on Santorini, Greece

The investigation comprises pulses (Lens culinaris and Lathyrus cicera), figs (Ficus carica) barley (Hordeum vulgare) and almonds (Amygdalus sp.).

A series of new radiocarbon datings from the Radiocarbon Dating Laboratory of Copenhagen based on short-lived samples from the storage vessels and a twig of Tamarix with 10 year rings gives the weighted (p =1/σ²) mean of four dates: 3355 ± 32 C-14 years BP. If calibrated by means of the high precision calibration curves of Pearson and Stuiver (1986), this corresponds to 1675 BC in calendar years, with a 1σ band from 1690-1625 BC, and a 2σ band from 1740-1530 BC in calendar years.

The dates thus agree with the previously published C-14 dates for the destruction level, and rather support a 'high' age for the Minoan eruption as suggested by ice-core dating (Hammer et al. 1987).

INTRODUCTION

The following observations and investigations are a part of an ongoing palaeobotanical study on Santorini by one of us (WLF) that has been under way since 1975, and which has led to the description of some plant-bearing layers in a volcanic sequence ranging in age from 60000 BP to the time of the Minoan eruption. Some of the results were published in 1980 in the proceedings of the second Thera Congress (Friedrich 1980). The present study, however, is the result of several visits to the excavations at Akrotiri, some shorter periods of work, and contact and discussions with archaeological colleagues, namely C. Doumas and Ch. Sigalas. A much more comprehensive study of the plant findings from the Akrotiri Excavation will be given by A. Sarpaki. A treatment of Vicia faba var. minor which also occurs in the material from Akrotiri, has been omitted in the present paper because it would involve a major revision, including earlier published archaeological material from the Mediterranean.

FOOD REMAINS FROM THE STORAGE VESSELS AT AKROTIRI

• Lathyrus cicera L.:

The seeds of Lathyrus cicera from Akrotiri resemble in size those of Vicia faba; however, they deviate in their characteristic trigonal shape. Since the seeds only form a single row in a pod, it is very easy to recognize seeds which were placed at end-positions in a pod from those positioned between neighbouring seeds. L. cicera is considered to be the progenitor of the cultivated grass pea (L. sativus). The seeds are also smaller than the cultivated form (Zohary and Hopf 1988) (Fig. 1 and Table 1). L. sativus is today used as animal food and only in times of hardship used as human nutrition.

• Lentils (Lens culinaris Medik.):

During excavation of the so-called 'House of the Ladies', as named by Spyridon Marinatos for a house containing some beautiful wall-paintings of women, some jars were unearthed which contained both faba/Lathyrus and lentils (Fig. 2). The latter were very small, measuring 1.80-2.79 mm (Table 2 and Fig. 3).

Lentils and other pulses are amongst the oldest crops known from the lands bordering the Mediterranean. They are named on numerous occasions in old Egyptian documents, e.g. in the 12th dynasty (2000-1167 BC) and on an inscription from the reign of Ramses III (1198-1167 BC). Lentils are also mentioned in the Old Testament. The story of Esau and Jacob (Book of Genesis) is well known. Theophrastus (Kirchner 1874), a student of Aristotle from Eresos on the island of Lesbos, wrote in the year 320 BC about how to grow lentils. It is known from several classical sources that lentils were part of the poor man's diet, and a mixture of barley- and lentil-flour was used in the baking of bread during times of hardship. A citizen of Athens narrates in a comedy written by the Attic poet Aristophanes (c. 455-387 BC) that 'Now that he has become rich he does not like lentils any more, whereas he previously, when he was poor, ate everything he could get'. The naturalist Dioscorides (De Materia Medica, c. AD 60) alleged that too many lentils could result in poor eyesight and that they could also result in a blown-up stomach. He further suggested that they were healthier if they were first boiled in vinegar. It has been suggested that the Romans brought lentils with them during their victorious campaigns, thus introducing them to regions north of the Alps. Several varieties are cultivated in Mediterranean lands today, most of them being larger than those found at Akrotiri.

• Figs (Ficus carica L.):

An earth-filled jar containing a black material was also found (Fig. 4). In order to study this material it was impregnated with synthetic resin, thin-sectioned, ground and polished. On examination under the microscope it proved to be composed of individual fig fruits containing distinguishable fig seeds. The dark fig-mass was also removed from one of the cavities which was subsequently infilled with plaster, resulting in the characteristic shape of a fig.

Figs have been found in several other Bronze Age settlements, including those of Tiryns (Willerding 1973) in Greece where they are of the same small size as those found on Santorini. Figs are common on Santorini today.

• Almonds (Prunus amygdalus L.):

One of the jars (excavation diary 33, p. 13, 18.08.77) was filled with light-grey pumice-like earth containing irregular almond-shaped cavities (Fig. 5). Most of them were completely empty, but the remains of an almond kernel was present in one of them (Fig. 6). In reality all that remained was the thin brown husk which encloses the white kernel of the almond. This husk, the testa of the endocarp, was subjected to fluorescent light-microscopy, which revealed the characteristic oil cells that are present in an almond's testa. In addition, on infilling the cavities with plaster, a small almond shape resulted.

Two possibilities exist which may explain how the almonds came to rest in the earth: one possibility is that the earth was pressed between the almonds during or just after the eruption; alternatively, the almonds may have been stored in an earth-filled jar in order to preserve their moistness. The latter suggestion is not altogether unrealistic as both pomegranates and walnuts were stored in the same way in Roman times (Lenz 1859).

• Barley (Hordeum vulgare L.):

During inspection of the contents of a little jar from the Akrotiri excavations (Fig. 7), the archaeologists came across a light-grey powder which possessed some form of structure. On conservation of this powder in a thin synthetic resin they were able to establish that it comprised crushed ears (triplets) of the barley Hordeum vulgare L. It is known from the excavation of other Bronze Age settlements that both barley and wheat were part of the staple diet of the people of that time, which obviously also was the case for the inhabitants of Stronghyle. Single grains of barley have been observed as contaminants in the samples of Lathyrus and faba. Both barley and faba motifs are often found as decoration on the pottery from Akrotiri (Fig. 8).

WOOD FROM AKROTIRI

In general the preservation of wood in the Akrotiri excavation is very bad. Most of the wooden constructions have disintegrated. However, the moulds are left. By filling these moulds with plaster several pieces of furniture and beams of the wooden construction could be recognized.

• Description of the wood (Tamarix sp., Fig. 9-12):

Archaeological data of the wood sample.

A small piece of wood was found in the pumice outside house 3 at Akrotiri in structure Delta on 25th July 1980. (Notebook 32, p. 233). In the pumice there was left a hole of 1.30 m length and 0.18 m diameter.

The growth rings are distinct. The rays are not tangentially distended at the boundary between the growth rings. The wood is diffuse, porous and the vessels are solitary, in radial multiples of 2-4 and clusters (2-3). The vessels are probably rounded in cross section (the bad state of preservation does not make it possible to ascertain this). Their tangential diameter is 12-90 μm and their radial diameter 20-80 μm. The vessel length is 60-? μm (in only a very few instances could the lengths of the vessels be measured). The perforations are simple in horizontal end-walls. The intra-vessel pits alternate, are rounded and 2-5 μm in diameter with slit-like, often coalescent apertures. The parenchyma is vasicentric, fusiform, storied with the vessel members, and the lengths of the fusiform cells are 50-80 μm. Fibres are present but almost disintegrated; no measures could be made. The rays are 4-16 cells wide - at least up to 1mm high, heterogeneous, mostly procumbent cells, marginal cells square. Crystals (silica bodies) were not found. Although several anatomical details cannot be seen, the specimen is diagnosed as Tamarix sp. The vessel members and vasicentric parenchyma cells are shorter than normal for Tamarix species in the area (Fahn et al. 1986), a fact which gives no reason for changing the diagnosis.

The piece (No. 25-07-79 from Xeste 3) is a branching point of a twig, in a bad state of preservation. Apparently the vessels are small, in radial rows or clusters. Due to the deformation of structure in the branching point, the form and size of rays could not be determined. The piece cannot be diagnosed, but the presence of vessels and the form of rays indicate a dicotyledonous wood; it is unlikely that the piece is Tamarix.

RADIOMETRIC DATING OF THE FINDINGS

• Selection of the material for C-14 dating:

There have been several attempts to date the Minoan eruption using samples from the Akrotiri excavations. The best material for this purpose is short-lived plant material that can be directly tied to the destruction level. Charred plant remains from the storage vessels in the excavated houses have been considered the most suitable material in this respect, together with thin twigs and branches from shrubs, but even with this type of material internally consistent series of dates can usually only be obtained by excluding obvious outliers.

Part of the scatter of the dates may be due to the effect of volcanic CO₂ emanating from fissures in the ground and assimilated in the plants during photosynthesis. Such an effect has also been measured on Thera (Bruns et al. 1980), but due to the high mixing rate in the atmosphere it is restricted to narrow stretches downwind from the points of emanation, and can only be expected to influence a minor number of samples. In cases where series of dates of short-lived samples, found in different structures in Akrotiri and presumably originating from different fields around the town, exhibit a normal statistical scatter, the volcanic effect must be considered negligible (e.g. Michael 1980, Hammer et al. 1987). But the effect may explain a few definite outliers.

Another source of error may come from contamination of the short-lived plant remains by organic material of a different age, as could happen even in the case of carbonized plant material found in storage vessels. In one case it could be seen that the excavated pulses had been kept in a sack placed into a jar (Fig. 13). In another case a wooden lid was found on a storage vessel, and it may be assumed that some jars were covered in this way to prevent animals from having access to the contents. Such lids could have been made of wood with a considerable age relative to the time of destruction. In both cases plant material of a higher age may have been charred during the eruption and mixed, as a kind of organic dust, into the contents of the storage vessels, and unspecified organic dust is known to have been part of some of the dated short-lived samples.

Younger material in the form of humic matter may also be admixed into the sample. If the seeds and the plant material are well carbonized, a standard extraction of humic matter can be performed with alkaline solutions. However, part of the Akrotiri plant material has only been lightly charred without really being carbonized. This may be a question of the position of the seeds in the storage jars, and the temperature to which the jar has been heated.

If the sample is not really carbonized, the whole sample material will be dissolved when treated with alkaline solutions, with the result that an effective extraction of humic matter is excluded. Such samples therefore cannot be considered safe. Extraction with ultrasonication in acid solutions has been tried on not fully carbonized seeds (Nelson et al. 1990), but this can hardly be considered a safe extraction method for humic matter.

In order to avoid such contaminants, it was ensured that the samples dated in this study consisted only of pure, single faba/Lathyrus seeds, all fully carbonized (K-5352, K-5353, and K-3228). These samples of pulses were most probably harvested within a year from the time of destruction, and should be the most suitable material that can be obtained for radiocarbon dating of the destruction level.

Included for C-14 dating was also a part of the Tamarix twig with only ten growth rings found in the pumice outside House 3, in structure Delta 1 (K-4255), whereas a piece of wood, which had been prepared with artificial resin, was considered unsuited for precise dating.

• Dating results:

Before combustion and counting as CO₂ gas in a proportional counter, the samples were treated with hydrochloric acid, rinsed in distilled water, given a repeated (up to 4 times) mild treatment with lukewarm 0.5 % NaOH, rinsed, and treated again with hydrochloric acid.

The dating results, which have been corrected for isotopic fractionation, are given in Table 3.

The weighted (p =1σ²) mean of the four dates is 3355 ± 32 C-14 years BP. If calibrated by means of the high precision calibration curves of Pearson and Stuiver (1986), this corresponds to 1675 BC in calendar years, with a 1σ band from 1690-1625 BC, and a 2σ band from 1740-1530 BC in calendar years (Fig. 14).

The dates thus agree with the previously published C-14 dates for the destruction level, and rather support a high age for the Minoan eruption as suggested by ice-core dating (Hammer et al. 1987).

CONCLUSIONS

It has been shown that storage vessels from the Akrotiri excavation contained carbonized remains of pulses mainly: Lens culinaris and Lathyrus cicera, figs (Ficus carica), barley (Hordeum vulgare) and almonds (Amygdalus sp.).

These seeds and fruits have been measured and described. In all cases the pulses were smaller than the modern forms. In addition a piece of wood was identified as Tamarix sp. on the basis of wood anatomical investigations.

Three samples of pure pulses and a sample of a Tamarix piece have been subjected to radiocarbon dating giving a weighted mean of 1675 BC in calendar years. This supports a 'high' age for the Minoan eruption.

To our knowledge the Akrotiri excavation has only yielded a single vessel with a content of barley ears, and barley grain as contaminants in other seed samples. This could possibly indicate that the main part of the most important foodstuffs were saved before the eruption. Alternatively, the inhabitants could have been surprised by the volcanic eruption at a time of the year when the previous year's harvest was largely used up and the new harvest was yet to begin. This point may be elucidated by a further analysis of the acidity peak from the Minoan eruption. The main part of the acidity signal in the Greenland ice core appeared early in the year 1644 BC, but the increase of the acidity began already in the later part of the year 1645 BC.

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

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