Fossil plants from Weichselian interstadials, Santorini (Greece) II

A description of fossils of Tamarix, Pistacia, Olea, Phoenix and Chamaerops from the Phira palaeosol (ca. 37,000 years B.P.) is given. The fossil plants indicate that the climatic conditions at that time were nearly the same as at present.

INTRODUCTION

The upper part of the caldera walls of the ring-islands Thera and Therasia of the Santorini volcanic group are characterised by three striking pumice series: the Lower, the Middle and the Upper Pumice Series.

According to fisson-track dates (Seward et al., this volume) the Lower Pumice Series was produced about 100,000 years B.P. The age of the Middle Pumice Series is not yet known, but must be older than 37,000 years (Pichler & Friedrich 1976). The Upper Pumice Series covers most of the old ring-islands and late Minoan settlements which are still being excavated (Marinatos 1968 - 1974; Doumas 1974). It is also called the Minoan tephra or Bo (abbreviation from the German "oberer Bimsstein") and has been dated by ceramic findings (Marinatos 1968, 56) to around 1500 B.C.

A series of radiocarbon dates, however, seems to support an earlier chronology by some 100 - 150 years (Michael 1976; Betancourt & Weinstein 1976). Eight of ten short-lived samples from the excavations at Akrotiri have an average age of 1660 B.C. (Michael 1979, 794). This time the discrepancy, which is greater than a century but less than two centuries (Weinstein & Betancourt 1978), has been one of the main topics of the 2nd scientific congress on the volcano of Thera. The discussion ended with a general plea for more radiocarbon samples carefully collected and firmly tied to the archaeological context.

The geology of the Santorini volcanic group has been studied in recent years especially by Pichler and his co-workers Kussmaul, Günther, Schneider, and many others (for references see Pichler et al. 1972 and Pichler & Friedrich 1976). In 1976 Pichler & Kussmaul presented a geological map of Santorini (in this volume). In connection with this mapping I joined this project in order to work on the fossil plants from this area.

The first report of fossil plants from Santorini was given by Lacroix in 1896. He collected plant material from quarries near the town of Fira on Thera together with the French volcanologist Fouqué. Knowledge about the existence of fossil plants from Santorini must be much older than indicated by this publication, because Lacroix and Fouqué heard about the plant fossils from the monks of the island. Plant fossils from this locality in the neighbourhood of Fira must have been known at least since the time when the pumice and the underlying ashes were used for building purposes. The Lacroix locality was at Katophira, 15 meters below the end of the lava-flow upon which the town of Fira is built. This is near the site where later collections were made by Schuster (1936) and the present author.

The plant fossils described were all collected from the pumice quarries south of the town of Fira on Thera island, Santorini. The material described by Schuster is kept in the Museum of Natural History in Berlin, GDR. Further material is to be found in the Naturmuseum Senckenberg, Frankfurt/Main, the "exposée" at Fira (Peter Delendas), and the collection of the Roman Catholic Church (Nicolas Kokkalakis, Santorini). It is intended to place the fossil plants described in the following at the disposal of the paleontological Museum at Athens.

THE VEGATATION OF SANTORINI BEFORE THE MINOAN ERUPTION

Plant remains occur in several strata between the Middle and the Upper Pumice Series on Thera and Therasia. They are found in palaeosols which were formed during periods of long volcanic quiescence when the vegetation could become re-established in the areas which were devastated during periods of volcanic activity.

The traces of former vegetation are found as imprints of leaves, occasionally also of fruits and seeds, tree moulds and very rarely also charcoal. The latter allows absolute dating after the radiocarbon method.

Based on radiocarbon data from charcoal of trees and shrubs, three different palaeosols of Weichselian age from Thera and Therasia were dated and named: The Fira palaeosol (ca. 37,000 yr. B.P.), the Akrotiri/Millo palaeosol (ca. 18,000 yr. B.P.), and the Therasia palaeosol (ca. 13,000 yr. B.P.) (Friedrich et al. 1977) (Fig. 1, 17).

FOSSIL PLANTS FROM THE FIRA PALAEOSOL

The Fira palaeosol has been dated to about 37,000 yr. B.P. on the basis of three radiocarbon dates from a carbonized trunk found in layer 14 in the quarries south of Fira (Fig. 1).

It can be time-correlated with the Kalabaki II and I interstadials in Macedonia and the Hengelo complex in NW-Europe (Friedrich et al. 1977).

Dicotyledones

- Family Tamaricaceae

Genus Tamarix

Tamarix sp. Fig. 2.

Only a small branched twig with scale-like leaves of Tamarix sp. has been found in layer 16.

- Family Anacardiaceae

Genus Pistacia L.

Pistacia lentiscus L. Fig. 3 - 6

1896: Pistacia lentiscus L. - Lacroix p. 657, nomen nudum.

1936: Pistacia lentiscus L. - Schuster p. 79, pl. 14, fig. 4 - 5.

1978: Pistacia lentiscus L. - Friedrich et al. p. 128, fig 13, 14.

Five leaves and numerous leaflets were collected from the pumice quarries south of the town of Fira. They are from layers 7, 13 and 16.

The leaves are compound, 5 - 10 cm long, paripinnate. The rachis is winged. The leaflets are oblong lanceolate or elliptical, 2 - 4 (rarely 5 - 7) - paired, leathery, 10.4 - 26.4 mm long, 5.3 - 12.6 mm broad.

The leaflets have entire margins. The venation is pinnate, mixed craspedrodromous. Eleven to thirteen secondaries branch from a stout, markedly bended midvein at an angle of ca. 45^o, branching close to the margin and the branches form loops with the adjacent branches. The stomata of paracytic type are confined to the lower epidermis. They are randomly oriented, fig. 4.

The size and morphology of the leaves, the paripinnate leaflets (figs. 3, 4, also visible in Schuster's pl. 14, fig. 5), and the winged rachis (fig. 3) are characteristic of Pistacia lentiscus L. Furthermore the SEM-micrographs of the cuticles of the fossil leaflets correspond quite well with those of the extant species. According to Zohary (1952) Pistacia lentiscus is very common all over the Mediterranean countires with the exception of Egypt and Sinai. It is a typical representative of the maquis and garriques and a leading species of the Oleetum-Lentiscetum in the Western and of the Ceratonieto-Lentiscetum in the Eastern Mediterranean. In the East Mediterranean it is limited to lower altitudes.

Pistacia lentiscus L. is listed among the plants growing on Thera today (A. Hansen 1971, 128), and I observed it also on Palea Kameni too.

Pistacia cf. terebinthus L. Fig. 3, A.

? 1936: Pistacia lentiscus L. var. latifolia Coss. - Schuster p. 79, pl 14, fig. 6.

One imprint of a fragmentary leaf is consisting of the apical end of a compound, imparipinnate leaf with three leaflets, probably from layer 16 from the Fira quarries, was kindly lent to me by Nicolas Kokkalakis of the Roman Catholic Church at Fira.

The two leaflets are 4.0 - 4.2 mm long and 14 mm broad, narrow elliptical-lanceolate slightly a-symmetrical, with a stout, bended midvein, entire margin and pinnate venation.

The speciemen figured by Schuster (1936, pl. 14, fig. 6) and nammed P. lentiscus L. var. latifolia Coss. consists of two isolated, ca. 5 cm long leaflets, differing in size from the typical P. lentiscus leaflets (10.4 - 26.4 mm). They might belong to the same species as the three leaflets described in the present paper. Concerning the variety "latifolia Coss" this variety of P. lentiscus is not accepted by Zohary (1950, p. 201) the monographer of the genus Pistacia. "After examination of about 600 specimens of this species in various herbaria from all over the Mediterranean region I was led to the conclusion that the Mediterranean material should not be subdivided into varieties". Schuster's leaflets could well belong to P. terebinthus L. which has imparipinnate leaves with larger leaflets than P. lentiscus. The fragmentary specimen of the present paper (fig. 3 A) shows the apical end of a compound imparipinnate leaf including the terminal leaflet which is the largest. The size and shape of the leaflets resemble those of the extant P. terebinthus L. which has imparipinnate leaves. Since the material available is so poor, only a provisional determination can be given. More and better material is needed.

The verisimilitude of the fact that the leaf described above is a P. terebinthus is supported by the fact that both species (P. lentiscus and P. terebinthus) often grow together and that both have similar range of distribution in the Mediterranean area.

P. terebinthus is growing in the West Mediterranean countries from Portugal to Turkey and from Morocco to Cyrenaica. It penetrates deeply into the S. Alps. According to Zohary (1950, 200) P. terebinthus L. has a wider ecological range than P. lentiscus L. and ascends the mountain up to 1200 m and penetrates more deeply into the adjascent territories.

P. terebinthus is also observed on Thera today (Hansen 1971, 128).

- Family Oleaceae

Genus Olea L.

Olea europaea L. Fig. 6 - 10.

1896: Olea europaea L. - Lacroix p. 657, nomen nudum.

1936: Olea europaea L. - Schuster p. 79, pl. 14, fig. 6 - 7.

1978: Olea europaea L. - Friedrich et al., p. 129, fig. 16.

More than 100 isolated leaves of Olea were collected during the fieldwork in 1975 - 1978.

The leaves are narrowly elliptical to linear-lanceolate with an acute apex and an obtuse base. They are 2.54 to 6.64 cm long and 0.84 to 1.46 cm broad. The petiole is about 0.5 long and 0.1 cm thick. The leaves are macroscopically identical with modern leaves of Olea europaea L. Microscopically, especially under the SEM, even numerous peltate, umbrella-like hairs with a diameter of ca. 0.150 mm are visible (fig. 9 A) on both sides of the leaves, on the lower (abaxial) side, however, they are much more numerous than on the upper side. These trichromes are very characteristics of Olea. The difference in number of the trichromes on the two leaf sides of the modern Olea-leaves causes a whitish grey colour on the lower and a subdued green on the upper side. The phenomenon of the different colour of the Olea-leaves has already been mentioned by Theophrastus and other early naturalists (Pliny, Aristotle). The white peltate hairs reflect the sunlight and reduce the transpiration of the leaves.

The olive tree is the most characteristic tree of the Mediterranean. The range of the wild and cultivated trees is identical with the limit of the Mediterranean Flora Province (fig. 10) thus providing a good basis for climatic considerations and conclusions. During summer the tree requires warm, well-drained soil and a long, warm dry period for the ripening of the fruits, yet humidity in winter.

According to Turill (1951, 449) "The place, time and immediate ancestry of the cultivated olive are unknown. It is frequently referred to in the Old Testament and in ancient Greek literature. The legends of Greek mythology point to its origin outside Greece and its introduction as a plant already in full cultivation. In ancient Egypt the Olive was certainly known, even in predynastic times, and recent archaeological evidence suggests that it was of greater importance than was formerly thought. Unfortunately, there is no known reliable evidence of Olea as fossils, and the few cytological studies published throw no light on the origin of the cultivated olive. After a careful consideration of various possibilities, it is concluded that Olea europaea may have arisen from O. chrysophylla in northern tropical Africa and that it was introduced into the countries of the Mediterranean Basin via Egypt and then Crete or Palestine, Syria and Asia Minor."

The fossil olive-leaves described in the present paper are the first macrofossils of Olea from an interstadial in Greece. Olea-pollen has been reported from several interstadials from Macedonia by Wijmstra (1969). They are also known from other localities in the Mediterranean area thus indicating that Olea is an original element of the Mediteranean flora.

There are only few olive trees growing on Thera and Therasia today.

Larvae of the olive whitefly Aleurolobus (Aleurodes) olivinus Silvestri

(Fig. 1 B, 11, 12)

The most interesting finding, however, are 4 imprints of larvae of Aleurolobus (Aleurodes) olivinus Silvestri, the olive whitefly.

They were observed on the upper surface of a fossil olive-leaf from Santorini. The arrangement of the larvae on the leaf surface, their shape and size (0.625 mm in the longest diameter), and even the imprint of the parallel striated rim of the larvae, visible in one specimen (fig. 1 B), resemble the extant forms (fig. 12) so closely that one can be almost certain that they belong to the same species. The larvae of the olive whitefly are today very common on olive leaves (Silvestri 1911). Thus the plant-animal (Olea-Aleurolobus) relationship of the 37,000 years old horizon shows that the co-evolutionary relations have not changed since that time. This is an example of the fact that behaviour is remarkably conservative through time, a fact which was demonstrated by Boucot (in press) in similar examples.

- Family Palmae

Genus Phoenix L.

Phoenix cf. theophrasti Greuter. Fig. 13 A B.

? 1896: Phoenix dactylifera L. - Lacroix, p. 657, nomen nudum.

During the field-work in 1975 - 1978 only fragments of the pinnate Phoenix-like foliage, casts of the spines, and a single impression of a Phoenix fruit were found (fig. 13 B). The impression is 12 mm long and 5.4 mm wide. On the inner side of the cast a white, reticulate impression of the strands is visible. The pointed shape of the apex might be due to the impression of the perianth connected with the fruit.

The fossil findings of Phoenix from Santorini are still so poor that a proper identification at the species level is not yet possible. Much more material is needed.

In Lacroix's (1.c.) list of fossils from Santorini the date-palm keeps the first place as the most frequent.

Unfortunately neither Lacroix nor Schuster (1936) published any illustration of date-palms. Fossils of the date-palm, however, especially of the foliage, are easily recognisable. Thus there is good reliance on this early observation.

The small fruit, however, (fig. 13 B) resembles in its shape, size, and the arrangement of the strands the small fruits of Phoenix theophrasti Greuter much more than those of Phoenix dactylifera L. For reasons of comparison I collected disseminated fruits and seeds from the palms (P. theophrasti) at Vai/Crete.

Another argument for my comparison between the fossils from Santorini and the living Phoenix theophrasti Greuter from Crete is the similarity of the plant communities at Vai and Preveli, Crete, with the fossils from Santorini. At both places in Crete I observed the occurrence of Phoenix theophrasti G. alongside Pistacia lentiscus L. Into the bargain at Preveli even Tamarix and Pistacia lentiscus were growing together with this palm. All these genera are found as fossils on Santorini.

- Genus Chamaerops L.

Chamaerops humilis L. Fig. 14 - 16.

1896: Chamaerops humilis L. - Lacroix, p. 657, nomen nudum.

1936: Chamaerops humilis L. - Schuster, p. 79, fig. 1 - 2.

About twenty imprints of the fanshaped foliage and several moulds of the stems of Chamaerops humilis were observed during the field-work, especially in layers no. 13 and 7 in the quarries of the town of Fira.

According to the observations in the field, the palms were small, reaching only about 30 cm to 1 m in height. Remnants of the leaf cuticle are only found in rare cases. They show the typical parallel nervation of palm leaves.

In layer no. 13, a light grey hardened ash, several upright standing palms with fanshaped leaves were found with only 2 - 3 m distance between each palm. The stems had disintegrated, and the moulds had been filled again with ash from later ashfalls. The palms were confined to an area of a few hunded m²; obviously, in this very place they had grown in a small grove on the volcano slope when they were embedded by the ashfall (fig. 16). The palms observed in layer no. 7, a reworked ash, show signs of transportation. They were lying on the ground when they were embedded in the sediment (fig. 15).

The dwarf-palm, the only European palm, is very common in the Mediterranean area today (fig. 14). It is a typical element of the palmito-formation in Spain and characteristic of the maqui - and garriques - associations. Commonly it is also associated with Pistacia lentiscus L. (Rikli 1948, 540). Today Chamaerops humilis L. is confined to the western part of the Mediterranean, but there is good evidence that it had formerly also grown in the eastern part. Theophrastus (350 - 286 B.C.) mentioned its occurrence in Crete: "For the leaf is broad and flexible, and so they weave their baskets and mats out of it. It is common in Crete and still more so in Sicily" (Hort 1948, 143). There is also a note by Martius (1884; cf. Gessner 1933) about the occurrence of the dwarf-palm in Greece from Korfu and Zacynthos, which, however, was never proved by later investigators. Fliche (1898) reported fossil wood of Chamaerops also from Mytilene.

The reduction of its range in the eastern part might well be due to human activity, since several early historians mention its use for many purposes.

Ancestors of the dwarf-palm are known from several places in the Tertiary of Europe (Béguinot 1938).

Fossil leaves of Chamaerops humilis L. are also reported from a volcanic layer from Lipari, which has an age of about 35,000 years (Friedrich et al. 1977).

AKROTIRI / MILLO PALAEOSOL (ca. 18,000 yr. B.P.)

Concerning the vegetation of Akrotiri/Millo palaeosol, which on Thera and Therasia is overlain by ignimbrite sheets, moulds of smaller trees containing charchoal were observed. On the Akrotiri peninsula ten tree moulds with diameters of 5 - 7 cm with charchoal were found in the lowermost part of the ignimbrite.

During the 1978 fieldwork on Therasia in the Millo bay imprints of twigs and charchoal fragments were observed in the contact zone between the ignimbrite and the underlying palaeosol. The Akrotiri/Millo palaeosol can be time-correlated to the Philippi Interstadial in Macedonia and the Lascaux interstadial in W. Europe (Friedrich et al. 1977).

THERASIA PALAEOSOL (ca. 13,000 yr. B.P.)

From the Therasia palaeosol only charchoal remains are reported (Günther & Pichler 1973).

MINOAN PALAEOSOL (ca. 3,000 (or 3,500?) yr. B.P.)

The uppermost part of the Therasia palaeosol corresponds to late Minoan time. The palaeosol in which late Minoan houses occur is spread all over Thera and Therasia. This so-called Minoan palaeosol is overlain by the Upper Pumice Series which was produced around 3,300 B.P. This palaeosol represents the old landsurface of the once circular island, also called Stronghyle. It is not a continuous soil but consists of volcanic products of various kinds such as lavas, ashes and ignimbrites, deeply weathered and locally intersected by erosional channels with their younger fillings.

Direct evidence of the vegetation of the Minoan landsurface is very rare. The organic matter has been decomposed and only whitish permineralised roots are seen in some places in the quarries at Fira and Akrotiri. Carbonized trees are extremely rarely observed in the pumice; however, a single piece has been reported from the quarries near Fira (Marinatos 1968, 56).

Indirect evidence of the former vegetation, however, is given in part by the beautiful frescoes and the painted pottery and the content of the storage vessels from the excavations (Höckmann; Iliakis; Douskos 1980). The fresco of the lilies is thought to represent the nearest approach to a representation of what part of Thera looked like before the eruption (Rackham 1978). Also the wall-painting showing young women gathering crocuses for the colouring essence contained in the stigmas (Douskos 1980) might contribute to the knowledge of the former flora. Many of the plants depicted by the Minoan artists are so real that they might partly reflect the flora of Santorini of that time. A study of the vegetation and resources of Minoan Santorini based on the archaeological findings has been published by Rackham (1978). Turner (1978) using her palynological studies of the lake Kopais/Greece as a guide concludes that the islands were originally covered with forests, at first open oak woods with juniper and Pistacia and then with more dense oak woods containing few other tree species.

CONCLUSIONS

The radiocarbon-dated palaeosols from Santorini represent periods when trees or shrubs grew on the volcanic island (Fig. 17). Tamarix, Pistacia, the olive tree, date-palm and the dwarf-palm grew on Santorini about 37,000 years ago, indicating that the climatic conditions in the Mediterranean were as they are today since all these plants still exist among the Mediterranean flora (Friedrich et al. 1977).

From a detailed study of a pyroclastic section at the pumice quarries near the town of Fira, which spans the radiocarbon-dated time interval from about 37,000 B.P. to 3,300 years B.P., traces of trees were only found in four horizons (excluding the Minoan palaeosol). Moulds of roots were observed in 11 horizons which, based on their general appearance, might be from herbs.

Thus the suggestion of a vegetational change from trees to minor plants, which has been repeated several times, seems likely. A probable explanation of such a change is that the vegetation was destroyed several times by volcanic activity, and that the observed plant-bearing horizons represent only the beginning stages of the reviving vegetation. But even more probably this change is due to variations of the climate. The period indicated by the radiocarbon dates ranges from the middle part of the Würm or Weichselian glaciations to the Holocene. Several climatic changes occurred in this period in the Mediterranean as indicated by pollen diagrams from Syria (Niklewski 1970), Iraq (Zeist 1969), Italy (Frank 1969), and Spain (Florschütz 1971) as already stated by Wijmstra (1972).

The palynological study of the first 30 m of a 120 m deep peat section in Macedonia, Greece, by Wijmstra (1969) showed that in the stadial phases within the Weichselian an open Artemisia Chenopodiaceae steppe existed in that area; during interstadials the vegetation changed into shrub steppe phases in which pine and oak were present. Climatic changes were also evidenced both through micropalaeontological studies of deep-sea cores from the Mediterranean (Blanc-Vernet 1972) and palynological work (Rossignol-Strick 1972). Many other places in the northern hemisphere reveal climatic fluctuations in the Weichselian. It is therefore very probable that the Fira, Akrotiri/Millo, and Therasia palaeosols from Santorini represent interstadials with trees in the Weichselian, while during the stadials only herbs (steppe?) grew on Santorini.

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

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