Observations on the Historical Ecology of Santorini

In the Bronze Age it is likely that the island was distinctly less arid than it is now, but probably not fertile enough to sustain the size and quality of the settlements. By Classical times the island would seem to have been much as it is today, and seems even less appropriate as the base for a quite important city.

INTRODUCTION

In the eleven years since the last Thera Congress, developments in three directions have affected our estimates of the ancient vegetation and land use of Santorini. Much more is now known about how the modern vegetation of the drier parts of Greece in general functions and is limited; in particular, the many road-cuts made in recent years have exposed the root systems which could not previously be studied. Although direct information on the ancient vegetation of the island remains meagre, new pollen diagrams from the dry side of Greece have added greatly to our general knowledge of ancient vegetation and climate. And estimates of the magnitude of the great eruption, and therefore of the shape and height of the island before the eruption, have been revised.

MODERN VEGETATION OF SANTORINI

Even in a highly-cultivated island, scraps of wild vegetation exist which give clues as to what the potential natural vegetation might be. Bushes and small trees grow on terrace walls, and cliffs afford a refuge where trees and plants of palatable species can grow out of reach of goats and sheep. For example, holm-oak (Quercus ilex), a relatively rare tree, grows on the great south-eastern cliff of Amorgos and on many cliffs in Crete.

As I remarked at the previous Congress, the flora of Santorini - that is, the list of species - is surprisingly rich for an island of this size, to a degree which suggests that the 'Minoan' eruption was not big enough to sterilize the whole island (Rackham 1978). In quantity, however, the vegetation is meagre indeed.

The wild vegetation of the drier parts of Greece can be divided into: maquis (λόγγος), garigue or phrygana (φρύγανα) and steppe (λειβάδια). Maquis consists of trees or shrubs; most Greek shrubs, such as prickly-oak (Quercus coccifera) and phillyrea (Phillyrea media), are trees bitten into a shrubby form by goats and sheep. Phrygana consists of undershrubs such as species of Cistus and Salvia, which are usually grey-green, aromatic and short-lived, and cannot grow into trees. Steppe consists of grasses and herbaceous plants, including many annuals and bulbous species. A typical Greek hillside consists of a mosaic of patches of all three. The predominant factor controlling them is the water available; the less water the less maquis and the more steppe there will be. Browsing controls the height of the maquis: if browsing declines, the bitten-down shrubs turn into trees (Rackham 1982; 1983).

Santorini is a very arid island. There are probably no wild trees. Prickly-oak (and all other oaks), juniper (Juniperus phoenicea) and even the very drought-resistant wild pear (Pyrus amygdaliformis) are absent. The only maquis shrub is lentisk (Pistacia lentiscus), one of the most drought-resistant of shrubs. Lentisk has been recorded (Rechinger 1943) - proving that its near-absence can not be due to the effects of the eruption - but is rare. The only frequent cultivated tree is carob (Ceratonia siliqua), which is the most drought-resistant. Pines and cypresses have been introduced, but only just survive.

The lack of trees and shrubs is not due to browsing. Santorini is no more browsed than many other Cyclades - probably less browsed, considering its history of vineyards. The common maquis trees are absent altogether, not reduced to a very thin bitten form as, for example, often on Amorgos. Lentisk is very unpalatable and prospers under browsing. Santorini has cliffs on which even sensitive species might survive. There can be no doubt that it is one of the most arid Greek islands, comparable to Chalki off Rhodes (Rackham and Vernicos, forthcoming) or the driest, south-eastern, corner of Crete.

Cycladic architecture in general is adapted to using short lengths of timber, where a tree 5 m high is a big tree. Santorini has two distinctive types of vernacular building: the cave-house, dug into a tephra cliff; and the shell-vault, a thin, rigid shell made of small boulders strongly cemented with tephra. These are the architecture of an island with no timber at all.

     Climate, vegetation and soils:     Water is likely to limit the growth of vegetation in an island as dry as Santorini. From the plant's point of view, this depends not only on the rainfall but on the extent to which the soil and bedrock retain moisture, on the ability of roots to penetrate the rock and get at the moisture, and on the evaporativeness of the climate. Root penetration is critical, especially for maquis shrubs which often extend much more widely below ground than above. For example, the marl country of middle Crete has a higher rainfall than Santorini, and water-retaining soils; yet the marl is too compact for the roots of trees and shrubs to penetrate it except where there are geological faults or it has been broken up artificially. The natural vegetation of this area is phrygana and steppe. (I leave out of account plants that depend on springs and ground-water, such as oleander and plane).

The 30-year mean rainfall of Santorini is given as 378 mm (Marinos and Marinos 1978); if typical, this makes the island one of the drier parts of the Aegean, comparable to Athens or the south coast of Crete. However, the climate is reputed to be less evaporative than most of the Aegean, with high humidity in summer and sometimes sea mists (as were observed during this Congress). The volcanic soil - little-altered bedrock - are coarse in texture (Davidson 1978), but being made of porous grains of pumice and tephra retain water well. This explains why there is so little ground-water on the island. It also makes cave-houses possible: provided that there is 1.5 m of tephra overhead, water does not come through the ceiling. However, the subsoil is compact and impenetrable to the roots of most plants. Plant roots are confined to a thin surface layer, and there are not the casts of old roots which might show that there had once been trees with deeper roots. (I have found the ghosts of roots, forking and about 2 cm thick, in the soil buried by the great eruption of 18,000 BP in the Phira quarry). The limestone massif of Santorini, in contrast, has good root penetration into fissures but retains little moisture.

It is therefore hardly surprising that the wild vegetation of the island should be limited to phrygana and steppe. Under cultivation, the volcanic soils can be made to produce quite passable corps, but this needs careful management to increase root penetration and conserve moisture. The historic crops have been barley, vines, and formerly cotton. Barley is the least moisture-demanding of the cereals. The many mills and threshing-floors bear witness to the importance of cereal-growing in the recent past.

The olive, which grows in all but the driest parts of Crete, is not a traditional crop on Santorini; it has recently been planted, but except in sheltered north-facing localities generally dies. The vine is a surprising crop to find where olive fails, but it is one of those few woody plants (like the fig and Vitex agnus-castus) whose roots can penetrate the tephra, and it is cultivated in a special way. The vines are widely spaced, low-growing (which reduces transpiration), and cover only one-fifth to one-third of the ground. The ground is terraced and broken up, which doubtless increases the root-run. The vines fill all the available space below the ground with a root-mat some 40 cm deep; this has the effect of multiplying the effective rainfall three- to five- fold. They grow best on the less compact ignimbrite deposits, containing boulders which help roots to penetrate. All weeds, which would compete for moisture, are scrupulously removed. Any vineyard which is neglected becomes droughted. (In much the same way the arid marl country of Crete has been made a good vineyard area).

The island has recently supported a huge population, but not for farming. Nature has made Santorini suitable for a port; its advantages as a harbor have at times outweighted even the lack of hinterland.

ANCIENT VEGETATION OF SANTORINI

Our knowledge of the vegetation of Santorini before the great eruption depends partly on plant remains and other materials excavated on the island. These do not add up to a systematic record of the plant life such as would come from a pollen deposit. I have little hope that a pollen deposit will ever be found in Santorini itself. We can, however, extrapolate from pollen analyses elsewhere in southern Greece. To do this we need to know how the climate and vegetation in antiquity differed in general from those of today, and also what were the special differences relating to Santorini.

       Vegetation and climate of southern Greece in antiquity: At the previous Thera Congress, Dr. Judith Turner summarized the evidence then available from pollen cores from Lake Copais in Boeotia, and from sites in the western Peloponnese (Turner 1978). Since that time, other cores have been published from the Argolid (Sheehan 1979), from Ayia Galini in southern Crete (Bottema 1980), and from two sites in the Akrotiri on the north coast of Crete (Moody et al., forthcoming). Several others have been taken by my colleagues and myself and are being worked on. Dr. J.A. Moody and I have also reinterpreted the earlier pollen diagrams in the light of what we now know about the structure and functioning of southern Greek plant communities (Rackham 1983; Moody 1987). The following is a summary of our conclusions:

1. Southern Greek vegetation in and before the Neolithic period was not continuous forest, but a mosaic of woodland and steppe, corresponding to the present maquis and steppe. All the pollen diagrams contain pollen of plants such as asphodel which do not flower in shade. The proportion of these plants varies according to the dryness of the present climate, but they are never quite absent.
2. The trees were predominantly oaks, in which deciduous oaks outnumbered the evergreen species - today it is the other way round.
3. The garigue component was local, well-represented only in the Cretan Akrotiri diagrams.
4. There were deciduous trees of kinds that now grow in north Europe - lime (Tilia), birch (Betula), alder (Alnus) etc. Now these are found only in special situations, but then they grew in small quantities throughout southern Greece and Crete. Lime, for example, occurs at that period in nearly all the pollen diagrams, including those in Crete; it is now absent from Crete, and in southern Greece is confined to north-facing cliffs at high altitudes.
5. Pine - sometimes conjectured to be the naturally dominant tree in southern Greece - was very local.

We infer that some of the differences between the ancient and modern vegetation are due to climate and some to cultivation and pasturage. For example, deciduous oaks (which are shallow-rooted) were evidently the dominant trees of the more water-retaining soils. They were largely destroyed by cultivation, thus creating the modern illusion that the Mediterranean oaks are necessarily evergreen. Deciduous oaks are now returning in many areas as cultivation retreats. But no amount of not cultivating would induce lime to grow in the present climate of the Cretan Akrotiri or of Ayia Galini.

The pre-Neolithic climate of southern Greece was evidently less arid than it is now, though still not wet enough for continuous forest. Dr Moody proposed that it was less strongly seasonal. Owing to shortcomings in the radiocarbon dates, we do not know with certainty when the change to something like the modern climate took place; Moody considers that it was gradual, and may not have been completed by the Late Bronze Age.

       The special climate of Santorini: All Aegean travelers will know that coastal fringes tend to be more arid than the rest of the country. Maquis is often absent, and where it exists at all is of drought-resistant species: prickly-oak rarely reaches the sea, being replaced by Juniperus phoenicea. This is very obvious in Crete, and is supported by rainfall records: for example Arkhanes, 10 km inland, has half as much rain again as Herakleion, on the coast. Small islands are more arid still, even tough many of them, having no landing-place or no water supply, are not browsed as is the mainland. The effect is partly orographic, but is not entirely dependent on altitude. In the rainy season one often sees clouds missing islets and coasts and deposing their rain inland.

The higher islands all have patches of lush vegetation, showing that their mountains abstract rain from clouds. For example the peak of Naxos, 1001 m, has drought-sensitive plants such as holm-oak, valonia oak (Quercus macrolepis) and maple (Acer orientale), of which there is no trace on the peak of Santorini, 566 m. There is evidently a critical height, between these two elevations, at which an island mountain makes sufficient rain to alter the vegetation.

Table 1 compares some of the high islands and isolated mountains of the Aegean. I have arranged them roughly in order of aridity as estimated from their vegetation. For example, the Aleppo pine of the mainland (Pinus halepensis) is rather more drought-tolerant than P. brutia of Crete and the eastern islands. Holm-oak and valonia oak indicated a lesser degree of aridity than prickly-oak, with arbutus in between. Water-retaining soils are a complicating factor. So are changes with time. A century ago Methana, the volcano accross the gulf from Piraeus, impressed Frazer (1898) by its arid barrenness:

"The general character of the scenery is one of barren desolation, the whole peninsula, with the exception of a few narrow strips on the coast, being occupied by the sharp mountain-ridges which radiate from Mt. Chelona... Water is scarce, and the air dry and hot. The inhabitants, however, contrive to cultivate patches of ground, supported by terraces, high up on the mountain sides. The contrast is great between this desolate and arid mountain-mass, and the rich and well-watered plain of Troezen"

Today, much less browsed, Methana gives the impression of almost tropical luxuriance.

Table 1: Some Aegean islands and mountains, in order of increasing aridity of vegetation:

Santorini today counts as a 'low' island, not high enough to attract rain. Has this always been so? In 1978 I could reasonably have supposed that the whole of the present caldera resulted from the 'Minoan' eruption, before which Santorini had a cone perhaps 1000 m high, sufficient to make it a 'high' island like Naxos today. More recent studies have tended to reduce that height. Pichler and Friedrich (1980), though retaining the theory of a single caldera, reconstruct the pre-eruption island as only 500-600 m high. Heiken and McCoy (1984) demonstrate that the 'Minoan' eruption created only the northern part of the present caldera; the southern part was formed by eruptions before human history. They reconstruct the pre-eruption island as almost identical to the present island in its southern half, and having a volcano no more than 350 m high in its northern half (see also Aston and Hardy 1990). This estimate may have to be revised upwards slightly, following the discovery of 'Minoan' tephra in Turkey (Sullivan 1988).

In the present climate, an island needs to be at least 200 m higher than Santorini to have lush vegetation, unless its soils are very favorable. It is now unlikely that Thera was so before the 'Minoan' eruption. The highest point of the island, then as now, was probably not the volcano but the limestone peak of Profitis Ilias.

       Direct evidence: Although there is no pollen record, evidence for leaves, wood and other plant remains from excavations is fuller for Santorini than for the other Cyclades. Unfortunately, for some of the identifications, the evidence has not been fully published; in absence of botanical details, it is hard to know how much weight the proposed identifications will bear. (For examples of how identifications ought to be published, see Friedrich et al. 1990).

Many Pleistocene plant remains have been found on Thera and Therasia - impressions and charcoal of plants killed in pre-Minoan eruptions. Identifications so far reported include the palm Phoenix, presumably Ph. theophrasti which is now almost confined to Crete, and Chamaerops, another palm, now confined to the western Mediterranean. (I have now found Phoenix theophrasti in several places on Santorini; usually it is a cultivated tree, but at Phoenix in the north of the island it may be wild. This palm is often still to be found at places named after it, 'Phoenix', in the Aegean). Whether it could have persisted from before the Minoan eruption is uncertain: fossils so far found seem to be related to the pre-Minoan or earlier eruptions. If any plant could survive being buried in red-hot tephra and struggle to the surface again it would be this palm, which grows from suckers and is adapted to withstand fire. Olive (Olea) is now common in coastal maquis in Greece and Crete, although the date from Santorini (c. 45.000 BP) is by far the earliest record in the Aegean. Pistacia and Tamarix are also reported (Friedrich 1978; Velitzelos 1990). All these are trees of moderately dry Mediterranean coasts. They indicate that Santorini, at this remote period, was more tree'd, and a little less arid, than it is today. However, many of the vegetation-bearing horizons show only herbs and undershrubs.

The Bronze Age excavations on Santorini have produced a few indications. An 'oak stump' was found in the first excavation of Akrotiri (Mamet 1874), and 'branches of vine and olive' as interlacing for stone walls on Therasia. (Vine could easily have been an error for juniper, which in recent centuries has been the favorite wood for this purpose). From the modern Akrotiri excavation the charcoal of pine, tamarisk, vine and olive have been identified, and the casts of reeds (Arundo donax) (S. Marinatos 1968-75; Friedrich et al. 1990; Kuniholm 1990). Pine and reed are no longer native on the island, but could have been imported. However, much of the material was firewood, and tree branches, with an attached heel rent from the trunk, were used in floor structures; these are less likely to have been worth transporting. Tamarisk grows on many Aegean coasts, but is often an introduction.

Crop plants of which the seeds have been excavated are barley (Hordeum spontaneum and vulgare), fig (Ficus carica), almond (Prunus amygdalus), and five annual legumes: beans (Vicia faba), lentil (Lens culinaris), chickpea (Lathyrus cicera), L. clymenum, and a species of Lupinus. There is also a report of Sherardia arvensis as a weed of arable crops (Friedrich et al. 1990; Nelson et al. 1990; Sarpaki 1990). Of these, the almond is the most moisture-demanding, but also the most likely to have been imported. All the others could well have grown locally (although I have no record of lupin or Sherardia on Santorini today).

The buildings in Akrotiri use timber lavishly, though the species cannot be identified. However, they are polite, rather than vernacular, houses, and tell us nothing about the source of supply. Imported timber could easily have been a status symbol.

Pictures on walls and pots tend to be stylized; where they can be identified, they represent lilies and other garden plants, or foreign plants such as cultivated date-palms, rather than the local wild vegetation. Occasionally wild plants such as Biscutella didyma or the grass Bromus madritensis appear, but these are widespread in steppe or cultivated ground and are not distinctive. The one picture that seems to be influenced by the actual landscape of Thera is the Fresco of the Lilies, with its fantastic rocks, vividly coloured and cavernously decayed, that are such an arresting feature of the island (Rackham 1978).

The animal bones from Akrotiri, with a predominance of sheep or goat and a large minority of pig and ox, suggest a usual type of pasturage for the Bronze Age Cyclades (Gamble 1978). They make it unlikely that the whole island was cultivated. Hare bones are the only evidence for what could have been wild (though not native) mammal. Three bones of red deer suggest that this animal may have been kept in semi-domestication.

CONCLUSIONS

Santorini would originally have been less arid than today, if only because pre-Neolithic Greece in general was less arid. However, on present evidence, it is unlikely to have been high enough to attract rain.

As is well known, the large Mediterranean islands such as Crete had a peculiar native fauna of elephants, hippopotamuses and special deer, but little is known of whether the small islands had mammals. Santorini can hardly have had a native fauna, because the eruption in 18,000 BP would not have given enough time for animals to colonize. All the mammals of its Bronze Age, or since, would have to have come in boats.

I conjecture that the island, before it had human inhabitants, would have been less arid than today, and unbrowsed. It would have contrasted with the Greek mainland (which had wild herbivores and carnivores), and especially in Crete (which had herbivores only). I suggest patchy woodland of lentisk and oak, with big patches of steppe. Deciduous oaks are a possibility if they could spread their wide shallow roots in a water-retaining soil.

What did Bronze Age Thera do for a living? It is no longer possible to argue, as I suggested at the last Congress, that the pre-eruption island was specially fertile. On the contrary, it would have been a little less arid than it is now, with trees at least among farmland, and with a wider range of crops as well as pasturage. The soils would have been better developed, being 14,500 years old instead of 3,500; but in the circumstances of Santorini this would not have made much a difference to what would grow. The island would have been somewhat more rewarding than it is today - both olives and vines were grown - but even the most favorable reconstruction brings it up to merely average fertility for the Cyclades. Ordinary agriculture could not account for the special prosperity that created Akrotiri - and Akrotiri was by no means the only, and may not have been the chief, settlement on the island (Davis and Cherry 1990).

A possible, though untested, explanation is the growing of some special, high-value crop, such as the famous mastic of Chios. Alternatively, the island may then have had a specially attractive harbor, where trading and seafaring could have kept a population well beyond the agricultural capacity of a relatively dry island. An exact parallel is the more recent history of Chalki (Rackham and Vernicos, forthcoming). Metal-working is another possibility (Stos-Gale and Gale 1990), though this would require wood for fuel, which would have grown slowly in this climate and could easily have taken up the whole island.

The whole southern half of Santorini was not directly involved in the 'Minoan' eruption. Although ash-fall would have destroyed most of the vegetation, it is quite possible that some plants survived on the cliffs of Mounts Porfitis Ilias and Mesa Vouno. This may explain why the flora today is not specially impoverished.

By Classical times the island would have been much the same shape and size as it is now, and the climate and vegetation of Greece in general would not have been very different from today. How did Classical Thera make its living? The city seems too important to be sustained by the meagre area and rather precarious cultivation of the island. Thera should be regarded as one of those Classical cities, like Tarrha in Crete, which must have been supported by something other than ordinary farming - perhaps seafaring or the growing of some specialized crop.

Possibly the Minoan and Classical ages of Thera were not so very different from the island's prosperity of just over a century ago. For about thirty years the circumstances of trade made it one of the richest places in the Aegean, when it rose to grand architecture in a very distinctive style. This brief and little-known heyday has left its mark on most of the villages of Santorini today.

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

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