The Dodecanese are a group of 12 islands (and a larger number of islets) in the eastern Aegean Sea, which closely follow the Turkish coast from Patmos in the north to Karpathos and Kassos in the south. Although Greek since 1948, following the Crusades, the islands were under Ottoman rule from 1478 - 1912, before becoming Italian possessions until the 2nd World War. The legacy of these former times is to be seen in the castles built by the knights, the mosques dating from the Ottoman period and the Italian architecture to be found in such places as Rhodes Town.
Botanically they combine a Mediterranean flora with species which have migrated from the Pontic and Anatolian regions to the east. At one time a part of the tertiary Aegean land mass which stretched in a crescent from the Peloponnese through Crete to the Anatolian Plateau, the islands are in fact the summits of this ancient mountain chain, and contain many endemic species which had migrated from east and west, and were isolated when the water level in the Mediterranean basin rose.
Cyclamen repandum ssp. rhodense has long been known from the southern Dodecanese island of Rhodos (Rhodes) where it is widespread and was thought to be endemic. In April 1996, Dr David Thomas of the Biology Department, Cardiff University, whilst on a family holiday, found an unknown (to him) Cyclamen species on Kos and reported it to Chris Grey-Wilson. From the photographs supplied by Dr Thomas, Grey-Wilson identified the plants as Cyclamen repandum ssp. rhodense. It now seems likely that the reports by Hansen (1980) of Cyclamen graecum on Mt Latra refer to C. repandum ssp. rhodense.
The principal objective of the Society's expedition was to locate and survey both the populations found by Dr Thomas, and any others, and to positively identify the plants. Following this, the expedition was to extend the Society's knowledge of other Cyclamen species in the Dodecanese Islands, by surveying the rest of the island of Kos, and the island of Symi. Symi was potentially interesting due to its proximity (11 km) to the C. trochopteranthum site (97/05T) found by the Society in 1997 on the Turkish Datça Peninsula.
The team consisting of Martyn Denney (leader & dogs-body), Chris Clennett (botanist) and Brian Cresswell (photographer) arrived on Kos on Saturday 18th April and spent 51/2 days on the island before transferring to Symi. About 3 days were spent surveying C. repandum ssp. rhodense, and 21/2 days surveying C. hederifolium. The team then spent 41/2 days on Symi surveying C. persicum before transferring to Rhodos.
On Kos, the team were based in a small, inexpensive, and very friendly hotel in the port of Mastichari on the north-west coast, where the ferries leave for the neighbouring island of Kalymnos. This central location allowed easy access to both the Kefalos-Kamari-Latra area in the extreme south, and the Oros Dikaio ridge.
C. repandum ssp. rhodense was found at eight sites at the south-west end of Kos just south of Kefalos, largely between Mt. Latra and Mt. Zini, and at one site on the north side of Mt. Zini. These nine sites, all basically north facing, contained four different habitats: i) Sparse or dense Pinus brutia woodland with phrygana on a brown sandy loam (sites 98/02D - 98/05D); ii) A steep wash-out gully on the north side of Mt. Zini, just above high water mark, on a sandy loam (site 98/07D); iii) Quercus coccifera and Pistacia lentiscus phrygana with isolated groups of Pinus brutia on 'terra rossa' (site 98/08D); iv) In and around a steep north-west facing gully under mature Pinus brutia, Arbutus andrachne and Pistacia lentiscus, on a brown clay loam (site 98/09D).
The differences between these habitats are significant, but two sites stood out from the others.
Site 98/07D was on the north side of Mt. Zini in a steep gully running down to the sea, with plants growing predominantly under Quercus coccifera phrygana in a thin sandy loam. Although generally the plants were not found in the gully bottom, a small colony was growing in an isolated group of more mature Q. coccifera at the top of the beach. - probably as a result of a landslip. This gully was explored upwards until it became too steep to proceed, and at this point the plants were thinning out. From the summit of Mt. Zini the gully was not accessible, but there were no Cyclamen growing in the area above it on the silica sand which makes up the summit.. We concluded therefore that the Cyclamen were confined to the area where the sand gave way to a topsoil with a degree of loam content. It was interesting that on both sides of Mt. Zini the Cyclamen populations were clearly limited by the well defined change in the soil from a brown sandy loam to a silica sand containing black crystals - possibly a form of quartzite.
Site 98/09D was on the north-west flank of Mt. Latra. Here C. repandum ssp. rhodense was clearly at its 'happiest', growing in a gully in open shade beneath mature trees. This was reflected in the vigour and density of the plants.
At the other sites, Cyclamen repandum ssp. rhodense was growing with the tubers 15-30cm beneath the surface, generally with a thick semi-decayed humus layer, principally comprised of both new and decaying Pinus brutia needle litter.
A total of 295 plants were surveyed at 5 sites. Unfortunately, due to an extremely mild winter and a hot period in February, the season was very advanced. Although isolated plants were in flower to allow the identity of the plants to be proved, and for herbarium specimens, there were insufficient flowers to carry out surveys of petal characteristics and dimensions. Of the plants seen in flower, a significant proportion had an entirely white corolla (98059D, 98060D, 98061D, 98178D, 98181D) - a higher proportion than was seen on Rhodos according to the surveys carried out during the 1991 expedition. One of these albino plants had an upward facing inflorescence of 'stargazer' type, with quite narrow petals (98059D). A number of plants with pale pink petals (RHS 76B, 76C & 76D) and a darker pink nose were found at a couple of sites (98002D, 98003D & 98057D).
On 1st May the expedition visited populations of C. repandum ssp. rhodense on Mt. Profitas Ilias and Filerimos, on the island of Rhodos. The purpose was to compare the apparent early nature of the season on Kos with that on Rhodos in order to establish whether the plants there were at a different stage in their growth cycle and if so, whether this was due to meteorological or other factors. It was concluded that the season on Rhodos was advanced in a similar way to Kos, and that at similar altitudes the condition of the plants (lack of flowers) was also similar. The expedition only found significant numbers of rhodense in flower above 580m, and these were starting to 'go over' with many flaccid petals.
The following are initial brief results of analysing the five surveys of Cyclamen repandum ssp. rhodense on Kos, and comparing them with a similar analysis of the results of the surveys carried out on the island of Rhodos in 1991. Flower statistics are not given, as insufficient flowers were found on Kos for these to be measured.
| . | Kos, 1998 | Rhodos, 1991 |
| Altitude Range | 35-260 m | 120-500 m |
| Shade | 38-83% | 0-100% |
| Slope | level - 1 in 0.5 | 1 in 3 - 1in 1 |
| Aspect | NW-N-NE | NW-N-E |
| Soil pH | 6.4 - 7.7 | . |
| Average Leaf length | 71.44 mm | 70.48 mm |
| Average Leaf width | 60.9 mm | 60.6 mm |
| Average Leaf Length:Width ratio | 1.17 to 1 | 1.16 to 1 |
| Leaf Patterns | A | B | C | D | E | F | Plain Green |
| 1998 | 36% | 57% | 5% | . | . | . | 2% |
| 1991 | 33% | 31% | 23% | 6% | 3% | . | 4% |
| Leaf Shapes | 1&2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
| 1998 | 30% | 37% | 5% | 1% | 4% | . | . | . | 23% |
| 1991 | 93% | 6.5% | 0.5% | . | . | . | . | . | . |
Note: The same leaf pattern and shape standards have been used by all Greek expeditions between 1991 and 1998, and were published in the Journal Vol 20 No. 2 (December 1996) pages 41 & 42.
Remarkably the size of leaves found on Kos and Rhodos are very similar, however, a subjective view is that the leaves on Rhodos exhibit a far wider variation in leaf pattern, and that those patterns are far more clearly defined than on Kos. This is borne out to some extent by the leaf pattern statistics.
Cyclamen hederifolium has been known from the island of Kos for some years, and was reported from Asfendiou by Meikle (1978). Its presence was also confirmed by Dr Thomas, although without specific identification. It is almost certain that the reports by Hansen (1980) of C. graecum at Zia relate to C. hederifolium.
The team found that C. hederifolium was confined to the northern side of the Oros Dikaio ridge approximately between Pili and Assomatos, although it was not to be found throughout this area, being largely confined to Pinus brutia and Cupressus sempervirens var. horizontalis woodland. Above Pili, near Lagoudi, and on the north-western flank of Mt Dikaio above Zia there were extensive populations, but otherwise the species was found growing as small isolated groups of plants in gullies, at the base of rock outcrops, under phrygana or in hedgerows.
Grey-Wilson (1997) describes two varieties of C. hederifolium and refers to var. confusum as growing in the southern part of the species' range "...from southern Sicily, Crete, central and southern Mani peninsula (Greek Peloponnese) and reported from several Greek Islands, including Zante, south of Cephalonia." He describes it as having "... leaves bright green with a poorly defined hastate pattern, thicker and fleshier, shiny, particularly beneath ...". The C. hederifolium found on Kos broadly fits into this description, some forms particularly so, although at the south-western end of its distribution, many of the leaves appeared to be thinner and less fleshy. Designating these plants as var. confusum extends the distribution mapped by Grey-Wilson further to the East. Near Zia, some of the leaf forms encountered showed considerable contrast between well silvered hastate patterns and dark green backgrounds, with an all-over glossy sheen.
One of the less rewarding aspects of field study is the need to map those parts of the study area where Cyclamen are not found - and from a research point of view, this is equally important as mapping where they do grow. In reality, one has to limit careful searching to areas which contain at least promising habitat. It would simply waste time, for example, to search open agricultural lowlands.
The team spent time looking in all likely and accessible parts of the Mt Dikaio ridge, particularly above Tolari (Kardamena), Empros Thermi, and on the north side of Sympetros, failing to find C. hederifolium or C. graecum. The majority of the southern side of Dikaio is inaccessible, falling steeply into the sea, and in any case is too exposed to provide reasonable Cyclamen habitat.
The area around and south-west of the airport has a number of gorges which in planning appeared to be worth investigation. However, this area of the island is entirely covered with a top soil comprising almost pure sand, and has virtually no shade.
Inter-island ferry services in most island groups in Greece are excellent in high season, but in the spring, even after orthodox Easter, they can be erratic and unpredictable. It is a situation which fits entirely with the relaxed nature of Greek island life, and is either accepted (entaxi!) as an inevitable part of life's rich pattern, or becomes an incomprehensible frustration (dhen katalaveno!) - it just depends upon your point of view. The island of Symi lies between Kos and Rhodos (Rhodes), but covering this short distance involved a hydrofoil journey from Kos to Rhodos (passing within 200 metres of Symi), spending a night in Rhodos Town, and then returning to Symi on one of the Trip Boat-Ferries the following morning. Even this journey could not be planned, as hydrofoil timetables do not exist at that time of year - they just fix a time to leave some time during the previous 24 hours. As a result of this, more than a day was spent travelling a mere 56km (35 miles).
Symi is essentially a limestone rock which rises to 617m (2023ft) at its highest point. There are four main centres of population: Symi Town (comprising Yialos & Chorio), Pedi Bay, and Emporios in the north; and Panormitis with its monastery in the extreme south. The centre of the island is covered by a forest of Pinus brutia and Cupressus sempervirens var. horizontalis whilst the Kourkouma-Diapori ridge which stretches north towards the islet of Nimos, is largely a tree-less moonscape covered with limestone boulders.
During the planning phase of the expedition, Symi had great potential. C. persicum was known to grow there, and was a certainty. The Datça Peninsula with C. trochopteranthum lies just 11km to the north, and rather optimistically, if one follows the ancient Aegean landmass from the south, from Crete and Karpathos where C. creticum is found, then Symi is the next limestone island in the chain - there is an obvious extrapolation. The reality turned out to be different. Although the soil on the island is either a brown loam or terra rossa which at the time of the field study, was very moist, there was no suitable habitat for C. creticum and in any case the light levels were generally too high. Although the season was clearly too advanced for the team to find C. trochopteranthum in flower the leaves would have been visible, there were no suitable habitats for the plant, especially in the north of the island, close to the 1997 site.
C. persicum is reported from Nanou Bay by Meikle (1978), and generally from the island by Carlstrom (1987). The expedition found C. persicum all over the island and nine sites were formally recorded. There appeared to be two forms: 'standard' persicum around Symi town and the coast towards Emporios; and a diminutive form with often only a single leaf throughout the rest of the island, and particularly in the extreme south beyond Panormitis. The reason for this diminutive form caused much speculation, one hypothesis being that the plants conserved their resources, so that if their leaf was eaten by goats then they had the resources to provide a replacement! On a more serious note, the team found that the tubers were often out of proportion (excessively large) to the size of leaf at the surface (significantly small).
Whilst the leaf patterns of the diminutive form were on the whole uninteresting to non-existent, the standard form found around the northern coast had a full range of patterns with some interesting silver centred forms, but nothing outstanding. A few flowers were seen, generally the common form with white petals and a deep pink nose, but a single pink flowered form was found.
Generally, C. persicum was common throughout the areas covered by Pinus and Cupressus forest, but it was surprising to find that in very exposed open areas on the Kourkouma ridge in the north of the island, plants were growing either in full sun, or in the minimal shade of rocks.
The team had intended to visit the islet of Nimos, however, local enquiries disclosed that C. persicum is common on Nimos, but that due to the size of the goat population the plants were largely destroyed (above ground), along with all other vegetation.