Cyclamen coum

The Society's Expeditions, 1987-1990

The Society has sent four expeditions into the field to study Cyclamen coum: in 1987 and 1988 to northern Turkey, in 1990 to Israel, and again in 1990 to southern Turkey with a brief visit to extreme north-west Turkey. This report brings together the field observations on Cyclamen coum made during the course of the four expeditions, and subsequent observations on plants that were collected on the expeditions and grown in a greenhouse.

The dates of the expeditions, were as follows:

Objectives

The general aim was to carry out field studies on C. coum over a wide geographical area, gathering both information and material as a means of answering some of the many questions surrounding this very variable species. Variations are known to occur in three characters of taxonomic importance - leaf shape, flower size, petal shape - and in other characters of importance to the grower, particularly leaf markings and flower colour. (Although in strict botanical terms the "petals" of cyclamen are corolla lobes, being united at the base, it has been decided to retain the word "petal" in this report, as "petal" is more readily understood.) It was hoped that the study would show whether variation in any of these characters is related to geographical, environmental or genetic factors. Meikle (1978) recognised certain systematic differences between east and west in respect of leaf and floral characters. By conducting the study over the widest possible west to east span within Turkey, with suitable intermediate points, it was hoped to shed further light on this aspect. Ideally, it would be desirable to extend the studies further eastward into Georgia, Azerbaijan and Iran, but up till now these areas have been inaccessible. The literature makes no mention of any systematic variation in a north to south direction, and the main purpose of the two expeditions in 1990 was to see whether such variation could be detected. In addition to studies conducted in the field it was the intention to bring back live material for cultivation under standard conditions. This material would serve two purposes: observing its behaviour over at least two seasons in cultivation would throw further light on the nature of the variation in the species; and it would serve as a source of seed for distribution to members of the Society. With the second of these aims in mind it was planned that a good proportion of the collected plants would be chosen on the basis of horticultural merit.

Areas visited

The spread of sites across northern Turkey, spanned a distance from west to east of some 720 miles (1155 km). This northern belt was distant from the two rather isolated areas in southern Turkey and Israel respectively, giving a north to south span of 620 miles (990 km).

Recording of information

At the majority of sites where cyclamen were found in sufficient quantity a sample of plants, intended to be representative of the population, was selected and the following details of each plant were entered on a proforma: length and maximum width of a typical leaf; shape of the same leaf on a 5-point scale by reference to standards; hastate pattern of the leaf on a 6-point scale by reference to standards (from 1 = plain green to 6 = overall silver); length and maximum width of a petal; shape of a petal on a 5-point scale by reference to standards; flower colour by comparison with the RHS Colour Chart; and the presence of any unusual features such as scented flowers, fimbriated (fringed) petals, bicoloured flowers, scalloped leaf edge. The environmental details of each site were also recorded: latitude, longitude and altitude, aspect, steepness of slope, shade density, type of tree cover, accompanying plant species and soil type.

The standards used for judging leaf shape and petal shape are shown in Figure 1. Elongated, pointed leaves that would lie beyond the limits of the scale shown are known to occur in parts of the USSR, and further east in Iran. When any such leaves are encountered in the field the scale will be extended beyond 5 as far as necessary.

Figure 1.- Leaf Shapes (above) and Petal Shapes (below).

Herbarium specimens and live plants

At most sites samples of detached leaves and flowers were labelled and pressed, for subsequent appraisal by taxonomists and as a permanent record of the characteristics of cyclamen at the different sites. Some of these herbarium specimens were donated to Gazi and Istanbul Universities in Turkey, some were sent to Kew, and some were retained by the Society.

Where permission from the relevant authorities was forthcoming under the terms of the CITES convention, live cyclamen plants were collected and brought back to England. They were cultivated for an initial period of at least two years in a greenhouse at Tile Barn Nursery in Kent, and their behaviour in cultivation was recorded at intervals over this period. A photographic record of the plants was also made. Subsequently, the majority of the collected plants were moved to other locations where greenhouse culture was continued by members of the Society. These plants form part of the National Collection of cyclamen, and seed from them is sent every year to the Society's seed exchange.

At a very few locations in Turkey a small amount of viable seed was collected. Surviving seedlings are being treated in the same way as collected plants.

Flowering season

The flowering season of C. coum is dependent on latitude, habitat (woodland populations coming into flower later than those in the open) and altitude. There are also variations from one year to another influenced by weather conditions. It appears that 1987 was a late season, and the expedition was fortunate to find so many plants still flowering in late March and early April, even at low altitude. In 1988 on the other hand mid-March was too late for some coastal sites in extreme north-east Turkey, and these would undoubtedly have been flowering in February. In fact the end of February was not too early for the inland sites in European Turkey visited in 1990, even though one of these was at the comparatively high altitude of 850 m (2,800 ft). Further south the season is somewhat earlier, as might be expected. Mid-February in Israel saw cyclamen in flower at 1,200 m (3,900 ft), and in southern Turkey some populations in the open had finished flowering by early March.

Variability

All accounts of C. coum stress that it is a very variable species. What the accounts do not make clear is that the variation is apparent not only between widely separated parts of its distribution but even at individual locations. The expeditions in 1987 and 1988 found an astonishing degree of variability at many of the sites where systematic recording took place, with round leaves and ovate leaves, small flowers and large flowers all mixed together in a single population. The variation observed in 1990 was not so marked, either in Israel or in Turkey, the populations in extreme north-west Turkey being comparatively uniform. The Society's decision to examine a generous-sized sample, if possible 50 plants, at each site is clearly justified, particularly for populations showing a high degree of variability. In fact even a sample of 50 plants cannot be expected to encompass the entire range of variation, and it may be necessary to take herbarium specimens of variants that are not represented in the random sample.

Views of Taxonomists

The view expressed by Meikle (1978) is that C. coum can be split into two varieties, which he called var. coum and var. caucasicum. It is worth quoting his precise wording in the key:

He went on to say that C. coum exhibits a clinal pattern of distribution: plants from north-west Turkey, around Istanbul, tend to answer to the above description of var. coum; further east the leaves gradually become more acute and cordate, and the flowers larger with more acute corolla lobes. He appears to include, under var. caucasicum, the various forms found in the Caucasus region of the USSR which Russian botanists divide into five distinct species (see for instance Komarov, 1965), but not the extreme eastern variant from northern Iran, usually termed C. elegans, with markedly elongate-acute leaves and very large flowers. Other British botanists, including Grey-Wilson (1988), follow Meikle's treatment of the species, except that Grey-Wilson elevates the two varieties to subspecific rank and gives more precise limits for petal length, namely 8-13 mm for subsp. coum and 10-16 mm for subsp. caucasicum. He also states that the pair of "eyes" at the base of each petal is white in the former, pink in the latter. Although Meikle emphasises differences between populations in the west and those in the east he makes no mention of any difference in a north-south direction. Grey-Wilson on the other hand states: "Southern forms of C. coum vary very little and match most closely those from the western end of the range of the species".

Expedition results

Meikle's view of the species was based largely on the examination of herbarium material. One of the aims of the expeditions was to see to what extent this view could be substantiated by a detailed examination of populations in the wild. That is why the first expedition, starting from Istanbul, travelled as far eastwards as it is possible to go without entering the USSR, stopping at intervals to make a detailed record of the cyclamen. Subsequent expeditions were planned partly to seek confirmation of some of the tentative, and unexpected, earlier findings, partly to extend the geographical area of search, particularly in a southerly direction.

Because the taxonomy of the species, as outlined above, is based mainly on the shape of the leaves, and the length and shape of the petals, it is convenient to deal first with this aspect of the expedition findings.

As previously reported (Wood, 1990), computer studies were carried out, aimed at establishing correlations between certain plant characters and the geographical coordinates of the various sites. Of the other environmental details recorded at each site, altitude alone was found to give significant correlations. Based on the combined data from all four expeditions, the computer findings can be summarised as follows:

In order to give a bit more detail to the bald statements above, and to give some idea of the magnitude of the effects mentioned, the results are shown below as a series of tables. In these tables, northern Turkey has been divided, quite arbitrarily, into three bands by longitude: "NW" is all points to the west of 34°E; "NC" extends from 34°E (just east of Kastamonu) to 39°E just east of Tirebolu); "NE" is all points to the east of 39°E. Southern Turkey is treated as a separate band, as is Israel. In a similar manner, the altitude has been divided into two bands: "low" is from sea level up to 2,000 ft, "high" is anything above 2,000 ft. With five bands by geographical position and two by altitude this gives a possible total of ten "boxes". In practice one of these is missing, since in Israel all three sites were at high altitude.

Petal length

Because petal length was actually measured, to the nearest mm, whereas leaf shape and petal shape were a matter of judgement, the results for petal length are given in more detail. Table 1 shows the average petal length for each individual site at which a minimum of ten plants were measured. The overall average for each box includes sites at which fewer than ten plants were measured. "Base" means the number of plants from which the average in each box was calculated.

It can be seen from Table 1 that individual samples in north-west Turkey had an average petal length ranging from 11.6 mm to 13.1 mm. With only one exception, samples from further east had an average petal length ranging from 13.4 mm to 17.8 mm. The overall averages show a progressive increase from west to east. Samples from southern Turkey and Israel were mostly comparable with those from north-west Turkey, the exceptions being two populations from a low altitude. The overall averages are slightly greater at low than at high altitude in three of the bands, but in north-west Turkey the reverse is the case. At the individual plant level there is a noticeable difference between the longest petal recorded in north-west Turkey (15 mm) and the longest petal recorded in the two more easterly bands (22, 26 mm). This pattern seems not to apply to the length of the shortest petal, which in most cases is 8-9 mm.

Additional information can be obtained by considering the proportions of short and of long petals in each of the nine boxes. For this purpose short petals are defined as those measuring 13 mm or less, long petals are those measuring 17 mm or more. The results are shown in Table 2.

In northern Turkey the percentage of short petals is high in the west, decreasing progressively further east. The percentage of short petals is also high in southern Turkey and Israel. No long petals were recorded in north-west Turkey or in Israel, nor in southern Turkey above the 2,000 ft level. As well as increasing from west to east there is evidence of an altitude effect.

Leaf shape

Because leaf shape was judged by comparison with a set of standards, rather than being measured in absolute units, it is inadmissible to calculate arithmetic averages. Instead, the proportions of different leaf shapes have been calculated, in an analogous manner to the treatment of petal length in Table 2. The results are summarised in Table 3, where round leaves means the sum of shapes 1 and 2, ovate leaves means the sum of shapes 4 and 5. Shape 3 is deliberately omitted from the calculations.

In northern Turkey the figures suggest that the percentage of round leaves is very high in the west, decreasing progressively towards the east, and is consistently greater at high than at low altitude. As would be expected, the percentage of ovate leaves shows exactly the opposite trend. Surprisingly, the greatest percentage of ovate leaves was recorded in southern Turkey, where at low altitude not a single round leaf was recorded. Israel appears to be intermediate between southern and north-east Turkey with respect to leaf shape.

Petal shape

In the treatment of petal shape, and referring to the standards shown in Figure 1, shapes 1, 2, 3 and 5 have been taken as equating with the descriptions "orbicular", "suborbicular", "subacute" and "acute" respectively. It should also be noted that the shape of the petals for C. coum subsp. coum shown by Grey-Wilson (1988) in the line drawing (p 82) is close to shape 2. So it is logical to combine shapes 1 and 2 as typical of var. coum, and to combine shapes 3 and 5 as typical of var. caucasicum. Shape 4 is somewhat anomalous, being narrow but rounded. On the advice of Desmond Meikle it was included with shapes 3 and 5: not that this had much effect on the overall result as there were very few recorded cases of petals of shape 4. Table 4 shows the percentages of rounded petals in the 9 boxes.

Expressed in this way, it is difficult to discern any overall pattern in the results. This is partly because of the inconsistent effect of altitude. When the distinction between altitude bands is eliminated the overall percentage of rounded petals in north-east Turkey becomes 32%, there being a greater number of plants in the sample at high altitude (241) than in that at low altitude (64). So the figures indicate a marked reduction in the proportion of rounded petals from west to east. In southern Turkey rounded petals are in a slight minority. Israel is remarkable for the figure of 97% rounded petals. (There is no need to show, separately, the percentages of shapes 3 + 4 + 5 since these are equal to 100 minus the figures given in the Table).

Combination of petal length, leaf shape and petal shape

The last three sections have dealt, separately, with those characters on which the taxonomy of the species is based, namely petal length, leaf shape and petal shape. It is implicit in Meikle's key that plants with short petals will tend to have orbicular leaves and rounded petals, and that plants with long petals will tend to have ovate leaves and acute or subacute petals. So one would expect a measure of correlation between these three characters. After the first two expeditions to northern Turkey, a computer was used to see whether this was the case. What was found, surprisingly, was that no two characters were correlated with one another. This result, combined with inspection of the field record sheets right down to the individual plant level, shows that although there are distinct trends, as one moves from east to west, for the leaves to change from ovate to orbicular and for the petals to become shorter and more rounded, these three things are not necessarily happening in the same plants. In fact the characters combine together in all eight possible ways. In individual plants, round leaves may combine with short round petals, short acute petals, long round petals or long acute petals; ovate leaves may also combine with short round petals, short acute petals, long round petals or long acute petals.

Colour of "eye"

According to Grey-Wilson (1988) C. coum subsp. coum is characterised by a white eye while subsp. caucasicum is characterised by a pink eye. The expedition results show that, within Turkey and Israel, a plant with a pink eye is a comparative rarity. In 1987 the incidence of a pink eye across the three longitude bands was as follows: 7% in the west, 3% in the centre, less than 1% in the east. In 1988 only 4 plants were recorded as having a pink eye, all from the Belgrad Forest near Istanbul. In 1990 just one plant with a pink eye was recorded in Israel, but 7% of those recorded in southern Turkey had a pink eye.

Leaf pattern

The remaining plant characters play no part in the taxonomy of the species, but they are included here since they are important in other respects.

The hastate pattern on the leaves, sometimes referred to as marbling, is a character to which growers pay a great deal of attention, and is often the first feature to catch one's eye in the field. The proportions of totally plain leaves (pattern 1) and of marbled leaves (patterns 4 + 5 + 6) are shown in Table 5

Modest numbers of plain leaves were recorded in north-west Turkey, elsewhere very few. Marbled leaves were particularly in evidence at low altitude in central northern and in north-east Turkey. The overall effect of altitude on the proportion of marbled leaves was shown to be statistically significant.

Leaf size

On both the first two expeditions variations in leaf size were particularly obvious. Whereas 80% of all leaves measured had a diameter of 50 mm or less, some populations had leaves measuring 90 mm or more, one giant leaf having a diameter of 147 mm. (As a simple index of leaf size, diameter has been chosen in preference to length to avoid any possible bias due to elongated, pointed leaves.) Inspection of the data failed to reveal any convincing relationships involving average leaf size. However the incidence of large leaves, with a diameter of 50 mm and over, appears to be related to a combination of longitude and shade density, as shown in Table 6.

The percentage of large leaves increases from west to east and is considerably higher in dense shade than in light shade. The effect of shade is not unexpected, and confirms previous suspicions based on visual observation.

Flower colour

This was determined in the field by matching each flower, as far as was possible, to the nearest sheet in the RHS Colour Chart. Difficulty was found in achieving this, and many flowers seemed not to match any of the available colours in the chart, the intervals in which are uncomfortably large over the range covered by cyclamen flower colours. Another problem, which became apparent later, is that two observers may have very different ideas as to the closest match, so that the recorded colour depends critically on the particular observer making the judgement. In any case the arrangement of the chart makes it impossible to treat colours mathematically. What can be said is that the commonest flower colour on all four expeditions was a purple or red-purple of medium intensity (saturation). Flowers that were more intense than the average could often be found, particularly among some populations recorded by the 1988 expedition. Very pale flowers were infrequent, except in three of the populations recorded in southern Turkey where they formed 50% or more of the sample. These three populations, alone among all those recorded by the four expeditions, contained a few white flowers with a dark basal blotch, a form frequently seen in cultivation. Two plants in Israel, and one in north-east Turkey, were complete albinos with no basal blotch, while two more in north-east Turkey were white with a faint pink blotch. Bicoloured flowers were present in some populations, and at one site near Artvin 40% of the sample was judged to have bicoloured flowers. Apart from a slight tendency for bicoloured flowers to be found at high altitude, inspection of the records has failed to find any link between flower colour and location.

Other plant characters

A number of what are loosely termed "unusual" characters crop up sporadically in populations of C. coum. Some appear to be linked with particular geographical areas, for others there is no discernible pattern.

A toothed, or scalloped, leaf is characteristic of var. caucasicum according to Meikle (1978). No scalloped leaves were recorded in north-west Turkey or in Israel, but quite a few populations in other parts had leaves that were either subentire (ie with a wavy edge) or scalloped, the precise distinction between the two being a matter of judgement. The most pronounced examples of scalloped leaves were seen at some of the southern Turkish locations.

Flowers with fimbriated petals seemed to be concentrated at some sites and virtually absent at adjacent sites. None were recorded from European Turkey whereas in two out of the three populations in Israel 95% of the sample had fimbriated petals.

A slightly exserted style was one of the commonest features noted, particularly on the first two expeditions, without any obvious geographical pattern. It is not thought to have any taxonomic significance.

Scented flowers were detected at some sites in northern Turkey, but none in southern Turkey or Israel. The scent, where present, resembled that of C. trochopteranthum, and could not be confused with that of C. parviflorum which, as reported by Bailey (1987), was found in northern Turkey close to populations of C. coum and thus afforded a direct comparison.

Behaviour in cultivation

Plants collected in northern Turkey on both the first two expeditions were cultivated in a greenhouse at Tile Barn Nursery. The measurements and comparisons made in the field were repeated over two seasons, in order to see whether the characteristics noted in the field would be maintained in cultivation.

Inspection of the results shows that the hastate pattern on the leaves is, with few exceptions, maintained in cultivation. The same cannot be said about leaf shape: while a fair proportion of leaves were more or less the original shape after two years in cultivation there were many exceptions. More attention will have to be paid to this in the future.

One clear-cut result that has emerged concerns leaf size, and this is shown in Table 7. The plants collected in 1987 have been grouped, by 10 mm bands, according to their leaf diameters as measured in the field. The average diameter of each band, to the nearest mm, is compared with the average for the same plants after 6 months in cultivation.

The overall average, at 39 mm, is precisely the same in the greenhouse as in the field. But the smallest leaves have become larger in cultivation and the largest leaves have become a lot smaller. This reinforces the view that size differences in the field are partly due to environmental effects.

Individual plants sometimes showed considerable differences in petal length between the field and the greenhouse, but when grouped into boxes as in Table 2 the same overall trends emerged with a tendency for petals to become longer both to the east and at low altitude.

Petal shape was rather inconsistent between field and greenhouse, and so far no clear picture has emerged.

Conclusions from the scientific results

It has to be said that the expedition results throw some doubt on a simple separation of C. coum into two varieties, or subspecies. Although, when examined separately, the three characters of leaf shape, petal length and petal shape on which Meikle's key is based were found, in northern Turkey, to exhibit the trends with longitude which he postulated, at the individual plant level there was little association between them. More importantly, the populations in southern Turkey, and to a lesser extent those in Israel, combine ovate leaves with short petals, a combination that does not fit into either variety. Meikle (1990), who inspected the entire herbarium collection from southern Turkey, judged the plants to be intermediate between the two varieties, but coming closer to var. caucasicum. He was particularly struck by the toothed leaves of some specimens, and at first glance took them to be leaves of C. pseudibericum. It would seem possible that the populations in southern Turkey have evolved in comparative isolation over a considerable period, and in the process have developed the particular characteristics that can be seen today.

There is little more that can be said about the taxonomy of the species until populations in various parts of the USSR, and eventually those in Iran, have been examined in the same amount of detail as those in Turkey and Israel. But Grey-Wilson's assertion that subsp. coum has a while eye and subsp. caucasicum has a pink eye would seem to be an over-simplification. A pink eye is rare in Turkey, but is in fact more prevalent in the north-west and the south than in the north-east, a fact that is at variance with Grey-Wilson's view. Further east, in the USSR and Iran, only some of the forms that might be grouped together under subsp. caucasicum are reported by Komarov (1965) as having a pink eye.

One unexpected finding of the expeditions was the effect of altitude on leaf shape, leaf marbling and flower size. In the literature, altitude is not mentioned as influencing these or any other characters of cyclamen. At present no explanation of the effect can be advanced. It does not appear to be simply due to a difference of temperature between high and low altitudes. as any such effect would presumably disappear when plants from different altitudes are brought together in the same greenhouse. As to the broader distinction between environmental and genetic effects, an environmental influence has only been demonstrated beyond doubt in the case of leaf size. Where a character such as petal length or leaf marbling remains more or less unchanged when plants are removed from their natural environment and cultivated in a greenhouse the character in question must be presumed to be largely if not wholly genetic in origin.

Horticultural considerations

One of the most striking features of the plants seen during the first expedition was the marbling of the leaves. From Unye eastwards some really fine leaf forms were found, one of the choicest examples consisting of a dark green centre of "Christmas tree" shape surrounded by a contrasting area of bright silver, with in some cases a narrow darker green rim. On this expedition the leaf pattern was the chief factor that determined which plants were collected. Nowhere else in Turkey or in Israel were such leaves seen, the intensity as well as the extent of the marbling setting them apart. The 1988 expedition concentrated largely on this part of northern Turkey, from Unye to Artvin. On this occasion the collecting was deliberately structured to include other features of the plants, and a greater proportion of the time in the field was spent searching for unusual or choice examples. This approach paid off, as not only was a plant with pure while flowers found, but also one with a pink hastate pattern on the leaves. This last plant was one of those collected, but unfortunately neither the original plant nor any of its seedlings have shown the pink marking in cultivation. So in this particular case the environmental conditions must have played a part in determining the leaf pattern, despite what was said on this subject earlier. The 1988 expedition also found some very good examples of cyclamen with deep coloured flowers.

Regrettably, in 1990 the Turkish authorities refused permission for any collecting; however a number of plants were brought back from Israel under licence, and these included two with pure white flowers.

One form of C. coum that is sometimes seen in cultivation has a plain, dark green glossy leaf and small, squat, deep coloured flowers. No such form has been seen in the wild by any of the expeditions, so its geographical origin is unknown. But because a particular form is common in cultivation it does not necessarily follow that there must be plenty of it in the wild. Bearing in mind the amount of seed produced by a healthy plant, and the relative ease with which the seed often germinates, a single plant that breeds true can, in a few years, give rise to a large number of identical offspring. So it seems quite possible that the forms of C. coum that were widely offered by British nurserymen over a number of years were derived from comparatively few plants in the early importations of the species, and do not reflect the relative abundance of those forms in the wild. Seven of the plants collected in north-east Turkey in 1987 and 1988, and considered to be of outstanding merit, have been set aside for special treatment. By isolating and close-pollinating each plant, growing its seedlings to maturity, selecting and inter-pollinating only those that mimic the parent, and repeating the process over a number of generations, it is hoped to develop a true-breeding strain from each of the seven plants. Their descriptions are as follows

References

• Bailey, R H (1987). Some notes on Cyclamen parviflorum Pobed. Cyclamen Soc Jour. XI, 43 - 55.

• Grey-Wilson, C (1988). The Genus Cyclamen. Royal Botanical Gardens, Kew.

• Komarov, V L (ed). Flora of the USSR. English translation, Jerusalem, 1965.

• Meikle, R D (1978). In Davis, P H (ed) Flora of Turkey and the Aegean Islands, Vol 6, 128 - 135.

• Meikle, R D (1990). Personal discussion with D.J. Wood.

• Wood, D J (1990). Some Scientific Aspects of the Expeditions. Cyclamen Soc Jour. XIV, 43 - 46.