Arriving in the Taygetos for the first time a few years ago, I came armed with a short list of likely places to look for snowdrops, given to me by kind friends (1) or gleaned from books and papers. Time was short and, although I duly found countless snowdrops, what I gained was a no more than a sideways glance into the marvellous diversity of Galanthus reginae-olgae in the deep valleys of these rugged and spectacular mountains.

Henry Kissenger got about as well acquainted with Dolly Parton as I did with G. reginae-olgae in the Taygetos on our first encounters.

The ethereal quality of the light under the Plane Trees; the delights of sitting by a stream and eating smoked Spartan ham and olives with chunks torn from a loaf of bread and a bottle of local wine; the sweet, honey scent of snowdrops drifting on the breeze on warm autumn afternoons; the white limestone peaks rising stark and crisp above the tree line and the almost perfect solitude all worked their way under my skin and I have been unable to resist going back year-after-year, ever since.

Light under Plane Trees, eastern Taygetos, October 2017
Profitis Ilias, October 2017

Reading again the blog post I wrote after that first trip, I find that it has aged better than some of my other early ruminations and I don’t think there’s anything in it that I now wish I could disown. It was just incomplete, a rough pencil sketch of a riotously rich, colourful landscape. After my fifth autumn visit, supplemented by one in spring, it is time for an update, though this post is far from the final word.

To understand the Taygetos snowdrops, you need to get to know the rivers and streams that have carved deep channels through the softer limestones and schists that built the mountains. Although elsewhere in Greece and in other parts of its wide range this species (2) has emancipated itself from close proximity to flowing water, that is not the case in the Taygetos, where it is almost always tightly associated with streams and rivers. Many of these are seasonal and contain little water in their upper reaches when the plants flower in autumn but the winter rains and spring snowmelt supply plentiful water to the snowdrops in their growing season.

The Taygetos Mountain range is a horst, thrust up between the Sparta fault and the Kalamata faults (3, 4); faults shown in black. Two major rivers, the Evrotas to the east and the Pamisos to the west (shown in red), flow across the flanking grabens.

One reason that the Taygetos is such a spectacular range is that it rises abruptly from the Evrotas River valley in the east and from the Messenian Gulf and the Pamisos River valley in the west. Especially when viewed from the east, the mountains look as though they have been hurled up from the surrounding flatlands by a giant hand. This image is not so far from the truth. The Sparta fault runs NNW-SSE along the eastern margin of the mountains and the parallel Kalamata fault along the western margin (see image above). You would not have wanted to be in Sparta in 424 BC when the Sparta fault slipped abruptly and destroyed most of the city. The next big one is long overdue (3), something you may want to bear in mind if considering a property investment there. In any event, the Taygetos is a ‘horst’, a big chunk of the earth’s crust, thrust skyward between two adjacent ‘grabens’, across which the Evrotas and Pamisos rivers flow. The result is a ridge of high peaks, separating two watersheds.

The Taygetos Mountains rise abruptly from the Evrotas River valley.

All of the rivers and streams that flow down the steep eastern flank of the mountains join the Evrotas River, which rises north of the Taygetos and flows into the Gulf of Laconia south of Skala. While the larger tributaries themselves have smaller tributaries, each valley is separated by a ridge, which is all but uncrossable by snowdrops.

Some tributaries of the Evrotas River. Left bank tributaries not shown.

The drainage of the western flanks of the mountains is more complex, with separate rivers running directly to the sea, without merging, providing even fewer opportunities for snowdrop populations in adjacent valleys to mix. The largest river on the western flanks is the Nedontas and its tributaries. While snowdrops are less common to the west of the ridge than to the east, they are fairly frequent in the Nedontas catchment.

Some of the rivers that drain the western Taygetos. The Nedontas is the most northerly catchment shown.

Looking at a view of the Taygetos from the east, with some snowdrop populations shown (and some places I have yet to investigate), it immediately becomes clear that, while the populations in each tributary of the Evrotas are linked via the river into which they flow, for all practical purposes they are as separate as populations of finches on different islands of the Galapagos, linked only by accidental dispersal. The same is true, perhaps even more so, for the less frequent populations on the other side of the ridge.

View of the Taygetos from the east, showing sites for G. reginae-olgae (green), plausible places (blue) and plausible places where snowdrops are absent (red). Rivers are in red.

In fact, it turns out that most snowdrops in the Peloponnese at large are linked by rivers. The Alfeios River, the longest in the Peloponnese, drains the northern slopes of the Taygetos and flows north before turning east and entering the Ionian Sea not far from Ancient Olympia. In this river and its tributaries, which flow into it from the mountains of the central Peloponnese are to be found many populations of G. reginae-olgae, but most flower in late February and early March, when there are many demands on the time of a galanthophile with a passport and an irrepressible urge to use it. I hope that these plants will be the subject of a future post.

The Alfeios River and some of its tributaries. Numerous spring-flowering populations and a handful of autumn-flowering ones are to be found in its valleys.

It is easy to see how snowdrops can spread rapidly downstream, both as seeds and bulbs. This is presumably how G. nivalis came to reoccupy such large regions of Europe after retreating to refugia in successive glaciations. It is harder to understand upstream spread. Snowdrops seeds are designed to be dispersed by ants and even the doughtiest ant is unlikely to be motivated to haul these burdens very far uphill. Evolutionary time is deceptive though. Suppose ants spread a snowdrop population upstream by an average of 10cm per year (I made up this distance, to illustrate the point), it would only take 100,000 years to get from the bottom to the top of most Taygetos streams. Presumably the snowdrops are occasionally assisted by mammals accidentally or deliberately  consuming the seed capsules or moving seeds stuck on hooves.

Little fruit had been set in this particular spring but it is not hard to see how both snowdrop seeds and bulbs could be dispersed downstream by water.
Snowdrops are highly toxic to mammals but that doesn’t always stop accidental or deliberate browsing. Could snowdrop seeds occasionally be dispersed on the hooves or in the guts of large mammals?

Why have I gone to such lengths to show where snowdrops occur and where they don’t in the Taygetos Mountains? My answer is that I’m fascinated by biological diversity and its origins. I have noticed over my frequent trips that the snowdrops in each river or stream system are distinct. There is considerable overlap, of course, and all the plants are clearly ‘of’ the same species, but the plants in each stream have their own characteristics. I hope to discover how the diversity in this one genus came to be, one species at a time.

It is obvious (5) to anyone who has thought about evolution that individual plants have parents that in turn had parents and so on, in an unbroken chain of ancestors that stretches back over space and time. In the case of snowdrops the trail seems to lead back to what is now the Caucasus and beyond that to Africa, whence amaryllids came, long before any human ancestor thought to leave that continent (6).

In the bad old days, we inferred relationships among species on the basis of how similar they looked. This so-called phenetic approach worked reasonably well – it got us a long way towards systematising biological diversity – and, when it was the best we could do, it was the best we could do. It resulted, however, in some absurd species delimitations. Stern, for example, in his well-known monograph of Galanthus and Leucojum (7) recognised G. corcyrensis (now G. reginae-olgae) and gave its distribution as Corfu and Sicily. Really? How did he suppose it got from one place to the other. What happened to the intermediate populations?

G. reginae-olgae (2), Corfu, October 2015
G. reginae-olgae (2), Sicily, February 2020

Humans have a deep-seated tendency to group things based on superficial similarities, from which much trouble has resulted and not just in taxonomy. Huge advances have been made recently in inferring evolutionary history from divergences in rapidly evolving genes. Fieldwork had, until the recent pandemic, become much easier and cheaper. We can now realistically anticipate reconstructing the true evolutionary history of groups of organisms and classify them based on phylogeny. We can do better now and we should.

Three years ago an old friend and I took our sons to Greece for a week in late October. We dragged them up the Acropolis in Athens, visited Delphi and watched Western Rock Nuthatches foraging among the fallen columns of ancient temples. These sedentary birds were presumably the descendants of those that Rob and I had seen together more than 30 years previously as eager teenagers on our first foreign adventure without parents. My son Pieter and Rob’s son Alec had a race on the original track at Olympia (dead heat). We twitched Pygmy Cormorant in the Messolonghi Lagoons National Park and logged our record in eBird, a clever App that allows birders to record and share their sightings (you can read the full list of what we saw here). Rob is the real deal in the birding universe and I was reminded, for the umpteenth time, what a joy it is to be in the field with a great naturalist, at the top of his game.

Two fat blokes go birding at Delphi, October 2017
“Why are you making me look at this yellow flower, Dad?” Pieter Mitchell, Delphi, October 2017
Pygmy Cormorant. Photo courtesy of Rob Williams.

The following morning, 17 October 2017, I ‘persuaded’ the others to detour from the long drive to Monemvasia via a small population of G. reginae-olgae near Smerna in the western Peloponnese. It had been in full flower on 30 October the previous year and I was curious to see whether the plants were yet emerging. ‘We’ve dipped Tom’, said Rob, in his most annoying glass-half-full voice as I peered disconsolately at the parched and desolate roadside bank that I knew would soon be covered with flowering snowdrops.

G. reginae-olgae, western Peloponnese, late October

This population is at 440m, relatively low for this species. As a rule, autumn-flowering snowdrops in the Peloponnese flower earlier at higher elevations, and later at progressively lower ones. This very sensible arrangement – presumably mediated by lower night temperatures or greater diurnal temperature fluctuations higher up – means that the higher elevation populations flower as late as possible but when there are still a few insects on the wing to pollinate them.

Flowering times of some populations of Taygetos snowdrops, ranked from highest to lowest elevation. See paragraph below for further discusssion.

The chart above shows flowering times for 18 Taygetos snowdrop populations that I have visited frequently enough to have a shrewd idea of timings, sorted from the highest to the lowest. Dark green represents peak flowering and light green early and late flowering. The red line indicates the elevation of each population. The highest is at 1100m and the lowest at 300m. This is a work in progress. It takes many visits over many years to establish accurate flowering times which, in any case, vary from year-to-year and can never be precisely pinned down. It is clear that there is not a perfect relationship between flowering time and elevation. Some populations break this general rule, presumably because different populations at similar elevations experience different microclimates. The very late flowering outlier, for example, is on the warmer western flank of the mountains, where it flowers three or four weeks later than populations at similar elevations on the eastern side. I hope, however, that the general shape of the relationship between elevation and flowering time is clear.

Several days after our dip at Smerna we drove the spectacular road from Kalamata to Sparta that twists and turns for its eastern half down the gorge carved by the Langada River. At the higher elevations of this well-known snowdrop site the flowers were just emerging. The next day Rob and the boys flew back to London, while I stayed on for a week of snowdropping.

It seems treacherous to say it, but snowdrops are not the most beautiful plant in the Langada Gorge. Platanus orientalis gets my vote.

While Rob selflessly shepherded our kids through customs trying, with limited success, to impress on Pieter the importance of avoiding jokes about bombs in his hand luggage, I headed for the high elevation populations that I knew would be flowering. And they were.

A small clump of G. reginae-olgae, flowering in a dry stream bed near Anavryti, early November

The remainder of this post is a description of the variation within and among different populations in the Taygetos. I have not tried to describe every population, in part because I don’t know many of them very well and in part because I don’t want to drive my few readers to drink.


Various dirt roads fan out from the village of Anavryti. They follow the contours of the mountains, occasionally switch-backing down to join a lower road, and cross several valleys carved by rivers and streams flowing down to the Evrotas. There is usually a trickle of water in these channels, even in autumn but the streams don’t come to life until the spring snow melt, when coincidentally – or perhaps not – the snowdrop fruits are ripening.

Typical valley flowing down the eastern face of the Taygetos near Anavryti. The watercourses are marked out by Platanus orientalis in their autumn colours.

Most of these valleys have populations of snowdrops that start flowering in early October, at their highest elevations of around 1050m. They are generally at their peak in the third week of October and by the beginning of November are already fading at the top of the population while still going strong a few hundred vertical metres lower. Because the slopes are so steep and the influence of elevation so strong, there can be only limited gene exchange among plants at opposite ends of the population, even in a single tributary of the Evrotas.

G. reginae-olgae, near Anavryti, late October

The habitat here is typical of that in which snowdrops grow in the Taygetos. The rocks are limestone mixed with schist (I’m not enough of a geologist to be more specific). Away from the rivers and streams the dominant tree species is Pinus nigra. Close to the watercourses Platanus orientalis takes over, with Abies cephalonica occurring regularly in the understorey and occasionally as a canopy tree. When the snowdrops are in flower the leaves of Cyclamen hederifolium subsp. crassifolium are prominent, though most have finished flowering. In spring Cyclamen rhodium subsp. peloponnesiacum also flowers in these woodlands. I have noted Crocus boryi too but it is either rare or largely finished flowering by late October.

Limestone rocks in one of the valleys near Anavryti. This photo was taken in spring, after snowmelt, which is why you see running water.
Platanus orientalis with an understorey of Abies cephalonica, near Anavryti, late October
A magnificent, old example of Abies cephalonica. Many mature specimens of this tree have been destroyed in fires, notably a devastating one in 2007. This species regenerates very slowly after fire (8).
Cyclamen hederifolium subsp. crassifolium, near Anavryti, late October. Most individuals had already finished flowering at this elevation.
Cyclamen rhodium subsp. pelopponesiacum, near Anavryti, late March
Crocus boryi, near Anavryti, late October

The populations near to Anavryti are fairly sparse and most clones grow either as single bulbs or very small clumps. There are exceptions but they are unusual. The plants are relatively small in stature, averaging about 16cm tall at peak flower, but with many shorter and taller exceptions (9). Leaves are completely absent at flowering time, in almost all cases. Virtually all clones produce a single scape per bulb.

A typical scene near Anavryti, late October

The flowers hang from pedicels that are, on average, the same length as the spathes that subtend them, but there are exceptions (10). The ovary is spherical or, more often, obconic (that is a cone shape, with the narrow end at the base of the structure), more-or-less elongated and usually pronouncedly constricted at the junction with the perianth segments.

G. reginae-olgae, an example with short pedicels relative to the spathe and a small, pea-like ovary, near Anavryti, late October
G. reginae-olgae, an example with long pedicels relative to the spathe and an obconic, elongated ovary, near Anavryti, late October

The outer segments are typically about 25mm long (11), rather narrow, with long claws, and concave, like the hull of an open kayak, with an acute apex. A few are heavily textured. Although clones with green markings on the outer segments are not uncommon in the Taygetos, they are vanishingly rare in the populations near Anavryti.

G. reginae-olgae, note long claws and acute apex of outers, near Anavryti, late October
G. reginae-olgae, an example with narrow outers, near Anavryti, late October
G. reginae-olgae, an example with heavily seersucker-textured outer segments, near Anavryti, late October
G. reginae-olgae, an example with longitudinally-grooved outer segments, near Anavryti, late October

The inner segments are narrowly obovate, in other words like a long, narrow egg, again with the ‘ob’ meaning that the narrowest part is at the base, obtriangular or pyriform (pear shaped, again with the stalk end at the base). I’m sorry, incidentally, about all this jargon, but it saves a lot of words. The sinus is usually rather prominent. The apex of the inners is either flat or slightly flared.

G. reginae-olgae, note obovate inners with flat apices, near Anavryti, late October
G. reginae-olgae, note prominent sinus, near Anavryti, late October

The outward-facing surface of the inners has a single green mark at the apex of the segment, usually covering less than a third of the segment’s length but most of its width. Its shape is variable, from an inverted U or V to a heart or arrow-head. In these populations near Anavryti, the inner segment mark is often reduced to a small bridge over the sinus or, rarely, to two small dots either side of it. In even rarer cases the mark is absent altogether (though these individuals are rare, they are commoner here than in other Taygetos populations). In most clones the inner face of the inner segments is entirely green, save a narrow white margin but in individuals with small marks on the outer surface, the green on the inner surface is also reduced. True albinos, with no green on either face, are very rare.

G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, variation in inner segment markings, late October
G. reginae-olgae, mostly green inner surfaces of inner segments, late October

A handful of plants in these populations have four outer and four inner segments. In my experience of other snowdrop species, this particular mutation is often ‘unstable’.

G. reginae-olgae, an example with four inner and four outer segments, late October

I cannot detect any scent in many plants in these populations, though some are distinctly scented. Walking through this, or any other snowdrop population in the Taygetos on a warm day, a mild but delicious honey scent drifts up from the flowering plants. Like every other trait, however, this one varies from clone to clone. I’m sure that, if I were a bee, I’d have much more to say on this subject (12).

The above description of G. reginae-olgae serves pretty well to cover most of the snowdrops in the Taygetos, except where I noted that the Anavryti populations are distinct. When I discuss other populations below I will not repeat this basic description but merely note how these other populations vary from the notional ‘average’.


West of the village of Paleopanagia is a popular picnic site that also serves as a base for one of the trails up to Profitis Illias, the highest peak in the Taygetos. Unlike George Mallory, who wanted to climb Everest ‘because it’s there’, I’ve never met a mountain that didn’t look better from the bottom, so while others head intrepidly for the heights, I crawl around on my hands and knees in the snowdrop population (a well-known pilgrimage place for botanical tours) that flourishes both up and downstream of the barbeque pits. Although at a similar elevation to the populations nearer Anavryti, this one flowers a good ten days later.

G. reginae-olgae, a robust, clump-forming clone, Paleopanagia, early November

The snowdrops in this valley are considerably more robust than those described above. Some are twice as tall, at about 30cm at peak flower. Many form small clumps. The valley contains many low cliffs, dripping with moss and ferns and G. reginae-olgae often grows here as a chasmophyte, rooting in crevices and pockets of soil and leaf litter in the rock face. A very small proportion of plants in this population produce two scapes per bulb.

G. reginae-olgae growing as a chasmophyte, Paleopanagia, early November
G. reginae-olgae growing as a chasmophyte, Paleopanagia, early November
G. reginae-olgae growing in the base of a Plane Tree, Paleopanagia, early November

Langada Gorge

The main ridge of the Taygetos is, for most of its length, an impassable barrier, if you are a snowdrop seed, travelling in the mandibles of an ant, which your distant ancestors foolishly chose as their primary disperser. Snowdrops occur on both sides of the ridge, though they are presently more common to the east. On which side did they start and how did they cross? Alternatively, did they spread from the north, colonising both sides of the ridge simultaneously?

Places where catchments on east and west come close shown as yellow circles, with diameter 2km. Confirmed snowdrop sites are in green, plausible sites in blue and sites where snowdrops appear to be absent in red. There are many such sites in the west. Rivers are marked in red, faults in black.

I’m going to dodge trying to answer these questions, for now, though I have some tentative ideas. It seems to me that G. reginae-olgae must be descended either from G. nivalis or from a common ancestor, now extinct, of both species. That ancestor must have spread south, crossing into the Peloponnese either near present-day Corinth or, more likely, near present-day Patras. More fieldwork is needed, I’m happy to say, to resolve this question.

One gorge, carved by the Langada River, cuts through the main ridge of the Taygetos and snowdrops occur high up in its catchment. Not far away, in the Nedontas catchment west of the ridge, snowdrops show up again. Further to the north other tributaries of the Evrotas also come tantalisingly close to the Nedontas catchment in their higher reaches. I speculate – and I must emphasise that this really is speculation for the time being – that snowdrop seeds got from west to east or vice-versa on the hooves or in the guts of wild boar (which I see occasionally in these mountains) or deer, crossing the ridge at these relatively low points. There are techniques, similar to those used to infer the routes of human migrations out of Africa, that could resolve these questions but I need first to start playing the lottery and then to win it at least once, if I am to access them.

Langada Gorge, looking east from near the pass, late October
Limestone rocks in the Langada Gorge, late October

The road from Mystras to Kalamata , winding up the Langada Gorge and down the western flanks of the Taygetos is one of the world’s great drives. Don’t miss it before you die. If you happen to be a galanthophile or perhaps even if you’re not, stop often and explore the river banks, for snowdrops grow in this valley from at least as low as 300m to at least as high as 1000m and they are quite different at either end of the population.

The upper end of the population is at about the same elevation as those around Anavryti but reliably flowers about ten days later. Perhaps warm air from the Evrotas valley is drawn up the gorge, slightly delaying flowering? The snowdrops are much more prolific on the north-facing bank of the river, tending to cling tightly to it but growing up bracken-covered rock falls too.

G. reginae-olgae growing on the riverbank, Langada Gorge, late October
G. reginae-olgae growing on a brackeny slope, well above the river, Langada Gorge, late October

The snowdrops here are similar to those near Anavryti in habit and appearance but most are slightly taller on average. Many plants in this population have unusually long outer segments, both in absolute terms and relative to the inner segments.

G. reginae-olgae, an example with long outers, Langada Gorge, late October
G. reginae-olgae, an example with long outers, Langada Gorge, late October

Much farther down the valley, not far from Sparta, the snowdrops are sparser but show some tendency to become clump-forming. It is often the case that snowdrops growing on fertile floodplains or which receive run-off from agricultural land appear more robust and are more likely to form clumps than relatives on poorer land, and I am not sure whether these differences between upper and lower populations are significant.

G. reginae-olgae, clump forming in the lower Langada Gorge, early November

The commonest mutation – of the kind that excites the admiration of galanthophiles – in wild snowdrop populations is the presence of variable green markings on the outer segments. This mutation is present to different extents in every snowdrop species. It is a common trait in G. reginae-olgae. It’s frequency varies dramatically among populations, however, from zero to perhaps one percent. It is quite clear that these markings are heritable. Find one plant with green on the outers and the chances are that a close examination of nearby clones will reveal others. One of many questions about snowdrop diversity that remain unanswered currently relates to the trade-offs natural selection makes between the advantage of having a bit of extra photosynthetic tissue and the arguable disadvantage of being less attractive to pollinators (13).

G. reginae-olgae, an example with green markings on the outers, lower Langada Gorge, early November
G. reginae-olgae, one of the largest snowdrop flowers I have ever seen, with 47mm outers lower Langada Gorge, early November

I have never seen a snowdrop with green markings on the outers in the upper part of the Langada Gorge. These populations are regularly visited by groups of plant enthusiasts and I can’t help wondering whether some selective editing has taken place over the years (14). Lower down such plants are relatively common but still rare.


Before I figured out how to search systematically for snowdrop populations in the Taygetos, I drove around more-or-less at random and this was extremely pleasant but not very productive, in terms of snowdrop populations found per ruined clutch. I must have driven half a dozen times up and down the beautiful gorge that runs out onto the Evrotas plain near Xiropotamos before I saw my first snowdrop in that valley. Having found the key, I quickly realised that snowdrops occur from the top almost to the bottom of this valley. The valley is particularly deep and steep sided and it is arduous, in most places, to get down to the river but there are snowdrops growing sparsely in the riverside rubble, wherever you look.

The gorge west of Xiropotamos, early November

High up in this valley the snowdrops are tiny, even smaller than those around Anavryti. They grow under Platanus orientalis but surrounded by dark forests of Pinus nigra, almost exclusively as single bulbs and they give every impression of being thoroughly cheesed off with having drawn this bleak location in which to eke out their lives. Lower down the scenery becomes ravishingly lovely and the snowdrops commoner and lustier. I have not yet spent enough time in these populations to form an accurate impression of the variation present but I have seen examples of many of the variants present at low frequencies in other valleys.

G. reginae-olgae, an almost albino example, west of Xiropotamos, early November
G. reginae-olgae, an example with green on the outers, west of Xiropotamos, early November

Many of these high valleys in the Taygetos were cultivated in the more-or-less recent past, probably with chestnuts, walnuts and some vegetables, as evidenced by the crumbling retaining walls of old terraces. I frequently come across active walnut plantations in the mountains, apparently in the middle of nowhere. I have often wondered to what extent humans abetted the spread of snowdrops between adjacent valleys, either deliberately or, more likely, inadvertently, carrying seeds on muddy boots as they tended their walnut trees in spring.

Snowdrops growing on abandoned terraces, west of Xiropotamos, early November 2017


Most of the right bank tributaries of the Evrotas are small and often flow only seasonally, at least in their higher reaches. One larger river near the northern limit of G. reginae-olgae in the Taygetos has permanent water at least up to 600m. Snowdrops are common in this valley and, indeed, this is the largest population of G. reginae-olgae I know in these mountains.

G. reginae-olgae, forming large, tall clumps (this one 40cm), Kastorio, late October

These snowdrops are rather different to any of those described above. They grow on both banks of the river, which is large enough to have formed a floodplain with deep, rich soil in places. Many of the clones form clumps, some containing dozens of bulbs and some are very tall, up to 40cm in some cases. Whereas the leaves are completely absent at flowering time in other populations, they are frequently just beginning to emerge in this population, although they are only rarely more than two or three centimetres long. Some clones have very large flowers. Outer segments of 40mm are not uncommon. Clones with green markings on the outer segments are also present at a low frequency.

G. reginae-olgae, forming large, tall clumps, Kastorio, late October
G. reginae-olgae, an example with green markings on the outers Kastorio, late October. Note also the leaves just beginning to emerge.
G. reginae-olgae, still only just emerging, Kastorio, early November, but note that the leaves are also beginning to appear.
G. reginae-olgae, forming large, tall clumps, Kastorio, late October. This example has particularly narrow outers.

The scapes of G. reginae-olgae are always slightly flattened in the same plane as the leaves and longitudinally ridged. At Kastorio this trait is sometimes taken to extremes and the scapes resemble bootlaces. I visited this population once when the fruits were almost ripe and I found that all the leaves and scapes had been swept in a downstream direction by a recent flood. I wonder whether these bootlace scapes are an adaptation to fast-flowing spring rivers.

G. reginae-olgae, Kastorio, early November. I cannot find a particularly convincing image to illustrate the flattened cross-section of the scapes, but you can just about see it here.

This valley is also a spectacular place to see Cyclamen hederifolium subsp. crassifolium at its very best. It flowers here in October and into November, whereas populations at higher elevations have mostly finished, just as with the Galanthus.

Cyclamen hederifolium subsp. crassifolium and G. reginae-olgae, Kastorio, late October

Nedontas River

I have spent many happy but fruitless hours driving around the Taygetos west of the main ridge looking for snowdrops. The only place I have so far seen them in large numbers is in the Nedontas River catchment, where several scattered populations are to be found. They occur at mid elevations and flower exactly when I’d expect them to, in the middle of the season.

These snowdrop populations are, by far, the most variable in flower morphology of any that I know in the Taygetos. Almost all the plants grow as single bulbs or very small groups. These are small plants, typically no more than 15cm at peak flower. As well as growing in their usual habitat, under Platanus orientalis on rocky river banks, parts of the population also spread well away from the river, in turf on abandoned terraces where walnuts were formerly cultivated.

G. reginae-olgae growing with grasses on abandoned walnut terraces, Nedontas catchment, early November

The diversity in these populations is quite spectacular. Essentially it encompasses all the variation in flowers described above and more. There are plants with very narrow outers and others with broad outers. A small but noticeable fraction of the clones have four outers and four inners, or some other unusual combination. A tiny fraction of plants are albino or nearly so. A larger fraction has inner segment marks that extend over more than half of the inner segment, a trait I haven’t seen anywhere else in the Taygetos. In rare cases the entire outer surface of the inner segments is green. Plants with green markings on the outer segments are relatively common and sometimes mostly green inners combine with green outers to pleasing effect. A very few plants appear to have yellowish ovaries, another unique trait and many more have rather pale ovaries and inner segment markings, compared with the typical dark green. Some plants have very long pedicels, whereas in other cases they are short. The ovary ranges from small and pea-like to elongated, like a guardsman’s hat.

G. reginae-olgae, an example with narrow outers, Nedontas catchment, early November
G. reginae-olgae, an example with broad outers, Nedontas catchment, early November
G. reginae-olgae, an example with four outers and four inners, Nedontas catchment, early November
G. reginae-olgae, an example with two outers and two inners, Nedontas catchment, early November
G. reginae-olgae, an example with a very reduced inner segment mark, Nedontas catchment, early November
G. reginae-olgae, an example with a large inner segment mark, Nedontas catchment, early November
G. reginae-olgae, an example with a large inner segment mark, Nedontas catchment, early November
G. reginae-olgae, an example with a large inner segment mark, Nedontas catchment, early November
G. reginae-olgae, an example with green marking on the outers, Nedontas catchment, early November
G. reginae-olgae, an example with a large inner segment mark combined with green markings on the outers, Nedontas catchment, early November
G. reginae-olgae, an example with a yellowish ovary, Nedontas catchment, early November
G. reginae-olgae, an example with a yellowish ovary and inner segment mark, Nedontas catchment, early November
G. reginae-olgae, an example with a particularly long pedicel relative to the spathe, Nedontas catchment, early November
G. reginae-olgae, an example with an unusually long ovary, Nedontas catchment, early November

It is often the case that diversity in a species is highest in the place where it originated, becoming progressively more impoverished further from the origin, as new populations are founded by individuals bearing only a fraction of the alleles in the original populations. This is why, for example, humans are vastly more genetically diverse in Africa than elsewhere (15). It is tempting to speculate on these grounds that the populations in the Nedontas catchment are ancestral to those to the east of the main ridge, but the evidence just isn’t there yet to be sure.


The last population that I want to discuss in this post is also in the western Taygetos and is the most southerly population I know. It flowers much later than any other population, starting in mid-November and going on well into December. I have only been in the region once late enough to see it flowering, so don’t have much to say, save that the plants remind me, in stature (huge), leaves (leaves somewhat developed at flowering) and clumping habit (making huge clumps) of those in the Kastorio valley, which is further away than any other snowdrop population in the Taygetos. It is not impossible that the small population at Milea resulted from a human introduction and it is indeed not very variable and sets very little seed, in my limited experience. I would very much like to see the populations of snowdrops that are said to occur near the head of the Mani peninsula, flowering in January. There are always more travels to plan.

I cannot locate my photos of this population in flower. This image, taken in early April, illustrates the size of some of the clumps and the full development of the leaves.

As mentioned I have only visited the Taygetos once in spring and my observations of G. reginae-olgae in fruit are therefore tentative. In the year that I visited most populations had set very little fruit. I have no idea whether this is typical. Although, as explained above, the various populations reach peak flowering over a period of perhaps five weeks, fruiting times are more compressed. The capsules seem to reach ripeness in the last week of March and the first two weeks of April. Although the higher elevation populations flower earlier, they ripen fruit slightly later.

G. reginae-olgae, almost ripe fruit, Paleopanagia, late March
G. reginae-olgae, almost ripe fruit, Paleopanagia, late March

As always, I would very much like to hear from anyone who has an opinion on any of the things I write on this blog. I try not to be anodyne, which may sometimes mean I come across as strident but that is not to say I don’t welcome dissent. Having been burnt by social media, my only rule is the one that hangs above the counter in my local greengrocer: be nice or go away.

(1) Many thanks to Melvyn Jope and Kurt Vickery for giving me my first leads.

(2) I’ll come back in another post to the question of whether G. reginae-olgae should be considered one species or several.

(3) Papanikolaοu, I. D., et al. (2013). “The Sparta Fault, Southern Greece: From segmentation and tectonic geomorphology to seismic hazard mapping and time dependent probabilities.” Tectonophysics 597-598: 85-105.

(4) Lyon-Caen, H., et al. (1988). “The 1986 Kalamata (South Peloponnesus) Earthquake: Detailed study of a normal fault, evidences for east-west extension in the Hellenic Arc.” Journal of Geophysical Research: Solid Earth 93(B12): 14967-15000.

(5) Though, as T.H. Huxley is said to have remarked when Darwin explained natural selection to him, “How extremely stupid not to have thought of that!” Most brilliant ideas are ‘obvious’ only in hindsight.

(6) Meerow, A. W., et al. (1999). “Systematics of Amaryllidaceae based on cladistic analysis of plastid RBCL and TRNL-F sequence data1.” Am J Bot 86(9): 1325–1345.

(7) Stern, F.C. (1956) Snowdrops and Snowflakes. London: The Royal Horticultural Society, Vincent Square.

(8) Christopoulou, A., et al. (2019). “Assessing the impact of different landscape features on post-fire forest recovery with multitemporal remote sensing data: the case of Mount Taygetos (southern Greece).” International Journal of Wildland Fire 28 (7).

(9) By way of comparison, the median height at flowering (from emergence point to top of spathe) of 170 individuals from Greek and Albanian, autumn-flowering populations that I have measured is 19.8cm.

(10) The median ratio of spathe to pedicel length of 153 individuals from Greek and Albanian, autumn-flowering populations that I have measured is 1.1.

(11) The median outer segment length of 156 individuals from Greek and Albanian, autumn-flowering populations that I have measured is 25.9mm.

(12) According to the interesting article linked to here, humans are “a species of cerebral, hypervisual snobs…One 2011 poll found that 53 percent of people ages 16 to 22 would rather give up their sense of smell than give up their smartphones and computers.”

(13) The following paper makes an interesting start in addressing this question. Aschan, G. and H. Pfanz (2006). “Why Snowdrop (Galanthus nivalis L.) tepals have green marks?” Flora – Morphology, Distribution, Functional Ecology of Plants 201(8): 623-632.

(14) I’ve thought long and hard about sharing the results of years of fieldwork on this blog. I believe that people are more likely to value nature if they feel they have a stake in it, which they are more likely to do if they understand it. I defy anyone to spend an afternoon beneath the Plane Trees in the Langada Gorge and come away uncommitted to preserving these snowdrop populations and their habitats for ever. On the other hand, unscrupulous collectors could use this article to shortcut the work of years and go straight to these populations. Unlike the rare birds Rob and I saw, snowdrops can’t fly. I don’t know how to answer this dilemma (and would appreciate advice), which is essentially the same one that faces every travel writer who shares a ‘secret’ destination. Who gets to play God? When I was a kid the island of Koh Samui, off the eastern coast of Thailand was as far flung as it got. By the time I made it to that part of the world it was necessary to go to the next island, Koh Phangan to get away from the crowds. Soon after then Kho Tao was the only place to avoid the raves and the rapes and now, sadly, there are no more islands.

(15) For example, see Amos W. and Hoffman J. I. (2010) Evidence that two main bottleneck events shaped modern human genetic diversity. Proc. R. Soc. B. 277131–137

Post a Comment:

Your email address will not be published. Required fields are marked *

10 thoughts

  1. Tom, Many thanks for an article which I have enjoyed immensely – and read for the second time! I feel it is unlikely I will ever physically follow your footsteps so it is especially good to be able to read of your travels and observations.

    I have, very gratefully, received some bulbs from Melvyn Jope which he grew from seed collected in the Langada Valley and value them hugely, a special gift of kindness and to be treasured.

    I also grow, another kind gift, Galanthus ‘Rachelae’ one of the very early introductions and this year, not previously, produced flowers with a light hint of green lines on the outer segments which might make one wonder if these markings arise as a result of the compost or conditions in which the bulbs are grown.

    Your note above, Number 14, re the dissemination of information regarding these snowdrop is, unfortunately, a genuine concern. I have been amazed this year especially at the number of postings on social media showing the only recently discovered and described Galanthus bursanus in gardens. It is beyond my belief that a plant could have been distributed so widely and so quickly without the removal of bulbs from the wild.

    Best wishes in further travels and studies and I look forward to reading more from you.

    1. I’m so glad you are enjoying the posts Paddy.

      On the subject of green lines appearing on the outers some years but not others, that’s something I’ve noticed in several cultivars. ‘Fly Fishing’ is a particularly good example. I’ve also seen pictures of the same cultivar grown in different gardens with dramatically different intensities of green. I am confident that the cause is a mutation (probably many different mutations), because of the obviously patchy distribution of such clones in wild populations, but I don’t doubt that the expression of the relevant gene could be modulated by different environmental conditions.

      Unfortunately I don’t have a garden these days, so I can’t do the experiments that might help to resolve questions like these.

      Thanks for the feedback re conservation. The truth is that any sufficiently determined individual can find snowdrop populations with a combination of internet searches, Google Earth and old-fashioned fieldwork. That’s what I do. There are many threats to wild snowdrop populations greater than illegal collecting. Reassuringly, the Langada Gorge is still full of snowdrops, despite being perhaps the best known population in the Taygetos. The recently described species that you mention is in an area of marble quarries. So far only a couple of populations are known (though there are surely others) and it is vulnerable to sudden extinction for that reason. I hope it will be the subject of my next post and I certainly won’t be disclosing any locations. There is a strong moral argument in favour of efforts at ex situ conservation in cases like these but I’ll expand on that idea in due course. What to do is a difficult, perhaps almost intractable problem. I greatly appreciate your thoughts.

      1. Yes, I’ve read of the quarries in the Bursa area which threated the environment of G. bursanus and, says he sarcastically, that we have so many environmentally aware people who have dashed in to rescue bulbs in advance of any threat. Timing is important, I suppose, but so too is motive and moral authority. Rescue for profit seems to be rife and it doesn’t lie easily with me.

        I’m looking forward to your next instalment.

        Many thanks.

        1. Paddy, We’ve all seen this week the consequences of living in an echo chamber and there’s no more acoustically perfect echo chamber than one’s own mind. I spend a lot of solitary time in there. I share my trenchant opinions about conservation in the hope of being challenged and made to think and revise my opinions accordingly. I understand your cynicism and agree – to some extent – that it is justified. I will try to expand on these ideas in future posts. Thanks for reading and I hope you will continue to tell me when you smell a rat! Tom

  2. Hi Tom,

    Another fascinating and beautiful post, thank you. Great pictures.

    Is there really any greater pleasure than being able to travel? Some people think it frivolous, ultimately pointless even, and no doubt it is down to some personal psychological defect but if I had the money I would travel continuously. One of the most painful wounds of the Brexit disasterpiece… Again, I am jealous of your ability to combine your greatest passion with (perhaps another?) travel. Fingers crossed for 2021.

    Your subscription widget is fixed!

    1. Thank you again Leadbelly. One could argue that travelling is the most distinctively human of traits. There are various candidates for this title – language, tool use, bipedalism, art – but travelling, literally, got us where we are, which is essentially everywhere. Other adaptations, for example sociability, may have liberated our travel genes but we’d be long extinct if some of us hadn’t had that urge to move. There are various dopamine receptor genes, variants of which seem to confer different propensities to wander (Chen, C., Burton, M., & Greenberger, E. D., Julia (1999). Population Migration and the Variation of Dopamine D4 Receptor (DRD4) Allele Frequencies Around the Globe. Evolution and Human Behavior, 20, 309-324.). Perhaps you and I carry the traveller gene?