Early Marsh Orchid, Dactylorhiza incarnata

This species seems to have been misunderstood by enthusiasts and experts alike. The field guides put an emphasis on flower colour, rather than the bigger picture with habitat and size playing an equal if not more important part in sub-classifying the varieties or subspecies. At least that is the message I get from John Haggar's series of articles published in the Journal of the Hardy Orchid Society. Combining his observations and research with Richard Bateman's published work we find that, despite very few genetic differences within the species, the basic Early Marsh Orchid is a rather polymorphic orchid. A few small genetic differences provide a range of flower colours. The Early Marsh Orchids of Gotland and Oland grow untainted by the presence of other Dactylorhiza, and are moderately sized plants growing in damp meadows with flower colour ranging from pale pink through reds to purple. This contrasts with the British view that Dactylorhiza incarnata incarnata is pink, and that purple colour indicates D. incarnata pulchella. Other plants of the genus, such as D. sambucina, exists in two colour forms which grow together. However those have no intermediates. Colour in D. incarnata appears to be the relative expression of genes controlling the production of the xanthocyanin pigments. The big question is why continental botanists regard D. incarnata incarnata to be purple (and pink) while in Britain it is the pink forms which predominate.
In Scotland pink forms of D. incarnata incarnata are rarer, and purple forms can predominate. These grow in alkaline meadows and therefore cannot be considered as D. incarnata pulchella just because of their colouration. Polymorphic colonies with wide colour variation, including purples, are showing primitive ancestral characteristics.

Flowering times may be not much use in helping to differentiate the different subspecies and varieties. You cannot compare those on the western coast of Ireland to those growing on an Island in the Baltic Sea. The species grows over many different climatic areas. However, the English common name of Early Marsh Orchid is a rather parochial view. Even here the Southern Marsh Orchid can be in flower earlier, and some D. incarnata in southern Sweden may still be in peak flowering in early July. I have a personal preference for the Swedish common name - Meadow Orchid, which describes its habitat. Perhaps that could be modified to Meadow Marsh Orchid to standardise.

The answer lies in the behaviour of the bees that pollinate the Marsh Orchids. The orchids attract bees, but do not reward them with nectar. Purple colours attract insects, and bees will visit purple Marsh Orchids without regard to species. The lack of nectar works because the bees with visit more flowers, transferring pollen as they do so, in search of a meal. It gets more subtle than that though, because the bees then `learn' that purple colour may not be good for them, so switch the colour of the flowers they visit. In a polymorphic population with a range of colours that polymorphism is maintained. Where the D. incarnata grow in the company of other polyploid Marsh Orchids other dynamics come into play.
Bee habits with regard to flower colour and height are important factors. If they are young naive bees they visit polyploids such as D. purpurella and the purple forms of D. incarnata in preference, leading to hybrids. These hybrids have one set of chromosomes from D. incarnata and two sets from D. purpurella. They are triploid (3n) and largely sterile. They are a dead-end which saps the D. incarnata population of the genes resulting in purple flower colouration. The effect is not as great in the D. purpurella population because their polyploidy makes them more robust and they mature to flowering 1-3 years sooner. This may be a reflection of D. incarnata being more adapted to cooler climates, suggesting it arrived on our shores soon after the ice retreat. The D. incarnata will suffer more losses before maturation. Add to this that bees getting wise to no return from purple flowers will switch their attention to Marsh Orchids that are not purple; the red and pink forms - ensuring the genes for those are accentuated.
Bees tend to travel between flowers of the same height. Hence hybrid loss tends to affect the taller purple flowered forms of D. incarnata to the benefit of the dwarf red and pink forms. The populations of D. incarnata growing in Gotland without any polyploids do not have these pressures on the purple forms.

So, if D. incarnata incarnata include purple forms, where does that leave D. incarnata pulchella? That taxon does still exist, but is restricted to those purple D. incarnata growing in acid bogs such as those found in the New Forest. These plants have a distinctive morphology that is not shared by other purple flowered plants elsewhere in southern Britain; these seem to be simply purple flowered Dactylorhiza incarnata incarnata. Many of the pulchella reports will have been erroneous, making this variety over-represented on the various databases. I have yet to see this variety, though I can possibly claim to have seen D. incarnata incarnata purple forms. Similar plants to the New Forest plants from similar habitats can be found in Holland and northern France, and this suggest that D. incarnata pulchella is of recent origin. Further to this, one can add that not all pulchella are purple; pale or even white forms can be found in the New Forest populations. There is no sense calling these by any other varietal name. That D. incarnata pulchella is invariably purple adds to the evidence that the ancestral stock for the various D. incarnata varieties has purple forms.

What about purple forms in northern Britain? While some found in mixed coloured flowering colonies are morphologically indistinguishable from D. incarnata incarnata, there are others with both pink and purple flowered plants that have a distinctive look. They have noticeably narrow leaves and have a diamond-shaped weakly lobed labellum. Similar populations can be found in Ireland and in northern Scandinavia. This as yet named variety could be of a more ancient branch from the main form than var pulchella. Or has it?
There have been slender, purple flowered, late flowering varieties described from Lappland (Var borealis) and the Alps (Var serotina), and these appear to be synonymous. Furthermore some Scottish examples match the description. Then a dwarf Swiss variety, Var drudei, with short spreading leaves and purple or occasionally pink flowers has been reported from Skye, and similar plants grow in Lappland too. The reality is possibly that there is a continuum of forms slowly changing from typical D. incarnata incarnata from south to north, and the northern forms have equivalents in the Alps and Scandinavia. Do the extremes of this spectrum deserve varietal names?

Not mentioned yet are the blotched leaf Early Marsh Orchids normally considered Var. cruenta. These can be found, yet again in Ireland, Scotland, Scandinavia and alpine regions. Not all plants in blotched leaf colonies have blotched leaves, and some only on the upper surface. Haggar has stated that blotching in itself is of little taxonomic value. This feature also occurs in southern Scandinavia in a variety referred to as `southern cruenta'. This bears little resemblance to the `northern cruenta' and has been shown to have its own specific enzyme markers. Like flower colour, this blotching is under the control of just a few genes, and molecular studies have yet revealed any difference between those with and those without leaf markings. D. incarnata cruenta (and/or Var. haematodes with only upper side affected) perhaps only deserve forma status! Remember this has previously had both subspecies and full species status. These plants are likely a part of the continuum mentioned in the above paragraph. It is likely that such hyper-pigmentation is an evolutionary adaptation to cooler, wetter climates, but the exact mechanism is yet to be revealed.

My initiation to wild orchids and the Early Marsh Orchid in particular was too many years ago to contemplate at Ynys Las. In those early days the plants there gave me the impression that this species would always be short, robust and compact. So I was quite surprised when I saw my first non-dune forms. That difference and the fact that the dunes held a deep red form marked that population out as something different. How true is this?
In my last two recent visits to Ynys Las and other Welsh dune sites have left me a bit baffled. There are pale pink plants and deep red ones .... and all shades in between. Where does Var incarnata end and Var coccinea begin? Is this a case of inbreeding producing intermediates between the two forms, or merely that it is a single variety displaying the range of shades of pink to red. The latter seems more likely. Why would two varieties exist for the same habitat? Clearly the dwarf D. incarnata in the dune habitats are a sub-population of the species, but too often it seems that excuses are found to explain two sub-species growing together. However, there may be some mileage in the notion that some colour intensity may be down to soil moisture levels or pH. I have noted that the pinker forms grow on the periphery of Ynys Las dune slacks, with the red forms in the centre.
The dwarf dune forms were originally described as Var dunensis, but then the red forms were labelled atrirubra. Similar plants from northern Europe with flowers shading from pink to lilac have been labelled Var. lobellii. These dune loving populations would have arrived from the east or evolved in northern Europe when the early North Sea and Irish Sea coasts were being formed - before Britain was split from the continent. After that split they would have started to diverge and go their separate evolutionary ways, but not far enough yet to be considered different subspecies or even varieties. The British dunes boast the deep red dwarf forms, but that is perhaps all they are - forma coccinea. The continental populations still retain relict characteristic regarding polymorphic colouration. Both populations also harbour plants that are intermediate in height between typical D. incarnata and the dwarf plants. This may be another such relict. There are populations comparable to those termed Var lobellii in NE England and SE Scotland, while the continental populations stretch from Belgium, past Denmark, and onto southern Scandinavia. Clearly there is an East-West split, with those populations with deep red coccinea on the western coasts of the UK. Why Britain alone harbours the red forms is unknown. That there can be red flowered forms found in a few D. incarnata incarnata populations not only suggests not only that positive selection by pollinators may play a part, but also that this colouration is a relict from ancestral forms that spread across Europe. This can explain the so-called unusual Var coccinea growing in Rhos-y-Gadd fen on Anglesey. They are D. incarnata incarnata f. coccinea.
The remaining question is what to call these dwarf dune forms. Precedent says dunensis, but are they a variety or subspecies. For the latter it would have to be shown that they were isolated by flowering time or habitat. That may well be true. then are the North Sea and Baltic coastal populations separate enough to retain the lobellii tag? Perhaps not.

So having decimated the numbers of pulchella in the UK, reducing Var cruenta and Var coccinea to forms, but re-instating dunensis to the list of British variations of the Early Marsh Orchid, what about Var. ochroleuca? These is the large and robust, pale-yellow or ivory coloured plants that now only grow (in the UK) in less than a handful of East Anglian fens. Apart from size and flower colour they have unique features such as a labellum with a central ridge giving it a marked three-lobed shape.
D. incarnata ochroleuca should not be confused with white flowering forms found in typical incarnata or dunensis populations which should be called forma ochrantha.
D. incarnata ochroleuca is also found on the continent, in comparable habitats from eastern France, north of the Alps, through Poland and to the Baltic Sea. A notable population grows on Oland. The East Anglian outpost looks incongruous with that distribution, and leads to a non-monophyletic origin, reducing it to a form. In fact the UK specimens, from molecular studies, may be more closely related to UK incarnata ss. and even UK cruenta than to the Swedish plants. Its rarity may be associated with hybridisation with D. praetermissa. F1 offspring of a yellow flowered parent invariably has the purple flowers (and general morphology) of the other. The Oland specimens are typically more the size of D. incarnata incarnata, while on the northern side of the island is a morphologically comparable pale purple form called lilacina. The possible explanation is that occasionally D. incarnata produces seeds with a genetic make-up that gives rise to ochroleuca forms that are able to colonise fens insuitable for the typical species. That said, there is opinion that ochroleuca has unique pollinators. At dawn and dusk the flowers seem to take on a glow that may attract moths for this purpose. If so, variety status could be appropriate.

All that remains is the plants found by Heslop-Harrison in Sutton Fen and somewhere Sussex as Var. gemmana. These were exceptionally tall plants, purple flowered and found to be diploid in chromosome numbers. Thus despite their rather D. praetissima like labaellum patterns they must have been D. incarnata. Pink flowered forms have also been found more recently. These seem to remain a mystery, defying any logic as to their true nature. The size suggests hybrid vigour, but the 2n chromosome count rules out a D. incarnata x D. praetermissa hybrid, unless something unusual is happening. Don't expect the jury to return soon on this!

Obviously much of this is derived from John Haggar's series of article in the Journal of the Hardy Orchid Society. I like that he has taken all factors into consideration as evidence for his work. He hasn't published a paper purely based on morphometrics or molecular analyses, slavishly followed the colour determines the subspecies route, but has looked at pollinators, habitats, re-colonisation post Ice Age, as well as the evidence. He has then put forward logical deductions from all the evidence. Some is inferred, and some could be refuted, but overall he describes a species as it is found in the field.

   Dune populations are dunensis; coccinea is a form of this.
   Not all purple flowered incarnata are pulchella
   Leaf blotching does is form cruenta
   Very few pale yellow forms are ochroleuca; and no pure white are
   All the rest are incarnata

Top of Page The Polyploids