At a recent club meeting (shout out to Twickenham bonsai club) the subject was azaleas. During a critique session it was noted that on Satsuki azaleas, different coloured flowers appear on different branches so you need to be careful when you prune not to remove the colours you want to keep. This got me wondering about Satsuki azaleas, why they have the colours they do, and what this might mean for their bonsai custodians.

Starting from the start, flower colour is the “result of pigment molecules accumulating in cells”^ref and depends on which pigment molecules are present, where they are distributed, and the shape of the host cell.^ref

There are several different types of molecules responsible for colours in flowers – the main ones are carotenoids, flavonoids and alkaloids. There are 700 known naturally-occurring carotenoids,^ref including substances in the carotene and lutein families like beta-carotene which is in carrots and lycopene which is in tomatoes – these produce yellows and some oranges. Flavonoids produce the widest spectrum of colours^ref, and these include anthocyanins which make pink to blue-violet colours, chalcone/aurone which make deep yellows and flavones/flavonols/flavanones which make white and light yellows. Betalains don’t get as much press as the previous two pigments, but these are responsible for reds and deep purples (such as in beetroot) – and interestingly they never appear in the same place as anthocyanins, so it is either one or the other in a given flower.^ref

Some colour examples are provided below from an article by Utah University.

And here are some betalain-containing fruits from this research paper:

In the case of azaleas, anthocyanins and flavonols are the major pigments.^ref One study analysed azalea colours by their pigments, and showed that the different colours fit into three pigment groups – red, purple and white. They found that the red group contained two to four major anthocyanins, and the purple group had two to six – therefore more colour variation is possible in the purple group. They found also that white group flowers did not contain any anthocyanin, but did contain the precursors to this pigment, so their lack of colour is probably a genetic defect in the pigment biosynthesis pathway which causes pigment creation to fail.

Pigment molecules are what are known as ‘secondary metabolites’ in plants – substances which they synthesise to help them in some way or other.^ref Aside from those mentioned above, chlorophyll is another pigment that plants make, this appears green and is responsible for photosynthesis. Without going into the detail of the chemical structure of each pigment, the key point for this article is that plant cells need to synthesise a pigment or usually a combination of pigments in order to display a particular colour. The synthesis of pigments is a multi-stage process which is mediated by enzymes, and the production of enzymes is under genetic control (see a cute animation of this here). So genes – contained in the DNA within chromosomes in every cell – determine what pigments are produced where when cells develop. To learn more about the biosynthesis pathway for azalea pigments you can read this paper (a reasonable understanding of genetics is needed to make sense of it). This seems like a good moment to state that I am not a geneticist! So the below is my citizen-scientist interpretation of what I have read – if there are errors I would be happy to correct them.

Let’s have a look at a couple of images of Satsuki azaleas from the Omiya Bonsai Museum website:

The most obvious phenomenon illustrated in these examples is that each plant develops flowers of different colours and patterns. Some flowers are one colour (red), some another (white), some have different coloured petals, some have coloured stripes and others have speckled colours. Not shown here are the different flower shapes, which can also be different on the same plant. Each one of these ‘phenotypes’ (observable traits) has its own explanation.

The ability of a single plant to produce different phenotypes for flowers is known as ‘sporting’ and the branches which have this variation from the ‘base’ colouring are known as ‘sports’. Sports result from “sudden variations in gene expression of somatic cells…[which]…results in plants having a different phenotype” – in plain English this means that one or more cells which produce a shoot or flower bud all of a sudden start to use different instructions for creating pigments.^ref According to an excellent South African study by S de Schepper et al, which I have used extensively in writing this post, “all the information required to sport (the ‘sporting capacity signal’?) is present in leaves”, which suggests the capacity for sporting is across the tree – it’s a characteristic possessed by the plant as a whole, which is triggered in a specific cell as a ‘sporting event’.

How a sporting event happens is that the meristem (the growing tip of the plant) spontaneously develops a mutation in a cell, and every cell that divides from this cell carries the same mutation. The South African researchers investigated this for azaleas^ref and found that the majority of these changes in azaleas are ‘epigenetic’ events such as those caused by ‘transposons’ or so-called ‘jumping genes’^ref. These types of sports can only be passed on to new plants by clonal propagation, since epigenetic events are by definition not captured in seed or pollen.

But what is clearly passed on genetically is the ability to produce sports in the first place, this is obvious by the fact that so many Satsuki azalea varieties which are created by sexual reproduction have sporting behaviour. Probably the best hypothesis I have come across to explain this is that polyploid parent plants are partly responsible. Mark Nijland notes in an article in 2022 that ‘Suisen’ is a dominant parent plant in Satsuki breeding.^ref It turns out that ‘Suisen’ is not only a tetraploid (ie. it has four sets of chromosomes instead of the normal two), it’s also a mixaploid, which means it has different genotypes in different parts of the meristem – some areas are diploid (with two sets of chromosomes) and some are tetraploid (four sets of chromosomes).^ref Other tetraploid/mixaploids include ‘Shinsen’, ‘Miharu’ and ‘Koyo’ and there are also a number of known Satsuki azalea triploids (having three sets of chromosomes).^ref What polyploidy/mixaploidy brings to the plant is a larger number of options (known in genetics as ‘alleles’) for development – so that when a flower is created it can have multiple different colour or pattern options.

Another factor is that polyploid organisms have to deal with what is called ‘genomic shock’. When more than two sets of chromosomes are combined in a single cell there is more potential for abnormal development, which can negatively affect an organism’s chances of survival. Plants are known to use transposons to adjust their genes to avoid this genomic shock.^ref (This is usually the moment in a research paper or article where the author calls out the remarkable Barbara McClintock for her discovery of transposons, she also proposed the idea of genetic shock. She presented her work at a symposium in 1951 to great derision and wasn’t believed for over a decade – her Nobel Prize wasn’t awarded until 1983.)

Anyway, transposons jump into a gene and change its function – switch it on or off or cause it to do something different. Due to managing genetic shock a polyploid plant is more likely to be making use of transposons than a genetically stable diploid (the ‘normal’ two sets of chromosomes) plant.

The theory goes that in Satsuki azaleas, the presence of polyploidy and mixaploidy result in plants which are more likely to spontaneously adjust their genetic expression via transposons, and when this happens, a wider range of colour and form options are available due to the larger set of genetic material from which to choose.

A prevalence of polyploid/mixaploid parents would undoubtedly be down to selection by Japanese breeders over the centuries. Azaleas have been intensively hybridised and selected since the 1600s, and in 2020 there were 1400 varieties listed in the Japanese Satsuki dictionary.^ref The polyploid parents such as ‘Suisen’ must have been observed to produce sporting child plants over the long term, and were retained for breeding (this article notes that Suisen is the most dominant parent over three decades of cultivar production). It has been found by genome analysis that not only did the ancestor of azaleas have a whole genome duplication event 75 million years ago, but also that the genome of Rhododendron simsii, a parallel branch of rhododendron to Satsuki azaleas (which are mainly Rhododendron indicum) is 47.48% comprised of repeat elements, the majority of which are transposable elements (transposons).^ref So it’s not surprising that somewhere along the way azalea breeders came across varieties which dialled up the ‘sporting’ aspect of their genetic makeup.

Let’s take a look at some of the different types of sports which are observed on Satsuki azaleas and why they form. First we need to understand one key attribute of sporting events which took me several days to get my head around – once you do it is a bit of an ‘a-ha!’ moment. This key insight is that sporting events can happen anywhere on the meristem – for example half-way up a shoot as the meristem is happily dividing – they don’t necessarily neatly happen just as a flower or shoot bud is emerging. Wherever a cell is dividing and generating a new cell, a sporting event can take place. If a sporting event related to flower pigment happens halfway up a shoot, you won’t notice it until a flower emerges but that sport will be there on the stem and on any leaves which bud from that stem. So to understand how flowers on a stem might be affected by sporting events you need to imagine the time-sequence of your azalea developing and how a sporting event would propagate along a stem as the plant grows.

Different but uniform colours on each flower occur when a sporting event changes the gene expression for pigments across the entire flower – so all cells are affected. For example, a pigment isn’t expressed any more, or a pigment which was suppressed is now expressed. It is said that darker colours in azaleas dominate genetically^ref. So in the image above, the ‘cancel pigment’ genetic defect was already present in the tree – to produce white flowers. As the tree grew and new stems developed, some transposons jumped into the defect in a cell which was a precursor to an entire stem or branch, and deactivated it. From then on the flowers would be red due to the red pigment being produced. Some branches haven’t had the transposons jump yet, so they stay white.

Stripes are caused by ‘genetic mosaics’ or chimeras where different cell layers or areas of cells within the meristem have different genotypes. So part of the stem where this flower bud developed has the sport and part does not – and both of those sets of cells contributed to this flower. In the example above you can see a flower with a clearly delineated section of dark pink on the left. This is an indication that the sporting event (in this case reactivating the pigment production) happened at a point which you can trace back along the two edges of pink. These flowers may be relatively solitary on the tree, since it’s their unique position and timing of development which gives them these patterns. You can imagine a small stripe of pink along the stem where this flower bud came from, back to the location of the sporting event, and one might continue along the stem above this flower as well.

Flecks are another version of chimeras but the mutation is not persistent in the same way as stripes – you could call them ‘micro-sports’. One explanation for this behaviour could be that the transposon which reactivated the red pigmentation ‘flickered’ on and off as the flower developed. Unlike stripe chimeras, flecks can be passed down to child plants, indicating that ‘flickering’ behaviour is a genetic trait and not an ephemeral one like most other sports.^ref

Petal borders or margins of a different colour on the same flower (also known as picotees) can arise in different ways. In flowers with large margins (7mm or more) this is due to different cell lines being mixaploid, resulting in the petal edges containing 4 sets of chromosomes and the rest of the flower containing 2 sets of chromosomes. In one study this made petal edges white and the centre red as in the picture above.^ref Other picotees are caused by ‘positional’ differences in gene expression where all cells have the genes necessary to make pigments, but only those in a specific position are ‘instructed’ to do so.^ref In both cases these are due to somatic/spontaneous changes and cannot be passed down through sexual reproduction.^ref

Combinations of effects are seen on many Satsuki azalea varieties – the one above has stripes, picotees, flecks and three different colours – intense pink, mid-pink and white. Having all these effects together comes down to the mastery of the breeder, in combining the right parents.

Effectively Satsuki azaleas are like giant 4-dimensional genetic puzzles, with gene switching happening all over the plant at different moments in time to create a completely unique plant! No wonder people get obsessed with them. So finally, if you’ve reached this far through my total geekfest immersion into Satsuki azaleas and their genetics, what does it mean for bonsai?

Well, hopefully the above will help you understand that ‘sporting events’ or spontaneous mutations are what cause many of the different flower colours and patterns seen on Satsukis. And that the type and duration of the sporting event will determine what colour/pattern change happens upstream. You might also be able to work out roughly where these events have taken place on the plant which will allow you to trace where they will go in the future and to either encourage that growth or prune it away, depending on your goal. This information might also help you choose the right variety if you’re buying a Satsuki azalea – especially if you want lots of variation, choose a polyploid/mixaploid or one of its children.

A final question I had for this article was if there was any way to increase epigenetic events to generate more variation on the plant and increase sporting. The answer is yes. “Stresses such as wounding, pruning, viral infection, and tissue culture are all known to induce the movement and/or activity of Transposable Elements, as well as DNA damage”^ref One relevant stress in our case is pruning. Pruning back hard (as is advocated by azalea specialists such as Peter Warren of Surayama Bonsai) should generate variation since it creates lots of new growth (ie. cell division and opportunities for mutation) as well as a certain level of stress. DNA damage is another one, which can be caused by exposure to ultraviolet light.^ref So in theory keeping your azaleas in full sunshine could also encourage sports. However note that as mentioned above pruning back hard might have the effect of removing a sport! If you have a really excellent sport, make sure to note where it is to avoid removing it.

I want to mention one other resource I used for this article – there is a BonsaiNut member – Harunobu – who although not recently active appears to have excellent understanding of the genetics of Satsuki. Some of their advice I found interesting although I could not verify it through other means.

Aside from confirming the theories above about transposons causing sports, Harunobu says that in azaleas “colour is dominant over whiteness, intense colour is dominant over paleness, purple is dominant over red, and solid coloured flowers are dominant over bicolour/variegated flowers”.^ref This means that over time, as the meristem expands and divides, the dominant colours or regions may start to – well – dominate. That implies that assuming you like the less dominant characteristics, you should cultivate them as much as possible and not remove or dilute them. Similarly the more dominant characteristics should be pruned back to avoid them taking over all the subsequent growth. Of course – a new sport can arise spontaneously anyway to give you something new to work with.

Another comment from Harunobu was “a tip that might be marginally useful for when you are trying to ID an azalea by flower colour when it is not actually in flower. If the old leaves go yellow: it is white flowering. If the old leaves go orange; it is pink. If they go red; it is red or purple. And if they have sectors on the leaves, they have sectors on the flowers.”^ref I assume this is because the pigmentation in the flowers also carries through to the leaves and when the chlorophyll dies away the pigments become visible. Theoretically you might be able to use this to visualise sports on a plant without it being in flower. However when I tried this on two of my own Satsuki, I could not see sectors (stripes) on the leaves, and all the old leaves were yellow although the trees are mainly pink. So – see if this works for you.

A weird idea I came across researching this article was the idea that somehow dominant characteristics could ‘take over’ a plant and actually influence branches which previously had been different colours or patterns. I don’t see how characteristics can literally go backwards and influence cells which are behind them in the developmental sequence. And surely new sports would appear? I can’t explain this idea and don’t think it’s true (however, I am open to being wrong). The only way I can see for this to happen is if you don’t prune at all and let everything grow out, then maybe it might be the case that the dominant characteristics take over all of the new growth.