Sunday 18 August 2019

Pink-flowered peas (Alderman x Salmon Flowered)

It was more than a decade ago that I made a cross between Alderman and Salmon Flowered, two heritage peas, with the vague dream of producing a good tasting culinary pea with pink flowers. But as my other three major pea-breeding projects took up more and more of my time, it kind of went by the wayside.

Just to recap: Alderman is a tall, elegant shelling pea with white flowers, introduced in 1891. Its flavour is outstanding, which is probably why it's one of the only Victorian tall peas which is still commonly available today. It's far and away my favourite shelling pea, and while there are a few others which can rival it for flavour when they're young, Alderman stays exquisitely delicious even when the peas are at full size and maturity, and I haven't found anything else that can match it. It has large, well-filled pods on large plants, its only disadvantage (if you consider it a disadvantage) is that it's quite late maturing. Salmon Flowered is a real rarity, whose seed I got from the Heritage Seed Library many years ago. It's an umbellatum-type pea, which means it has a heavily fasciated (thickened) stem with all the flower buds borne in a great clump at the top. The flowers bloom more or less all at once and the pods form in a big clump, sticking out in all directions. There were a few of these varieties around in the 19th century but they're no longer commercially available – and in terms of flavour and yield they can't really compete with modern varieties. But Salmon Flowered (not its real name, which has been lost*) has really beautiful and unusual bicolour pink flowers which I haven't seen in any 'normal' pea at all. The wing petals are a peachy salmon pink and the standard is a very pale blush pink.

*A Swedish heritage pea called Rosakrone is now available from Real Seeds and is very, very similar to Salmon Flowered. I grew some Rosakrone this year to see just how similar it was, and while I'd say it's not absolutely identical, it is similar enough that they're most likely different stocks of the same original variety. The pink flowers are very much the same colour.

So the purpose of hybridising Alderman with Salmon Flowered was to see if I could breed the pink flower trait into a crop of otherwise normal garden peas – using Alderman as the benchmark because of its exceptional flavour.

Back in 2010 I grew out the F1 seeds from my Alderman x Salmon Flowered cross and wrote about it here. The F1 plants all had bicolour purple flowers, which might seem like an absolutely bizarre thing to get from a cross between a white flowered and a pink flowered pea, but actually it's what I'd expected. Purple bicolour is the ancestral default flower colour for peas (not just culinary peas but also the sweet pea Lathyrus odoratus) and the only reason why most garden peas DON'T have purple flowers is that they've all been bred to have a recessive gene which suppresses the production of anthocyanin pigment in the plant. I went into some detail about that in my original post about this cross, so I won't repeat it all here, but suffice to say that Alderman's pure snow white blossoms are not due to any 'white flower' gene as such, they're caused by the presence of this gene which switches off the expression of purple colour so that the flowers are white by default. Salmon Flowered doesn't carry this colour-suppressing gene – it can't do, or it wouldn't have pink flowers – so when you make a cross between a variety which has the pigment-suppressing gene and one which doesn't, the F1 generation will default to the dominant condition – which is for colour to be expressed. Being the dominant ancestral trait, purple flowers prevail. But on a genetic level, the recessive half of the gene pair, which forces the flowers to turn white, is still there and ready to be passed on to a large proportion of the F2 offspring.

I only have a modest sized garden and very limited free time so the focus on my breeding for coloured pods (edible fibreless coloured pods at that) took up more attention and space for the next few years and the Alderman x Salmon Flowered went by the wayside. I had the bag of F2 seed which I produced in 2010, but hadn't sown them. So this 2019 crop was another "back from the dead" miracle story.

I found the F2 seeds in a box, and thought about how nice it would be to work on breeding for flower colour for a change, rather than pesky pod colour. But seeds from 2010 were only fit for the bin, surely? I had to do a quick test though rather than lob them straight out, so they went into a tray of water, alongside some other decrepit sideline projects.

I don't normally soak pea seeds or recommend soaking pea seeds. I was doing it here because it was just a germination test on 9-year-old seeds, and they were not expected to sprout. However, almost all of them did.

This goes against everything I've been taught about pea seeds. 1-2 years is commonly given as a pea seed lifespan. Some people report them lasting a bit longer if you store them in the fridge or the freezer. But mine hadn't had any special storage conditions. They were in plastic bags inside a cat food box on top of my bookcase. The 16 seeds (of quite diverse size and colour, being F2s) in the top right of the photo above became 16 sprightly little plants in a frame in the garden.

I didn't see any loss of vigour from the seed being nine years old. In fact one of the plants turned into the most productive pea I've ever grown, producing large numbers of pods on a multitude of sideshoots. That one will be covered in another blog post later, along with a couple of other interesting things which emerged from this growout.

I labelled the plants individually and numbered them from 1 to 16, and took notes on them as they grew, paying particular attention to the colour and density of the axillary pigmentation (the pinky purple splodges where the leaves join the stem) as I've increasingly noticed a significant correlation between axil splodges and the eventual flower colour and/or pod colour.

Well, I already knew that the F1 had produced purple flowers, so what was I expecting to get from the F2? In a sample size of only 16, I wouldn't expect to get perfect Mendelian ratios for anything, but still there are general trends to look out for. The first would be the colour-suppressing gene discussed above. That should, in theory, turn up as white flowers in about one in four of the F2 plants. In the event, I only had one plant with white flowers (plant no.7), but that's enough to show that the inheritance of that trait is working as expected. In this project, the colour-suppressing gene is not what I want, and I will have to select against it in future years as well, as it will be lurking as a hidden recessive in many of the other plants even though they didn't have white flowers.

Of those which showed coloured flowers, I was expecting a majority to have purple bicolour flowers like the F1, as that's controlled by dominant genes and is the trait most likely to express itself. I know very little about the genetics behind pink flowers in peas, if I'm honest, but I was pretty sure it would be a recessive trait and so I had my fingers crossed that it would turn up in a proportion (maybe one in four) of the non-white plants.

And voilĂ !

It actually turned up in exactly four plants, numbers 3, 6, 10 and 15, which is close to being a Mendelian ratio. These all produced flowers which were a very consistent shade of pink, that is, they were all bicolour pinks with peachy salmon wings and a pale blush pink standard. There was no variation in the expression of the pink colour, other than some differences in how the colour changed over time as the pigment broke down in the fading flowers, where some turned a more dusky, rosy pink than others. But in terms of the essential flower colour, they were identical to one another and identical to the colour of the parent plant, Salmon Flowered, which provided the pink gene. They did all have somewhat larger flowers than Salmon Flowered – flower size was more consistent with that of Alderman – but that's controlled by a different genetic locus.

They were absolutely beautiful and I was thrilled with them.

A second recessive trait which I was expecting to show up in one in four plants was the umbellatum form – which as far as I know is a simple recessive gene which causes the flowers and pods all to bunch together in a cluster on top of a fattened stem. But having never done breeding work with umbellatum-type peas before, I couldn't be sure how this would work out in practice. Again though, it turned out much as predicted. Only two plants, plants 5 and 9, were of the umbellatum type, which is a bit short of a Mendelian ratio but close enough to show the principle of it. There were no intermediate types, all the plants were either umbellatum or non-umbellatum. As far as this project is concerned, the umbellatum trait is not a desirable one and so I'll be having to select against it until all the hidden recessives are eliminated from future generations. The two plants which had this trait were both purple flowered anyway. It may not be a desired trait in this cross but it does look quite spectacular!

All of the pink-flowered plants had what you might call 'normal' form. They produced two large flowers per node, which gave rise to large green pods, much like Alderman. The umbellatum types did tend towards very slightly smaller pods, whereas all the non-umbellatum types had normal size pods, which suggests that smaller pods are a byproduct of the fasciation trait in umbellatum peas. It may simply be that the plant doesn't have enough energy to produce the flowers and pods all together at the same time without compromising on size a little bit. If there was a genetic cause, i.e. a gene in Salmon Flowered which made its pods smaller (as they ARE quite small) then I would expect that trait to segregate randomly through the F2 plants – but it didn't.

Axillary pigmentation is another interesting one. It occurs in all (in my experience) peas with coloured flowers. A purple-pink splodge at the point where leaf meets stem is invariably seen in a plant with purple bicolour flowers. The purple blotch is made from anthocyanin pigment, so it's subject to the action of the same colour-suppressing gene which causes white flowers. A pea with white flowers will have no pigment in the axils at all, because this gene switches off all production of anthocyanin pigment throughout the whole plant. So in a cross between a white-flowered and a coloured-flowered pea, like this one, the presence or absence of colour in the leaf axils is a very early indicator of whether the plant will eventually have white or coloured flowers. The axil pigment usually shows up quite early in the seedling stage, when they've produced their first couple of sets of true leaves.

In the case of this hybrid, the Salmon Flowered parent variety has an unusual kind of axillary pigmentation. It's lighter than the usual type – in fact it's a dusky rose pink, and very soft and subtle. I was thinking it's probably not a coincidence that a variety with unusual pink flowers also has unusual pink axillary pigmentation – there's probably a meaningful correlation between the two. So I was watching my F2 seedlings to see if there was any sign of this correlation, and there was.

The photo shows the subtle dusky pink pigment in the axil of ASF 06, one of the pink-flowered phenotypes. All the other pink-flowered ones had this as well, while the purple-flowered ones had the more normal blotch of purple colour in the leaf axils, and the white-flowered plant had none at all, as expected.

This is bloody useful, actually. If the colour in the leaf axil is a reliable indicator of flower colour, which it does seem to be, subject to errors of interpretation when the shade is a bit ambiguous, then it means you can identify the flower colour a good month or two before they flower.

Right then, so the appearance of pink flowers in roughly a quarter of the F2 plants suggests a fairly straightforward recessive gene at work. I'm sure there is some info out there somewhere about how this works, but the details of which genes control which traits are often buried in papers in academic journals which I don't readily have access to – and even when I do get hold of them, I struggle to make sense of the scientific jargon and it makes my brain hurt. So in my layperson ignorance I'm going to make a speculative guess about what's happening with this pink flower business.

I think that a pink-flowered pea is essentially a purple-flowered pea which has come under the influence of a modifier gene – probably just a single, recessive modifier gene. I think this modifier gene acts on the chemical makeup of the anthocyanin pigment, suppressing the production of blue pigment while leaving red pigment unaffected, so that the flower comes out pink instead of purple.

There are two reasons why I think that. The first is to do with the axillary pigmentation. If there was a gene specifically coding for pink flowers, I can't see why it would affect the colour of the axils as well. But clearly it does, because pink flowers and pink axils go together. Which suggests a modifier gene having a blanket effect on anthocyanin production throughout the whole plant.

The second reason I think this is the case is because of a study which has been done on sweet peas, which are a different genus from edible peas but have a lot in common with them. As I mentioned earlier, the default ancestral colour for sweet peas is a purple bicolour, but some time in the 18th century a mutation occurred which gave us the lovely pink-and-white bicolour known as Painted Lady, which is still widely available today. A study was published in 2017 in the Canadian Journal of Plant Science on the genetic basis of this mutation, and although I don't have access to the paper itself there was enough information in the abstract to tell me what I needed to know: a single base pair mutation means that the flower is lacking the blue pigment known as delphinidin, which is one of the anthocyanins which make up the purple colour in sweet peas. In the absence of delphinidin, the flower becomes pink. In simple terms, if you imagine that the colour of the purple flower is made from layers of translucent blue and pink, removing the blue layer leaves you with just pink.

So that's what I think is probably happening in Pisum sativum as well. If you look at the pink flowers in my F2 plants, they are all bicolours. They are all essentially the same, there's no variation in the colouring. So it seems quite plausible that they are meant to be the default purple bicolours, and that a recessive gene has come along and deleted the production of delphinidin (or whatever blue pigment they're supposed to have) and this salmon-pink bicolour is the result.

As I said, this is just my speculation! I'm sure there are people out there who know more about it than me.

This is turning into a very long post, but I have learned such a lot from growing these sixteen F2 plants!

So let's finish up with a bit about the pods and peas. This project is not seeking to produce edible pods or coloured pods: both parents are green-podded shelling peas, and all the offspring are green-podded shelling peas as well. My aim was to get the kind of big, plump green pods and fat peas found in Alderman, and not so much of the small pods and small peas of Salmon Flowered. In this, the F2 generation has given me what I wanted, because all the non-umbellatum type plants produced pretty good pods and most had good sized peas.

I did some taste tests as well. I was hoping to get as close as possible to the sweet and complex flavour of Alderman and not so much of the pleasant but rather mealy taste of Salmon Flowered. In this I was also very lucky. I tasted three out of the four pink-flowered phenotypes and they all had very good tasting peas, with ASF 06 being the best. Unfortunately ASF 10, which was a lovely plant with beautiful flowers, died prematurely after getting its main stem damaged in a storm. It had only just begun setting pods at that stage and the peas inside were still very immature. I thought I had nothing to lose by leaving the pods on the plant as long as possible in the hope that they would use the residual energy of the plant to carry on maturing a bit. And they did. When I finally harvested the pods, the peas were still quite small but they look like they might, just might, be mature enough to germinate. I didn't eat any of these – I wanted to conserve as many as I possibly could.

Here are the seeds from the four pink-flowered plants after being harvested and dried. As you can see there are a few differences between them. They all have subtle purple speckles on them, except for the salvaged seeds of ASF 10, which were not fully mature. ASF 15 has more of a green colour to its seeds, while ASF 06 and ASF 15 have a mixture of green and tan. The tan seeds are a trait inherited from Salmon Flowered, which also seems to be related, albeit loosely, to the pink flower trait. You may also notice that ASF 06 in particular has some variation between wrinkled seeds and rounded, dimpled seeds. The wrinkled ones are a rule-of-thumb indication of sweetness in peas, because sugar shrinks more than starch does. I probably won't select out the wrinkled ones next year though, I'll grow a bit of both, but I might possibly separate them out into different halves of the seed tray so that I can keep track of whether there's any correlation between wrinkled seeds and sweeter flavour.

That's it now, until next year when I can grow the F3. I would expect all four of the pink-flowered peas to breed true for the pink colour, as it's a recessive trait and they must be homozygous for that trait. It's likely that some of the purple-flowered phenotypes are heterozygous for pink flowers, and will produce a few of them in their offspring – so I will probably grow out some of the best of those to see if I can get some more pinks. But either way, I'm extremely pleased with what has come out of this F2 crop and I'm feeling quite optimistic about the prospect of getting peas with pink blossoms and lovely flavour, within a couple more years.

Friday 9 August 2019

Red-podded pea update: the F6 crop and beyond

Ever since I first had a red-podded pea turn up (completely unexpectedly) in a breeding project in 2008, I’ve had a bit of a frustrating time trying to get it into a form which is worth releasing into the world. Not that I’ve been trying continually, because I haven’t been able to maintain the garden every year during the last few years, but it has had me scratching my head trying to come up with a way of producing edible pods with this rich crimson colour. Finally, in 2019, I had a significant breakthrough.

The original red-podded pea plant, numbered GSC 15 in my breeding project, was an F2 from a cross between Golden Sweet and Carruthers’ Purple Podded. The purpose of the cross was to breed a new purple mangetout pea for Ben Gabel at Real Seeds, but neither Ben or I had any idea that crossing a yellow pea with a purple one would result in a red colour. It was only this one plant which produced red pods, and it happened because it was a yellow-podded pea with a solid layer of purple pigment on top. The pigment is translucent, so the yellow shining through from under the purple creates a beautiful glowing red, like stained glass.

This breeding line, which is currently somewhere around an F6 or an F7, is now a stable, true-breeding variety. It’s reliable and consistent for all its traits, except one: it still turns up a small minority of offspring whose pods are not fully red. It doesn’t tend to throw any plain yellows any more, at least not in the growouts I’ve done, but it will produce some semi-red plants which look like the red pigment has just been sprayed around the edge, and some which have a jagged, flamed pattern of solid red on a yellow background. I think these are caused when one or other of the two dominant genes for purple pods is unable to express itself. But WHY it can’t express itself I don’t really know. All I know is that it happens often enough to be a thing, and on that basis I don’t think the semi-reds will cease appearing no matter how many years are spent trying to rogue them out. They just have to be accepted as the nature of the beast, and besides, I had one semi-red this year which I liked enough to want to keep it, if I can persuade it to become a variety in its own right.

One of the semi-red phenotypes. I believe it's probably caused by co-dominance between purple and non-purple genes. Because of that, no amount of 'roguing out' will eradicate these oddities, which show up in a small minority of offspring.

But anyway, having reached the F6 generation this stable red-podded variety still has a problem, and the problem is not going to go away. It doesn’t taste good, either as a shelling pea or a mangetout. It has a gristly layer of fibre inside the pod which makes it unsuitable for eating even when it’s small, and if you shell out the peas to eat fresh they are starchy as hell, with a bitter aftertaste. The genes controlling flavour in peas are complicated, but the situation here is very simple – it is true-breeding for crap flavour.

It does, however, look truly stunning.

It’s possible that it will find a use as a soup pea, maybe becoming edible when it’s dried and then boiled for ages. The jury’s still out on that one. But in its current form it’s not suitable for unleashing on the world, and will have to be considered a breeding line. To that end, I grew out a batch of seeds in 2019 which had been in storage for 4 or 5 years, just to replenish my stock of seed. I saved seeds from the best ones (which was most of them, as it’s certainly a fine-looking variety) and saved the rest for taste-testing in the kitchen, some time when I can be arsed to boil up a vat of pease pudding.

So, what to do? I need red-podded peas with edible pods. And ones which taste nice.

When I did my initial report on the edible-pods frustration nearly ten years ago, I said there were five things I could do about it. Here they are again, with the results of my efforts.

1. Stand in the middle of the garden and shout "BOLLOCKS!" in a really loud voice. Yep, done that.

2. Grow out the remaining F2 seeds (about 20 left) in the hope of finding another red-podded phenotype but with edible pods. This didn’t provide what I was looking for, but I did find an extremely nice purple mangetout (the whole point of the breeding project in the first place) which I selected and re-selected in the F3 and named Barcarolle.

3. Grow out the remaining F1 seeds to get more F2 seed. I tried this, as part of a belt and braces approach. Growing out more of the F1 took a year but it yielded plenty of seed. I grew a batch of F2s in 2019 and some of them were quite nice but no, I didn’t get the tasty-edible-red-podder I wanted.

4. Switch to another line of F3 seed, even though none of them are proper red-podders. This was always a good bet, because some of the necessary traits for red mangetout pods are recessive genes which may be hiding in the DNA of a plant which isn’t showing it. For example, yellow is the base colour for red pods and you can’t get red pods without the recessive ‘golden pod’ gene known as gp. The laws of probability suggest that two out of three of the F3 lines should carry the gp gene, even though they don’t themselves have yellow pods. Also, the two genes which produce edible (fibreless) pods are both recessive, so a goodly proportion of inedble-podded lines could produce edible-podded offspring.

Although only one of my original batch had properly red pods, I did get a couple of peachy-coloured or semi-red mangetout lines. My efforts revolved around these, as they were so “nearly there” – they had sweet-flavoured edible pods and a beautiful fiery blush. But sadly the fully red colour didn’t show up in their offspring either.

But although I didn’t get anywhere with the ‘peachy’ F3 lines, it was another F3 line which gave me my breakthrough in 2019, in a completely unexpected place: my Barcarolle purple mangetouts.

5. Grow more of the original red-podded F2 and cross it with something else. I tried a lot of this as well, but it's the most time-consuming option and I don't have many results yet, so I’ll report on it another time.

Barcarolle purple mangetout
Right then, this is what happened with the breakthrough. Alongside my growout of red-podded peas this year, I planted 16 seeds of my newly developed purple mangetout. Sixteen plants is not very many, but I’m doing all my breeding work on a very small scale in my back garden, so that’s all I have room for.

Barcarolle is a very promising variety which originated from the same cross as the red podders – Golden Sweet x Carruthers’ Purple Podded. I selected an F2 plant a few years ago which had beautiful dark velvety purple pods – a really solid dark purple, which is not something you get all that often in peas. As the peas start to swell inside the pod, they bulge through the pod wall and look like a row of little blackcurrants. The reason this happens is because there’s no fibrous membrane inside the pod, like you’d get with a shelling pea. It takes two recessive genes to get the pod walls completely fibreless like this, and it’s a desirable trait to have because it means the pods are fully edible at all stages of development – they don’t go coarse and gristly. I also got lucky with the flavour, as the pods tasted sweet and juicy and were pretty much stringless. Again, this is an uncommon thing to find in a purple-podded pea, because the purple colour is often associated with a slightly bitter flavour. It’s not that the pigment itself is bitter tasting, but there’s some kind of gene linkage going on which means that the purple pod genes are usually inherited alongside some less-desirable flavour genes. This was by far the best purple podded pea I’d ever tasted so I was well chuffed with this plant.

The following year I grew out the F3 generation, and to my delight, it came almost completely true to type. I had expected a few unwanted recessives to show up, but they didn’t. The F3 plants were as uniformly purple and bobbly and sweet and juicy as the previous year’s plant. I don’t always name my breeding projects at this relatively early stage, but it seemed like it was going to be quick and easy to make it into a stabilised variety so I gave it the name Barcarolle. (In case you’re wondering, I play the piano and I’m a big fan of Chopin.)

So then I grew out this batch of F4 plants in 2019, and, well, to say they began to segregate is an understatement. They segregated like buggery. Fortunately they all had the bobbly edible pods (a recessive feature), but they varied in size, flavour, number of pods per node, and most of all they varied for pod colour. In my batch of 16 plants, I ended up with four different pod colours – purple, green, yellow and red!

My 2019 crop of Barcarolle F4 purple mangetouts, segregating for four different pod colours ... aaargh!

I must admit I don’t quite know why this happened. It doesn’t make sense for a variety which is true-to-type in the F3 to suddenly start segregating like mad in the F4. The most likely explanation, I think, is that I happened to grow out (by chance) only seeds which were heterozygous for pod colour in the F3 generation, so they all showed the dominant purple colour but they still had the recessive genes squirrelled away in their genome. Given that I work with such small sample sizes, this is entirely possible. But anyway, who cares – the thing that matters is that I ended up with no less than four red-podded plants, all of them with lovely juicy edible pods. This is such a holy grail for me, some 13 years after I made the original cross, that I spent an awful lot of time just standing in the garden gawping at it.

The purple pods on the left are what Barcarolle is supposed to look like, but it came out with these red podded variants as well.

In some ways I probably shouldn’t be surprised to be getting red pods segregating out from Barcarolle. The only difference between red pods and purple pods is the base colour of the pod – green for purple-podders and yellow for red-podders. Barcarolle is derived from a yellow-podded parent, so the presence of the recessive gp (golden pod) gene shouldn’t be a surprise. I might have saved myself some grief if I’d tried looking for it here earlier.

As well as being exactly what I’ve been looking for for the last decade, the Barcarolle red-podders have some other advantages over my original red-podded pea. The red colour itself is a bit brighter – more scarlet than crimson. Both colours are equally nice when the pods are young, but one slight flaw in the original variety is that the red tends to darken as the pod matures, until they start to look purple rather than red. The reason for this is not a change in the pigment itself, it’s because of the natural darkening of the base colour of the pod. As yellow pods age, they tend to go more green – a trait seen in the parent variety Golden Sweet – and a greener pod means the pigment appears more purple. In the Barcarolle-derived red-podders, the pods stay yellow right through to maturity, and so the pigment continues to look red for a lot longer. I don’t know what subtle genes are responsible for this difference, but it was consistent across all four of the plants in this batch.

The best of them was the plant known as BRC 16 (above), which was hugely productive. It grew to around 7ft tall and not only produced two pods per node, it also developed fully productive sideshoots – a rare thing in peas – and bore good quality pods on those as well. It also did very well in the looks department, having beautiful colour-changing flowers borne on stems with a lot of bright scarlet colour and pretty pink mottling on the calyx. BRC 14 and BRC 06 were both very similar but not quite as vigorous. The best one for flavour was BRC 11 which had a fantastic juicy raw flavour, like apples! 

I do want to develop a red-podded pea which tastes good raw, because the colour is spoiled by cooking. With any red-podded or purple-podded pea, if you cook it any way other than the lightest steaming, the colour just disappears into the cooking water. No amount of careful breeding is going to resolve this, because it’s in the nature of the pigment itself – anthocyanin, which is water-soluble, and that’s just the way it is. So the only thing a plant breeder can do to get around this is to develop varieties which taste so good raw you won’t need to cook them.

Now, of course, I will have to wait until next year before I can grow out the offspring and see how they look. But I’m very optimistic. And I still have the purple-podded Barcarolle line which, once I’ve stabilised it, will be a wonderful variety in its own right. The red-podded line is not named yet, but I have another year or two to think about it.

The Barcarolle-derived red peas have beautiful bicolour flowers which turn blue as they fade, like these on BRC 16.