How to breed your own potatoes

Breeding new potato varieties is easy. You can hand-pollinate potato flowers in far less time than it'll take you to read this article, but I'm going to attempt a reasonably thorough explanation, so I hope you find it helpful.

Potato breeding is done through sexual reproduction, i.e. pollinating flowers to produce berries which contain true seeds (TPS).  Normally when you plant potatoes you propagate them from tubers, confusingly called seed potatoes but which are not actually seeds, but root cuttings. You can't cross tubers. They can only reproduce themselves as they are. Occasionally a plant may produce a spontaneous mutation but it doesn't happen often enough to be useful as a breeding method. Flowers are the way to go, because they give you the option to combine and reshuffle genes from the parent varieties of your choice.

There's a lot to be grateful for in the anatomy of a potato flower. Hand-pollinating them is very easy. The flowers are large and easy to work with, and the individual parts are easy to manipulate.

What's not so easy is making careful plans and predictions for what you might get out of it, and that's because potatoes are tetraploid. If you've no idea what I'm talking about then have a look at my previous post about TPS for a simplified explanation. To give a one-sentence summary: a tetraploid has double the amount of genetic material that a normal (diploid) organism has, which is a bit like inheriting traits from four parents rather than two. Tetraploids are a quirk of nature but in potatoes they are a very successful one, and the vast majority of cultivated potatoes in Europe and North America are tetraploid.

You may still come across the occasional diploid. Mayan Gold and its associated varieties are diploid, and those who are growing TPS from Tom Wagner may have a few diploid lines from him. Diploid potatoes can be recognised by a tendency to have smaller and less fleshy leaves, but the most distinctive feature is the berry. A diploid potato berry has a distinctively pointed end, kind of strawberry shaped, while tetraploid berries are more rounded and tomato-like. If you're feeling experimental you can try crossing a diploid with a tetraploid. At best you will only get a few viable seeds out of it, but it's a brilliant way of introducing new diversity into potatoes. At some point soon I will give it a whole article of its own, as it's too elaborate a subject to go into here.


Potato berries (these are tetraploid ones) in development.

However, assuming the potatoes you want to cross are tetraploid, since most of them are, it's very difficult to predict what the resulting offspring will be like because of the genetic variability involved. With tetraploids, the convenient order of the Mendelian ratio is thrown out the window and replaced by something more akin to a gene tombola. F1 hybrids are not uniform as they are in most other types of breeding project. If you grew a lot of offspring from your cross you'd find that many traits show continuous degrees of variation through the population, rather than segregating into Mendel's either/or groups … which happens because there are so many different ways the alleles can arrange themselves. To quote a research paper by Scotland's premier spudmeisters, Meyer et al (1998) "[Tetraploid] inheritance implies the random pairing of four homologous chromosomes at meiosis, and in a highly heterozygous outbreeding species results in a large number of possible allelic combinations at a single locus. In the most extreme case, eight different alleles could segregate independently in a population, resulting in 36 possible genotypic classes in the progeny." In other words, potatoes naturally have a mixed up genepool (from outbreeding) and when they pollinate and set seed those alleles can arrange themselves in any order - with each different combination having a unique effect on how that trait is expressed. And we're just talking about an individual locus here … the same is happening at every other locus throughout the whole genome. Yowza!

In short, tetraploids are complex and contain a lot of genetic material which can be immensely variable. Scientists doing genetic research on potatoes often choose to work with diploid lines instead, because tetraploids make such a muddle of their data it's hard to interpret anything.

So where does that leave you as a home gardener or small-scale farmer wanting to develop your own potato varieties? It leaves you in a position where you may as well have fun, experiment, use your imagination, be creative. As the results can't easily be predicted, you don't actually need to know anything about genetics. Think more along the lines of what you might get if you cross this colour with that colour, or this flavour with that shape - and then be prepared to be surprised!

One piece of misinformation I see spread all over the internet is a belief that you won't get anything worthwhile out of a home-made potato hybrid because producing just one good variety takes thousands of plants and many many years. This myth has arisen from under the slow-grinding wheels of the potato industry, which does work like that. Sure, if you want to breed a variety which will be listed in all the catalogues and sold in Tesco's and will make you rich from the royalties, your chances are very slim. The selection criteria for commercial potato varieties are immensely restrictive - and largely at odds with what most gardeners would want. Commercial breeders may well churn through (and reject) 200,000 seedlings to find one with commercial potential, then spend the next eight years doing field trials with it before it's ready for release. But don't let that put you off. You can breed a good variety within two years - easily. The majority of your home-made potatoes will be worthwhile, at least decent enough to eat and enjoy and feel proud of. A few will be exciting and wonderful. Even if you're only working with a very small patch of garden, you will almost certainly get some tubers that are worth saving and growing on next year.

The beautiful thing about potatoes is that it only takes two seasons to get a completely stable new variety. So it's actually quicker than most other vegetables. You make a pollination the first year and produce the F1 seeds, which are all unique individuals because of the genetic diversity. The second year you grow plants from those seeds and they make tubers. If you like the tubers, you simply propagate them by saving and replanting them. As the tubers are basically root cuttings of the parent plant (clones) they are genetically identical. There's no arsing about trying to make F2 and F3 hybrids (unless you want to) or years spent roguing out unwanted recessives. Once you've got something interesting, it's instantly a new variety.


Hybrid potato grown from TPS. This is one of Tom Wagner's hybrids, an F3 of Pirampo x Khuchi Akita. The parent varieties are Bolivian landraces, and are diploid.

If you've dipped a toe into plant breeding before, you'll know that plants tend to be either inbreeders or outbreeders - though that's more of a sliding scale than a polar absolute. Potatoes are really a bit of both. The natural status of potato is outbreeder. The various (diploid) landrace species from which cultivated potatoes are derived have a self-incompatibility mechanism which prevents them from pollinating themselves. The majority of diploid varieties are self-incompatible, although there are exceptions. This forces them to hybridise and mix their genes up in every generation, hence the wondrous diversity found among diploid landraces. However, when potatoes went tetraploid the compatibility barrier got screwed up somewhat. Many tetraploid potatoes have sterile pollen which can't fertilise anything at all, but others can fertilise themselves as well as each other. So they're designed to be outbreeders, but in practice a lot of flowers simply get knocked up by their own pollen.

Which gives you a choice: you can make hybrid seeds by crossing two different varieties, or you can make self-pollinated seeds which are the product of just that one variety.


Choosing parents: hybrid or OP?
Both are worth experimenting with, but for different reasons.

When a potato plant sets berries naturally without your intervention, it's most likely that it self-pollinated, but it may also have made hybrids with other potatoes flowering nearby, and you may have a mixture of selfed and hybrid seeds in the same berry. This is called open pollination (OP) … and the results are basically pot luck.

Making a deliberate hybrid is the most usual way to breed a new variety, as it introduces a lot more diversity. The basic method is to emasculate the flower to stop it from pollinating itself (which you don't even need to do if it's one of the many varieties with sterile pollen) and fertilise the female part of the flower with some pollen from a different variety. The offspring will be very varied, but that's the fun part and you should also get some hybrid vigour which makes for healthy and abundant plants. The only problem with this is that so many varieties of potato are poor berry setters, so not all varieties can be hybridised.

If you want to grow seeds from one specific variety it can be as simple as saving naturally pollinated berries from it, but if you want to be sure of getting self-pollinated seeds, it's easy enough to do (as long as it's one of the fertile varieties). Just dab a flower with pollen scraped from its own anthers, or other flowers on the same plant, or from other plants of the same variety. Bear in mind though that you will not get a true-breeding offspring of the parent variety by doing so. As potatoes are very heterozygous and have four lots of genetic material to throw around with cheerful abandon, even when they're self-pollinated they segregate into many different phenotypes. If you grow self-pollinated seed from Salad Blue, for example, you will not end up with a lot of spuds which look like Salad Blue. You will get varying shades of blue flesh, some much lighter than the original, some darker, and a few with pinky skin. If you grow selfed seeds from Congo, another blue variety, you may end up with a baffling range of purples, pinks and pure snowy whites, with considerable variation in tuber shape. What's happening is that all the genetic material which has been funnelled into the variety from various ancestors is segregating. Recessive traits emerge which weren't apparent in the variety you started with. If you grow enough self-pollinated offspring, you can start to build up a picture of the variety's pedigree, as many of the ancestral characteristics magically come back to life. So it can be a really fascinating thing to do.


A seedling grown from self-pollinated seed of Salad Blue. It's the only plant in the batch which has this striking black tinged foliage and black stems. I'm hoping it'll produce some dark tubers to go with it.

It's important to remember though that nature designed potatoes to be outbreeders, and if they self-pollinate they may show some degree of inbreeding depression. Only a bit though. As most spuds have such a rich and diverse genepool they can get away with a certain amount of inbreeding, but you may find self-pollinated seeds grow less vigorously than hybrids. That's not a problem and shouldn't put you off trying self-pollinated seed … but it's better to sow a few more than you need and then select the seedlings which show the most vigour and whoomph.


Variety differences
Differences in the fertility of individual varieties will most likely dictate what crosses you make, and how you do them. Over many years of being propagated by tubers, cultivated potatoes have moved away from the idea of flowering and producing seeds, and many of them can't be bothered to do it any more. Ironically for such a naturally variable and heterozygous plant, a historic lack of genetic diversity is thought to be the cause of the potato's fertility issues. The vast majority of modern cultivated potatoes are descended from one single Chilean spud, which had what is known as T-type cytoplasm, a genetic predisposition to making offspring with infertile pollen. Over the years many of these semi-infertile lines have been selected deliberately, as the male-sterility makes the process of hybridising them much easier. Consequently an awful lot of modern spuds have infertile pollen, and some are female-sterile too. Some can't be arsed to flower at all, and just dump their buds as soon as they appear. There are things you can do to force a reluctant variety to produce flowers, but it's a lot of hassle which I won't go into here, and from the point of view of future breeding work it makes more sense to choose varieties that at least show some willingness to come up with the goods.

I wish I could give a simple list of which variety does what, but I only know about the ones I've grown and observed myself, and it can vary from garden to garden anyway. Different countries have different varieties - import restrictions have affected exchange of material - so the ones I work with in the UK may not be available to people in the USA (just as most popular US varieties are strangers to me). So you will have to experiment with whatever you have available. As far as I can see, varieties fall roughly into four categories.

Some potatoes are very fertile and make excellent male or female parent varieties. Salad Blue is the Cassanova of the potato world - it only has to look at another potato and a berry starts swelling. You can usually tell a fertile variety because it naturally sets its own berries in profusion. Desirée is another very fertile one, and so is Mayan Gold, although the latter is a diploid so it needs to find the right kind of partner, or get lucky mating with a tetraploid.

Most cultivars fall into the male-sterile or almost-male-sterile category - these are the ones which flower happily enough but don't tend to set berries. Highland Burgundy Red is a good example of this, as is British Queen. It gamely produces a mass of dainty little flowers but in years of growing it I've never had a single berry. Give it a dab of pollen from a fertile variety though, and it sets berries very readily. So it makes an extremely good female parent. The advantage of male-sterile varieties is that you don't have to emasculate them, which makes it much quicker and easier to hand-pollinate them. Some varieties which appear to be male-sterile may actually be female-sterile. So it's worth trying the pollen on another variety to see if it will take. The disadvantage of using these partially sterile varieties is that it perpetuates the poor fertility of potatoes. If you want to do Solanum tuberosum a real favour in your breeding projects, select the progeny for good berry production. Because good berry production is what will keep its genetic heritage alive, as well as enabling some much needed new diversity to come in.

Then you have what you might call the awkward buggers category. These include Pink Fir Apple (syn. Rose Finn Apple) which not only has male sterility issues, it often can't be bothered to set a berry even when it's given fertile pollen. What usually happens is that the flower opens happily enough and you carefully pollinate it two or three times and on the third day the whole bloody thing drops off. Or worse, it starts to set a berry and then it falls off before it's mature. It pays to try again though, because there's a good chance that one of the pollinations will take eventually, when the plant is in the right mood and the planets are in the right alignment or there's an 'r' in the month. It's a pain in the backside to have to keep pollinating more flowers, but bearing in mind that each berry can easily produce 100 seeds or more, it only takes one successful pollination to give you loads of future breeding material - so it's worth persevering. Again, with a variety like this you don't need to waste time emasculating. I just go through the whole crop each day dabbing fertile pollen on every stigma I can find and saying "come on, set a bloody berry you sod!"

And finally you have the total refuseniks. There is a wonderful Victorian potato called Witch Hill which is reputed to be one of the best flavoured potatoes around - it is truly delicious. I would love to use it in breeding work. But every year the flower buds appear, and just as they're starting to look promising they drop off. All of them. Little dessicated posies cast to the ground. Now, unless it changes its mind, I cannot breed from it. If a variety won't flower, there is no breeding possibility, it's as simple as that. I could grow fields of the stuff and hope for a spontaneous somatic mutation, but that may never happen. Witch Hill is a genetic dead end. This is why breeders like Tom Wagner select breeding lines from varieties which are good berry setters. If a variety won't flower or won't set berries, it has no future.

The annoying thing is, Witch Hill did flower for me a couple of times when I first got it (it came to me as a laboratory-grown microplant) but I hadn't got into potato breeding at that time so I didn't think to make any crosses with it. D'oh!


Hybridising potatoes: the practical bit
Let's be grateful for small mercies: potato flowers are nice and simple and easy to work with. They are 'perfect' flowers which contain both male and female parts.


Anatomy of a potato flower. Each of the anthers is a double sac, both halves containing pollen. When the anthers mature they develop little holes in the ends (like a salt cellar) and the pollen falls out onto the stigma.

The Mayan Gold blossom shown above is fairly typical, but there are variety differences in the exact shape of the flower. Some have a long style where the stigma protrudes some way out of the flower (be aware that a sticky-outy stigma has more chance of being cross-pollinated by passing insects than one where the stigma is hidden away). Some produce a neat little fused anther cone, others produce a rather grotty collection of misshapen anthers which don't hold together properly. Some (like Pink Fir Apple) do weird things where anthers and petals morph into one another. None of this matters - the principle is the same. You'll get to know the individual character of the flowers in your own garden as you work with them.

In order to control what pollen fertilises the flower, you have to stop the flower from fertilising itself, so that means removing the male parts of the flower before they mature. As I explained above, with some varieties you don't need to do this - if the variety produces sterile pollen or none at all, you can save yourself the trouble. The instructions shown here are for if you have a fertile variety or want to be sure of getting hybrid rather than selfed seed.

Potato flowers are produced in cymes - bunches of flowers which open consecutively, 2 or 3 at a time. The flowers last two to four days but tend to close up in late afternoon. The anthers develop holes in their tips when they're ready to dehisce, though they're not very glamorous - in fact they look more like some insect has had a go at them. Potato pollen is white, powdery and very fine. The stigma is receptive for about 2 days and the period of pollen shedding also lasts about 2 days. Fortunately for the garden dabbler, the female part tends to become receptive just before the pollen starts to shed, so you have a window of opportunity to intervene.

The best time for hand-pollination is in the morning when pollen is most abundant, and when the temperature is fairly cool. But I wouldn't worry too much about this, it works at other times too.

If you're using a variety with infertile pollen, or you aren't bothered about the chance of a few self-pollinated seeds, you can skip steps 2 to 4.


Step 1: Having chosen the variety you want to use as the female parent, find a blossom at the right stage. Potato pollen can be shed quite early, before the flower opens, so emasculation has to be done while it's still at the bud stage. What you're looking for is a nearly-ready bud where the calyx (outer green bit) has started to open but the petals are still shut. This is a variety with a sticky-outy stigma, but with many varieties it will still be hidden inside the petals. Doesn't matter either way, although a sticky-outy like this inevitably carries a small risk of picking up stray pollen from elsewhere.

You may notice a strand of mauve wool poking out underneath. I tied this around the stem of the flower (or in this instance the whole cyme, as I'm going to hand-pollinate all of them) as a marker, so I can be sure I know which ones I've hand-pollinated. I use a different colour of wool to indicate different pollen fathers.



Step 2: Peel back the petals and you'll find the anthers inside. They are still immature at this stage - with no holes in the ends. If they do have holes and are shedding pollen, try a slightly younger bud instead!



Step 3: Using a blunt scalpel blade, tweezers or similar, pull/scrape the anthers off, being very careful not to damage the style - the central stalk with the stigma at the end.



Step 4: After removing all the anthers you're left with a denuded female part, ready to be pollinated with the pollen of your choice.



Step 5: Next find the flower you want to use as the male parent. Choose a blossom which is newly opened, as those are the ones most likely to have a good pollen stash (the ends of the anthers should be open at this stage). Pull off a single anther using tweezers/scalpel/fingers.



Step 6: When you turn the anther over you'll see it has a seam down the back, separating the two pollen sacs. Additionally, each individual sac has a little slit down its centre. Carefully slip the tip of a blunt scalpel blade through the slit and slide it along. Note that the slit should be open so you can insert the blade freely ... you want to avoid cutting into the anther if you can. 

If there is pollen inside, you will see it on the tip of the blade. It's a very fine white powder. If you don't see any white powder, try another anther from a different flower. You don't have to collect all the pollen at once ... just scrape out enough to dab on the female flower, and use the rest for more pollinations. (When the first sac is empty you can do the same with the other side. You can often pollinate ten or a dozen flowers from the pollen in a ripe anther like this.)



Step 7: Armed with your pollen-tipped scalpel, go back to the bud you just emasculated and dab the pollen powder onto the stigma - which is the knobbly-bobbly thing at the end. The stigma is mildly sticky when it's receptive, so you should find the pollen grains sticking to it quite readily. No need to make a song and dance with it - just a gentle dabbing so as not to risk damaging the stigma.



Step 8: The next day, go back to the same flower and pollinate it again with pollen from another fresh anther. The stigma remains receptive for around two days in total but you don't know exactly when that is, so for best results give it a pollen dab on three consecutive days. You'll notice that the petals have opened on this flower now, although it looks a bit weird as it has no anthers. Once the petals have closed and wilted a bit, you can assume it's no longer receptive.



Step 9: The berry starts to form. Yay!


From this point on, patience is the order of the day. Potato berries seem to mature painfully slowly. Try to resist the temptation to prod and poke them, you don't want them to fall off as they dangle clumsily on their alarmingly scrawny stalks. After about four weeks you need to watch for them dropping off naturally. Ideally, tie a little cloth or paper bag over them at this stage so that they are caught safely if they drop. Alternatively, be very vigilant, and ready to rummage about on the ground if you notice them suddenly go AWOL. Fortunately they don't taste nice enough for animals to be interested in them, at least not in the way of UK garden wildlife, so they're unlikely to be carried off, but you don't want to chance it.

One key factor about potato berries, significantly unlike tomatoes (to which they're closely related), is that the seeds carry on developing even after they've detached from the plant, and the berries will stay firm for months. This has several advantages. For one thing it takes the pressure off you to extract the seed from them ... you can leave them until you've got the time and inclination, even weeks or months down the line. Secondly it means that all is not lost if the berry is dropped too early or the plant dies prematurely. This is very significant in the light of the blight problems we are all besieged with. The plants can be struck down by blight, wither and rot - and the berries will still survive. Save the berries and let them mature, and as long as you clean them thoroughly they will yield perfectly healthy seed.

The topic of cleaning and processing TPS from the berries is to be the subject of a separate post. In the mean time, go out and make some berries!

With thanks to Tom Wagner and friends at the TaterMater forum for advice and suggestions (any errors are entirely my own responsibility).

Gene genie

Genes are the most wondrous things. What flower colour do you suppose I got when I crossed this white blossomed beauty ...




... with this gorgeousness of rosy-pink?




Well, I got this:



It's my F1 hybrid of Alderman x Salmon Flowered. It has just blossomed and produced flowers of the 'standard' mauve and maroon two-tone, the colour you get in field peas. Neither parent shows this colouring. Alderman is a delicious late Victorian pea with the snowiest of snow white blossoms. Its partner in this liaison was the strange umbellatum variety Salmon Flowered, supplied by the Heritage Seed Library, which is the only pea in my collection to have the particularly lovely two-tone pale pink and salmon pink flowers.

I don't know all that much about the genetics of pink flowers, but I know there are at least a couple of genes which can produce them. I don't know which one(s) are present in Salmon Flowered, but the ones I'm aware of are both recessive. The white flowers of Alderman are also recessive. (Yes, despite the fact that most garden peas have white flowers, it is a recessive trait.) When you cross two varieties which both have different recessive alleles controlling the same trait, some weirdness can show up in the hybrid.

So am I surprised that my hybrid came out purple? Actually, no, not really. I did wonder if it might. And here's what I think the explanation is.

I've mentioned many times in my pea genetics posts (for those who don't glaze over while reading them) the existence of a gene called A. That's short for anthocyanin. This particular gene is an on-off switch which controls the production of anthocyanin, the pigment responsible for all pink and purple colouration in peas. The function of A is that simple - on or off. There are other genes which control which part of the plant the colour is expressed in ... flowers, pods, leaf axils, seeds. They are all separate genes which can inherit independently. But none of them can express themselves without the dominant A allele which switches on the pigment production. Without it, the colour genes are still there but they are mute.

With Salmon Flowered, it's obvious that it has genes for colour in various places. The rosy pink blossoms, the pink blush on the pods, the soft pale pink smudge in the leaf axil. The presence of all these colours tells me that it has the dominant AA genotype - in other words, anthocyanin is switched on. Conversely, Alderman shows no anthocyanin pigment whatsoever. It is entirely green leaved and white flowered. I can safely assume that it carries the recessive aa genotype - in other words, anthocyanin is switched off.

Because Alderman is genetically incapable of producing anthocyanin pigment, I have no way of knowing what other colour genes it has hidden away, clawing at their nucleotides and begging for release. It's very possible that it has a full palette of colour genes, and that it wants to express purple flowers, splodgy leaf axils, purple stems, the works. Even though all these colour genes are dominant, they are helpless, disempowered, in the presence of aa. It's a curious subversion of the usual law of inheritance, with a recessive allele suppressing the expression of several dominant alleles.

Now this is turning into a very wordy explanation, but it's all so gloriously simple. The mystery purple flowers in my hybrid have almost certainly come from Alderman. I believe Alderman has the dominant gene which makes purple flowers, but it's not normally expressed in Alderman plants because they have no pigment capability. When I made the cross with Salmon Flowered, I gave it the 'on' switch. In a cross between a plant which is aa and one which is AA, the hybrid is going to be aA. The dominant allele gets the upper hand, anthocyanin is switched on, and all the colour genes in both varieties are free to express themselves. My hybrid is showing colour traits from Alderman as well as from Salmon Flowered.

I suspected this might be the case when I saw the colour blotches on the leaf axils in the hybrid plants. They were very prominent, with the dark pink colour streaking right out into the stems. Although Salmon Flowered does have pink in the leaf axils, it is very pale and subtle. This was quite different and could only really have got there if it came from Alderman. So I knew there was a good chance that Alderman might have a purple blossom gene as well, especially as those two colour genes are closely linked and usually appear together.

What this means for the F2 generation next year is that I will get a quarter of the plants unable to produce anthocyanin, and therefore having white flowers. Of the remainder, I will get mostly purples but I'm hoping that there will also be a few rosy pinks. I don't know exactly how the pink-flower gene works, so I wouldn't want to predict anything more than that at this stage.

Last time I wrote about my Alderman x Salmon Flowered F1 hybrid I said it was expressing the recessive trait for fasciation (stem widening). Well it isn't. It did show some fasciation, but it turned out to be from environmental causes and the plants reverted to a more normal pattern of growth. They do have very thick stems, but this is common to nearly all my F1 hybrid peas and I think it's mostly just hybrid vigour. They are, however, saving their flowers for the top of the plant, and have grown to almost 6ft before showing any buds. This, coupled with the late-maturing trait from Alderman, makes them very slow to reach maturity. That's probably something I will have to select against in the F2.

Meanwhile, jolly solstice blessings to all who observe such things.

A triumph of carrots

There is, you'll agree, a certain 'je ne sais quoi' oh, so very special about a firm, young carrot.

So said Uncle Monty in Withnail and I, my favourite film of all time (though in all honesty I don't watch that many) and it's certainly a triumph to me because I'm absolutely rubbish at growing carrots.  I've grown the odd good one in the past, "one" being the operative word. To actually be able to hold up a bunch of homegrown carrots without them wilting in spindly shame is a first for me.

These photos are from the first batch I harvested. I dug up another load a week later and the colour was better on some of them in the second batch. They might have grown bigger if I'd left them a bit longer but I needed to vacate the greenhouse borders for my tomatoes.



So what did I do different this time? Well I started them off early and grew them in the greenhouse, which is probably what made the biggest difference to their fortunes. They were safe from the buffeting of spring weather and the oscula of marauding slugs. Also, instead of direct sowing the seeds in rows in the traditional way, I sowed tiny groups of seed in small modules and then transplanted them into the greenhouse border in baby clumps (no thinning needed).

I still need to perfect the method though, because some of them got into a bit of a cosy relationship under the ground, like these Purple Dragons.

Ménage à quatre

So let's have a little rundown of the varieties (see pic below).

Red Samurai - this one is an F1 hybrid. If you've read some of the cynical ranty stuff I've written about commercial F1 hybrids you might be a bit surprised that I'm growing one. But I always reckon if you're going to slag something off it's as well to be sure you know what you're talking about. So although I think commercial hybrids are largely a waste of money, I sometimes grow them to see how they compare with the OP varieties I sing the praises of ... and also to see whether they have potential for breeding work, because the F2 seed they produce can segregate into interesting new combinations. Anyway, Red Samurai is a red carrot - or at least a kind of brick red. The ones shown here are not very red, but I harvested some better coloured ones in the second batch. I also had some which didn't look right at all ... thin, spindly and whitish. The presence of the thin white dud in a carrot crop is usually a sign of an accidental cross-pollination with a wild type of carrot like Queen Anne's lace. Such crosses are not uncommon even in commercial seed because carrots are such tarty outbreeding slappers. But I did get enough decent red carrots to be able to evaluate it. It was sweet and tasty and pleasant. Did it have that special extra edge to justify the cost of an F1 hybrid? No. In a word. Nice, but nothing special and there are plenty of OPs which are better.

John's Purple - or John's Light Mauve, if my specimens are anything to go by. This is from the Heritage Seed Library and not commercially available. It was raised in the 1970s by a chap who spotted a purple carrot in a bag somebody gave him for his rabbits. The roots are neither large nor pretty, but they are certainly different from anything else I've seen. Eaten raw it has a strong, earthy, slightly soapy flavour and is very juicy with a really nice texture. The carrot is white inside but the mauve colour extends some way into the flesh. Worth giving this one another go, I think.

Cosmic Purple - hmmm ... I suspect I may have a packet of mislabelled seeds, because the carrots I dug up are neither purple nor particularly cosmic. Maybe that's a blessing, because I've heard nothing but negative things about the flavour and texture of Cosmic Purple. The thing I have here is a decent, nice looking orange carrot with a smooth skin, a sweet and juicy flavour (though rather mild in character) and lovely texture. God knows what it is though.

Purple Dragon - ahhh ... now this is purple. The ones I dug up in the second batch had an even stronger colour. Rich though the colour is, it's entirely on the surface. Slice it open and the carrot is a normal orange colour all the way through. Raw flavour is very sweet but with a hint of a bitter undertone, and the texture is firm and crunchy. It does however keep its colour pretty well when it's cooked, and loses the hint of bitterness. It's a nice variety and one I will grow again.



Left to right: Red Samurai F1, John's Purple, Cosmic Purple (supposedly), Purple Dragon.

Purple Dragon is certainly the pick of the bunch, and the one I'll continue to grow. It's not the best carrot I've ever tasted ... that particular accolade goes to a heritage yellow variety called Jaune Obtuse du Doubs. If I had time I might even try crossing them, but I shall need a couple of extra lifetimes to fulfil all my plant breeding ideas.


Purple Dragon, sliced.

And another nice little triumph. Yesterday the blog's hit counter clocked up the 100,000th visit. I managed to capture the moment in a screenshot. This has taken exactly four years ... I put the site meter on here in June 2006, a few months after I started the blog. It was depressing to have it there at first because I had bugger all readers - even the spammers couldn't be bothered. It has built up steadily since then. So I'd encourage anyone who's just starting out with a blog not to get despondent if it seems like nobody's reading it. It can take several months to a year before you really get noticed. Anyway, thank you to everyone who has supported and encouraged me and left comments over the last few years.

Beautiful Luna Trick

Everything's happening so fast in the garden at the moment I'm barely able to upload my photos onto the computer before they're out of date (however did I manage in the days when I had to send films to Bonusprint?) so I'm just going to chuck a few pictures of my Luna Trick F4 peas up here to show you how they're getting on.

I'm very happy with the way things are going with this variety. You may recall, dear reader, that I have two separate batches of F4 plants because last year's Luna Trick F3 segregated into mangetout (snow) types and sugarsnap types, and as I'm intending to develop them as different varieties I'm growing the two types separately. Or I thought I was. As it turns out I have sugarsnaps among the mangetouts and mangetouts among the sugarsnaps, so balls to that idea. The former situation was entirely expected - I knew some of the mangetout-podded plants I saved seed from last year would've sneakily closeted away the recessive n allele in their genome, ready to spring forth unbidden in future generations. This recessive allele singlehandedly creates the sugarsnap type simply by thickening the pod wall, transforming a wide flat pod into a slim, round, plump and juicy one. I love both types, so I'm open minded about what I get. But the recessive nature of this gene means that it will be hiding itself in a proportion of the mangetout plants for a generation or two yet, and I won't know which ones until I grow their offspring.


Upside down flowers? Well why not.

The most promising F3 plant I had last year was one called LT10. (There's nothing arcane about this name, it was simply the 10th plant in my Luna Trick trial to get a label stuck on it.) It was a tall mangetout-podded plant with a divine flavour. And the gene gods were smiling on me, because when I started growing out the F4 it turned out to be the only one of my mangetout lines which was true-breeding for tallness, which is what I want. All the others had hidden recessive dwarfing genes lurking in them. If I save seed only from LT10's offspring and not from the other lines I should have no more trouble with dwarves. I supposed it would be too much to expect LT10 to be true-breeding for mangetout pods as well, and indeed it has presented me with a goodly smattering of sugarsnaps. They may well be good enough to contribute to the development of my sugarsnap line, but in the breeding of a stable mangetout line they are a pain in the backside.

While I was expecting to get sugarsnaps in the mangetout batch, I certainly wasn't expecting to get mangetouts out of the sugarsnaps - that was a complete surprise. A recessive trait is by definition true-breeding. It can only express itself when a matched pair of recessives get it together and there are no dominants involved to spoil the party. As the mangetout-pod trait is dominant it couldn't possibly be present, hidden, in any of the sugarsnap plants I saved seed from last year. So I was intrigued to see not one but several of the F4 sugarsnap plants developing mangetout pods. How could this be? Is there some other gene at play which suppresses the sugarsnap gene? Did the faeries dig my plants up and move them in the night? The solution became obvious when I looked at the identification tags of the plants concerned. They were all from the same parent plant, LT2. I'm growing ten of LT2's offspring, and the majority of them have mangetout pods, with a few sugarsnaps. While I hesitate to declare a Mendelian ratio on such a small sample size, it seems obvious that there is one and I made a mistake in classifying LT2 as a sugarsnap when it was nothing of the sort. Sometimes mangetout pods look a bit sugarsnappy. It's a relief to find such a simple explanation, and it's one of the reasons I number my plants at the F2, F3 and sometimes F4 stage - as laborious as it is, it enables me to track the pedigree of each individual, to find patterns in weirdness and resolve mistakes.


The first Luna Trick F4 pod, still with its little petal nosebag on. This one is a sugarsnap type (honest).


The same pod five days later, now with a beautiful luna crescent shape.

In the last couple of days I've managed to do the first couple of taste tests, and so far they've all been very nice. Flavour is the most important thing I'm selecting for, apart from healthy vigorous plants which goes without saying. But I'm also looking out for crescent moon shaped pods, particularly among the sugarsnaps. I'm also mindful of the colour intensity. For the most part Luna Trick has a stronger and longer-lasting yellow colour than its yellow parent, Golden Sweet, but in last year's batch I did have some which faded to pale washy green a bit more quickly than I would like.

I'm also busy with the scalpel using Luna Trick for various hybrids, particularly as it represents a superior alternative to Golden Sweet, so rather than doing backcrosses to Golden Sweet in my Peachy and Red-Podded lines, I'm using Luna Trick instead. I'm making these crosses with both the mangetout and the sugarsnap types, which I'm hoping will give me a red sugarsnap as well as the decent flavoured red mangetout I've been coveting. I've also made some crosses with Salmon Flowered (the antique umbellatum pea) which is more of an experimental endeavour. What I have in mind is to extract the gene which makes the rosy pink flowers, which would look lovely on a yellow-podded pea. But my weird imagination also conjures up a vision of a yellow-podded umbellatum, crowned by a giant clump of sticky-outy golden sugarsnaps. What a spectacle that would be.


This is one of the mangetout type pods.

A wildflower interlude

A thunderstorm chucks its load on the Severn Valley (and shortly afterwards, on me) during my walk on Sunday. Normally you get a spectacular view of the Malvern hills from here, but they are swamped!

I rarely blog about anything that isn't in my own garden, but it's worth making an exception sometimes. Not least because I ought to mention occasionally that I have an interest in native British wildflowers (and have a Flickr set of my best photos). And that I'm also lucky enough to live in the Cotswolds, a designated Area of Outstanding Natural Beauty, so I have no excuse not to go out and gawp at them.

I don't have to go very far to find them. Cheltenham sits within a cosy semicircle of hills which form a continuous ridge around it. The ridge gives the town a beautifully sheltered climate and keeps frosts in my garden to a minimum. And the ridge itself is mostly grassland overlaying beautiful honey-coloured limestone which makes a wondrous habitat for wild flowers and butterflies. All of the Cotswold hills have their own unique character and variations in plant life, and when I feel the need for a botanical frenzy I tend to go to the quieter ones which the tourists (and even most locals) don't know about. My favourite is Nottingham Hill, which is a spur at the northern end of the ridge.

Nottingham Hill is a very special place to me. The whole of the top is an iron age hillfort, but you can't see it from the ground because it's too big to view in its entirety. The earth is full of natural holes, dips and openings. The lower slopes (particularly on the N and E sides) are liberally dotted with springs, some of which are the sources of local brooks and streams. Most of the trees around the fort are magical trees, elder, hazel or thorn. There's also a very strange grove on the southern edge just below the fort where almost all the trees are hollow or have holes in them, growing in the dips and hummocks left by ancient quarrying. Some are small finger-sized holes, some cup-shaped holes with grass and violets growing in them, some large oval holes right through the tree. With gorgeous views over the Severn Valley to go with it, it's an incredibly evocative place.

It's also lush with wild flowers.


Rock rose (Helianthemum nummularium) growing next to a common spotted orchid (Dactylorhiza fuchsii) on Nottingham Hill.

These pictures were all taken on Sunday, when I went out for a walk up there on what seemed like a charmingly pleasant early summer day. I was having a great time grovelling about in the undergrowth with my camera when it became apparent that the distant booming I could hear was not some bizarre rural motorsport as I'd assumed but a rather formidable thunderstorm approaching very stealthily. I can't resist rainstorms ... they are an exhilarating manifestation of nature's power and they make me feel wonderful. So once I'd got my camera to a place of safety (just as the rain was starting) I went off up another footpath and climbed up onto a mini-hillock on the hillside and embraced the storm while it pelted down and soaked me. It was bliss!

I had to drive home with a wet arse, but it was worth it.

None of these are particularly rare plants, but I don't care, it's always a joy to see them.


Wild thyme, which is a species called Thymus polytrichus. It looks like a miniature version of garden thyme and you certainly can take it home and sprinkle it on your pizza, but I've found Cotswold wild thyme to be decidedly lacking in the aromatic department, so I generally just admire it and use home-grown stuff for eating.



Yes I know this stuff is common as dirt. But it doesn't half look beautiful once you get down on your hands and knees for a proper look. Bird's foot trefoil, also known as Eggs and Bacon, and Lotus corniculatus.



The decidedly blue flower known as green alkanet, and also by the glorious botanical name of Pentaglossis sempervirens.



Although it looks a bit like lady's bedstraw, this is the slightly more subtle crosswort (Cruciata laevipes). Most things with 'wort' in their name have a history of medicinal use.



Another old healing herb, common milkwort (Polygala vulgaris) which I found in two shades of blue and a pink variant, all growing together in the same clump.



A particularly elegant blossom of creeping cinquefoil (Potentilla reptans).


One of the highlights of my walk was not a flower but a particularly beautiful moth. I don't know much about moths at all but these days all you need is a quick trawl of the interwebs to identify things.




This gorgeous chap (two views of the same moth) is a five-spot burnet, or Zygaena trifolii if you prefer.

Finally, here's something that is in my garden.

I have a few wild plants naturalised in my garden, either because they've always been there or because I've introduced them through locally collected seeds. My house was built in the 1930s on land that was formerly a damson orchard, and I still have some damson trees in what was originally a field boundary at the back of my garden, and it's quite possible that some of the native wild flowers have stuck around here too.

At any rate, when I moved here in 2004 and started weeding the garden, which had been fallow for a year, I found a scarlet pimpernel with purply-mauve flowers instead of the usual pale red. I carefully weeded round it to give it a chance to mature and set seed. I wasn't sure whether the flower colour would come true from seed, but sure enough the following year there were more.

It's not particularly unusual to get colour variants in scarlet pimpernels. But blue is a more common one to find, or white. I'm not sure how widespread the mauvey purple ones are. But it certainly is a heritable trait because there's no variation from year to year. I try to help them colonise by carefully preserving all the pimpernels when I'm weeding and then removing a lot of the scarlet ones once they show their colour. Sometimes if I'm out there with my tweezers I grab an anther off one of the purples and dab it on the stigma of another purple on a different plant, just to make sure. But in all honesty it doesn't need my help, it's surviving fine by itself.


Violet-purple petals on a scarlet pimpernel, Anagallis arvensis. Native to my garden!

TPS potatoes: transplanting and growing on

A young TPS seedling ready for a new home. This one is from an OP berry of Mayan Gold which I collected in my garden in 2008.

Well, I said I would give further instructions on cultivating potatoes from TPS (true potato seed) once I had some photos from my own crop, so it's about time I got on with it. These tips draw heavily from the videos of Tom Wagner's potato workshop posted on Patrick's blog, so if you want more details I recommend giving those a spin.

Last time I wrote about sowing and growing young seedlings from TPS. Once the seedlings get established, you essentially grow them the same way you would ordinary tuber-grown potatoes … you earth them up as they grow. The only difference is that they are smaller and more delicate than tuber-grown potatoes, at least until they get established, and need to be handled a bit more gently.

The first thing you will probably notice about potato seedlings is that they grow very fast. When they first germinate they're so tiny they almost look like little threads of cotton sticking out of the soil. But it doesn't take long for them to outgrow their seed tray and be elbowing each other for space. They also tend to be a little bit straggly and thin-stemmed (though less so if they've been given plenty of direct sun) and when they reach a couple of inches tall it's time to give them their first transplanting.

Find a container or module thingy deep enough for the whole seedling. That is, the depth of the soil in the seed tray plus the height of the plant. Make a deep well in the centre of the module/pot and drop the seedling into it. Bury the whole plant right up to its neck - stem, leaves and all. Just leave the growing tip and a couple of sets of leaves sticking out at the top.


Put the seedling into a deep hole in the compost and bury the stem and lower leaves, submerged up to its tip, like the one on the right.

Once they are potted up they will grow like rockets. The submerged stems will grow little feeder roots which will help the plant grow bigger and better. They are exactly like tomatoes in this respect. They have the additional advantage though that they can also set tubers from these roots ... and earthing them up in this way will increase their tuber production just as conventional potato crops are earthed up for the same reason. When transplanting them, there's no need to remove the lower leaves - just bury them. They're better left in place to wither naturally rather than causing unnecessary injury by pulling them off.

Keep the potted up seedlings well watered in a sunny place for a few more weeks and admire them as they turn into something like this:


A mature seedling ready to be planted out in the garden. This is a Pirampo x Khuchi Akita F3 plant grown from Tom Wagner's TPS, now starting to look properly potato-like. Ignore all the funny business at the base of the stem ... relics of the fused cotyledons which this plant produced before growing normally.

When there's a good bit of stem growth on the seedlings and all risk of frost is past, harden off the seedlings for a few days and nights and prepare a plot of ground for them. By now they are getting to a point where they can be treated more like tuber-grown potatoes, and you can plant them in the same kind of soil you would use for any other potato crop. Having said that, the lack of a tuber does make them more vulnerable. If a tuber-grown potato haulm is damaged, it usually regrows fairly easily from the tuber, which is full of stored energy and nourishment and capable of compensating for pretty much any degree of human ineptitude. TPS-grown potatoes don't have that backup - they are dependent solely on their youthful root system, which is somewhat fine and tangly. Add a thin and brittle stem into the equation and you have a rather more delicate plant than a normal spud. Even so, it's amazing what they will survive. One of my seedling batches newly planted out this week had a devastating visitation from the Arse of Doom, which left evil deposits followed by a savage ground raking in which several plants were damaged. I was pretty surprised to find that baby plants reduced to mangled muddy stumps with all their leaves ripped off were sprouting new green growth within two days. Them's tough little chaps.



Depending on how mature the plants get by the time you plant them out, you may find tiny tubers already forming when you take them out of their pots, as shown on this Carola hybrid above. The tubers form on special chunky roots called stolons. Sometimes you get stolons above soil level with little microspuds on the end, but usually when stolons grow above ground they develop leaf shoots, which is not what you want, so plant them out and get them earthed up. Something else I noticed was a very slight blue-ish tinge in the roots of some of my Salad Blue OP seedlings. I may have imagined that though, I was drinking some Belgian Leffe at the time, and that stuff is like LSD to me.



Dig a little trench. Tom recommends a north-south orientation to make the most of the sun passing overhead. Then dig out a hole at the bottom of the trench deep enough to accommodate the whole height of the plant and its rootball. Yes, once again you are going to bury most of the plant right up to its top leaves. The original seed level is going to end up some way underground, which allows lots of stem depth for making extra roots and tubers.

I have been planting my seedlings around 10 inches apart. That's slightly closer than I would with tubers. Plants grown from TPS are not realistically going to produce as big a yield as a tuber-grown crop in their first year (they will make up for it next season) so I squeeze them up a little. Space is at a premium in my garden so I sometimes need to plant things a bit on the close side.

If waterlogging is a problem in your soil, you can plant the seedlings slightly to one side of the trench rather than in the bottom (choose the side where they will get most sun) to ensure the roots aren't sitting in a puddle where they might rot. My soil is sandy and well drained so I don't worry about it.


The seedlings planted along the base of the trench, planted deeply with just their growing tips exposed. They will soon grow up and away, and as they do I will be able to fill in the trench to keep them earthed up.

The earthing up process continues as the plants grow. The filling in of the trench is followed by a gradual mounding up, so that you end up with a long hillock around the plants. Just as you would with a conventional potato crop.

Next instalment will be an explanation of hand-crossing potato flowers to make your own TPS. Wahey!