Observations on breeding Damasks

This aren’t a whole lot of people breeding these anymore, but I’d hope that others who have bred them will contribute what they can, so that those who want to try it in the future will have a good place to get started.

We don’t know for sure which was the original R. damascena, that probably originated somewhere between the western slopes of the Himalayas and northeastern Persia a few thousand years ago, but most all of the thorny primitive types have been found to be genetically indistinguishable, so they are either the original, or sports of the original. All of them pose problems for the breeder, but they aren’t insurmountable.

Early Damasks are good pollen producers, but barely lukewarm as seed bearers. Not all pollinations will set, and of those that do, some will produce empty hips. Almost all hips which are not empty will contain one achene, and only once have I seen three in one hip. Plants that are not well established, or underwatered, may not produce any seed. Big, established plants, on the other hand, are hard to pollinate without running into a lot of thorns. When left to the bees, mine produced quite a few selfings, but others which clearly were hybrids. Maybe a quarter of them mildewed badly, as is a problem with a great many kinds of rose seedlings, but the majority were reasonably healthy, and survived conditions which killed the babies of most other roses I was breeding. Healthy seedlings is probably the only aspect of using old Damasks as seed parents which seemed very encouraging to me.

Reblooming Damasks don’t pass along their remontancy like more modern, juvenile rebloomers do, and I know of no rule of thumb to say how many seedlings will have to be grown out to get a repeater. Generally speaking, the answer seems to be LOTS. Paul Barden mentioned a dwarf seedling of Autumn Damask he’d been given, which may have been a selfing, and which repeated, but most of my Autumn Damask selfings look much like the original, so if it is possible to get repeaters that way, it’s hard to see why anyone should prefer them to their parent.

Portlands and other hybrids of the original damascena are often better seed parents, but most aren’t by very much. Duchess of Portland, for example, sets hips a bit more frequently (though many abort), and they’re empty less often, but still only produces 2-3 seeds per hip, and reblooming seedlings are too rare for me, or any living person I know of, to have grown any. Pickering Four Seasons gave me a lot of impressively big hips, orange and showing Damask influence in their shape, but most have been empty. A large and happy specimen of Duchesse de Rohan produces a useful amount of pollen, but will never set a hip. I’ve heard that Indigo will set hips, but mine hasn’t yet. Stanwell produces a little pollen, but almost never any seed.

Of once-blooming Damask hybrids, Mme Hardy is famously sterile, producing only empty hips, and never more than a tiny trace of pollen. Ispahan makes plenty of fertile pollen, but no hips. Hebe’s Lip is relatively fertile in both directions; almost all pollinations set, it rarely aborted, and it never made empty hips for me, but how many seeds you will get seems to depend a lot on luck and the pollen used, and it may not be very many. On some pollinations I got as many as half a dozen seeds, but usually only two or three. Saint Nicholas, which (like Hebe’s Lip) shows eglantine traits in its foliage and thorns, produces fewer flowers, and thus fewer hips, but set is reliable and the seed count per hip is more like 15-20. Its hybrid seedlings were not especially healthy for me, though a few weren’t bad, and its selfings were often disease resistant. It doesn’t smell much like a Damask however, and I don’t think it really fits into the class. Celsiana seems as much an Alba as a Damask, but produces plenty of pollen and seed, averaging around 5 per hip with decent germination, though seedling health can be expected to be variable. Petite Lisette produces little pollen and no hips. I haven’t yet tried breeding Leda, but suspect it would act like pure damascena. I also haven’t tried breeding my Botzaris or Blush Damask, but have no reason to believe they’ll be very fertile.

To sum up, unless one wants to use Celsiana, or maybe Hebe’s Lip, Damasks are probably best used as pollen parents. But on what?

In days of old, they would use gallicas, which are mostly better seed producers. Hortensia can be counted on for 1-2 seeds per pollination, Ipsilante and Alika for 2-3. Apothecary’s Rose/Rosa Mundi and Tuscany Superb did more like 4-8 seeds/hip for me, but seedling health was no better than with Celsiana, and most people are probably not wanting to recreate Duchess of Portland from scratch.

Albas are fairly fertile in some climates, but not so much where I live, plus they usually take a long time to germinate, and then you’ve got ploidy issues to deal with. Canina and eglanteria are more fertile, but the germination and ploidy problems are similar.

Centifolias make seed in Pakistan, but in more temperate climates produce very few hips which aren’t empty.

Musk x Damask is something I’ve just tried, but I don’t expect that to be any easier than using damascena as a seed parent, if it works out at all.

I’ve heard that some moderns don’t take damascena pollen very readily, but haven’t really tried that route myself. (I figured that, with my luck, I’d get tea scented once-bloomers with the thorns and growing habit of damascena. Those who are less pessimistic might want to give it a go.)

One possibility is to try backcrossing it with fedtschenkoana, which is more hip fertile in some environments, and would give a chance of rebloom in F1s. Another, which I’ve been doing a lot of this year, is to use Damask/Portland pollen on HPs, some of which are quite fertile, and most of which have much better chances of passing along rebloom, though it will require two generations.

I’ve put Damask and/or Portland pollen on the following, with reasonable hip set, at 8-20 seeds per hip:
Rose du Roi “(of commerce)”
La Reine (also “Barbara’s Pasture Rose,” which is very similar)
Reine des Violettes
Alfred Colomb

While one of the old authors noted that a breeder had grown out thousands of seedlings of Rose du Roi without producing any rebloomers, it isn’t 100% clear whether his rose was identical to what gets called that nowadays, nor was anything said about what pollen he was using. Were he, for example, treating an early HP as having Damask rebloom, when its rebloom was from chinensis, he could certainly cross it with something like The Portland Rose and get no rebloomers. The rose that is usually sold as Rose du Roi in the US will bloom for me in December, and produces ovoid or spherical hips that are quite large and purple, traits which make me suspect lurking chinensis. I won’t be able to comment on rebloom among descendants for a few years, however.

Marbree also makes way too many seeds for a Damask, or even a Gallica, and Paul Barden said he got one rebloomer out of five or six of its seedlings, so I think, given its late introduction date, that it’s got some chinensis ancestry. Its scent is not terribly strong, or exactly the typical Portland fragrance, but using it to get more rebloom and female fertility into Damasks without going too far out of class seems like a real possibility, and I’m enthusiastic about trying it for that purpose.

The others are clearly much purer chinensis, but all have gotten decent hip set using pollen from Portlands, and make a lot of seed. Though La Reine is finicky and sickly in many environments, it does have a reputation for lots of rebloomers among its offspring, so try it if you dare. Reine des Violettes is less disease prone but nearly as demanding, less prolific, and set hips less consistently. Alfred Colomb has been healthy for me, and produced 10-12 seeds from nearly every pollination, if you don’t mind growing yet more descendants of Gloire des Rosomanes.

Comte de Chambord produced somewhat less seed than those, and Baronne Prevost only a few seeds per hip, but both are worth a try. Vick’s Caprice also took Portland pollen, but didn’t set hips very consistently, though that may have been the drought at work. Eugene de Beauharnais is only just now deciding to bloom, but I’m hoping it will like Duchesse de Rohan pollen, because that’s what it’s getting. It hasn’t produced much seed for me in the past, but I haven’t quite given up on it yet. Rose de Rescht has also been a disappointing seed producer for me, and has a reputation for sickly seedlings, but I’m still trying.

The only HPs which utterly failed were Glendora, which is about as sterile as Mme Hardy, and the rose I got as Jacques Cartier, which produces some pollen but no hips. In neither case can I fault Damask pollen, they’re just not made to be mothers.

Other possibilities which I haven’t tried much, since the seed parents don’t do very well here, are Moss x damascena, and rugosa x damascena. The latter could make an interesting security rose, and for all I know, might even have a chance of rebloom in the first generation.

I did 46 crosses involving roses discussed in this post this year, so hopefully will have more to add several months from now.

Thank you for such a detailed account of your efforts. Do crosses of the oldest damasks produce some seedlings with fedtschenkoana traits dominating? Since that species is supposed to be a big part of the ancestry of the original damasks, I speculate, based purely on research, not experience, that it may express itself in some damask seedlings. It sounds like that speculation isn’t borne out by experience.

Has anyone ever tried early damask pollen on R. kordesii?

Research has shown that there is considerable genetic variation in the old Damask roses of Iran. This seems to contradict the 3-species hypothesis because the distinctive characters of this “species” apparently hang together in subsequent generations.

IRANIAN JOURNAL of BIOTECHNOLOGY, Vol. 3, No. 4, October 2005
Analysis of the genetic diversity 12 Iranian Damask rose (> Rosa damascena > Mill.) genotypes using amplified fragment length polymorphism markers
S Mostaffa Pirseyedi, Mohsen Mardi, Saied Davazdahemami, Maryam J. Kermani, S Abolghasem Mohammadi

Our results suggest that the AFLP approach is a reliable, rapid and sensitive technique to estimate genetic diversity of Damask rose genotypes. The results showed an extreme variability and genetic complexity among 12 Iranian Damask rose (> R. damascena > Mill.) genotypes grown in the famous rose gardens of Tabriz, Meymand and Kashan regions of Iran. For breeding of Damask roses, it is necessary to conjunct molecular data with agronomic characters such as yield of rose oil and novel scent compounds. This strategy may benefit the breeding of > R. damascena > plants as proposed by Debener et al. (2003).

Perhaps. I saw your post while previewing an answer to Mr. Cook, which I’ve delayed posting because I want to take some better pics first, but one snippet out of my draft was:

(When I read an article recently about doubling ploidy of roses in an attempt to improve yield for the perfume industry, I wondered why they hadn’t simply grown out a few hundred damascena x damascena and selected the best, but I suppose that’s too low tech and boring.)

Previous articles, like these oft-cited warhorses which probably belong in this thread…
…have pointed to common damascena ancestry across the world. If one looks at probable origins, there was a major (for 1800 BCE) civilization in the area where one might expect to find both moschata and fedtschenkoana, and it interacted extensively with Persia: https://en.wikipedia.org/wiki/Bactria_Margiana This fits with the age geneticists guessed for Summer and Autumn Damasks, Kazanlik, and York and Lancaster. Bactria also had trade with India and the rising Andronovo culture to the north. Much later, Alexander the Great conquered the area, which eventually became home to important Chinese silk routes, so it was in a fine position to distribute suckers over a period spanning millenia, before becoming the total backwater it is today. If we assume that the original plant was exported without attempts to hybridize it, which seems perfectly reasonable, the single plant triparental origin theory looks good.

Though perfumer’s Damasks in Turkey and Bulgaria have been found to be identical, several studies in the Middle East and South Asia have shown genetic differences in what their various commercial growers produced. Comments in some of the papers indicated that the academics were unsure whether differences in climate and conditions accounted for the variations, so started checking the genetics to see. (Ispahan would have stood out as an obvious hybrid, but isn’t used by the perfume industry, which prefers a short blooming period and the highest possible yield.) Genetic diversity has been found in Syrian damascena: http://scialert.net/fulltext/?doi=ijar.2011.429.436&org=10 http://jonnsaromatherapy.com/pdf/Mirali_Genetic_Characterization_of_Rosa_damascena_2012.pdf And in Indian damascena: https://www.researchgate.net/publication/272172373_Characterization_and_analysis_of_genetic_diversity_among_different_species_of_rose_Rosa_species_using_morphological_and_molecular_markers Some diversity is found between different Middle Eastern growing areas, but most of it is found within particular growing areas. http://www.sciencedirect.com/science/article/pii/S0304423807003482

It wouldn’t be logical to assume the variety in Middle Eastern and South Asian genetics is from mutation, since it seems (on the 3500-4000 year time scale) to be a relatively modern thing, more recent than the exportation of perfumer’s stock to Bulgaria and Turkey, which had happened by the 1500s. My thinking on the subject is that a few fields of seedlings (even if they were all selfings) by farmers in India and the Middle East, at any point between 1500 and 1900, could account for a lot of the genetic diversity in those areas, and that the Turks and Bulgarians have shown a relative lack of willingness to experiment, though it’s also possible that some of the studies showing genetic homogeneity weren’t as rigorous as they might have been.

What are your thoughts on the subject?

Let me quote an earlier post of yours for context:

So your general idea, as I understand it, is to get much of the good out of species without necessarily having to start over from scratch. My opinion is that the answer to your question is, to some extent, probably so, but not unconditionally so.

A natural species happens when a community of related roses interbreed for many generations, with factors like blooming time and geography letting them stay distinct from more distant relatives that might be within bee range. When you cross one of them with one of their own kind, the seedlings will bear a consistent resemblance to their parents. If you cross one of them with a different species, you can expect variation, with some looking more like one parent, and others like the other, and some may be more vigorous than either parent, but none of them are likely to look unlike either one. If you really like the cross, you can cross it with itself, and/or siblings, and in several generations, it will start breeding consistently, almost like a natural species, though the vigor from the initial cross may quickly fade.

But if, instead of doing that, you start crossing them with all kinds of different things, where nothing is close to a backcross, you will start getting some seedlings which don’t look very much like either parent. Instead, they may look like something from a couple of generations back, or farther. And they will not come close to breeding true, like a species. Although garden roses do have some ways in which they’re very inbred, i.e., one might be descended from a single ancestor several times over, no garden rose has ever been inbred to the point where it was a stabilized strain that would breed true. One trait or another from the recurrent ancestor(s) may show up in almost all of them, but the rest will be a real crap shoot. The fact that most garden roses are polyploid, just makes it all the messier.

Okay, so, back to damascena. The oldest are said to be ((musk x gallica) x fedtschenkoana), making them an F1 of an F1. They may breed a bit more consistently than a more complex hybrid, but there will be obvious variation. It’s not usually very visible when the seedlings are little, but after a couple of years obvious differences may develop. (When I read an article recently about doubling ploidy of roses in an attempt to improve yield for the perfume industry, I wondered why they hadn’t simply grown out a few hundred damascena x damascena and selected the best, but I suppose that’s too low tech and boring.) Time for some pics.

Here are a couple of hybrids, Saint Nicholas x Pickering Four Seasons, and Autumn Damask x Grandmother’s Hat. They look about halfway between their parents. Subtract the traits from one parent of either, and you’re left with something approximating the other parent.
Here are some Autumn Damask seedlings that appear to be selfings. As you can see, they don’t look identical to each other, but there aren’t any signs of influence from other species or cultivars. They’re not identical to mom either, but if you subtract the Autumn Damask, you’re left with much less than the expected influence of a second parent.
Since I can only put three pics in a post, refer to the following post for a couple more examples, one of which looks like a pure Damask (but not exactly like its siblings or parent), and another which looks like light green fedtschenkoana without the hooked shape to the thorns. Subtract the damascena looks from it, and you get something resembling an R. pendulina hybrid which doesn’t exist. If it has flowers that are reminiscent of pendulina I’ll eat my words, but for right now I am assuming it’s a selfing where the genes recombined in an unusual way, and made it look more like a fedtschenkoana-musk hybrid than a typical Damask.

If that sort of range appeals to you, intentionally selfing them may do the trick. It’s not a substitute for the genetic diversity one could get by repeating the (Musk x Gallica) x fedtschenkoana cross, nor will it give you what crossing with fedtschenkoana would, but the seedlings do well here, and I’ve found it an enjoyable and interesting experiment.


Wow. Thanks for such a great and informative post. There is so much more to roses than meets the eye. Just looking at those seedlings, though, I can tell why people fall in love with it. I agree with you about a common aversion to low-tech methods. In a lot of fields, though, the low-tech, or at least the simple, solutions often are the most enduring.

I guess I shouldn’t be but I’m surprised at the range of diversity among selfings from the same plant. I imagine that would happen in an F-1 or F-2 parent that is carrying around a lot of unexpressed genetics from its parents.

This post was so informative! I was wondering if anyone has more experience since this was written?
So far I only have Banshee to work with. I have been focused more on Albas and on hybrids that had centifolia or alba and maybe Gallica.
But have been debating whether fragrance and cold hardiness would make a damask worth the pursuit?
So far I haven’t had any success with Banshee pollen, although I will continue trying: a few seeds set on shrubs, but they didn’t germinate, even with extraction. Banshee set a few hips, but dropped them. (It is young yet).
I’ve debated Hebe’s Lip or Leda. Any thoughts?
Still plan on using Portland roses when I get them.
And Merry Christmas!

There seems to have been more than a little accidental crossing going on in old gardens, before botanists started sticking garden varieties into books.

l’Obel (Kruydtboeck, 1581) informed us that Rosa damascena was the Latin name for the plants called Roses de Provence or Roses incarnates in France.

Gerard (1597) listed two Damasks. The “common” form was Rosa provincialis sive Damascena. The Province, or Damaske Roses. “The common Damaske Rose in stature, prickley branches, and in other respectes is like the White Rose; the especiall difference consisteth in the colour and smell of the flowers; for these are of a pale red colour, and of a more pleasant smell, and fitter for meate or medicine.”
Is this Maiden’s Blush?

The Lesser Damaske “differeth not, but is altogither lesser: the flowers and fruit are like: the use in phisicke also agreeth with the precedent.”

Under Musk roses we find a possible hybrid.
“Of these Roses we have another in our London gardens, which of most is called the blush Rose: it flowreth when the Damask Rose doth. The flowers heerof are very single, greater then the other Muske Roses, and of a white colour, dasht over with a light wash of carnation, which maketh that colour, which we call a blush colour. The proportion of the whole plant, as also the smell of the flowers, are like the precedent.”

And just one more.
“The great Rose, which is generally called the great Province Rose, which the Dutch men cannot endure; for say they it came first out of Holland, and therefore to be called the Holland Rose; but by all likelyhood it came from the Damaske Rose, as a kinde thereof, made better and fairer by an art, which seemeth to agree with truth.”

I’m with the Dutch men on this one. We know that the Dutch were doing some serious improving of vegetables at this time. The progress was preserved by painters of the period. Much might be accomplished by selection of seedlings, but I have a suspicion that the Albas also contributed some pollen.

Over time, and through selection, the pink roses diversified into groups that some folks insisted on identifying as species. I

So how has it gone Hardy?