Thanks for the response David. That’s too bad. I’d be really interested in knowing the answers to some of the above questions!
Mark
How disappointing. If we were to believe the plot lines of TV crime shows, a quick DNA test can reveal the eye color, shoe size and political inclinations of a victim or suspect. But out here in the real world we can’t even get the ancestries of our “mystery” roses.
1 Like
Not likely. So far as I have been able to learn, Autumn Damasks were not cultivated on Reunion or the neighboring Mauritius. And no Summer Damasks, either.
http://bulbnrose.x10.mx/Roses/breeding/Mauritius1816.html
The story you mention seems to have been based on the observation that hybrids of Chinas and Teas with once-blooming types were always once-blooming. It seemed “logical” to assume that both parents were recurrent. Folks back then were not clear on the concept of recessive traits reappearing in later generations.
E. H. Wilson (1915) studied the ‘Rose Edward’ and concluded that R. centifolia was the other parent. And as it happens, Centifolias were cultivated on the islands, as well as in India.
http://bulbnrose.x10.mx/Roses/breeding/WilsonRoses1915.html
I’m also curious to know where some of the “china” roses with strong yellow tones (‘Mutabilis’, etc.) fall in placement among other china and tea roses, since Rosa chinensis sensu stricto doesn’t really include plants with yellow or orange petals.
‘Mutabilis’ is another rose that is supposedly linked to Reunion. I note that Rosa macrocarpa was grown there, but no description was given. If this was the yellow-flowered species that was later renamed R. xanthocarpa, we might have one of the parents. The other might have been the one described by Ventenat.
http://bulbnrose.x10.mx/Roses/Rose_Pictures/I/RedouteIndica1835.html
Lacking the species, one might attempt to duplicate some of the qualities of ‘Mutabilis’ by crossing ‘Emmie Grey’ with a yellow Tea.
Would it be possible to attempt to duplicate the Japanese study indicating the triparental origin of the damask rose (moschata, gallica, and fedtschenkoana)? It would be nice to see that one put to the test.
Stefan
Tests are good!
Iwata et al. (2000) wrote, “If the second hybrid was self-pollinated, only one-half of the offspring of the second hybrid could have rRNA gene clusters of all three species. That is, the second hybridization seems to have happened soon after the first hybridization, after which the second hybrid soon started to propagate vegetatively, otherwise hybrid plants would not keep rRNA genes clusters of all three ancestral species for a long time.”
This could be tested by raising self-seedlings of the Damasks mentioned in the report. Do any of the seedlings inherit all three rRNA clusters? Are these clusters linked to other desirable traits (e.g., perfume, hardiness) that could account for their survival? In fact, the first question could be answered by examining embryos.
In particular, I am curious to know what sort of “rRNA” cluster might be found in Rosa alba, which was cultivated along with the Gallicas, Musks, and Damasks.
Here is a further challenge to the hypothesis that all damasks are sports derived from an ancient hybrid of three species.
BMC Plant Biol. 2007; 7: 12.
Microsatellite analysis of Damask rose (Rosa damascena Mill.) accessions from various regions in Iran reveals multiple genotypes
Babaei et al.
Conclusion
Our analysis showed for the first time the existence of multiple genotypes within Rosa damascena. We are currently performing an analysis of oil production across several years, in order to determine whether different genotypes also have a qualitative difference in production and/or composition of essential oil. If so, these genotypes may be used to broaden the production of rose oil, and they can also be used as the basis of a breeding program. As these nine genotypes were found after sampling only 40 large and small production fields, we expect that a more intensive sampling will be valuable in order to find more genetic diversity. For this, we will focus on the areas where we have found the unique genotypes, i.e., the Western and Northern provinces.
I was at a seminar today where someone was talking about identifying specific genes for interesting traits in populations of mosquitoes of a particular species. He said that the costs of complete sequencing for one insect is now down to about $1300. That’s a 5-fold drop over what he paid a few years ago. Once you have a complete genome sequence of one species or cultivar of rose, everything else can be compared to that very broadly, to assess inter-relatedness. Of course your do need to have some powerful bioinformatics tools to sort out the terabytes of data that you generate. So the actual cost is considerably more than the simple sequencing cost. Prices keep dropping on the sequencing but the really cheap DNA testing is the mass-produced simple stuff like identifying one suspect from another by a fixed number of DNA pieces being used for the whole world human population.
If you can focus in on a select region you can probably design a way to do informative testing of several CV at one time by the sequence all strategy mentioned above.That gets teh price per CV down near something reasonable.
I’m as interested in the academics of rose phylogenetics as anyone else here but, really, how useful can it actually be for the serious breeder to know if one rose is related somehow to another one beyond avoiding the obvious problem of introgression?
Every time we make a wide cross we are doing our own bit of dna comparison testing, and it’s the bit that counts the most.
“…beyond avoiding the obvious problem of introgression?”
Sorry, it’s not obvious to me, Don. ;0)
What is the problem with introgression that needs to be avoided?
Thanks, Tom
Tom,
I probably should have said that consanguinity is the obvious problem rather than introgression. I’m all in favor of introgression - most of my own work is dedicated to it.
It’s a spitting of the hair because repeated introgression, more specifically the repeated back-crossing that is part of the introgressive process, leads to consanguinity. It’s important to balance back-crosses with out-crosses so as to limit the degree of consanguinity hence one reason for knowing the ancestry of your foundation breeders.
My own breeding strategy is to get the target trait into whatever modern roses will accept it then to criss-cross the various hybrids against each other in a way that minimizes consanguinity.
I published this chart some time ago that illustrates the problem of consanguinity in modern roses:
The Major Descendencies of Modern Roses
To put this chart in perspective read my article titled “Fun with Numbers” in the Fall 2009 RHA Newsletter.
You will see from this chart how difficult it really is to avoid consanguinity - modern roses are very nearly a monoculture.
Thanks Don!
That does make sense to me now. :0)
I disagree with that chart – there are a few other species which were not mentioned but which are also in the ancestry of most modern roses.
-
R. gigantea – Teas and many Chinas have this species as an ancestor.
-
R. gallica – Found in Damasks, which led to Bourbons and Hybrid Perpetuals, and then Hybrid Teas and Floribundas, among others
-
R. fedtschenkoana – Ditto above
-
R. wichurana – Many of the more vigorous Hybrid Teas and others have the Hybrid Wichurana ‘New Dawn’ in their ancestry, and many Polyanthas also have infusions of this species
~Christopher
Did the link to my article not work?
The only one I don’t address is gigantea. I agree, as it seems obvious but, despite much effort, I never found a single reference to gigantea in any breeder’s program or in the literature. Maybe you have some?
Sorry, I didn’t read the article – but am about to now. I was responding to the graphic. My disagreement is not that breeders used the species directly, but that the species factored in the ancestry of hybrid cultivars used. So, for example, in the chart where one sees a Hybrid Perpetual, one must also include the species that went into it – Gallica, Moschata, Chinensis, Gigantea, and Fedtschenkoana would appear in the background of most HPs.
~Christopher
I read the article.
OK, I guess we’re interpreting things differently. You’re looking at major rose ancestry by measuring descendants listed on HelpMeFind of rose species. But I look at rose ancestry by genetic contribution. Rose development had two basic origins – the West, and the East. In the West, the major rose groups were based upon Gallica – its genes are in Damasks, Albas, Centifolias, and Mosses, with further contribution from Fedtschenkoana, Canina, Moschata, etc. In the East, Chinensis and/or Gigantea were the basis. When European rose breeders got a hold of repeat-blooming Eastern roses, they sought to transfer that repeat-blooming trait into their Gallica-based roses. It didn’t go much the other way around, except for expanding the Teas through breeding with Bourbons and Noisettes. It seemed as though a preference for tetraploidy was carried through the years, being as the European Gallica “base” was also tetraploid. While you look at Hybrid Teas and Floribundas and focus on the Eastern species’ traits of rebloom and flower shape, I also see the Gallica-based origin of stiff, straight stems and holding flowers upright. I see that you go further by mentioning the contribution of Gallica, for example, in the article, but this didn’t come through on the graphic, which I saw first.
Oh, and one other source for error is HelpMeFind itself. When you search a rose’s ancestry and look at percentages of ancestors’ individual contributions, the numbers won’t be accurate because HelpMeFind doesn’t take into account differences in ploidy. So take, for example, ‘Souvenir de la Malmaison’. HelpMeFind will show it as being half Tea and half Bourbon. But it’s triploid, and 2/3 of its chromosomes came from its Bourbon parent. So – do you measure ancestry with diploid and tetraploid parents having an equal contribution to their triploid offspring? If so, then I can understand how the continued influx of diploid ancestry from your four main species seems to swamp all the others.
A second shortcoming of using HelpMeFind this way is that you skip over all the roses with no stated ancestry but which are obvious descendants of particular species. Again, I go back to the Gallica connection. We can accept that all Damasks have Gallica in them – as well as Fedtschenkoana and Moschata. Damasks are in the ancestry of Bourbons and Hybrid Perpetuals – therefore, these two groups ALSO have Gallica and Fedtschenkoana and Moschata in them. Hybrid Teas are descendants of Hybrid Perpetuals and Teas – so they, too, will have some Gallica and Fedtschenkoana and Moschata in them, even if HelpMeFind can’t trace them back that far based on stated parentages. Not coincidentally, around the time that Multiflora and Wichurana came into the picture, there was more of an interest in publishing parentages – and this may be why there are more descendants listed on HelpMeFind for those species than there are for Gallica and Fedtschenkoana, even though if one were to look for DNA markers, one would surely find Gallica and Fedtschenkoana genes in most modern roses today.
I did very much enjoy the article, and I understand and agree with your idea that there’s been a lot of inbreeding going on. But it’s not necessarily just that so few species were used – it’s also that until relatively recently, roses were raised from OP hips, and were often selfs. Whenever I see “seedling of…” on HelpMeFind, I assume it’s a self, unless something dramatically different is apparent between offspring and known parent. I’ve been wondering about what could arise if some of the original crosses which led to new classes were repeated, but with different cultivars. I’m still getting things set up here, but I’d like to see “new Noisettes” that don’t have ‘Old Blush’ in their ancestry for that very reason. I’m looking at ‘Miss Lowe’s Variety’ as just one “other China” to try.
In any case, thanks for linking the article. I’ve saved it for future reference.
~Christopher
Your comments are sound Christopher and highlight how HMF percent ancestry feature is unreliable.
Don’s chart is very interesting as it evidence how much roses pedigrees are convergent.
This convergence is illustration of an all breeders constant breeding strategy over centuries.
The strategy being repeated backcrosses to the best just as it was done for many animals whose pedigree was a strong argument such as race horses or show dogs.
With the same result: standardization of progenies for failures. Most obvious failure being deseases susceptibility for roses.
The major fault of this HMF percent feature is that it does not consider the selection breeders effect at each generation. Just as if every seedling was selected. On the contrary elimination at each generation is very drastic for roses. Major breeders do grow some 10.000 seedlings to get one to introduce.
With a much lower selective pressure most genes are eliminated with a single backcross. In this case 25% = 0.
The more so as roses are complex multispecies hybrids with a lot of non mendelian genetical behaviour.
And yes fundamental bias is not considering gallica contribution just because early hybrids lacked pedigree.
In my opinion it is not unlikely that all but very few garden roses cytoplasm is gallica’s HMF ignore.
Here is a question to answer through DNA analysis: from which species garden roses cytoplasm went from?
I look at rose ancestry by genetic contribution. Rose development had two basic origins – the West, and the East
Up to this point you are correct. My advice is to avoid the romantic lay literature that purports to document the heritage of roses which tends to be a conflated jumble of confusion that mostly feeds arcane disputes among dilettantes.
Have a look at this paper.
Martin et al.pdf (93.6 KB)
one other source for error is HelpMeFind itself…percentages of ancestors’ individual contributions, the numbers won’t be accurate
True but not for any reason having to do with ploidy. The HMF percentage contributions have the more obvious problem of missing ancestral data. Regardless, percentage parentage is a meaningless metric of no use to a hybridizer but bountiful fodder for arcane disputes among dilettantes. Nor is ploidy itself of much significance to rose hybridizers except as a key indicator of the work required to accomplish a particular cross successfully; and as a target, specifically triploidy, that sometimes lends itself to producing abundant blooms.
Helpmefind.com is, in my personal opinion which I offer humbly, the best source for data on any rose cultivar outside of the primary (scientific) literature.
A second shortcoming of using HelpMeFind this way is that you skip over all the roses with no stated ancestry but which are obvious descendants of particular species. Again, I go back to the Gallica connection…if one were to look for DNA markers, one would surely find Gallica and Fedtschenkoana genes in most modern roses today.
Possibly but probably not and, if so, heavily diluted with the residual bearing the obvious trait of perfume which had been under continuous selection pressure until the advent of the hybrid tea at which point color and form became the dominant selection pressures.
Whenever I see “seedling of…” on HelpMeFind, I assume it’s a self
A foolish assumption. Take a look at the ancestry of Tropicana as an extreme case but be assured that a great many unnamed deliberate hybrids are listed as ‘seedling of’.
I’ve been wondering about what could arise if some of the original crosses which led to new classes were repeated, but with different cultivars
Free yourself of the notion of ‘classes’ which are nothing more than marketing schemes, but by all means explore the paths not taken. Perhaps the principal descendency chart could serve as a model in that effort.
I’m still getting things set up here
Where is 'here?
The “romantic lay literature” is what is referenced on HelpMeFind to build its rose lineages. The Martin et al paper constructs a dendrogram, showing relatedness between selected cultivars, and extrapolated to give an idea about the respective classes to which those cultivars belong. I don’t see how that can be used as a reference to support your claim.
Ploidy may not be something which some hybridizers worry about much in their efforts, but it is significant with respect to ancestry. A triploid derived from a diploid X tetraploid will still have 2/3 of its genes coming from its tetraploid parent, despite HelpMeFind showing 50% ancestry from each. If you’re discussing genetic contributions of ancestors, ploidy is significant.
My argument with your conclusions is in the method you used to derive them. You searched rose species on HelpMeFind and counted descendents, and because some species have more listed descendants, you count them as being more significant “founders.” I say that there is a bias here because HelpMeFind does not have complete ancestries for its roses – going only as far as references from “romantic lay literature” can be traced together. As a result, there are numerous gaps between species and cultivars derived from those species – and as such, if you look at the numbers of descendants of those species, you won’t see among the numbers all the cultivars with NO ancestry listed for it or any of its ancestors – unknowns form broken links between cultivars and species ancestors the way you search descendants, even though we “know” there is a chain of descent. You did acknowledge that the species ancestors of the old Europeans were used before detailed breeding records were written, and that Gallica forms the basis of all the old European groups. So let’s use that species as an example.
R. gallica on HelpMeFind has 787 unique descendants listed. Now look up ‘Charles de Mills’ – what’s its ancestry? Unknown. It’s clearly a Gallica, but HelpMeFind has no connection between it and ‘R. gallica’ via its lineage feature. So even though we “know” it’s a descendant of ‘R. gallica’, according to HelpMeFind, it’s not among the 787 unique descendants of that species. And many, many other Gallicas will follow this pattern.
You also acknowledge that this species is in the make-up of Damasks, Centifolias, Albas, besides, of course, Gallicas themselves. And that Hybrid Chinas, Bourbons, Portlands, Hybrid Perpetuals, Hybrid Bourbons, etc. descend from one or more of the previous groups. So, thus, each and every cultivar of those groups should also descend from Gallica. Let’s pick an influential Hybrid Perpetual from an ancestry standpoint – ‘La Reine’, which has 16,343 unique descendants listed on HelpMeFind. Its HelpMeFind lineage goes back as far as ‘William Jesse’ – thus no connection to ‘R. gallica’, and as a consequence, none of the offspring of ‘La Reine’ will be tallied among descendants of ‘R. Gallica’ even though we know that the species is an ancestor of ‘William Jesse’.
The same story will follow when you look at the other species in the ancestry of the old Europeans. So although HelpMeFind lists 725 unique descendants R. fedtschenkoana, for example, that number doesn’t include the thousands whose ancestry contains unknowns.
I do not argue that some species have been used more often recently, and that their usages are better documented. But they were more often bred into existing hybrid cultivars than selected for pure-species variations. And most of those existing hybrid cultivars had European OGR ancestry – the “unknown” wasn’t that they descended from the species that went into that pool, but the exact details of which cultivars were the intermediates.
Oh, and I should have specified about the “whenever I see ‘Seedling of’ I assume selfs”. I was thinking of the context of the old cultivars from the 19th Century and beyond. Those are the types which fascinate me more, so that’s where my mind was focused. As to the notion of “classes”, I actually prefer to stick to them because what I’m interested in reproducing were the older classes that were more about ancestry than function. So, to me, a Noisette will be a China X Moschata cross, but not necessarily involving ‘Old Blush’.
I’m in Middlesex County in NJ.
Thanks for the discussion. I do thank you for providing that further information and article – it’s definitely something I find fascinating.
~Christopher
Don,
I enjoy such charts, but there are always gaps. For instance, you listed ‘Colour Wonder’ without noting its wildly introgressed background.
Tracing the ancestry of its seed parent, ‘Perfecta’, through ‘Karl Herbst’ and ‘Peace’ we come to ‘Margaret McGredy’. This important variety was one of numerous members of the McGredy stable that were derived from Sam II’s work with Rugosa and Cinnamomea.
http://bulbnrose.x10.mx/Roses/breeding/monosom.htm
The pollen parent, ‘Tropicana’, was the culmination of Mathias Tantau Sr.'s work of infusing his stock with the “blood” of Multibracteata and Roxburghii.
http://bulbnrose.x10.mx/Roses/breeding/Wulff/Wulff_roxburghii.html
Aside from its “official” Hybrid Tea background (Tea and Hybrid Perpetual), ‘Colour Wonder’ is descended from Multibracteata, Roxburghii, Rugosa, Cinnamomea, Foetida, and two races of Multiflora (Japanese and Chinese).
Tantau’s ‘Floradora’, also descended from Roxburghii, was a parent of ‘Queen Elizabeth’. And Ralph Moore made use of ‘Floradora’ in breeding Minis.
http://bulbnrose.x10.mx/Roses/breeding/Moore/MOORE.html
Karl
Don,
Schoener made extensive use of various “Giganteas”.
http://bulbnrose.x10.mx/Roses/breeding/Schoener/SchoenerGigantea1932.html
Sadly, I have not been able to find any of them still in existence. It is possible, of course, that some of his productions were used by other breeders.
‘Improved Cecile Brunner’ (Rosy Morn) was raised from ‘Dainty Bess’ pollinated by a double flowered Gigantea hybrid descended in part from ‘Mlle Cecile Brunner’. This variety, growing at the Heritage garden in San Jose, did not impress me.
http://bulbnrose.x10.mx/Roses/Rose_Pictures/I/improvedcecile.html
On the other hand, ‘Susan Louise’ is a beautiful everblooming variety that has the flowers of its mother (and father?) ‘Belle Portugaise’, but a shade darker.
http://bulbnrose.x10.mx/Roses/Rose_Pictures/S/SusanLouise.html
‘Belle Portugaise’ is one of Cayeux’s Gigantea hybrids raised while he was director of the Botanical Gardens of Lisbon. It grows well in California. And in suitable locations it can flower intermittently from October through April or May.
http://bulbnrose.x10.mx/Roses/breeding/gigantea.htm
Karl
Karl, thanks for these notes.
It’s true that other rose species appear in the ancestry of some modern roses. These are great examples. Hopefully we can add to them. We should also mention Rudolph Geschwind who did a lot with species - his Gruss an Teplitz lurks in Tropicana for instance; Michael Horvath and his work with American species roses; and the Canadians Robert Simonet, Robert Erskine, George Bugnet, Percy Wright and Frank L. Skinner.
The work with species roses continues. The proteges of Ralph Moore figure prominently in that work among whom Rob Rippetoe (the person who first drew my attention to the obvious contribution of gigantea), Kim Rupert, Jim Sproul and Paul Barden stand out. Warren Millington’s species work has progressed to great depth.