Tantau, roxburghii and Baby Chateau

I am sort of thinking out loud in this post but I do get to the point at the end of it so please bear with me.

I have been wondering: what was Mathias Tantau up to when he crossed R. roxburghii with Baby Chateau? The cross yielded two important ancestors of Tropicana, Cinnabar and Kathe Devigneau, and he saw fit to backcross these several times in the development of Tropicana.

From the results it appears that the combination gives just the right biochemistry to favor the formation of pelargonin (scarlet red) while supressing the formation of cyanin (pink). However, there must be more to it because the yellow carotenoids are a big part of the orange color of Tropicana too and there is, of course, more to Tropicana than just the color - such as form, fitness and the like.

Roxburghii’s form is a curious sort of zig-zag lattice, and it has a characteristic peeling bark. You can see these in, for instance, the canes of Little Bo Peep. Tropicana shows none of this. The integration of roxburghii with Baby Chateau must therefore have involved selection for form as well as color.

I have a young roxburghii which I expect (hope) will bloom next year. It would be very interesting and maybe even instructive to see what comes out of a cross with Baby Chateau, but HMF lists only the Cranford (Brooklyn) as having it. If you know of someone else who has it, or have it yourself, please let me know. We could swap pollen and see the results for ourselves.

Of all of the modern bush roses I see thriving in barren areas in the countryside, poor garden environments and abandoned homes, it is Tropicanna that I see the most. That plant can survive horrible conditions and manage to bloom in full force. It is amazing lol.

I have seen that the ploidy is supposedly triploid, which is interesting.

I have one seedling from open pollination of r. roxburghii normalis in 2006.

In 2007 I pollinated with r.h. Cantabrigiensis.

At coarse view all 2008 seedlings look like the one from 2007.

2008 seedlings from different hips differ in foliage color and all seedlings have less leaflets than the seedparent.

Some are good growers with one strong stem, others cover the ground, having more than one shoot.

When I took pollen in 2009, I was wondering that the pollen was already dry and dusty.

Most likely r. roxburghii normalis is selfing at a moment when the hip is still tightly closed, thus all my 2008 seedlings are not from r. cantabrigiensis.

See this article



My experience with R.roxburghii normalis is that in a pollen rich environment, it seems to prefer foreign pollen. When my plant was growing alone, it produced about 5 or 6 very large round seeds per hip from self pollination. After several years when its diploid neighbours matured, it produced very large numbers of squashed, smaller seeds per hip. These seeds germinated normally & produced seedlings with intermediate characteristics. Many were exceptionally thorny. In recent years when I’ve applied foreign pollen, only a small number of seeds per hip were produced. So far there are no blooms.

I’m finding that while roxburghii normalis is unfailingly healthy, most seedlings, even from healthy partners are not. So I’m constantly culling for health, in addition to excessive thorniness. Winter takes care of hardiness issues. The early selfs didn’t survive over winter.

My climate is borderline for roxburghii so it needs to be severely pruned every few years. I’m hoping to get a hybrid that resembles the parent plant that is cane hardy in my zone.

There may not be any roxburghii in Tantau’s seedlings from the cross Baby Chateau X R. roxburghii. One would expect that most of the seedlings from that cross would be triploid, while all of Tantau’s are tetraploid. Also, Tantau’s seedlings have no discernible roxburghii characteristics. Other roxburghii seedlings, like Coryana, have obvious roxburghii characteristics. Tantau’s seedings might be apomictic seedlings of Baby Chateau, or perhaps some other pollen crept in.

Link: www.rosehybridizers.org/forum/message.php?topid=6402#6404

Most likely r. roxburghii normalis is selfing at a moment when the hip is still tightly closed

Rene, thanks for this observation. Wulff’s article on roxburghii is informative, thanks for mentioning it.

My climate is borderline for roxburghii

Lydia, I am in a 5B/6A zone. I lost a plant of roxburghii in its first winter in the ground (1986-7 iirc) so I haven’t ventured to try it again until now. I’m glad to hear it seems to be an effective seed parent, hopefully I can keep it alive long enough to find out how useful it might be this time around.

There may not be any roxburghii in Tantau’s seedlings from the cross Baby Chateau X R. roxburghii.

Jim, this seems to be a valid concern. I can’t help thinking that Mathias Tantau knew what he was doing, though. As you mention in your earlier comments, DNA would be definitive but, lacking that, it really comes down to making the crosses to duplicate his work and seeing what transpires. The hard part is getting Baby Chateau, or its pollen. An alternative would be to identify surrogates based on parentage or pigmentation, which is something I’m pondering now.


“DNA would be definitive but, lacking that, it really comes down to making the crosses to duplicate his work and seeing what transpires.”

Actually, DNA testing isn’t always capable of providing a definitive answer. Case in point; Dr. David Byrne tested ‘Basye’s Purple’ years ago to see what species were involved in its pedigree (some people had their doubts about Basye’s stated parentage) and his results suggested that there was no R. foliolosa in it! Of course, that isn’t right because myself and others have obtained very similar looking plants using R. foliolosa as a seed parent. Somehow the plant is masking its genetic pedigree in ways we do not yet understand. I have done enough work with an assortment of species to know that odd things can (and often do) happen when crossed into modern breeding lines. ‘Floradora’ might really be a R. roxburghii hybrid that simply doesn’t exhibit any of the species characteristics.

A modern cross of similar breeding, See below:

Link: www.helpmefind.com/plant/l.php?l=2.41380.0

It would certainly be news if DNA evidence, could in fact be found inaccurate. Such a finding would have broad implications.

I’ve certainly come up with some odd results in terms of hybridizing but until proved otherwise, I’d abide by any findings obtained from DNA testing.

That’s a great point about DNA testing. I wouldn’t say I wouldn’t trust it, but we just need to have the limitations at hand in mind. It is very powerful if we know the supposed clones that produced a hybrid in question. For instance, I tested DayDream with AFLP markers and its supposed parents Lavender Dream and Henry Kelsey and learned that there was no Henry Kelsey contribution in it and it is very convincing that DayDream is a diplospory apomict.

When we try to expand things to see if a whole species may be involved we need to have an idea of some species specific DNA markers. If that work has not been done before, we really don’t know if the R. foliolosa for instance we have access to or not would have the same DNA markers as the one that could have been the parent plant used by Dr. Basye. If we can get a wide sampling of a species and discover some markers that seem unique or frequently common to the species that helps when we try to make these wider extrapolations.

I like the respect that was extended to Tantau in that “he knew what he was doing”. I like to error on the side of respect and give the benefit of the doubt unless we have strong evidence otherwise.

Here are just a couple examples for fun to consider:

Rosa x kordesii has significantly more petals than 5. Rosa x kordesii is supposedly a self-pollination of ‘Max Graf’, as single. More than 5 petals (it is semi-double to just barely double) is dominant. So, there is incredibly strong evidence that although Kordes may have self pollinated ‘Max Graf’ the male is likely some other, double-flowered parent. I think that in this case we are justified not to perpetuate this story as fact without also bringing this major descrepancy into question. We can easily do some marker analysis with ‘Max Graf’ and ‘Rosa x kordesii’ to see if R. x kordesii has markers not present in ‘Max Graf’. We may not be able to say who the dad was, but at least that it wasn’t a self if it is filled with a lot of markers not present in Max Graf.

Another case in point is in the recent issue of Floriculture and Ornamental Biotechnology I edited. There is a great reproduce biology paper. The authors articulated Dr. Eileen Erlanson must have been in error with her data presented on reproductive biology of Rosa setigera. They claim it is a dioecious species (separate male and female plants) and her self polliation data must be in error. Well, even though the senior author has done much to understand R. setigera and its dioecous nature, there are some hermaphrodites out there. In some of his writings I believe he claims none were reported so far that had both male and female fertility. One of them at the MN Landscape Arboretum that is a fully functional male does set some hips and has limited female fertility with viable seed inside. I shared my photographs of it with the authors and information of some of its seedlings also being hermaphrodites and they became more gentle in their description of Dr. Erlanson.

It would sure be fun to learn more about the contribution or lack of from R. roxb. in these hybrids from Tantau. Hopefully someday people with access to some good DNA and other tools will be willing to help answer this question.



I was in a hurry earlier and misspelled a few things above. I meant reproductive biology paper.

Paul wrote:

Somehow the plant is masking its genetic pedigree in ways we do not yet understand.

I guess that’s really my point, too. Partly I think that roxburghii contributed a unique bit of biochemistry to the color machinery, but also there seems to be something about crosses with roxburghii that mask its primitive features. I think it may be that they are one and the same phenomenon.

We know from Yokoi (see reference 7a in the link below) that Floradora, and by inference and observation Cinnabar and Kathe Devigneau, produce lots of pelargonin and no peonin. Yet roxburghii makes peonin and no pelargonin. In theory, at least, first generation offspring of any cross with roxburghii should make peonin and/or-only cyanin, but no pelargonin. The following diagram from Eugster (reference 1 in the link below) shows why:

In order for a rose to make pelargonin the pathway from Flavanol A to Flavanol B must be blocked and it must also have the genes that complete the pathway to pelargonin (Pel).

Since roxburghii makes peonin, the Flavanol A to Flavanol B conversion is not blocked. Chances are very good that it is homozygous for this, too, since there don’t seem to be purely cyanin producing offspring of roxburghii, or any that produce pelargonin.

Yet we have at least three F1 progeny of roxburghii that produce pelargonin. Unless Tantau happened to have a variant of roxburghii that produced pelargonin, there must be a reason for this apparent anomoly.

I think that the evolutionary relationship of roxburghii to other roses holds the key.

(see http://holeman.org/images/rosatree.jpg)

The sequence of evolution starts with the most primitive rose, hulthemia. The first few branches yielded foetida, then sericea, then hugonis and then the spinosissimas, then all the North American species (and acicularis, the circumpolar rose).

At the next branch roxburghii splits from all other roses, including those which comprise the gene pool of modern roses. The genetic distance of roxburghii, which is indicated by the horizontal lines (but not the vertical ones) shows that it evolved much further than any of those which split off before or after it with the exception of canina.

Another way to put this is that it’s genes have the greatest number of differences from all other roses. It is not different enough that it can’t be mated with other roses, but it is possible (and I think, likely) that the versions of its genes are different enough that they don’t function ‘properly’ when paired up with those of other roses.

There are a lot of possible differences, ranging from producing enzymes with different rate constants and binding coeffecients to differences in gene regulation. Whatever the reason, pairings of other roses with roxburghii seem to give offspring in which the traits of the other rose dominate.

This phenomenon would not be unprecidented. As I mentioned, canina is the only other rose that has diverged farther than roxburghii. In this case, offspring are always dominated by the traits of canina rather than the other parent but, its unique meiosis notwithstanding, the principle is the same.

Link: holeman.org/Dons%20Handy%20Rose%20Pigment%20Color%20Chart.pdf

Over the three years (to the day) since I first started this thread I have continued to ponder why Tantau’s three hybrids of Baby Chateau with R. roxburghii appear to have no roxburghii in them (Cinnabar, Floradora and Kathe Duvugneau). I think the answer may be the phenomenon of dihaploidization. I dredged up the thread because I came across a wikipedia page that discusses dihaploidization.

Essentially, wide crosses can sometimes result in loss of the chromosomes of one parent which, if doubled either purposely or through serendipity, results in a seedling that is completely homozygous for all its genes.

If this is truly what happened with Tantau’s roxburghii hybrids then roxburghii would be a ‘magic bullet’ of sorts for producing dihaploids and, thus, would be a valuable breeding tool. It’s a bear to overwinter because it is so tender but I do have a couple of cultivars of roxburghii that I have been working with, so far with nothing exceptional to report.

Correct. That is what I have been implying that Rosa plimula did for me.

The current link for the Wulff (1954) article.

In addition to dihaploidization, there is the matter of partial hybridization.

An unusual mechanism in crosses between cultivated sunflower and perennial Helianthus may lead to partial hybrid in first generation plants

N. Faure, H. Serieys, A. Bervillé, E. Cazaux and F. Kaan

"Hybridisation between the annual diploid sunflower (Helianthus annuus) and the perennial diploid species Helianthus mollis and Helianthus orgyalis was studied. Cultivated sunflower was either used as female or male parent. Progenies were obtained by means of a normal crossing or embryo rescue procedures, but hybridisation success was low. All plants examined cytologically appeared to be diploid. Plants showed a great morphological variation, but they resembled the female parent-type predominantly. In crosses with cultivated sunflower pollinated by one of the two Helianthus perennial species, 35% of the plants had a sunflower phenotype. We applied RAPD and RFLP markers to determine the genetic constitution of progenies. It appeared that the contribution of the wild male parent was reduced in comparison with the female parent: hybridisation occurred at different levels depending on the plants. On average, only 5% of the minimum number of expected RAPD and RFLP bands from the male parents were recovered in plants produced from mature seeds after pollination of sunflower by H. mollis. "

This phenomenon has also been reported in crosses between Solanum tuberosum and S. phureja. Some clones of the latter species produce “clean” dihaploid S. tuberosum, while other clones leave a small “genetic footprint”.

Also, C. G. van Tubergen, Junr. (1906) wrote:

“LILIUM.— Very numerous crosses among various species have been effected, and many seedlings are still under observation; a good and noteworthy race has sprung from the crossing of Lilium Martagon album with L. Hansoni. It is of particular interest to note that whereas L. Martagon album, if raised from seed, almost always comes perfectly true, scarcely ten among a thousand plants reverting to the typical purple Martagon Lily; out of the mingling of L. Martagon album with L. Hansoni not a single white martagon occurred. All plants (several hundreds) that showed no influence of the pollen-parent (L. Hansoni) reverted to the typical purple Martagon Lily. Those that showed the influence of L. Hansoni developed into stately, tall-growing lilies with broad, dark green foliage in whorls and pyramidal spikes, composed of very numerous flowers.”

I could go on.

As for the weirdness of ‘Baby Chateau’ producing tetraploid offspring with no external signs of Roxburghii (Wulff wrote, “Only anatomical studies revealed a certain similarity and relationship to the latter species.”) we may note that its grandparent ‘Eva’ (= Robin Hood x J.C. Thornton) also has more chromosomes than expected (4x rather than 3x).

I admit that I am surprised that no peonin was found in Tantau’s hybrids. Peonin + pelargonin usually gives a brighter orange than cyanin + pelargonin. Peonin (non-blueing) is apparently produced from cyanin, which often “blues”, and thus dulls the orange.

The pelargonin presumably comes from ‘Baby Chateau’. Stoddard (1980) crossed ‘Orangeade’ with a broad assortment of roses. “Crosses by R. laxa, R. suffulta, and Victor Hugo unexpectedly produced strong and pure oranges among their very sparse populations.” Suffulta, at least, has peonin-colored flowers.

Has anyone seen ‘Baby Chateau’ in the U.S.? I saw Tantau’s three hybrids, and R. roxburghii, at the Heritage Rose Garden, San Jose, CA, but might have learned more if I could have compared them with ‘Baby Chateau’. I wonder if it is as vigorous as its offspring. They reportedly have more resistance to fungal infections than ‘BC’.


I’d always really wanted to grow Chateau, Karl and wondered what it must have been like. After seeing the photos on HMF, I’m honestly not surprised it fell out of commerce and disappeared here.


I agree that the HMF pictures are not impressive, but there must be something to ‘Baby Chateau’ that has attracted breeders. Some of the children and grandchildren are Gertrude Westphal (seed parent of Marlena), Independence, Karl Weinhausen, Little Darling, Winifred Coulter, Sparrieshoop, Florence Mary Morse, Governor Rosellini, Rudolph Timm, Siren, Mexico, Red Camellia, Johannes Boettner, Janida, Wilhelm Hansmann, Winifred Coulter, Major Gagarin, Kathy Fiscus, Sommermärchen, Chaperon Rouge, Sea of Fire, Sister Kenny, Serdce Danko, etc.

If it is really as weak as the one photo shows, then Rosa roxburghii must have contributed more than a little to Tantau’s vigorous hybrids.



Peonin + pelargonin usually gives a brighter orange than cyanin + pelargonin.

Karl, I’m surprised to see this because, iirc, peonin and pelarginin are mutually exclusive because of the biosynthetic pathway that produces them. What are the roses that you had in mind?

What I’ve read (in Roses, by Jack Harkness, if I recall correctly) was the Chateau de Clos Vougeot was a very popular, intensely fragrant, “black” HT. From growing it, it was (is) also one of the most terrible rose bushes you could ever be cursed with. Totally lacking vigor to Twentieth Century (and later) eyes; addicted to mildew and black spot; demanding to be budded to grow and produce any real quantity of flowers, much less quality; throwing its short, begrudgingly produced canes out parallel to the ground so the few flowers you receive can flop into the mud around the bush. Yet, it was popular because of the black, velvet, intensely fragrant blooms. Reportedly, it received its name to honor the Vougeot wine making house because the flower was the color of their wine.

That said, twenty-eight years later, Kordes raised Baby Chateau, giving it that name because it reportedly resembled Chateau de Clos Vougeot in flower. Ironically, it was raised from Ami Quinard, itself a Vougeot seedling. I’m sure to 1936 German eyes, this was quite a break through. Polys and hybrid polys (floribundas) were taking on greater HT traits, larger flowers and increasing in their fragrance. Suddenly, at this terrible economic and political time, a new seedling appeared with “that” flower and fragrance. It reportedly made quite a sensation and, as you can see from its lineage report on HMF, breeders jumped on it like hens on a June bug! Actually very logical considering what the rose represented at that time. Through Twenty-first Century eyes, it’s a dawg. Definitely something to consider as a museum piece. Kim


Peonin + pelargonin usually gives a brighter orange than cyanin + pelargonin.

Karl, I’m surprised to see this because, iirc, peonin and pelarginin are mutually exclusive because of the biosynthetic pathway that produces them. What are the roses that you had in mind?[/quote]


From what I’ve read, peonin is produced from cyanin or one of its precursors. Peonin is clear pink without blue tones (except when it is highly concentrated, as in ‘Hansa’). Cyanin, on the other hand, tends to form complexes with whatever co-pigment is present. So, with peonin synthesis depleting the cyanin, there is less “blue” to dull the color of any pelargonin that is present.

I’ve also noted that roses with pelargonin and cyanin tend to be brighter orange in cool weather or when they open in the shade. ‘Margo Koster’ is one example.