Fragrance heritability in Hybrid Tea roses




The link isn’t working. Is this a pay for article website?

I’m one who is interested in the fragrance of hybrid teas. (I have zero luck-- and I thought i had perfect crosses such as Sutter’s Gold X Compassion, which is quite a nifty and disease resistant seedling but lacks all fragrance.)

The link is almost as long as the abstract, and the space for the link was not long enough. Copy and paste the link that Henry provided in the text of his message. The abstract says, in essence, that non-fragrant roses don’t have much fragrance. Please correct me if I’m wrong.

But, if you have already learned that there’s no such thing as a perfect cross, you’re on the right path.

Fragrance is more likely in seedlings resulting from crosses of fragrant parents, but it is not assured.


The authors studied 107 offspring of the cross between the roses:

“Anna” Plant Search


“Bo” 'Bô' Rose . (There should be a ^ symbol above the o in Bo.)

The authors state: “Anna has a typical Chinese fragrance (dominated by DMT and exempt from monoterpene alcohols) whereas Bo is closer to Europen rose fragrance (presence of high quantities of monoterpene alcohols and low quantities of DMT).”

3 offspring were particularly “well balanced” in their combination of the above.

The final sentence of the paper states: This association has already been seen in some roses like Double Delight, a Hybrid Tea rose that may be used as a garden rose or for cut flowers."

I just got a catalog in the mail today for new (expensive) books. There is one on Biology of Floral Scent by Natalia Dudareva and Eran Pichersky. A student/friend, Michael Long, who got his Masters at the U of MN (where I am at) went to work in Natalia’s lab with some of her rose work. I haven’t talked with Michael in a while. It would be interesting to learn more about Natalia’s work and what practical information from her biochemical and genetics work can be applied to traditional rose breeding.

Here is a link to her book



Henry’s linked article has some really interesting articles with links in the bibliography.

One suggests a different china lineage based on the measured scents of china roses. Another measured scent output re time of day.

I’m not sure that I’m ready to think of roses and their scents with chemical terms attached, but if I can, it will broaden my view of roses and may suggest those that should be planted in cooler locations to expand the period of maximum scent release.

And the one thing that really surprised me was that Rosa multiflora’s scent was 100% a single chemical, whose name I haven’t yet learned.

The paper for which Dr. Kuska has provided this new abstract follows up on a theme introduced in an earlier paper by some of the same researchers. I discussed that paper, “The evolution of the modern rose”, in this thread:

The theme of that earlier paper was that modern roses can be characterized, more or less, as having varying degrees influence derived from both European and Chinese rose ancestry. In this newer paper, the authors present a specific example of these influences in the form of the heritability of certain (limited) classes of fragrance compounds.

European and Chinese roses are represented here by proxy. Rather than resorting to Old Garden Roses or even species roses, they have chosen instead two very recent hybrid tea introductions by the breeder Paul Pekmez. Anna represents Chinese influence and Bo represents European influence.

Rose fragrance results from the production of volatile organic compounds, or VOCs. The scents of VOCs from Chinese roses have ‘green’, ‘fresh’ and sometimes ‘tea’ characteristics, but not ‘sweet’ or ‘fruity’ notes. VOCs of Chinese roses are dominated by the methyl ethers of a chemical called oricinal and some other closely related compounds and these just don’t smell very appealing, and in fact do not smell very much at all.

The VOCs of European roses, conversely, are by dominated compounds called terpenes (geraniol, nerol) which give classic ‘rose’ characteristics to the fragrance and by low molecular weight alchohols and thier esters, which smell quite a strongly and tend to be sweet and fruity.

So Anna was chosen because it produces mostly oricinol methyl ethers and Bo was chosen because it makes mostly terpenes (but, oddly, does not make certain important alcohol esters).

Anna was crossed with Bo, 107 progeny were grown up and the compositions of the VOCs of these progeny were analyzed continuously over a period of three years. The data from these analyses was crunched with some fancy software. Some confusing looking graphs were created from the results of the data crunching. The graphs are presented as Figures 1 and 2.

Figure 1 (not illustrated here) is actually a set of six graphs each representing a single class of chemical compounds present in the VOCs of the progeny and the two parents. The X axis of each of these graphs represents some highly abstract statistical metric incomprehensible to mere mortals but which can be used to fix the position of each progeny relative to the others in some statistically meaningful way. The Y axis simply represents the amount of that compound made by the progeny. So each point represents the amount of that compound which a particular progeny made related to the other progeny in some statistically meaningful way.

The point of making the graphs in Figure 1 is to look for some regular pattern, usually a straight line, which would then indicate that there is a correlation between that compound and the progeny, hence having something to do with patterned inheritance.

Only one class of compounds, the monoterpenes, was found to have such a correlation - to segregate in a relatively simple way among progeny. That is to say some progeny made only tiny amounts of monoterpenes, some a little more, some a little more still, and so on. This graph supports thier finding that that monoterpenes segregated at about a 50% ratio among progeny.

Figure 2 addresses the more complex inheritance patterns of the remaining classes of compounds.

Some progeny cluster around Anna, others cluster around Bo. These progeny make about the same amount of each compound as the parent around which they cluster, with others falling in a gradient between the two. However, a significant number of progeny do not cluster around or between the parents. Instead, they take up “extreme positions”, making greater and lesser amounts of each compound than thier parents. This is especially true for benzyl alcohol and for the fatty acid derivitives hexanaldehyde and hexenaldehyde. And herein lies the rub.

It happens that benzyl alchohol is the presursor for 2-phenyl ethanol (2PE), one of the most abundant compounds expressed by modern garden roses and by the very important perfumery roses, the Damasks. 2PE is considered to be a European rose characteristic yet their European proxy parent Bo does not make 2PE. If it did, one could argue that some of the outlying progeny would be extremely fragrant and that the fragrance would be different from either parent.

Another glaring omission from this analysis are the esters of any of these various compounds because, while the base compounds themselves are odiferous, their various esters can be far more potent scents. For instance, a major scent component of the rose Fragrant Cloud is 2-phenylethyl acetate, the ester of 2PE (Gang, David R. “Evolution of flavors and scents”. Annual Review of Plant Biology, Volume 56 p.301 (2005).

This issue speaks to the root of the problem with the overall thesis that modern roses are basically dichotomous hybrids between European and Chinese roses.

There is no doubt that the gene pool of modern roses is astonishingly narrow, and the data here and in other related papers does show heavy influences of both these classes of roses but, as with the 2PE issue and the issue I mentioned in an earlier discussion regarding Champney’s Pink Cluster, there are some missing pieces.

They have demonstrated, though, that the inheritance pattern for a complex model system involving fragrance is itself complex, but can be defined in such a way as to be a useful tool for hybridizers. For breeders without access to a chemistry laboratory, though, the lesson reduces to a rule we already knew, related from Ralph Moore by Paul Barden: a given cross requires roughly a hundred progeny in order to illustrate the full spectrum of variability that will be expressed by the progeny.


The value added to Moore’s rule by this paper is that you can expect some extreme outliers among the progeny which have characteristics much beyond either parent.

Wow. Thank you, Don, again.

About the scents of those China and Tea VOC’s not being sweet or fruity…I suspect that is a subjective impression, as I find many Chinas both sweet and fruity, though subtle. Ducher is an example of the scent I describe.

It would be interesting to see the contested OGR class of Portlands/Damask Perpetuals evaluated for the presence of China VOC’s. Some believe their repeat blooming is not due to any China rose breeding. The trickiest part would be determining that any individual cultivar is actually a Portland or Damask Perpetual and not a compact Hybrid Perpetual.

I’ll second Cass’ description of ‘Ducher’ as fruity and sweet - to my nose it smells a lot like peaches.

'Ducher is fertile too. I have seedlings using it coming along now.

Since this thread got brought up again I’ll post a link to one of my ‘Ducher’ hybrids. It is fragrant.


Fragrance heritability in Hybrid Tea roses
Magali Cherri-Martin, Frédéric Jullien, Philippe Heizmanna Sylvie Baudino

Haven’t had time to read the article yet, but the comments alone have been very instructive.

Thank you!

Central NJ, zone 7a

The parent don’t have to have perfume to create perfumed offspring. I have two of mine Elara and Arahan , who have no perfume but yet produce the most highly perfumed seedlings. If I want perfume it is these two I will go to first and crossing these with other perfumed rose makes for a more interesting perfume note. These two are my very early roses and I have had time to work this out . A lot of times you don’t really need to go searching , its right under your nose all the time.


1 Like

Another paper concerning rose perfume:
Baser, K.H.C… Turkish Rose OiI. Perfumer & Flavorist 17: 45-52 (May/June 1992)

Figure 1. GC of factory-produced Turkish rose oil (1989)
Total components detected: 85; components identified: 68; 96% of total components detected.

It is interesting to compare this report to Leffingwell: β-Damascenone (0.14% of oil, 70.0% relative odor) β-Ionone (0.03% of oil, 19.2% relative oil).
That is, these two “Rose Ketones”, not listed by Baser, make up only 0.17% of the rose oil, but contribute 89.2% scent.

The degradation of carotene by Carotenoid dioxygenase is illustrated here:

Edwards (1960) discussed some of the more fragrant roses.

He drew attention to the fact that old roses were not universally fragrant. In fact, in a study of many roses, old and new, it turned out that in each era only about 1/3 were considered fragrant to very fragrant, 1/3 were not particularly fragrant, and the remaining 1/3 were only moderately fragrant.

Furthermore, “The panel discovered that few of the more fragrant old-time roses equalled the intensity and variety of fragrance given off by such modern roses as MIRANDY, MISS CLIPPER, HEART’S DESIRE, SUTTER’S GOLD, SAN FERNANDO, or TALLYHO.”

Much of the increased perfume of these modern roses can be traced to the introduction of Rosa foetida, through the Pernetianas. This was a welcome surprise, considering that the species’ only apparent virtue was its color.

This is a bit off topic, but I’m curious as to everyone’s opinion. What in your experience is the strongest citrus scented rose? The best I have smelled so far for a citrus scent is AF. I’ve been wondering if the citrus scent in AF is somehow derived from Rosa foetida, since I’ve also heard that Sutter’s Gold is citrus scented.

‘Intrigue’ (Warriner 1982) combines lemon with old rose. RICA ‘Sombreuil’ is lemon with sweet apple.

LeGrice (1976) thought ‘Harriny’ had a lemon scent, but to my nose it is distinctly grapefruit.
“Lemon is a common perfume mostly in combination. I would prefer to think of it as lemon-scented verbena (Lippia citriodora). A rose with this perfume is Sutter’s Gold, while City of Hereford, Harriny and Ophelia are a mixture of damask and lemon, the former being more evident in the younger flower.”

Prairie Star had the strongest grapefruit rind scent of any I’ve encountered. Many have lemon zest elements, at least at varying stages of bloom maturity.

I feel like kicking myself when I find that I have pictures of roses, such as ‘Prairie Star’, with no notes on scent.

I just found and posted a newspaper from 1966 with Dr. Lammerts’ comments on fragrant roses. I don’t recall ‘Golden Showers’ as having an orris scent, nor does “lemony” ring a bell when I think of ‘Tiffany’. But fragrance varies with environmental conditions, and noses differ in their sensitivities.