Rose Hybridizers Assoc. Newsletter 34(1): 18-19. 2003.
Rosa foetida and fragrance
Karl King

I've noticed some Rosa foetida hybrids that reportedly rebloom. These are listed as immediate offspring, rather than F2 or later generations. Rosa foetida is a dubious species at best, and we may wonder what its ancestors were. The occurrence of rebloomers such as 'Lawrence Johnston' (Mme Eugene Verdier x R. foetida bicolor), 'Geheimrat Dr. Mittweg [(Mme Norbert Levavasseur x Trier) x R. foetida bicolor] and 'Naïr' (R. foetida bicolor x R. wichuralana seedling) make the question even more intriguing. ('Geheinirat Dr. Mittweg' is listed as a diploid Polyantha, which is also a bit odd.)

Some other varieties with Poly and Foetida ancestry are:

'Ausonius' is listed as a Hybrid Musk, which makes as much sense as calling Foetida hybrids Polyanthas. It is available from Vintage Gardens. Of these, I've raised only 'Mevr. Nathalie Nypels', which has no obvious Foetida influence aside from the strong perfume. This variety is more readily available than the others.

'Bloomfield Dainty' is rather farther removed from its Foetida roots, but combines bright yellow color with a rich "musk" fragrance and deep green, healthy leaves.

Another hybrid involving Rosa foetida and a Poly was raised by the great Russian plant breeder, Ivan Michurin. He crossed the 'Persian Yellow' with a damask (I think it was 'Kazanlik'), then crossed one of the hybrids with 'Clothilde Soupert'. The new attar rose, called 'Slava Sveta', displayed an odd assortment of ancestral qualities. The leaves and shoots were identical with those of 'Persian Yellow', the flowers were similar in shape and structure to 'Clothilde Soupert', while the fragrance was that of the Damask, but much intensified. Of it, Michurin said, "A trial steam distillation by means of a small laboratory still showed that the flowers contained a much higher percentage of attar than the regular Damask rose" and "...it is interesting that the disagreeable odour of the flowers of the grandfather plant — the Capuchin Yellow Rose — far from spoiling this fragrance, considerably intensified and improved it."

We all know that heredity, to a great degree, involves discrete units. It is interesting, but should not be too surprising, that Foetida might contribute to the quantity of fragrant substances in the hybrids while not also contributing its own unpleasant scent.

Of the various Poly-Foetida derivatives, I have grown only 'Mevr. Nathalie Nypels' which does have an unusually strong and pleasant fragrance.

It would be useful to determine whether genes of Foetida that amplify fragrance are selectively preserved in crosses with Polys, particularly the diploids. It is also curious that none of Michurin's hybrids possessed the yellow color of Foetida. How many deep yellow HTs and Floribundas lack fragrance? Le Grice's 'Allgold' has unfading yellow flowers (in hot or cool weather) and very healthy foliage, but no scent worth noting.

It appears that the genes for the yellow color of Foetida are not linked to those which amplify fragrance. Furthermore, these different genes appear to be differentially retained in some crosses where chromosome loss occurs. When we cross a diploid with a tetraploid we usually expect the offspring to be triploids. This is not always the case. Some of the Poly-Foetida hybrids are diploid, while Tantau's three hybrids of 'Baby Chateau' x Rosa roxburghii (diploid) aretetraploid. These are partial hybrids, which are not uncommon.

Poly-Foetida hybrids and their derivatives may not inherit the deep yellow color, but the amplification of fragrance is a highly desirable quality. In addition, 'Mevr. Nathalie Nypels' is unusually tolerant of heat and drought, which could be an inheritance from R. foetida. Hardiness may be another virtue of Foetida that passes to its offspring, but I have no definite evidence for this.

Michurin commented that 'Slava Sveta' so closely resembled R. lutea (=R. foetida) that it was classified with that species. Similarly, William Paul (1848) described a form of R. lutea called 'Double Blush' or 'Victoria': "flowers light rosy pink, their centre buff; form, cupped. Very sweet." This other potential of Foetida to amplify fragrance has been ignored too long.

I think we have plenty of good yellow roses that are free from Foetida's worst qualities. On the other hand, strongly fragrant Polys are still in short supply — and some climbers (e.g. Hyb. Kordesii) might benefit from some fragrance enhancement.

It's nice to know that Foetida is not a one-trick pony.


Rose Hybridizers Assoc. Newsletter 34(3): 17-19. 2003.
Inheritance of Yellow Flower Color in Roses
Karl King

The inheritance of any trait becomes more complex as the number of independent factors increases. If we are unaware of one or more factors, the problem may be beyond analysis. The inheritance of yellow coloring in roses has long baffled rose breeders because they lacked one important piece of the puzzle. Dr. Lammerts (The Scientific Basis of Rose Breeding, American Rose Annual. 1945) wrote:

*Or so I thought in 2003. In fact, Harison's Yellow was in commerce nearly a decade before Persian Yellow reached the West.

"Deep yellow such as that found in Goldenes Mainz is recessive to light yellow. Hence to recover deep yellow colors, one must backcross to the deep yellow varieties. Also white is recessive to cream, buff, or light yellow." This seems easy enough, but Percy Wright raised his sun-resistant Hazeldean by crossing the white-flowered Rosa altaica x Harison's Yellow, the latter being from R. altaica x Persian Yellow.* In this case, at least, a strong yellow coloration survived two generations of out-crossing.

Ivan Michurin, the great Russian plant breeder, pollinated Kazanlik x Persian Yellow. None of the seedlings bore flowers with any yellow color, which persuaded him that the R. canina rootstocks had robbed the plants of their color. This may seem an odd conclusion, but he observed that almost all of the rootstocks of these grafted plants had developed without branches or fibrils. R. canina stocks grafted with other rose varieties had normally branched roots. If the scion could influence the stocks so markedly, it was not unreasonable for Michurin to suspect a reciprocal influence, even if he was wrong in this case.

Michurin pollinated one of the seedlings by Clothilde Soupert, which resulted in a new attar rose, Slava Sveta, with pink flowers similar to those of Clothilde Soupert, but with leaves and stems identical to those of Persian Yellow. The fragrance was that of a Damask, though much intensified. The flowers also yielded a greater proportion of attar than Kazanlik.

William Paul (The Rose Garden. 1848) described a rose reminiscent of Slava Sveta named Double Blush or Victoria:

"flowers light rosy pink, their centre buff, form, cupped. Very sweet." He listed this variety with the forms of R. lutea (=foetida).

Occasionally a yellow flowered variety will turn up unexpectedly. In a note to rec.gardens.roses (Feb 18, 1996) Sam McGredy wrote, "I never like to dispute a breeder's given parentage, but I have some trouble working out where my friend the late Alex Cocker got the beautiful bronzy colour [of B&H Gold]. The declared parentage is all pink."

Hybrids involving Dwarf Polyanthas or Synstyles with forms of Rosa foetida have given unexpected results. Mevr. Nathalie Nypels [Orleans Rose x (Comtesse du Cayla x R. foetida bicolor)] is bright pink and very fragrant. Equally surprising is Polaris [(R. wichuraiana x R. setigera) x R foetida bicolor): very fragrant, white. Tip-Top (Trier x R. foetida bicolor seedling), inherited some of the color of its seed parent but no fragrance worth noting.

I have observed that among the Tea-Noisettes the brightest yellows also have the largest flowers. Cluster-flowered varieties always (in my experience) quickly fade to cream or white, though these typically exceed the brighter varieties in fragrance. In fact, the larger yellows such as Chromatella, William Allen Richardson and Crepuscule have little perfume, and that is only the terpene scent of Tea roses rather than the potent Musk sweetness of other Noisettes.

At length, and quite by accident, I learned that many fragrant flowers derive their scent from the oxidative degradation of carotene. Yellow carotene is synthesized, but is broken down into fragrant substances often before the carotene can be seen as a color. The flowers of Rosa moschata are light yellow if you catch them early in the morning, or on cool and cloudy days. but fade to white as the carotene is degraded, oxidized and released as perfume. I have seen R. alba simplex with petals a solid pale yellow. Other Albas are tinted with pink and yellow in the morning, but quickly fade to white as the pigments are broken down.

To understand just how much carotene is produced in these fragrant white flowers we need only look at Rosa banksiae lutea, which has traded its carotene-derived perfume for color. So far as I have learned, this is the only yellow rose with clusters of small flowers, and thus the most promising parent for the development of yellow Polyanthas and Ramblers. In all other cases I've seen involving Synstyle hybrids, the ability to cleave carotene into fragrant essences appears to be tightly linked to the cluster-flowering habit, and strongly dominant. For example, Mevr. Nathalie Nypels was a seedling from Orleans Rose, and thus is probably heterozygous for cluster-flowering and carotene-cleaving. Helen Leenders (Orleans Rose x R. foetida bicolor) is another fragrant variety that inherited carotene from its pollen parent and the ability to cleave carotene only from its scentless seed parent.

The Chinas and original Tea-scented Chinas, have only simple terpenes for perfume. As these are not derived from carotene, they may co-exist with yellow coloring. And because Parson's Pink (=Old Blush) is neither fragrant nor yellow, we may assume that it lacks the "genes" necessary both for yellow petals and for carotene-cleaving. Fortunately, the "reblooming gene" of the Chinas does not appear to be linked to other traits that interest us.

Judging from the breeding behavior of the Dwarf Polyanthas it appears that the ability to synthesize yellow pigment in the petals is inherited independently from carotene-cleaving. Thus, the Chinas have contributed to a decline in fragrance in two ways: reduction in the synthesis of carotene in the petals, and the inability to cleave carotene into fragrant substances. Orleans Rose is not noticeably scented, presumably through. the loss of carotene synthesis in the petals, but still retains the carotene-cleaving enzyme(s) linked to the cluster-flowering "gene".

Rosa rugosa also possesses carotene-derived perfume, which explains the difficulty of combining the better qualities of that species with bright yellow coloring. Agnes (R. rugosa x R. foetida) and Grace (R. rugosa x Harison's Yellow) did not receive a full measure of yellow color, but are richly scented. At present I do not have enough breeding data to indicate how carotene-cleaving is linked in the Hybrid Rugosas.

As mentioned above, Rosa banksiae lutea shows us how much yellow carotene would be present in the fragrant, cluster-flowered species if they lost the ability to cleave-carotene. R. banksiae lutescens is presumably heterozygous for carotene-cleaving.

Quinto Mansuino (Breeding Miniatures in San Remo, American Rose Annual. 1960) crossed the Mini-China Tom Thumb with R. bankiae lutescens (in both directions) and raised a series of hybrids varying "from the dwarf ones of about eight inches to the big climbing ones; some thornless, all having good ornamental foliage. The blooms are white, in corymb, long lasting and very decorative. A climbing plant which seemed to have no everblooming habit [Purisissima] has now improved giving a profusion of sweetly scented double flowers which, from December till July, form a candid cascade. During the August-November period the luxuriant glossy foliage compensate for the lack of flowers."

The red pigment of Tom Thumb did not disappear entirely. Mansuino continued, "By crossing the seedling (R. banksiae lutescens x Tom Thumb) with the old Noisette Lamarque I have lately obtained a Banksiae type bearing beautiful deep rose colored flowers. Its open-pollinated seeds gave last year some Miniatures producing in profusion flowers of charming colors."

The quantity of yellow carotene produced in the petals varies from none (Old Blush) to moderate (R. banksiae lutea) to intense (R. foetida). The rich color of R. foetida is supplemented with co-pigments that make small but useful contributions to the brilliance: gallotannin as well as Quercitin 7- and 4'-glucosides. The inheritance of these must be studied separately if we wish to achieve the full intensity of Foetida yellow. But if we can be satisfied with the color of Harison's Yellow the problem of breeding color-fast yellows is not quite so difficult.

Carotene-cleaving ability, as well as the precise positions of the cuts, influences the degree and type of perfume. This will vary from species to species, and even among varieties of a single species. The white flowered forms of Rosa banksiae have a delicious violet-like perfume, probably rich in beta-ionone. R. moschata adds a note of eugenol (also found in cloves) to its sweet scent — beta-ionone and beta-damascenone, among many other components. The Damasks in general are predominantly scented by beta-damascenone and beta-ionone, with eugenol and many other fractions. The "old rose" scent of rose oxide (rosenoxide) is found in R. rugosa, R. nitida, R. gallica, Damasks and Albas, among others, but is not derived from carotene degradation. I have no data on its inheritance and linkage.

With the above facts in mind we may begin to organize some breeding programs. For example, we might cross Mevr. Nathalie Nypels and Old Blush, expecting the second generation to include white, yellow and various shades of pink and salmon. If we want a yellow Dwarf Polyantha-type we should consider R. banksiae lutea as the cluster-flowering parent, and a China for the other. Yellow ramblers might be had from crossing R. banksiae lutea with non-fading yellow Floribundas (e.g. Allgold) or Tea-Noisettes (e.g. William Allen Richardson).

On the other hand, if we wish to improve fragrance, we should -breed from the deepest yellow varieties available to provide an abundance of carotene for perfume production, with a Synstyle or other species with carotene-derived perfume. Crossing R. alba x Allgold should increase the perfume, perhaps beyond what commercially grown attar roses produce.

Another interesting possibility would be to cross Rosa banksiae lutea with R. hugonis, R. xanthina, or other yellow-flowered species. Such hybrids would be strikingly original, though perhaps not easily obtained.

Interestingly, the same problem of yellow color vs. perfume is found in violets, irises, sweetpeas, crocuses and many other plants. There was once, and may still be, a yellow Parma violet that apparently traded some of its fragrance for a degree of yellow coloring. And in crosses of sweetpeas with yellow-flowered Lathyrus species, those pigmented with carotenes did not inherit the vivid color. Only those colored by yellow chalcones were able to pass along their color. Alas, yellow chalcones are not found in roses.

Early iris breeders thought it would be an easy matter to combine the handsome flowers and growth habit of Iris pallida with the bright yellow pigment of I. variegata. Unfortunately, non-cleaving of carotene remained tightly linked to the zig-zag scape of I variegata, just as the desirable scape-form of I. pallida remained linked to carotene-cleaving. There are now many fine, bright yellow irises with excellent form and habit thanks to long years of effort. Perhaps rose breeders could breed yellow Dwarf Polyanthas if enough of us persisted in our efforts. But Rosa banksiae lutea offers a quicker route, and is the only obvious parent for the development of yellow cluster-flowered Dwarf Polyantha-types and Ramblers.

Most of the material in this article was previously posted to the Chez Vibert rose group, December 2002 and April-May 2003.