Genetic diversity and genetic similarities between Iranian rose species.

Title: Genetic diversity and genetic similarities between Iranian rose species.

Authors: Samiei, L.; Naderi, R.; Khalighi, A.; Shahnejat-Bushehri, A. A.; Mozaf-Farian, V.; Esselink, G. D.; Kazempour Osaloo, S.; Smulders, M. J. M.

Authors affiliation: Department of Horticulture, College of Agriculture and Natural Resources, University of Tehran, Karaj, Iran.

Published in: Journal of Horticultural Science & Biotechnology (2010), 85(3), pages 231-237.

Abstract: “Wild rose species were collected from different regions of Iran for a rose breeding program. They included accessions from Rosa persica, R. foetida, R. pimpinellifolia, R. hemisphaerica, R. canina, R. iberica, R. damascena, R. beggeriana, and R. orientalis. Ten microsatellite (simple sequence repeat; SSR) markers were used to analyze the genetic variation among these rose species. The SSR markers amplified alleles in all species, even if they were from different sections within the genus. An unweighted pair group method cluster anal. (UPGMA) based on similarity values revealed five main Groups. The data showed no support for any distinction between R. canina and R. iberica, as all the accessions were placed in one Group, and accessions of these two species were more closely-related to each other within a Province than to accessions of the same species in other Provinces. Accessions of sect. Pimpinellifoliae were combined with plants from sect. Rosa and Cinnamomeae in two different Groups. Genetically, R. persica clustered distinctly from all others, with few alleles shared with the other taxa. We discuss the use of SSR markers for phylogenetic anal. when these markers are amplified in all species of a genus.”

Some very interesting information. Seems like North American species are not the only ones that mix openly. It is also interesting on the subject of R. perscia I think this further suggest that R. persica should not be included with roses but as their own distinct section. I wonder if any one has done this with species from India and China? I would think the results would be similar.

Genetically, R. persica clustered distinctly from all others, with few alleles shared with the other taxa.

I strongly suspect that R. hulthemia persica is relatively more closely related to the R. omeniensis / Sericea / R. hugonis constellation than to other roses. My opinion is based on some apparently successful crosses I made this season that were, in turn, inspired by Koopman’s Bayesian phylogenetic chart.

R. persica should not be included with roses but as their own distinct section.

Koopman showed that R. hulthemia persica is truly a rose: “The phylogenetic analyses showed that (1) two of the four subgenera ( Hulthemia and Platyrhodon ) do not deserve subgeneric status;”

See “A case study of Rosa”. American Journal of Botany 95(3): 353

Don I have studied Koopman’s chart a lot over the last year or so. Koopman’s work is great I glad Henry sent me it. I am convinced that R. hulthemia can be used as a bridge species with some of the other roses on that side of the family. I have cutting coming from Jim with some of his hybrids that I want to try with R. stellata and R. minutefolia. I base this mostly on his chart and a few other charts of the species rosa. Even though some of the roses are missing in one chart to another I have short of guessed where they should belong. It sounds like your doing the same thing with R. hulthemia as I am planning on doing. In the end I figure Jim and a few others will have the blotch effect more than covered so I am less interested in this trait than in experimenting with it as a bridge species. So Don do you think the R. hulthemia hybrids could make a bridge with R. stelleta and R. minutefolia?

I will have to reread the paper however that Koopman did. I did not infer that R. hulthemia was a rose as much as it was on the outskirts of the family. So I will have to look at it again. I do however believe that most of the subgenera in the roses are wrong from the evidence at hand. It seems like the subgenera where based on physical features before DNA evidence and now with DNA evidence suggesting some plants are more closely related to this instead of that and vice verse it seems like the whole subgenera classification at least to me needs to be redone and roses need to be realigned with DNA evidence in mind.

do you think the R. hulthemia hybrids could make a bridge with R. stelleta and R. minutefolia?

Hmmm. Maybe, but New World roses might be better options.

Looking at Koopman’s chart - the third, Bayesian chart - you’ll see that the New World roses cluster together at the bottom. Since Koopman didn’t include Hesperhodos we are left without a measure of genetic distance for them from the other New World roses but we can infer that, since R. virginiana is the least evolved (shortest genetic distance from the beginning of the branch measuring horizontal lines only) that it must be the most closely related to the Hesperhodos.

My copy of virginiana seems to be fertile in crosses with some modern roses, so if I were chasing traits of stellata and minutifolia I would attempt to cross them with virginiana and close hybrids of virginiana. You would have the problem of imbalanced ploidy, virginiana being tetraploid and the Hesperhodos being diploid. However, we know from the study that Paul B. and David Z. did that some polyploid roses make pollen in a range of ploidys which, if true for virginiana, would increase the odds of success using that as the pollen parent.

Another way to approach this is to play the hunch that the Hesperhodos descend from some common ancestor with other, probably western, American roses such as R. californica, R. pisocarpa and R. nutkana (for minutifolia ) and maybe R. carolina/arkansana and R. laevigata (for stellata).

Finally, you could also play the wild card of the circumpolar rose, R. acicularis. In general, Koopman’s results cluster geographically. Acicularis is enigmatic in that it appears scattered within these regional clusters, probably indicating some degree of hybridization that may have occured with local native species. Regional variants of acicularis near to wild populations of the Hesperhodos thus might also be an option.

However you approach the challenge it is not going to be a trivial undertaking.

I do know that both have been successfully crossed onto R. rugosa.

I was guessing where Hesperhodos would fit in through other papers. Basically matching up the roses I knew with the ones I did not know.

Walter Lewis is still actively working on a revision of the roses in this part of the world. He’s at the MO Bot Garden and replies to e-mail. I expect he ahs some useful insights on the relationships in question. He’s been at it over 50 years.