Diploid x Higherploid or Higherploid x Diploid

Which way is better, given equal fertility in both directions of crossing tetraploids and diploids?

Can you reduce or enhance chances of desired outcomes by choosing the right direction?

Do tetraploid rose stigmas resist diploid rose pollen?

Always learning,


Which way is better, given equal fertility in both directions of crossing tetraploids and diploids?

It may depend on what you mean by better. Extrachromosomal organelles have their own dna and get inherited mostly maternally so the vigor of a seedling might mostly resemble mama, all other things being equal. I’m not sure to what extent chromoplasts other than chloroplasts have their own genetics but color of leaves and petals would likewise favor mama if they do. Perfessers?

Can you reduce or enhance chances of desired outcomes by choosing the right direction?

Theoretically not except for traits involving organelles but somebody is sure to post anecdotal evidence to the contrary. IIRC Ralph Moore had an opinion on this, or maybe I’m thinking of his rubrik about painting a bloom on a rose.

Do tetraploid rose stigmas resist diploid rose pollen?

Not in my experience. Some cultivars do have sterility mechanisms to prevent selflings which may present as directional sterility but in these cases not all pollens should be affected equally.

Crossing my fingers after I cross my roses seems to work pretty well for hard cases.

I’m hoping to learn something from this thread. I’ve been exploring Serafinii (pentaploid) (unsuccessfully), hoping using it with triploids might result in something useful. So far, nada with di, tri and tetra.

That’s a great question. I don’t know… I suspect it may be related to the genetic background and how precocious certain parents are with outcrossing and what they are being mated with. Tetraploid Moore mini roses that are very fertile (‘Splish Splash’, 'Rise ‘N Shine’, etc.) cross pretty well with diploid polyantha males. Maybe this is because Moore has some polys or at least Systylae section roses (R. wichurana) in the background of many of his roses. Diploid polyantha females haven’t crossed very easily with the hardy tetraploid shrubs (‘Champlain’, etc.) I’ve tried and visa versa too. I didn’t cross the Moore minis as males though onto polyantha females. Generally, I’ve found that the polyploid Caninae section roses (R. eglanteria, R. pomifera) cross better as females and the males just happened to be lower ploidy. They do have awkward meiosis though and their pollen is typically 1x.

There’s some discussion of this topic in this previous thread:


I"ve always wonder about this too. I tend to use the cultivated rose as the seed parent. The seed set is often higher and germination more predictable.
One thing I am having trouble with, is getting viable offspring from (wild sp. x Garden Rose) x Garden Rose.
I wonder if environmental conditions, ie:- Temperature, Altitude, Latitude can and does influence on seed set and fertility,

For example, Warren Millington seem to be take Prairie bred roses and create some astonishing hybrids from them in his New South Wales, AUS. garden. Therese Bugnet does not set
seeds here in my garden or greenhouse in MB.

I’ve read heat often increase chances but I don’t know how anecdotal that is. It could also just be the stress induced/survival mode thing, we (also in NSW) have had a lot of stupid hot days (40C+/104F+).

When you throw triploids into the mix one thing to consider is that, in my understanding, polyploid 2x pollen has a competitive advantage over 1X pollen. Triploids will theoretically produce some 2x pollen and some 1x pollen, as well as some 1x ovaries (eggs? what is the female counterpart of pollen?) and some 2x. Let’s hastily assume that each triploid produces equal amounts of 1x and 2x reproductive cells in both genders. Putting the 1x pollen of a diploid on such a triploid should result in 50% diploids and 50% triploids. However, reversing that cross, if for the sake of argument the 2x pollen has a 100% competitive advantage, then you’d end up with 100% triploid seedlings. Of course in the real world things vary widely, but it seems that there could be a significant overall difference in the characteristics of the seedlings based upon the direction of the cross.

Likewise when crossing this theoretically perfect triploid with a tetraploid, if the triploid was used as a female there should be half triploids and half tetraploids. If the triploid is the male, the seedlings would be primarily tetraploid.

Which way is better, given equal fertility in both directions of crossing tetraploids and diploids?..Can you reduce or enhance chances of desired outcomes by choosing the right direction?

I got to thinking about this a bit. Nobody mentioned epigentics which does actually influence the outcome of a cross based on direction. The problem at the moment is that not a lot is known about which particular traits can be affected in plants, though a little more is known about animals.

This papers sums up the situation pretty well.

If you want a really deep dive into epigenetics this lecture series by Edith Heard is the place to go. She did mention that vernalization in flowering plants - arabadopsis - has been shown to be affected by epigenetics, in lecture five iirc but maybe six.


Thanks Don and all others,

I think the answer to my question is a qualified, “besides avoiding complicated ploidy issues it does not, except”:

What matters genetically is the trace back to the original female. As Don pointed it out, due to non-nuclear DNA inheritance, you will always get the mitochondria and chloroplasts from the female side. I believe the rest is a genetic crap shoot, including epigenetic influences, so keeping that in mind, choose the seed parent wisely because it will have total influence from the mitochondria and chloroplasts.

That may not sound like much of a conclusion, but then, R. wichuraiana is the female root of lot’s of Moores and Bucks, and hence Radlers, and who is to say which way Max Graf occurred?

I’d wager that R. wichuraiana is the original female of all of the R. kordesii descendants as well. Any takers?


About those mitos and chloros. The actual amount of DNA that they carry, and hence proteins that they make is very small. But those few have to get along with what is carried in the nuclear genome. So it’s like a person who will only date left-handed, tall, thin persons with a full brown head of hair. That does affect the range of good choices. But it is realtively easy for someone to sequence the DNA in the mitos and chloros of some different rose species. There might be some slight differences that allow one to say with certainty that one is wichurana and another is canina, multiflora or calolina. But maybe not. If not, then the species differences don’t actually reside in those organelles.

This would be a great project for a class in DNA techniques. Relatively cheap these days too.

I’m not sure that it holds much water.

I read somewhere that Rosa gallica is a bit unusual in that it crosses about as easily with diploids as it does with higher polyploids. I don’t recall the source, but it might be this paper. I don’t have it, though I would be happy to scan and upload it if anyone has access to a copy.

J.C. Ratsek, W.S. Flory, Jr., and S.H. Yarnell. 1940. “Crossing Relations of Some Diploid and Polyploid Species of Roses,” American Society for Horticultural Science, Vol. 38, pp. 637-654.

Then there is an earlier paper that might also be useful:

J.C. Ratsek, S.H. Yarnell, and W.S. Flory, Jr. 1939. “Crossing Relations of Some Diploid Species of Roses,” American Society for Horticultural Science, Vol. 37, pp. 983-992.

I’ll get them retrieved from off-site storage if we have them at K-State, or do ILL. May take a few days.

The differences in reciprocal crosses can be dramatic. This paper deals with Epilobium. The first picture shows reciprocal hybrids of E. luteum and E. hirsutum. The second shows equally divergent reciprocal crosses between two geographical races of E. hirsutum.

I have a note about reciprocal crosses between two strains of Fragaria vesca. Mildew resistance was inherited maternally.

Karl et al,

I requested the articles from cold storage. Maybe 2 days to receive. Then all I have to do is figure how to get pdf files forwarded.
While checking for possible on-line presence of those articles I found an interesting one that makes me want to say Ralph Moore was right “the rose will find a way”. Somehow I skipped this one when it first came out. Now I see why. Taxonomic swamp is not yet drained. Article was free on-line from home computer so I think it is generally available.

M. Fougere-Danizan, S. Joly, A. Bruneau, X-F Gao and L-B Zhang (2012) Annals of Botany 115:275-291 Phylogeny and biogeography of wild roses with specific attention to polyploids. They used markers of chromosomes and nuclear genes to see how the various families, groups, clades, genera, species etc might be related and evolved. Most N.A. species came from Asia more recently than the first. Canina has complex origin. Many species are mixed up or mis-classified ancient hybrids and so on and on.

Assuming correct labeling in the Ratsek studies, it is a useful batch of information for practical breeders. Whether it tracks at all with phylogeny regarding what is related to what, we’ll have to see the data.

I came across another from the same group, then fiddled with the URL until I got to a directory of available papers.


I have become disillusioned with this DNA business. Every paper I’ve read seems to contradict another paper … sometimes papers by the same group.

Around 20 years ago I read: Using RAPDs to Study Phylogenetic Relationships in Rosa; Millan, Osuna, Cobos, Torres; Theor. Appl. Genet., 1996, 92:273-277)

Among other odd things, the authors concluded that Rosa cymosa and R. banksiae are as far removed from each other as both are from the rest of the genus.

In the paper by Fougere-Danizan, et al. (2012), we read that Rosa blanda and R. nitida are close cousins (node 64). Similarly, node 75 links R. majalis and R. rugosa.

Should we interpret these report to mean that Blanda and Nitida are recently separated, as are Rugosa and Majalis?

I think it is worth mentioning at this point that Boulenger (1937) had a dickens of a time distinguishing R. blanda and R. cinnamomea (majalis). Cinnamomea usually has paired, infrastipular prickles, but sometimes does not. Blanda usually lacks these prickles, but sometimes has them.
Did R. cinnamomea somehow sneak across Europe and Asia to reach the New World via the Bering Land Bridge?

Then there is Regel’s (1881) suggestion that R. nitida is the American form of R. rugosa.

It seem to me that the modern DNA folks are making the same mistakes as the neo-mendelists did a century ago. When we distinguish two species, can we really assume that the genes responsible for those distinctions are scattered among all the chromosomes? I don’t think so. And can we assume that the distinctions between two species will also be revealed in whatever structures happen to catch our attention?

For instance, Rosa nitida is stoloniferous. R. blanda is rhizomatous. There are other differences, of course, but can we assume that differences in the the number of glands on the sepals, or the diameter of the stomata will always be linked to the more obvious distinctions?

Joly and Brunneau (2007) wrote, “Morphological characters were selected to be applicable to as many specimens as possible.” No stolons or rhizomes.

Furthermore, the authors saw no reason to distinguish R. woodsii from R. blanda. They cited Erlanson (1934) but ignored her report that Woodsii has 35-85 stamens (av. 65), whereas Blanda has 85-140 (av. 110).

If the supposed relationship between Nitida and Blanda is the result of hybridization, rather than fairly recent speciation, where are the hybrid swarms? Or could it be that the traits that allow the species to be distinguished tend to be inherited together through linkage or correlated expression?

I have become disillusioned with this DNA business.

I think maybe your gripe should be with the taxonomists and not the geneticists. DNA dendrograms are based purely on unimpeachable empirical data - dna sequences - whereas the taxonomies are purely opinion based on subjective observations.

It is impossible to reconcile the differences comparatively and not only because of methodologies but mostly because they represent different philosophical schools. Taxonomic models are hierarchical while dna is proving actual relatedness to be a network.

Also need to consider that often both things, despite seemingly contradictory, could be true depending on the DNA sequences being looked at. I mean within a genus there’s a lot of genetic overlap, look at humans and orangutans share ~97% of DNA…that’s a lot of DNA that says we’re the same and only a tiny tiny bit that’s unique saying we are different.