Making an octoploid

Curious if there has been a sucessful attempt at doubling a tetraploid into a octoploid?

I think I remember it being said that a rose that is not fertile, will become fertile if it is gene doubled. What would that mean for a plant like Suzanne, that does not set hips very well but does have fertile pollen?

Why do I ask?

(Doubled Suzanne x Golden Glow) x R. Hugonis


(Doupled Suzanne x Golden Glow) x R. Xanthina


(Doupled Suzanne x Golden Glow) x Any really hardy diploid


(Doupled Suzanne x (Golden Glow x R. laxa)) x Any really hardy diploid

etc. etc. etc

Flights of fantasy I am sure but an interesting thought path for me.

Interesting ideas to end up with a tetraploid in the end!

I am treating some open-pollinated seed of ‘Carefree Beauty’ to see if I can recover an octaploid just to see what the plants will be like. I’ll let you know how it goes. I just want to see how viable induced octaploids would be.



Do you mean that ‘Suzanne’ doesn’t set hips well from controlled crosses? Because it sets a lot of hips if it is open pollinated, although I suspect a lot of them are selfed. I never think of using ‘Suzanne’ as a pistillate parent though; I guess because it is so effective as a staminate parent.

I would be inclined to cross the hexaploid Rosa acicularis with Rosa hugonis to get a tetraploid breeding line. By doing this, you should be able to substantially increase hardiness and you may not have to use ‘Suzanne’ for hardiness in the parentage of the other breeding line that you intend to hybridize with.


Well I’ve been doing this for the past 2 years with a few of my rose sprouts, but they usually die very early on. I blame this because I have no way of measuring surflan other then using parts. I haven’t applied this on buds on my hybrid teas, but I’m a bit scared. A few of my roses, good as they may be, have RMV, and I don’t want to mutate it into something worse if its possible.

But I’m 90% sure that my rugosa is a tetraploid because 1.) leaves show signs of chromosome doubling, 2.) sets seeds easily when this rugosa hybrid, which is Roseraie de l’H


I must have been mistaken. I don


Yes, it is likely that the Rosa acicularis you are obtaining is a hexaploid.

The easiest way to bring diploid rose characteristics to a tetraploid level is to cross them with hexaploids. Don’t make any extra work for yourself than you have to, unless of course you enjoy experimenting. Unfortunately there aren’t too many hexaploid species to work with, but there is no better rose to work with for cold hardiness than Rosa acicularis. It’s important to work with superior genotypes of Rosa acicularis and I have a couple of them. The best one for flower colour is Robert Erskine’s ‘Kinistino’; it is almost red and has superior foliage colour (deep green). Send me Rosa hugonis pollen and I’ll make the cross for you.


Steve, another way to introduce 25 % rugosa genes is to use a tetraploid rose of the Cainia section. On my web page I suggest R. glauca.

This past summer I used R. Pomifera (since it was grown from seeds, I have to call it (R. Pomifera X OP)).

9 hips of (R. Pomifera X OP)X ((Martin Frobisher X OP) X ((Therese Bugnet X OP) X OP))

2 hips of R. Pomifera X OP)X ((Therese Bugnet X OP) X OP)

1 hips of (R. Pomifera X OP)X Bonavista X OP)

3 hips of (R. Pomifera X OP) X Agnes

8 hips of (R. Pomifera X OP) X Mixed rugosa

I will have to modify my web page to include R. Pomifera.


Paul, when placing R. hugonis pollen on R. acicularis, would you not be worried about creating an endosperm imbalance? Wouldn’t it be better to place R. acicularis pollen on R. hugonis instead, or does it really work fine the other way around?


I’m not an expert on this subject, but yes if I had a choice I would use Rosa acicularis pollen on R. hugonis. What I had in mind was the difference in climates. By the time Rosa acicularis bloomed in Edmonton, Alberta where I live, Rosa hugonis would probably be long gone in the U.S. mid west where Steve lives.


FWIW, I think I’ve never had more than 3 seeds per hip on hugonis, and the germination percentage is less than one might desire. Steve McCulloch might have some comments on this thread if someone can get hold of him.