Step one: Cross a tender, everblooming rose with a hardy tetraploid species rose.
Step two: line out seedlings to observe that may never be able to bloom in my climate because they don’t have enough live wood.
Step three: select a seedling that has good health and vigor to dig up and store inside over winter so it can bloom the next year.
Step four: When that seedling blooms, cross the heck out of it back to everblooming roses and keep only the tiny minority of seedlings that show juvenile rebloom.

So far I have gone down this arduous path with three crosses:

Gina’s Rose x R. virginiana (resulting in a seedling I call GRVIRG)
First Impression x R. virginiana (FIVI3)
All A’Twitter x R. virginiana

With FIVI3 I put it so much effort and had so many seedlings and so few juvenile bloomers that I ended up keeping many of the non-blooming F2’s. There were only a tiny percentage of remontant F1’s, a phenomenon about which David Z mentioned “preferential pairing.” The non-remontant plants show good health and vigor and a few were able to bloom in this their second year. However, I’ve already diluted the R. virginiana down to 25% and those seedlings STILL might not have more than two out of four of the sets of reblooming genes.

So can I change my strategy? Use a slightly hardier reblooming rose so that crossing it with a species rose will result in seedlings that carry enough live wood to bloom when planted outside.

One annoying complication is that some reblooming roses seem to have the tendency of not passing on reblooming genes all of the time. Examples include Prairie Joy, Cherry Frost, Peter’s #42 (from Morden, L83 x polyantha), and my seedling #1019 (a rudely healthy and awkwardly vigorous plant that blooms non-stop). Even when crossed with modern rebloomers the aforementioned roses do not give a majority of juvenile remontant seedlings. This makes me question any assertion that remontancy is a straight recessive.

I’m hesitant to use these hardier roses to go down the species rose path. I will have no idea if the F1 with the species will carry even 50% reblooming genes, so any efforts to get a reblooming F2 might be futile.

Paul Barden’s words are a tickling echo in my mind: “Some of my most heavily remontant seedlings have not blooomed AT ALL the first season.” However, I have no experiences to match this and I have to assume that most of my seedlings that do not bloom the first year are never going to be remontant.

Except for those pesky rugosas. Rugosas are the only roses in my climate that are both tip hardy and remontant. That makes them a tempting pathway towards combining hardiness and rebloom. But their tendency of delayed remontancy is very frustrating. For example, I have a Rugosa x Cherry Frost seedling that is robust and vigorous, but in it’s second year is showing no signs of blooming. Remember that Cherry Frost doesn’t always give rebloomers anyways - this seedling might never bloom and it’s taking up field space. Some rugosas can produce juvi-blooming F1’s when crossed with moderns. I noticed about three or four out of twenty Will Alderman x OP seedlings that are blooming as youngsters. On the other hand, I have dozens or hundreds of seedlings from another remontant rugosa that are not blooming as youngsters. Another issue with using rugosas crossed with modern roses is that the resulting plants seem to be sterile triploids most of the time. (see Henry Hudson x Above & Beyond pic below.)

One possible success: I have a Lemon Fizz x R. carolina seedling that is able to produce a handful of yellow blossoms each year. From putting its pollen on Yellow Brick Road I have about three out of twelve seedlings that are blooming young this year. I will contain my excitement because YBR has shown some tendency toward accidental selfs, but ok I’m excited because they look like they might actually be from the intended cross. It would be really fun to have a reblooming yellow rose that was 25% R. carolina, was as healthy as Lemon Fizz, and hardier than LF or YBR.

See captions for parentage of the following. The Hannah Gordon x FIVI3 seedling is remontant and might actually be the intended cross…fingers crossed that it will be hardy and healthy.

The length of the juvenile period is inherited independently (so far as I can tell) from rebloom. Van Mons (1835) described his system of plant breeding that concentrated almost as much on shortening the juvenile period as it did on improving the quality fruit and flower. Wild Belgian pears, in Van Mons’ experience, grew 10 to 12 years before flowering. After five generations of selection, many of the seedlings flowered in four years from seed, and a few in only three.

Van Mons was Belgian, and was better known among French-speaking specialists. When Van Mons died, Vibert named a rose for him. The following note suggests that Vibert followed at least some of Van Mons’ methods.

"Let me add an important observation: Within this set of roses [Trianons] are plants which bloom the first year from seeds, which does not ordinarily happen with Chinas, Noisettes, or Bourbons; last year, 10 or 12 young plants bloomed in July. All of these roses coming from this Trianon seem to me to be very receptive to pollination from other varieties—or perhaps it is because of their own inherent qualities that they show so much variation, which seems to me to be the more likely explanation [signed: Vibert, Angers, June 28, 1846].”

According to a writer in The Florist and Pomologist (1882):
“Adverting to the influence of selection on precocity, M. Carrière mentions the fact that while the seedling Roses of fifty years ago took six or eight years to produce their flowers, it is not now unusual to see them flowering the first year. There is, however, great variation in this respect between seedlings derived even from the same fruit, some seedlings requiring four, and others from eight to twenty years to fruit. The Duchesse d’Angouléme Pear only began to produce fruit thirty years after it was raised from the seed.”

Much of the selection for precocity in garden roses was accomplished long ago. Most species, to the contrary, have not yet been selected for early bloom. I think it would be a useful investment to raise at least one generation of wild seedlings and select for precocity.

A wild plant that needed six or eight to get blooming is not likely to yield many everblooming grandchildren that bloom their first year. Similarly, a reblooming Rugosa that began flowering in its third year probably will not give so many first-year bloomers as one that was also a first-year bloomer.

Joe,Here are a couple of quick blooming successes i have had with Henry Hudson–I have several Rosa foliolosa x Henry Hudson crosses that seem to cross with anything and have had pretty good luck with heat and they also seem quite drought tolerant (not xeriscape material, however). But with foliolosa, rugosa, and woodsii genetics, this should be good for the cold. I am just surmising that they will (especially the woodsii) be quick to rebloom, or at least rebuild plant material upon which to bloom with their unusual speed of growth. The woodsii cross is assumed-it was an OP next to woodsii with juvenile foliage identical with woodsii. The rosa woodsii I have is a repeat bloomer-so far this year it has bloomed from end of March to now-still blooming, has mature red hips from some first crosses-has new blooms almost weekly. But this might be something you could get something out of with a few crosses to better caroliniana and virginiana offspring. I am going to try some of these crossed with Prairie Peace, altaica, and FIV offspring. Might try some of the rosa Caroliniana pollen there too. I am probably going back to an all pink rainbow. Would love to see if these could survive to rebloom in the colder zones.

Here is the presumed (Rosa foliolosa x Henry Hudson) x Rosa woodsii. The photo seems to have gotten bumped. This plant seems to have been injected with some of that “hybrid vigor” stuff. Plant is on steroids but the blooms are still small.

That Art Nouveau X Carolina is really neat, Joe!

On a theoretical level, wouldn’t those two things together be the path forward? Granted it’d be a bit of a numbers game for the first and second generation.

I mean if Will Alderman can produce juvenile bloomers then select those and try cross them together if any are capable of setting seed, or try the pollen (since many triploids seem to have at least a little fertile pollen) back to Will Alderman and select again for juvenile bloomers.

The F2 should see at least a few juvenile bloomers capable of setting seed that should make the path forward far easier.

Thanks everyone for your responses!

‘Conrad Ferdinand Meyer’ is a fertile tetraploid that has been used with some success in breeding hardy roses.

David Austin’s English Roses (1993) p. 28

The third line we pursued was by way of the Rugosa hybrid ‘Conrad Ferdinand Meyer.’ At first we harbored no great hopes of success, for we feared that the resulting seedlings from a cross with this excessively vigorous hybrid would be altogether too gross in character. ‘Conrad Ferdinand Meyer’ was itself a cross between the very popular and beautiful Climbing Noisette Rose ‘Gloire de Dijon,’ and an unknown Rugosa hybrid. It also had one of the most powerful and delicious fragrances. As before, we crossed with some of our better English Rose in particular ‘Chaucer,’ and had one of those pieces of luck that sometimes turn up in rose breeding. Some of the seedlings from this cross were of typical rugosa appearance, while others bore absolutely no resemblance to a Rugosa Rose. It seemed that some of our hybrids had taken the genes only from the ‘Gloire de Dijon’ half of ‘Conrad Ferdinand Meyer,’ while others had inherited those from the Rugosa side. What we had in many instances were in effect hybrids of ‘Gloire de Dijon.’

If it were easy…
I think you are taking the route that ultimately may lead to success. “Stirring the pot” offers little to those of us without the means to grow massive numbers of seedlings, I would think, and the harder paths will probably offer the best rewards.
I’m far enough south to have no clue as to what I’m talking about here, but several years ago, Joan Monteith shared seeds from a Mr. Nash x Rugelda cross. (She said she didn’t know what she was thinking when she made the cross, but I was certainly intrigued by the concept.) If Nash is Doubloons, as speculated, the potential for cold-hardiness was in the resultant seedlings. Germination was surprisingly good and quick, and the plants had warm colors, very double flowers, and good fertility. (Alas, I was unable to keep any, and I won’t pretend that the last of those seedlings appreciated central Texas very much.) Point being simply that I imagine there are probably some pretty less-than-self-evident mixes that could pan out to give you some interesting material.

Joe, have you considered switching to rhododendrons?

Unfortunately Doubloons isn’t actually hardy below 0 F and it is one of the worst black-spotters around. Can’t comment on Mr. Nash.

I’m still using a couple Doubloons offspring but most have expired naturally. Well over 100 of them.

That’s good to know, but disappointing. Fortunately, ‘Doubloons’ offspring ‘Goldilocks’ seems to be healthier, as is its grandchild, ‘Allgold’. I can’t vouch for the hardiness of either, though.

Goldilocks here is about the same as Doubloons, perhaps a bit worse. Allgold didn’t last long either. And Golden Slippers is maybe worse than either. I mean for hardiness and BS. Beautiful flower for about 5 hours. That’s why I produced Ruby Slippers (13-2).

Don, my soil is too alkaline for rhododendrons. Rugosa roses are borderline in this soil, too, which is another reason I want to cross them with moderns.

Plazbo, my OP Alderman seedlings that bloomed as juvies are seeming pretty weak. I doubt they will amount to anything. I do, however, have some juvi bloomers from crosses with Ann Endt, Belle Poitevine, and a rugosa seedling. We’ll see how that goes.

Combining hardiness and rebloom is like moving a fridge by yourself…it’s a slow walk forward, corner by corner.

I have seen ‘Goldilocks’, ‘Golden Slippers’ and ‘Allgold’ only at the San Jose Heritage. BS is not a problem there, and the summers are not as miserably hot as Kansas. So, maybe it’s the heat. Le Grice was aiming for a yellow rose that did not turn turnip-y white during cool, damp weather. ‘Allgold’ was the result. It isn’t the best shaped rose, but the color is stable … at least in San Jose.

‘Golden Showers’ is a fine rose in San Jose, but in Kansas the buds that began to open at breakfast were full blown by lunch, and dropping petals by supper. The canes stop growing in really hot weather.

I should have mentioned that ‘Golden Promise’ (floribunda, de Ruiter, 1972) behaves in San Jose, the way ‘Golden Showers’ does in Kansas. The canes stop growing in heat. San Jose heat is somewhat less scorching than Kansas heat, and ‘Golden Showers’ is quite happy in SJ.

‘Golden Promise’ caught my attention only when I was going through my photos and noted that it was blooming in December and January.

Another aspect of hardiness that is often overlooked is optimum growth temperature.

The subject of temperature and growth has been studied from time to time, but the needs of commercial growers of cut flowers are not the same as those of gardeners. A commercial grower wants a plant that continues growing and blooming at low temperatures (approaching freezing) to reduce heating costs. A northern gardener, on the other hand, is better off with a plant that refuses to grow at low temperatures when frost is possible.

Then again, many growers expect their rose bushes to start early and finish late. These expectations conflict with hardiness. A plant that begins growing too early in the season risks damage from late frosts. And a plant that continues growing too late in the year risks similar injury from early frosts.

Rosen (1956) found that varieties that break dormancy late suffered less damage from late March freezes than the earlier types. And plants that grew slowly were less harmed than the faster types that started at the same time.

The late season is just as important in the search for total hardiness as Bugnet (1941) observed.

“With a much longer experience, on a larger scale, in stone-fruits breeding, I am led to believe that a plant, in order to withstand our climate, needs a very early ripening of its tissues. Winter-killing, apparently, is not caused by extreme cold but rather by a too early cold snap catching immature wood, like the 30° below we had in the first part of November last. Once, at dawn, on October 12, 1930, we had 16° below zero. The next day was rather warm. None of my hardy hybrids and no native tree or shrub suffered. I have often noticed that half-hardy plum or apple trees here, unhurt by December 1, passed unharmed through the rest of the winter no matter how intense the cold.”

Risley (1958) pollinated ‘Skinner’s Rambler’ by an assortment of other roses. He chilled the seeds, but found that the time to germination varied widely (83 days to 173), according to the pollen parent. Two of these caught my attention. The offspring of ‘Persian Yellow’ started sprouting in 91 days, whereas those from ‘Diamond Jubilee’ took 160 days to come up. Both of these varieties are descended from Rosa foetida. However, ‘Diamond Jubilee’ is a seedling of ‘Marechal Niel’, a variety that loves heat and refuses to grow at low temperatures that other roses could enjoy.

The slowest seedlings to appear were from ‘Queen of the Lakes’, a Brownell “Sub-zero” rose.

I think that Risley got tangled in terminology. He assumed that the late appearance of some seedlings was due to a greater chilling requirement. But he was not watching to see when the seeds germinated, only when the seedlings emerged. The ability to grow (or not) at low temperature would skew the results. That is, one seedling might germinate early (short dormancy) but emerge late due to its slow growth at the stratification temperature. On the other hand, a different seedling might germinate late, then grow rapidly.

The point, here, is that growth rate can be assessed independently from dormancy-breaking. Plants that grow slowly or not at all at low temperatures will likely suffer less damage from late Spring frost and early Fall freezes.

Selection for growth temperature has been done successfully in the Calla, ‘Green Goddess’.

Bugnet: The search for total hardiness (1941)
Rosen: Resistance to Spring Freezes (1956)
Allen & Asai: How Frost Damage Occurs (1943)
Asai: Repeated freezing and thawing (1944)
Greeley: Temperature and Rose Bloom (1919)
Greeley: Night-Growth of Roses (1920)
Risley: Male controls sprouting (1958)

Possibly just the juvenile phase or investment in blooming over plant growth, I’ve got a few seedlings from last year that were very runty and wanted to bloom but a year later and a lot of them have grown out of it…although that may be because they grew through winter with a break from trying to flower.

The OP Will Alderman have surprised me with the amount of growth they’ve put on, which is still not much but they appear to want to survive at least. I’m a little disillusioned with rugosa crosses as the blossoms seem to be inevitably smaller and shorter lived than those of either parent and they scream “Sterile!” Might just have to find the right rugosa parent. Henry Hudson has done some nice things but I no longer have one growing and will have to wait until I get one established until I can start breeding with it again.

That’s kind of why I’m basing rugosa around Schneezwerg and Calocarpa. The former has demonstrated a small capacity of fertile crosses with tea/china and the latter is half china. Will see what happens in the season that’s approaching. Bound to be a lot of sterile but selection over a few gens will probably make that less of an issue.