Pedigree breeding and diversity.

A meeting place for rose breeders.
Post Reply
Pierre Rutten

Pedigree breeding and diversity.

Post: # 17011Post Pierre Rutten
Thu May 29, 2008 1:00 am

There were many rose DNA analysis in the last decade.

All agreeing on the little diversity found among modern roses.

This lack of diversity has two sources.

The first is the prominent role of a few priviledged progenitors among which Parks' Yellow Tea-scented China that has in HMF 12028 unique known descendants. We may infer that many others share PYT unveiled ancestry. And its repeated appearance in parentage trees shows how much its characteristics were sought after and selected over others.

Second is the constant backcross to the best HT breeding strategy that was applied to any outcross. A lot of outstanding vars have a close enough relation to some species. Backcrossing them to HT dilutes the species contribution to the point it vanish. Carruth said with his words that

Jim Sproul - (zone 9)

Re: Pedigree breeding and diversity.

Post: # 17028Post Jim Sproul - (zone 9)
Thu May 29, 2008 1:00 am

Hi Pierre,

Do you get downy mildew at all where you live? I have noted that 'Baby Love' and seedlings of 'Home Run' are particularly susceptible to it and will show it under the conditions that you describe.

Jim Sproul

Philip_LA

Re: Pedigree breeding and diversity.

Post: # 17029Post Philip_LA
Thu May 29, 2008 1:00 am

Outcrossing, yes, but in outcrossing with the extant lines, we will inevitably be maintaining and reinforcing the same genes.

I suppose all of us have, at some point or another, fantasized somewhat about reinventing the wheel, so to speak, by creating lines completely excluding the genes that provide the achilles heal. But much of the problem isn't from using inherently weak species in the mix so much as it is in diluting the diverse strategies different species may have to combat disease, I think.

In crossing an evergreen rose, for example, with a rose whose strategy is to quickly shed diseased leaves, one might obtain an offspring which tenaciously holds on to sick foliage. Both the stronger cuticle of the evergreen and the disease-shedding of the other would be compromised. A poor example, but I think some understanding of species' mechanisms for disease-resistance might aid in hybridizing strategies for disease-resistance, no?

Don
Posts: 1865
Joined: Wed Dec 31, 1969 7:00 pm

Re: Pedigree breeding and diversity.

Post: # 17030Post Don
Thu May 29, 2008 1:00 am

>>Backcrossing them to HT dilutes the species contribution to the point it vanish.... Outcrossing to restore genetical diversity and testing habilities in less than ideal conditions there is no other answer.

Well said.

I have speculated about starting fresh using only species and then outcrossing to some modern hybrids rather than the other way around, though.

I think I might start with R. chinensis spontanea, R. xanthinia, R. foetida, R. gallica, one or two Caninas and maybe one or two native American species. I would use several different cultivars of each, all started from seed. I would first put each of these species through several generations just crossing within species to select for disease resistance and vigor. Then I would proceed with reciprocal crosses, first one species to another, then back again to each species, to select for fragrance and plant form, over and over until I got that right. The last step would be outcrossing to modern roses to polish the flower form and color.

For the sake of argument, where would other people start?
What doesn't kill them makes them stronger.

Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17031Post Pierre Rutten
Fri May 30, 2008 1:00 am

Jim

Yes I get downy mildew rarely and only earlier in spring on all young foliage that "burns" suddenly. Apparently there is a strong strain-host relation as only a few vars are succeptible even if damage is very spectacular. As it is neither generalised nor long or usual I cannot say I am breeding for DM resistance.

Using tropical and sutropical asiatic sp should help.

Philip

I do not think modern roses genes are bad. Overall it is their lack diversity that favour so much deseases.

Why not then breed for other plant models such as close to wild sp more decorative shrubs. Instead of the usual strategy of backcrossing to mainstream Ht; why not try the reverse that is backcrossing desirable features of modern roses in a plant with most of one or more sp strong features.

In my opinion most sp adaptation to their original environment rely also strongly on diversity with variations around the specific theme.

That is why we should not rely too much on a single hybrid as it is with Muriel derived hybrid bracteatas or Tigris hulthemias.

Arno has good ideas on this point.

I agree fully when you say that all are not compatible and that adding weaknesses is easy. May be you experienced it?

Don

We have to make our options according to climate and possibilities.

Personally I put often new species at work as each has its qualities and faults. A long list now. Combine them for better results is quite exciting, discovering the unexpected results of combinations a bonus. The F1 from two sp is less homogen than one can think. Halfsibs or crossing different F1s may unveil fantastic diversity. Some outstanding.

Crossing new hybrids from never or not enough used sp with modern roses has possibilities one cannot imagine.

Just do it.

Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17608Post Pierre Rutten
Fri Jul 04, 2008 1:00 am

http://www.rosehybridizers.org/forum/message.php?topid=

16052&rc=10&ui=1998843342

Don stated:

"A major conclusion of the study is that modern roses retain only a small fraction of the genetic diversity of the original founder species."

Quite surprising when one consider modern roses origins!

Roses breeding history starting with gallica in Europe and the founder Chinas in Asia (whose ancestry has to be studied with actual knowledge) is a list of successive outcrosses.

Fedtchenkoana through Damasks, moschata through Noisettes, foetida through Pernettianas, multiflora through Polyanthas, wichuraiana and rugosa are the main contributing species.

All are in modern roses ancestry.

Instead of added diversity we are ending with considerably narrower variation range.

When the same parent appears many times in a genealogic tree it is called consanguinity. When the consanguines that are modern rose are cloned that is grafted or cutting raised no wonder they are neither strong nor healthy in less than ideal conditions everyone experiences in a less than ideal season.

Why it is and how was it done are questions.

To break the vicious circle the indispensable aspiration.

Any ideas?


Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17661Post Pierre Rutten
Tue Jul 08, 2008 1:00 am

One has to point the fact that almost everywhere most ancestral species are a lot healthier and stronger than modern roses.

After centuries the founder Chinas are quite healthy and strong in the south, gallica, fedtchenkoana, moschata, multiflora are quite easy to invasive and have wide adaptation as are to the highest point rugosa and wichuraiana.

All are infinitely better than their progeny we grow...

Only foetida is not the cleaner plant.

That its faults are not eliminated after up to 25 backcrosses is incredible. They will not continuing.

Obviously roses breeders did something wrong and as long as we do not consider why and accordingly revise our breeding we are carrying on.


Don
Posts: 1865
Joined: Wed Dec 31, 1969 7:00 pm

Re: Pedigree breeding and diversity.

Post: # 17662Post Don
Wed Jul 09, 2008 1:00 am

>>Obviously roses breeders did something wrong and as long as we do not consider why and accordingly revise our breeding we are carrying on.

It is a problem of philosophy. Hybridizers pay no attention to the workings of evolution within the fitness landscape. For a rough introduction to this concept see

http://en.wikipedia.org/wiki/Fitness_landscape,

but it is more informative to read Kaufman's "At Home in the Universe", see

http://books.google.com/books?id=FxvENH ... lt#PPA3,M1

Viewed in this context I think the major mistake is in selecting, then exclusively propagating, a single progeny of any given cross. There simply is not enough genetic diversity in a single cross for subsequent crosses to exploit for adaptation.

Mother Nature herself works with populations, not individuals. She also walks with small steps, not leaps and bounds.

In nature, any mutation or hybrid cross would be followed by many subsequent generations of introgression within that population, mixing and matching all the genes present in all the individuals. In this way Nature ensures that some individuals having newly introduced traits are stably adapted to their environment (a small area within the fitness landscape), some optimally so.

By discarding all but a few selected progeny from any cross we gain novelty at the expense of vigor and much else. Retaining only a small number of progeny even on the basis of individual vigor guarantees tossing out better combinations not yet made.

How can it be otherwise, though? Commerce demands novelty, so the philosophy of breeding for fitness competes with that of breeding for novelty. The need on the one hand of working systematically with thousands of progeny from crosses among many, very similar cultivars over many generations (that is to say, introgression) competes with the need of devoting those thousands of progeny to crosses among many different cultivars.
What doesn't kill them makes them stronger.

Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17668Post Pierre Rutten
Wed Jul 09, 2008 1:00 am

Thanks for the readings. I will need some time to go through and integrate...

Your words are fully agreed Don

It is choosing parents the key.

Doing so one presumptuously eliminate an infinite lot of possibilities about which nothing is known. And never will.

Choosing parents leads inavoidably to genetical impoverishment.

And breeding for decorative purpose the other key.

If we choose certain parents is because we have a goal in mind. A goal that is the consensus about what is (or will or should be) loved by others. The result is that we all use the same reduced set of parents and this leads to more genetical impoverishment.

To be continued later...

Fara Shimbo

Re: Pedigree breeding and diversity.

Post: # 17675Post Fara Shimbo
Thu Jul 10, 2008 1:00 am

Pierre,

Great post, and a claim I have heard repeatedly about Thoroughbred and Akhal-Teke horses, which are suffering the same problems due to the overuse of "commercial sires."

Your comments above about choosing parents bring to mind my breeding results of last year: all my deliberate crosses failed to sprout (I think this may have been in part that I didn't keep them cold enough while stratifying), but open-pollinated seeds sprouted to the point where I now need at least three new rose beds dug, and I have nowhere else to dig unless I start taking over parts of the horse pasture.

Among the roselings growing with wild abandon from the 2007 seeds are two which I am fairly sure are crosses to Rosa arkansana, which is a weed around here and blooms at the same time as the seed parent. All the leaves are 7 leaflets (mom's are 5 and Arkansana 9 to 11), and the color of the leaves and the shape of the thorns is the same as arkansana. I wish there were a paternity test for roses! While the other 12 roselings are growing on well, these two seedlings are roaring! And I have to wait until next year to see them bloom! AUGH!

Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17688Post Pierre Rutten
Thu Jul 10, 2008 1:00 am

Reduced genetical diversity ending to consanguinity and debilitation is a common consequence of pedigree breeding.

The quest for more beauty ending i.e. with dumb animals that have genetically related health problems. Or with universalized desease problems.

Apple breeders realized a few years ago that, all over the world, they all were breeding with the very same parents. Most up to date successful cvs as well as species resistance carrying hybrids.

Everyone duplicating others breeding work and hoping for the rare outstanding seedling.... with inbred problems for future problems linked to consanguinity....

Quite rose like is'nt it?

Main difference is that there are a lot more generations in rose history. That is much more consanguinity.

It was advised to use more unrelated and diverse breeding stock and as much as possible to "reinvent the wheel".

Don
Posts: 1865
Joined: Wed Dec 31, 1969 7:00 pm

Re: Pedigree breeding and diversity.

Post: # 17689Post Don
Thu Jul 10, 2008 1:00 am

>> It was advised to use more unrelated and diverse breeding stock and as much as possible to "reinvent the wheel".

Judging by the posts on this messageboard many people do seem to be working with one species or another, particularly New World species. However, even here there is the probability of repeating the processes that lead to consanguinity.

The danger lies with limiting breeding stock to one or a very few plants from unprovenanced, commercially available cultivars. It would be better to source species from the wild, using as many plants as possible collected from a variety of different geographic locations to decrease the likelihood of working with clones. This is much easier said than done, however.
What doesn't kill them makes them stronger.

Jim Sproul (zone 9)

Re: Pedigree breeding and diversity.

Post: # 17697Post Jim Sproul (zone 9)
Fri Jul 11, 2008 1:00 am

***This is just my 2 cents worth***

Okay, while I agree that increasing the size of the gene pool in modern roses is potentially a good thing, I have always felt that modern roses as a plant type were already an incredibly diverse group. That diversity is what attracted me to roses in the first place. The gene pool that I am working with includes climbers, large HT's, floribunda types, to minis and microminis. They include every color available in roses and diverse types of leaves (both of size and number of leaflets and surface - glossy to matte). Blooms may be solid colored, blends, bicolors or striped. Petal number is from 5 to more than 100. Fragrance range is incredible. And disease susceptibility is also very diverse and mixed - with respect to powdery mildew, blackspot and others. Canes are from very lax to stiff and carry one bloom or up to one hundred blooms.

With all of this diversity, I cannot help but believe that modern roses have a significantly large gene pool (yes, there are other species traits that could be added). I think the fact that modern roses are tetraploid, they are somewhat "protected" from losing their diversity.

In fact, sometimes some of the "diversity" that is brought in by using species roses is quite undesirable. The hulthemias as a group have been described as "weeds" (probably a good description!) They have nasty thorns, rambling and angular canes, thin leaves that are dull and disease prone, with small flowers that last only a day, and are non-remontant. Working with these has provided me with plenty of ugly, and very undesirable seedlings.

The concentrating of a collection of desirable traits will have an effect on eliminating other traits (genes), resulting in a decreased gene pool. Though that might be bad in some cases, it can also be good. No?

Jim Sproul

Don
Posts: 1865
Joined: Wed Dec 31, 1969 7:00 pm

Re: Pedigree breeding and diversity.

Post: # 17698Post Don
Fri Jul 11, 2008 1:00 am

>>The concentrating of a collection of desirable traits will have an effect on eliminating other traits (genes), resulting in a decreased gene pool. Though that might be bad in some cases, it can also be good. No?

If the roses offered up in commerce did indeed concentrate desireable traits and eliminate undesirable ones then we would not be having this discussion. Desirable traits need to come from somewhere other than modern roses all of which suffer at least some of the problems we wish to breed out.

You make a really good point that there is a lot of variability among modern roses, especially morphological differentiation which exists to a degree unknown in any other plant. This has come at a cost especially of disease susceptibility and decreased vitality, the inability to form roots readily and, with a few exceptions, fragrance.

>>sometimes some of the "diversity" that is brought in by using species roses is quite undesirable.

Therein lies the rub. Rose breeding is truly a high art.
What doesn't kill them makes them stronger.

Jadae (zone 8b)

Re: Pedigree breeding and diversity.

Post: # 17699Post Jadae (zone 8b)
Fri Jul 11, 2008 1:00 am

The issue with using the phrase, desireable traits, is that it is partially objective (agreeable by most, at least) and partially subjective (highly arguable).

Pierre Rutten

Re: Pedigree breeding and diversity.

Post: # 17704Post Pierre Rutten
Sat Jul 12, 2008 1:00 am

Jim

You are speaking about phenotype that is physical characters when we are dealing with genes that may be expressed or not. Apparent diversity may rely on very little DNA as are all the dominant genes i.e. for miniature or rambler.

Stated above is: "There were many rose DNA analysis in the last decade. All agreeing on the little diversity found among modern roses."

Diversity is not good in it self (although it is for resistances). It is only indispensable for the breeders to select this and not that. As Jadae shows desirable is subjective according to our tastes and goals.

The point is that ancestral species qualities such as strength and desease resistance / tolerance were lost.

And how to restore them.

Jim Sproul - (zone 9)

Re: Pedigree breeding and diversity.

Post: # 17710Post Jim Sproul - (zone 9)
Sat Jul 12, 2008 1:00 am

I hope no one minds a friendly discussion, so please accept the following as an attempt to just add another point of view to the discussion...

Don, I think that you might agree that the major rose breeders are now trying to breed in desirable traits. Historically that may not have been the case, but I think that they have heard "loud and clear" that the public wants disease resistant roses and roses that are fragrant. They are also, all working to produce varieties that do well on their own roots (an important goal that I think Mr. Ralph Moore had suggested years ago), and they are always looking for vigorous and floriferous seedlings. What else is there?

The trick is getting all of those traits into the same plant.

Jadae, you are right, "desirable traits", means something different for each of us. But, as you suggest many desirable traits can be agreed upon. For instance, I don't think that when given all other qualities being equal that anyone would choose a non-fragrant rose over a fragrant one; or a disease prone rose over a disease resistant one; or a rose that could only be grafted over one that could keep on blooming with excellent vigor on its own roots..... The arguable traits are more of the "art" that I think Don was talking about - traits like bloom style, plant habit and shape of foliage. For example, I like more rounded foliage, while a rose breeding friend prefers the foliage to be more narrow and tapered.

Pierre, I think where dominant genes are operative, it would be easier for a larger and more diverse gene pool to continue. In fact, I wonder how many desirable recessive traits there are in our own modern roses, just waiting to be "found". When recessive genes are expressed and bred for, I think that there is a greater likihood of shrinking the gene pool.

Now, I am sharing these thoughts just for the sake of discussion, because I think that we have a long way to go before we exhaust the incredible possibilities available in existing modern roses.

With regard to disease resistance - something that most, if not all of us hold as an important goal, there is a distinction made between horizontal and verticle resistance. Though there is a difference between the two, it is very difficult for me to imagine being able to, with purpose, come up with a breeding program to address horizontal resistance in any practical way. For instance, working with a particularly resistant species after a few generations of selecting for clean seedlings (and simultaneously also trying to combine this with other desirable traits), I suspect that what would happen would be that the more dominant resistant genes would be selected for and the more recessive resistant genes would become dilute. That would, in effect, result in resistance that would more resemble verticle rather than horizontal resistance. Where horizontal resistance is concerned, it would be very difficult to select hybrid seedlings based on whether or not the recessive resistant genes were or were not present.

So yes, while I agree that adding more genetic material to modern roses is desirable, I am not certain that it will cause us to arrive at the desired goal of increased horizontal disease resistance. What we really need is a way to artificially link as many of the resistant genes involved in horizontal resistance and splice them into existing modern roses.

Wouldn't that be a great trick?!

Jim Sproul

Don
Posts: 1865
Joined: Wed Dec 31, 1969 7:00 pm

Re: Pedigree breeding and diversity.

Post: # 17714Post Don
Sat Jul 12, 2008 1:00 am

>>I hope no one minds a friendly discussion, so please accept the following as an attempt to just add another point of view to the discussion.

Jim, your opinions are highly valued, and I hope my own haven't seemed less than friendly. I am really enjoying this discussion.

>>....rose breeders are now trying to breed in desirable traits

Agreed, though not much has been published about their philosophical approach. Liz wrote earlier about Kordes' efforts beginning in the 1990's (see http://www.rosehybridizers.org/forum/me ... 3009#13244), and from her description of Kordes' test beds we can guess that they are selecting rigorously for disease resistance, but against which particular gene pools seems to be a trade secret given that the Kordes are no longer disclosing their roses' parentage.

>>I wonder how many desirable recessive traits there are in our own modern roses, just waiting to be "found".

There may be a few opportunities left within the genes of modern roses. For instance, I think fragrance improvement can probably be accomplished without resort to species infusion, but will benefit by going back to old garden roses (which the institutional breeders are obviously doing, witness Austin's successes). However, we can see from both genetic studies and scrutiny of the database at helpmefind.com that modern roses spring from a mere handful of ancestral cultivars whose genes have been mixed over and over and over to the point where it is likely that most possible phenotypes have appeared and been selected for or against repeatedly.

I think there are many more opportunities lurking in the wild than in roses already under cultivation. As I have proved to myself recently, these opportunities are going to be very hard to find and reach.

>> vertical...horizontal resistance

The manipulation of traits like color and form were the first to be mastered for three reasons: they are relatively simple biochemical systems having correspondingly small numbers of genes; the phenotypes are profoundly obvious making selection a simple matter; and the market places high value on them.

Traits like fragrance, disease resistance and rooting behavior have far more complex biologies, require greater time and effort for evaluation, and are not critical for marketplace success. This is especially true of disease resistance. In fact, I disagree with any assessment that the marketplace is now demanding disease free roses.

While I applaud the efforts of the Earthkind folks, it is more the growers and marketers of roses who want improved disease resistance because it would decrease production costs and (they hope) give them a competitive advantage in the marketplace.

The same thing can even be said about hardiness. The public has proved that it is quite willing to treat roses as annuals, and as the retail costs continue to come down this will increasingly be the case. Who cares if you have to plant new rose bushes every year if they cost five bucks at Walmart? Add to this the fact that your lawn care service is quite willing to spray the rose bed for a small additional charge and you've conquered the biggest barriers to growing roses.

>>the more dominant resistant genes would be selected for and the more recessive resistant genes would become dilute...What we really need is a way to artificially link as many of the resistant genes involved in horizontal resistance and splice them into existing modern roses.

It is probably a mistake to regard disease resistance as having simple Mendelian inheritance, and it is anyway unnecessary to understand in detail the mechanisms involved in order to breed and select for it.

Nor do I hope for an engineering solution to the problem, in fact I hope this does not happen. Classical methods allow me to legally breed with the progeny, but an engineered solution would be put the "invention" of a disease resistance mechanism off limits, except by license, for 20 years. This is, incidentally, also true of Suntori's putative "blue" rose supposedly in the pipeline at J&P.

>> The issue with using the phrase, desirable traits, is that it is partially objective

How true, and the problem is compounded with systems in which the goals are highly complex and abstract. Yet sometimes opportunity lies in altering one's paradigm or even just thinking outside the box. For instance, there was a discussion a few weeks ago here about the phenomenon of extreme proliferation (see http://www.rosehybridizers.org/forum/me ... opid=17365). I have never seen this occur, but Timo's picture made me wonder: what might happen if this trait were selected for rather than against? It might be possible to take this trait to some limit beyond the grotesque and develop an entirely new flower form, say Rosa multicorolla timoflora.
What doesn't kill them makes them stronger.

Fara Shimbo

Re: Pedigree breeding and diversity.

Post: # 17716Post Fara Shimbo
Sun Jul 13, 2008 1:00 am

Can the retired prof of genetics chime in? (Domestic animal genetics, but hey, genes are genes...)

Couple of things that caught my eye:

"When recessive genes are expressed and bred for, I think that there is a greater likelihood of shrinking the gene pool."

This is absolutely true. A recessive trait can hide in a line for hundreds of generations, but once a simple dominant trait is gone, it's gone. I've wanted for years to do a study to find out just which colors (because I'm a color kind of person) are dominant and which are recessive and how they operate, but progress has been slow for a number of reasons, not the least of which is that there seem to be at least two color mechanisms in roses, one pigment based and one structural.

Don, you mentioned desirable recessive traits. There are probably tons of them! Unfortunately the only way to find them is to inbreed, and understand what masks what and so forth. This is, believe me, a much desirous goal to pursue, but you have to be willing to carefully study each plant and throw out a lot if you're going to do it this way.

"It is probably a mistake to regard disease resistance as having simple Mendelian inheritance"

I think I can say that it's more than "probably" a mistake.

And what's this about a "blue" rose?

This isn't on the topic of genetics really, but it strikes me all the same: "The public has proved that it is quite willing to treat roses as annuals..."

The public (grrrrr) is quite willing to treat anything from pets to ... well, anything as annuals. One thing I like about being a (very, very) small-time breeder is that I breed what I like, and if anyone else likes it too, wow, cool! I've been trying with an absolute dearth of success to get the members of the local rose society to come visit my OGR garden. They can't understand why I would even want to breed a rose that only blooms for a short time. Well, my reason is this: I melt at over 80 degrees. I like a rose that blooms in the relative cool, then says, "Thanks for all the care all spring, I'm going to rest now, go sit in the air conditioning." And then there's the matter of fragrance. Mmmm!

Oh well, that's all I have to say. Please feel free to ignore.

-Fa


Henry Kuska

Re: Pedigree breeding and diversity.

Post: # 17719Post Henry Kuska
Sun Jul 13, 2008 1:00 am

This study appears to be the type of study that is needed for roses.

http://journal.ashspublications.org/cgi ... /133/3/427

Or is one already available?

Link: journal.ashspublications.org/cgi/content/abstract/133/3/427

Post Reply