Rosa multiflora self-compatibility

Rose Genetics Ploidy Level
, , , , , , , , , , , , , , , , , , , , , , , ,
21

Hi Adam.

I know that in a sampling of 14 such seeds, only one germinated. This had me curious so I opened the other 13 seeds up only to find possibly all the rest were grub affected! So getting percentages might be a tricky thing to interpret with such nasty “biases in the field”, so to speak.

Still, I very much love your suggestion…I am putting your idea into place for a 2 year old grafted olive tree I have in my mom’s back yard…it is a variety that produces MASSIVE olives, and looks like it is throwing flower buds for the first time in its life, now! If you are at all interested in olive varieties, it is the “Bella di Cerignola” variety. I was kindly given a piece from this variety to use for grafting from my great aunt’s olive tree in Adelaide South Australia, 2 years ago. So I am going to get hold of one “Frantoio” olive plant, to act as pollinator and get more fruit set ina few years time!

22

though it’s spotty and inconsistent.

Short version for R. multiflora: pollen from a different plant is the best way to get viable seed, but self-pollination can occur, if only once in a blue moon. The odds may improve slightly with pollen from a different flower from the same plant. We have to expect that the seed of SI species will produce hybrids almost exclusively.

Link: www.uoguelph.ca/canpolin/Publications/2009%20MacPhail%20and%20Kevan%20Rosa.pdf

23

This is a great discussion. Gametic self-incompatibility is what roses have and the system tends to be leaky as suggested. The pollen germinates on the stigmatic surface and then proteins in the style interact with proteins in the growing pollen tube. When there are Self incompatibility (first called self sterility) alleles in common between the pollen and female, the pollen tube is hindered and typically does not reach the ovule.

The rate of self-sired offspring is influenced by a number of things. One is that some individuals have a more leaky incompatibility system. There is variation among genotypes for strength of the incompatibility system. In addition, environmental considerations also come into play. Very early in the duration of the flower and late the incompatibility proteins tend to be less active and can allow for incompatible pollen tubes to grow long enough to sire offspring. In addition, if the plant hasn’t set seed yet, some later blooms may have a greater tendency for some rate of selfing. Temperature is a big factor too. Warm temperatures can denature the SI proteins to some degree and let pollen tubes get through. I have found this to be the case with some diploid R. multiflora based polys. In a typical greenhouse they do not set seed without controlled pollination, but in a hot greenhouse (over 100F during the day), I got hip production. I raised seedlings and the seedlings were consistent with selfing with the traits in the seedlings (thornlessness is recessive, etc.).

In lilies, another genus with gametic SI, putting the long styles in 100F water for a period of time before typical pollination or stump pollination (decapitate the style to reduce distance and then pollinate the stump with applied stigmatic exudate are used to work around the SI system. I got some stump pollinations to work with embryo rescue after that point for intersectional hybrids, but for selfing, seeds can form and mature on the plant to facilitate inbreeding and parent building.

Plants have systems in place to help them survive. Outcrossing helps promote genetic diversity and SI helps aid that. Selfing helps to produce offspring even when plants are isolated. It’s definately nice when we have strong SI to help us avoid having to use emasculation when we know we don’t have pollinators around to “contaminate” our crosses.

24

Hi Cass.

Thanks for the interesting article you posted.

It was very interesting to read and ponder over!

25

Hi David, that is great information!

26

pollen from a different plant is the best way to get viable seed<<<<<<<<<<

Cass, do you mean pollen from various other nearby R. multiflora plants which are not genetically identical (eg. a population of R.multiflora-derived seedlings)?

27

Yes.

The following explanation is no doubt over-simplified or wrong. I read about population genetics in a Pop Sci kind of way. I hope those among us with modern training will correct me.

Think of SI as a self-fertilization-death-to-inbred-offspring gene. Self-incompatibility seems to be a process by which one precise allele - a form of a gene that must have many, different, variable forms within the population of the species - triggers a growth process that aborts the normal reproduction. [SI is more likely a complex of genes, but this will work for now.]

Stated another way, self-incompatibility is stored in an SI.a-form of a gene in a single plant. When a plant’s own pollen pollinates the pistils, the SI mechanism is activated, and the reproductive process is aborted: death to the potentially inbred offspring. The same gene, but without that specific SI.a-allele, is carried by other (but not all) Rosa multiflora plants. Pollination by plants without the specific SI-allele and instead carrying SI.b, SI.c, SI.d or SI.e…z-alleles (or even lacking any SI-allele!) results in normal reproduction.

In practice, you can pollinate with another plants (not another clone) of Rosa multiflora, and, depending on how many different SI-alleles exist, you’ll get a successful pollination that isn’t a hybrid. Note I didn’t say “pure” multiflora, but that’s how we think of it. Rosa multiflora is a population of individuals. Certain members of the population are going to have that particular SI.a-allele and the pollinations by those members are going to fail. Other individuals will have SI.b, SI.c, SI.d, etc. alleles, and pollinations by those members are going to succeed. The success rate will depend on how many different alleles exist among the available pollinators.

I don’t know if the research is clear how this system works in Rosa, let alone in R. multiflora. Maybe there are 4 or 20 different SI-alleles. Google gametophytic self incompatibility, and you’ll get many hits explaining hypotheses with nice graphs.

This is a favorable evolutionary strategy for a species with a large population in which there is no shortage of pollinators from diverse members of the population. It prevents inbreeding and promotes diversity. And as the Debener article shows, Rosa multiflora “recruits” pollen from other compatible roses when an appropriate mating doesn’t occur. One hypothesis in the invasive plants research is that resulting hybrids then back-cross into the species, contributing another form of the SI-allele, allowing pollination to recur from within the population.

Population bottlenecks and the Founder Effect are other population genetics theories that explain our inability to produce representative species roses when the only plants available for breeding are either clones or seedlings from a single hip.

28

Hi Cass.

So when we talk about hybrids versus selfs in R. multiflora, such hybrids could in fact be a R. multiflora of one genotype X R.multiflora of a slightly different genotype (eg the two parents of the hybrid R. multiflora plant could even be two sibling R. multiflora from the same hip?!).

Here is one example of where I think SI works in a vegetable plant I have tried to propagate by seed (note it can alternatively be propagated by root division leading to genetically identical clones):

My mom grows a type of endive in her back yard, which she uses in cooking. Last year I took one of her tiny endive seedlings and grew it as a lone seedling in a large crack in between a cemented pathway in my cemented back yard (rofl). It thrived, flowered and then I collected the dried flower heads for seed.

Anyways, later on in the season when I sowed the crushed flower head material to get the new seedling crop, I got like 3-4 seedling germinations instead of the usual thousands of seedlings for that amount of material that was sowed!!

I suspect it is likely in this example, that since this lone seedling had no nearby sibling (not clonal) pollinators, self-incompatibility issues led to near zero successful self-pollination / seed set.

Lucky she always keeps her own “good seed”, which she collects from several endives that have flowered together, and which are themselves seedling/siblings and not genetically identical clones (ie. they are not asexual propagules from root cuttings which could then cause SI issues at the time of flowering and pollination).

29

Yes, George, exactly right. chinese cabbage is the same way (at least most kinds). Sometimes the system can be forced to break down by strong selection. For instance your 3-4 seedlings, might now have some self-fertility. I worked several years to get a self-fertile Brassica rapa (turnip or Chinese cabbage) in the strain called Wisconsin Fastplants. Repeated selection of the most self-fertile single seedling in each generation over about 10 generations gave decent self-fertility. These are diploids. A tetraploid like Napa cabbage (B. rapa x B ???) is self-fertile. Parallels some observations with roses too. So some folks might have R multiflora that is truly self-fertile while most do not, if there was strong selection for self-fertility at some point in its history.

30

That’s also very interesting to know about, Larry.

31

Sometimes the system can be forced to break down by strong selection<<<<<<<<<<<<<<<<<

Simon’s multiflora clone seems to be one example where this may have happened.

32

Interesting: Simon, you did actually a test on apomixis in your multiflora:

I pollinated them and put a small draw string bag over the whole lot. I had intedned to return to pollinate the others too but when they took longer than a week to open I decided not to do them because my pollen had ‘expired’. I left the bag on until the risk of cross pollination had passed and then continued to leave it there waiting until the other buds had matured… they still formed hips even though I didn’t pollinate them and they had died by the time I removed the bag.<<<<<<<<<<<<<<<<<

I sometimes wonder, how many pedigrees might be incorrect due to those roses “that find a way”

33

No… I might expect apomixis to have occurred maybe once or twice, but not with 1000s of seeds and the crops is literally 1000s of seeds each year. I don’t know that I tested for apomixis (if I am understanding you correctly or are you asking whether any tests were carried out to determine whether the seeds were formed apomictically), but rather indirectly demonstrated self-fertility. The actual hips that formed in the bag were not collected as I had more than enough seed/seedlings to play with from previous years OP crops.

34

Yeah the article Cass posted here supports that there would be very little if any agamospermy in R. multiflora, unless I am reading it wrong?!

35

Hi Simon,

as far as I understand: seeds that are produced without any pollen contributing to the embryo must be apomicts.

I had to deal with this problem when I used male parents of very low fertility on one particular hip parent (sorry don’t know the name of the latter. I bought it as Queen of England, but I doubt that). My very few resulting seedlings did not show any paternal influence. Some looked extremely similar to the hip parent, most of them like a simplified version of it. So I checked what happened if I just demasculated flowers and put no pollen on it. The female parent produced hips and the resulting seedlings gave me the same range of looks.

Crossing QE with Teasing Georgia gave me seedlings that definitely showed influences of the pollen parent.

My suspect Queen of England produced also in other crosses and in op hips some seedlings of those apomictic looks.

In the next years I checked several other roses and in most cases demasculated flowers without pollen aborted sooner or later.



So, some roses seem to be very prone to produce seeds just on their own.

But what about others? They might occassionally produce apomicts…

36

they did have pollen… self pollen

37

ok,that is different then. I assumed they were demasculated to do controlled crosses.

Sorry

38

I suppose if I had a clone of R.multiflora which I suspected did appear to show lots of selfing (like Simon’s clone), then I might think about emasculating it to lessen selfings (if I wanted to do some controlled crosses with non-multiflora pollens).

Having said that, on the other hand, would this emasculation step create a problem in little/no hip-set, (as I have read that some diploid species “resent” emasculation and are less likely to cross if this is done)?

Any opinions on this?

39

Well… I now have 11 OP R.multiflora seedlings that are old enough to be characterised. Some have prickles, some are completely smooth, one is dwarfed, some have dark green leaves some have pale lime green leaves.

These all look like unique individuals to me.

40

What correlation, if any, is there between this type of observed variability, and the assumption that they are selfs or not selfs? Can anything be inferred on these findings alone?