re: Re:Auto-tetraploid versus amphidiploid was Hyb spots
iris@hort.net
  • Subject: re: Re:Auto-tetraploid versus amphidiploid was Hyb spots
  • From: C* C* <i*@aim.com>
  • Date: Mon, 11 Mar 2013 10:27:30 -0400 (EDT)
Sorry for the long delay on this. After doing all the research I then had to scramble to get caught up in everything else. Still scrambling.
The literature on all this is hard to track down, and rather heavy reading. 

With the TB iris we have  used the following species.
I. pallida
I. pallida subsp. cengialtii oor species, depending on who you are talking to 
I. illyrica
I. variegata
I. aphylla several clones, and there is a lot of chromosome variability here 
I. cypriana
I. kashmiriana
I. mesopotamica
I. trojana
So a lot of sets of 12 chromosomes. Mitra has published karyotypes of a number of these species. These are very well done. He used a chemical to fix chromosomes at division point, and drew chromosomes using 2500x . No banding studies have been done on this group of iris. It is seen on his studies, that some chromosomes match well across species, and a number donbt The satellite differences, as it turns out, are not critical to chromosome pairing, and usually donbt have coding regions. Which means, if not matched, (or lost) have no effect on the plant.
Of all these chromosome sets floating around, we probably mostly have pallida, variegata and aphylla. Exact matches in four chromosomes will mostly form tetravalents. Exact matches in two chromosomes will form bivalents. Very close matches will form bivalents. Exact matches in three chromosomes will form trivalents. All the rest remain asunivalents.
As to tetravalents, apparently they do go to one cell or the other during first stage of meiosis. But in the second stage, they resort, a process called "double reduction". The odds of sorting out two into each of the gamete cells turns out to be 50%. Much better odds then I had thought. So if we have 4 tetravalents, the odds of a fertile gamete would be (B=) to power of 4. Or one in 16. Randolph found between 1-4 tetravalents in the chromosomes he studied. So this would represent 1-4 chromosomes that are fully matched in each of four sets.
If we had 12 tetravalents, as in full tetraploid, then odds of a fertile gametes would be (B=) to power of 12. Or one in 4,096. In allotetraploids, formed from chromosome doubling, fertility varies from 0-87% of the diploid form. So there is some variability here.
The same situation applies to the other odd pairings. That is the trivalents and the univalents. That is they sort out in ratios of 50% in meiosis one and resort in meiosis two. For each one of these the probability of proper sorting out gamete formation is about 50%. So if four matching genes form a tetravalent, sorting out would be 50%. If instead of pairing in a tetravalent, or a bivalent, we have univalents, the proper sorting out of these two unpaired chromosomes would be B= x B= or 1/4 (25%). So univalents are much worse for fertility.
In regards to this variability, many of these allotetraploids were done on commercial crops, that had been outcrossed. That is were all ready hybrids. And most species are the result of joining of other species, and chromosome . When it comes to pairing up in meiosis and mitosis (normal growth cell division) a close match will work. Of course the poorer the match, the more likely there will be a duplication error or gamete error.
We do have some tetravalents in TB iris. Kidd showed this with some statistical studies on sorting out of plicata alleles .
A fully fertile pod parent produces around 100 + seeds. We do have a wide range of fertility with the 4x 48 plants. Ranging from the 100+ seeds to those which are fully sterile. This range of fertility probably matches with the number of tetravalents, of fully match chromosomes. That is , the more the plant is a tetraploid, the less fertile it is. The more it is an amphidiploid, the more fertile it is. Each tetravalent reduces fertility by B=, each trivalent or univalent reduces fertility by a factor of four, that is 1/4 of a bivalent.
What this all means, is that the farther a 48 chromosome plant is from an amphidiploid, the less fertile it is. This of course affects pod parents mor e then pollen parents. There is a limited number of eggs per flower (probably around 100-125 per TB flower) but basically unlimited pollen grains. So this is why we get more plants that are pod sterile then we get plants that are pollen fertile. Probably each pod/pollen combination that is not fertile reflects these problems to the degree of non seed set. The number of seeds set in a pod also probably reflects the degree of mismatch among the chromosomes sets. The reduced fertility of aphylla 4X MTB X SDB is also a reflection that most likely we have a mismatch between the 12 chromosome set from the SDB and the 12 set it was paired up with from the MTB.
A term used to describe the species with a few tetravalents along with bivalents during meiosis is "Segmental allopolyploids" this would seem to be what we have, for the most part with the 48 chromosome iris. Except for the univalents. Not sure if this term applies to plants witha tetravalents, bivalents plus univalents. 

Chuck Chapman


-----Original Message-----
From: Tom Waters <irises@telp.com>
To: iris <iris@hort.net>
Sent: Fri, Mar 1, 2013 6:22 pm
Subject: re: [iris] Re:Auto-tetraploid versus amphidiploid was Hyb spots

Chuck wrote:
"So fertile 4x 48 iris are should basically be considered
amphidiploids. The infertile ones are probably autotetraploids."
I believe I understand the point you are making, but I think "amphidiploid" is a very poor choice of terminology in this case. We know that TBs are not autotetraploids, and it makes sense, as you suggest, that two chromosomes 
that are more closely homologous than either is with the other
near-homologs would pair and form a bivalent, rather than a tetravalent
with the other two. (Although I have no idea if there are any observations 
that confirm or refute this conjecture.)
However, in a true amphidiploid, chromosomes from the "unlike" sets do not 
pair at all, or at best rarely and with difficulties. It is clear that
there are no such two "unlike" components in the TB chromosome sets. If
there were, the amphidiploid SDBs and CGW arilbreds could not be fertile, because the two TB sets they received from their supposed "amphidiploid TB" 
parent would not be able to pair.
It's very confusing to refer to the tetraploid TBs as amphidiploids when nearly 100 years of breeding demonstrates clearly that they do not function as such in outcrosses to other types. Amphidiploids function as diploids (hence the name), and so cannot produce fertile offspring when crossed with 
other types.
I make essentially the same point in my post about tet MTB parentages. If 
you routinely get fertile seedlings when backcrossing to either of the
parent types (for as many generations as you like), you have a functional 
tetraploid, not a functional amphidiploid.
Regards, Tom




Tom Waters


Telperion Oasis ~ www.telp.com/irises


Cuyamungue, New Mexico, USA (zone 6)

----------------------------------------
From: "Chuck Chapman" <irischapman@aim.com>
Sent: Thursday, February 28, 2013 9:51 AM
To: iris@hort.net
Subject: [iris] Re:Auto-tetraploid versus amphidiploid was Hyb spots

We had been discussing  aphylla  X TB hybrids and fertility.

Actually, it turns out that full tetraploids, (autotetraploids) with
four identical  genes, have much reduced fertility compared to
amphidiploids, or straight diploids.

This is because  during meiosis (forming of  gametes, eggs and pollen)
four sets of identical genes often form tetravalents. And during first
stage of meiosis,  either all of these four genes go into  one cell, or
the other, leaving no copies in the other cell. When there is  small
differences in the  four sets of genes, of a particular chromosome, you
get two  bivalents and  so proper  division during meiosis

Now remember,  most tetraploid iris, at least the TB iris , have
multiple  species involved  in their creation, so probably a number of
genes from different species, with enough of a difference, and enough
of a similarity, to pair up as fertile bivalents , rather then
tetravalents during  meiosis.

Also almost all ( I'd say all, but there probably an exception or
several out there somewhere) fertile tetraploids found in nature, are a
result of an inter  species hybrids, and thus are  amphidiploids,
rather then autotetraploids.

So fertile 4x 48  iris are should basically be considered
amphidiploids. The infertile ones are probably autotetraploids.

Look up "fertility in autotetraploids"

Chuck Chapman

---------------------------------------------------------------------
To sign-off this list, send email to majordomo@hort.net with the
message text UNSUBSCRIBE IRIS

---------------------------------------------------------------------
To sign-off this list, send email to majordomo@hort.net with the
message text UNSUBSCRIBE IRIS

---------------------------------------------------------------------
To sign-off this list, send email to majordomo@hort.net with the
message text UNSUBSCRIBE IRIS
Other Mailing lists | Author Index | Date Index | Subject Index | Thread Index