|
Post by gilbert on Dec 20, 2016 9:43:41 GMT -5
As a purely hypothetical case; let's say a critically important but underfunded seed genebank is on the rocks; there is no more money for fancy isolated grow outs, the land with the backup freezers is being foreclosed on, and the thousands of accessions are nearing the end of their viable lifespan. To make things worse, bureaucratic squabbling in the organization has crippled any constructive response.
An ambitious employee, however, has determined to save the genetics in their cabbage collection. There are very few OP cabbages left in the seed trade in the USA, and many of the hybrids probably contain similar genetics or male sterility. So he feels this is important, and obtains samples of all hundred accessions.
Having no funding and only an acre of land of his own, he plants two of each accession in a big evaluation and landrace plot. His goal is to identify any standout varieties, which will then be maintained as pure varieties for those who want a uniform variety, while forming the base of a hyperdiverse landrace to preserve as much genetic diversity as he can.
Assuming that each accession is truly different to some degree, and that his climate is favorable enough to cabbage so that none of them are eliminated before flowering, how many plants should he save seed from in each subsequent generation to maintain at least 99% of the original diversity?
|
|
|
Post by reed on Dec 20, 2016 10:40:55 GMT -5
I would guess at least in the first few seasons, all of them, which of course will rapidly increase the amount of land needed.
Is the attempt to identify and keep pure the standout varieties and develop the landrace simultaneously on one acre? I don't understand how that could be done. Wouldn't each variety have to be isolated? Not easy on one acre and counter to the landrace aspect.
If I was him, assuming the seed is at risk of being lost and I was able to get my hands on it I would try to find multiple collaborators in varying areas to pursue the landrace , not worrying about any being lost due to a particular climate. I would love to get my hands on something like that.
|
|
|
Post by gilbert on Dec 20, 2016 12:27:34 GMT -5
I guess my particular scenario was just confusing.
But the main question is; with initial diversity from 100 varieties in an outcrossing population, and without unusually heavy selection pressure from a harsh climate, what population size would retain 99% of the genetic diversity? I agree that in practice many collaborators would be better. But as far as theoretics go, what size will allow a landrace to retain all of its diversity?
In practice, does anybody know if there are collections of the type mentioned in danger, and could the folks on here band up to save at least the genes if not the varieties? Of course, the original curators may not like this idea; they might prefer that the varieties be saved, but I guess if we pooled money and bought the seeds instead of just asking for them, it would be legitimate.
I'm especially interested in preserving the species that seem to have been hit hardest by genetic loss; brassicas, carrots, some grains, beets, perennials, and some unusual crops. I really can't get too excited about loosing a few tomato or pepper varieties, seems there are plenty and to spare of those.
|
|
|
Post by walt on Dec 20, 2016 14:48:00 GMT -5
I do my internet at the public library, so I don't have my copy of Allard's book Principles of Plant Breeding. But I believe that the equation for loss of genes is 1/2n per generation, where n is the population size. So to only loose 1%, !%=1/100=1/(2x50). So in a population of 50, only 1% of genetic diversity is lost per generation. Now the questions are (1)do I remember the equation correctly, and (2) did I apply it correctly? The your hypothetical case, you state that each of 100 accessions has some unique trait. That unique trait could mean a unique gene combination, or it could be each having one or more unique genes. If each of these unique traits was due to a distinct unique gene, then there would be a lot of genes already at 1% gene frequency. It seems to me that some would get lost quickly due to genetic drift. With 50 plants per generation, you could prevent inbreeding depression forever, but not keep 99% of the genes for even a single generation. Going back to your original hypothetical case, you state that for starters, you would use 2 plant from each of 100 OP varieties. That is 200 plants. Hypothetical genes unique to 2 of those plants would seem to me to still be at risk. In other words, I have my doubts that the equation applies to populations that have many genes each with very low frequency. I'll point out that 100 unique varieties may not have 100 unique genes. You might have varieties with abcdefg, bcdefgh, acdefgh, abdefgh, etc. Hypothetically 100 unique traits may have 100 genes each at 99% frequency. In real life then, 100 unique traits could be due to anywhere between 0 to 100 unique genes. We just don't know without a lot of test crosses and/or gene mapping. I'll get out my book tonight and see what Allard had to say about it. Allards book Principles of Plant Breeding was the textbook for 3 of my graduate level classes in genetics-plant breeding. And in all the other classes it was just assumed we had read and understood it. That said, I try to re-read it every winter for the last 38 years. It is that full of information. It is also written clearly that a layman can start with it and learn a lot on the first reading. But there is more on each re-reading. When I was packing to go to the Peace Corps, I asked one of my graduate professors what books I should take. He said "You have Allard don't you?" I said "Yes." He said "He said it all." Sorry, nothing in it about gene-spicing. That hadn't been invented yet when my edition was printed. Amazon has it. Second hand it is affordable for most people.
|
|
|
Post by Joseph Lofthouse on Dec 20, 2016 19:27:01 GMT -5
Having no funding and only an acre of land of his own, he plants two of each accession That's a tremendously severe bottleneck. After doing that, I wouldn't consider the descendants to be the same variety any more...
|
|
|
Post by zeedman on Dec 20, 2016 20:28:59 GMT -5
Ditto on Joseph's comment. Brassicas are out-crossing, requiring insects (or other means of mechanical transfer) in order to pollinate. Furthermore, individual plants are generally self-infertile, so they actually require pollen from a different plant. You could do a mass cross planting, with the aim of developing a 'land race'. To save pure seed from that many varieties, though, borders on the realm of impossibility. In either case, with the small theoretical population stated, a great deal of generic diversity would be lost in the first generation.
|
|
|
Post by reed on Dec 21, 2016 6:24:23 GMT -5
... could the folks on here band up to save at least the genes if not the varieties? Of course, the original curators may not like this idea; they might prefer that the varieties be saved, but I guess if we pooled money and bought the seeds instead of just asking for them, it would be legitimate. I like that idea, if a bunch of people sent a few dollars to one person who then bought all available varieties of a particular thing, broke them down into smaller packets and sent back to all the contributors humm. It would take quite a commitment though. A quick check of just a couple sites came up with about twenty varieties of cabbage. At say $3.00 each, probably $75.00 or more since no one source has them all which increases shipping. Then shipping back to everybody, another few bucks depending on the type of seed, brassica would be pretty cheap to ship but still a couple, three bucks per. A more thorough check would come up with more so it would get pretty pricey. Still I like the idea, it might be doable at least to some degree.
|
|
|
Post by gilbert on Dec 21, 2016 11:00:06 GMT -5
Hi everyone,
Thanks for the answers!
The two plants a variety would be to evaluate and to start the landrace in a mass cross, not to preserve the individual varieties.
I'm going to start a new thread about the cabbage project, which is starting to move from theoretics to actuality, and keep this thread about the theoretic maintenance of genetic diversity in a landrace.
Walt, if one percent is lost per year at 50 plants, when would this loss stabilize? What percent would eventually be lost? I understand about the difficulty of figuring out how much diversity there is to start with; does anybody have any ideas on how diverse the genetics actually are in a given species? And you are probably right that the 1/2N equation wouldn't hold true for a landrace in the initial stages; all sorts of random accidents could lead to a loss of diversity before the genes had time to mix.
I think most of you are working with landraces as a way to adapt crops to your area, which is great; I'm working on this myself. But I also want to preserve as much genetic potential as I can in a given crop so that others can eventually use it. I don't think the current seed bank model is working. So I'd like to get a hold of as much diversity as I can, and, with collaborators in other climates, create a "jumbled landrace seedbank" containing all the diversity from the separate accessions in one big pool. Collaborators could swap seed back and forth to keep the pool large. I would assume that less rouging would be done then if we were working to adapt the landrace to our climates.
But for this to work, I'd have to figure out if all the genes will actually survive. It would be a pity to loose the rare genetics and end up with only the ones that are common anyway.
|
|
|
Post by walt on Dec 21, 2016 16:28:03 GMT -5
OK, I miss understood the original question. I thought that you were talking about starting a landrace by growing 2 each of 100 varieties. To keep 99% of the genes in an OP variety for one generation would require 50 plants during that one generation. I calculated that to keep 99% the genes for 2 generations for 2 generations requires 200 plants per generation I'm not even going to try to calculate for 3 generations. So where does it stop? To maintain a population for any length of time, you have to select. At somewhere under 50 plants mutation pressure, the constant creation of new mutations, can maintain genetic diversity. Most new mutations are undesirable, but with constant, even mild, selection, will keep a variety going, maintaining its identity as a given variety. Thus, varieties are never stagnant. They are either being selected, or they are declining. The selection can be entirely by nature at one extreme, or entirely by humans at the other extreme. Selection is best with both human and natural selection going on. When the big gene banks were being started, there was much discussion about how it isn't possible to keep a variety constant. But then everyone shut up because the big gene banks, while never perfect, are better than nothing. Even that is still debated sometimes. Ancient Greek philosophers said "You can't stick your hand in the same river twice" The idea was that by the time you stick your hand in again, the river has changed. Quantum physicists say "You can't even stick your hand in the same river once." Breeders need to accept this, and accept it as possibly a good thing, since we have to deal with it anyway. So, with an upper limit of 50 per plaants generation, one might save seeds from 10 pants per year, if you need that much seed. And freeze it while you grow another bunch and save seeds from another 5 plants the next year, etc. In 10 years you have seeds in the freezer from 50 selected plants. Mix all that seed and start again. Problem here is that it is 10 years per generation. So maybe go with fewer total plants, therefore fewer years per generation. There is no perfect solution here. There are solutions one can live with
|
|
|
Post by walt on Dec 21, 2016 16:37:34 GMT -5
Please note that I have edited my post from yesterday. I wrote the wrong name for a book. Principles of Plant Breeding is the one I highly recamend.
|
|
|
Post by keen101 (Biolumo / Andrew B.) on Dec 22, 2016 11:49:57 GMT -5
While i only tinker with adaption and landraces on a very small scale i'll add my two cents. I really don't think saving 99% of the unique genetics to be very realistic even if you were to have lots of land and enough plants of each variety. 50% might be more realistic, unless each variety is isolated, but even then there is still genetic drift when plants are inbred.
The second thing i will add is that with the pea breeding and evaluating sometimes i have gotten one year vastly different climatically from the year beefore and the varieties that did excellent the first year do not do well the next and vice versa. If you were to only be evaluating 2 plants for evaluation to find the best before starting the landrace i would recommend evaluating for three years first.
|
|
|
Post by walt on Dec 22, 2016 14:34:47 GMT -5
While i only tinker with adaption and landraces on a very small scale i'll add my two cents. I really don't think saving 99% of the unique genetics to be very realistic even if you were to have lots of land and enough plants of each variety. 50% might be more realistic, unless each variety is isolated, but even then there is still genetic drift when plants are inbred. The second thing i will add is that with the pea breeding and evaluating sometimes i have gotten one year vastly different climatically from the year beefore and the varieties that did excellent the first year do not do well the next and vice versa. If you were to only be evaluating 2 plants for evaluation to find the best before starting the landrace i would recommend evaluating for three years first. About a half century ago, Harlan and Martini published work with barley that showed exactly what you saw with peas. Wow! It is getting more like 3/4 century, but that isn't the point here. I also don't think saving all the genes is a good idea. What you don't want is loss of random genes due to too small populations. But we would like to loose genes for poor growth in given situations, poor flavor, poor keeping quality, etc. Most of those desired genes will have been selected for in several varieties. Some will be in all domestic varieties, like genes for heading, if we are limiting this to cabbage. Then there is the question of bringing in genes for heat tolerance from collards, etc. And then there is the question of whether I should even get involved, as Chinese cabbage does much better for me than does true cabbage. Interesting thread though. It addresses questions that apply to all crops.
|
|
|
Post by jocelyn on Dec 26, 2016 9:04:19 GMT -5
I'm trying to save chestnuts, american chestnuts from the eastern seaboard, specifically the more northern populations. Chestnuts are obligate outcrossers like cabbages. Would the same numbers apply? 2 grafts planted side by side, each seedling gets half from each parent, so half of genes left out. 2 seedlings get half and a half again, so 3/4 of available genes. 3 seedlings from that pair of trees captures 7/8, 4 captures 15/16....am I missing something, or is that mostly it?
|
|
|
Post by walt on Dec 27, 2016 14:23:18 GMT -5
You got it right.
|
|
|
Post by jocelyn on Dec 29, 2016 9:59:12 GMT -5
Thanks, Walt. When just learning, it's good to double check, grin.
|
|