ok, back at PC. Let me just copy and paste the post, it's long... but most important is the first sentence and the 3 different types of backcrossing at the end. The other two posts are essentially what inspired me to write the Backcross your own variety thread here in the breeders forum:
https://www.rollitup.org/t/backcrossing-your-own-variety.840329/
--------------------------------
What's the best method depends on the specific goals. Back crossing populations (what cubing basically is) is not some new technique, it's been documented nearly 50 years ago already and a very common method in modern plant breeding. It has its specific purposes which is kind of what I wanted to show. Cubing is still in a way "best to best and cull the rest" with the difference that one of those "best" is the same homozygous plant every generation and the other best are multiple plants. Instead of selecting genotypes you manipulate the gene frequency. Instead of selecting AA, you increase the number of A genes in a population, and hence the chance of two AA's being combined instead of Aa or aa.
"Best to best and cull the rest" is population breeding, also known as recurrent selection breeding (you hunt for and select the best
phenos every run). The goal there is to increase uniformity while still retaining a wide heterozygous gene pool (i.e. prevent inbreeding depression and increase the survival chance of the population). "Best to best and cull the rest" is what farmers have done for thousands of years. It's what the pioneers had to do and still applies well to acclimatizing and beefing up land races or an existing strain. It's basically what happens in nature too, the weak and ugly don't get to mate.
While population breeding reduces
phenotype variation to the better ones and can lead to a strain uniform enough for growing, for a true bred IBL one needs to go a few steps further, and reduce
genotype variation too.
Since about a hundred years ago plant we (mankind) use "modern plant breeding" techniques (partly based on Mendel's work). I can't put this any other way than:
that's what breeding
is about for nearly a hundred years. The real F1 hybrids and true bred homozygous lines used to create those are a result of that, fruits, veggies, ornamental plants, cows and pigs. Unlike heirlooms. Cannabis industry is a different story.
Please don't get me wrong, I'm not saying the old brute force methods don't work or can't lead to excellent results. Population breeding will by itself however not lead to a true homozygous variety (by genotype and not just phenotype). I'll spare you the punnet examples, but it's related to the Hardy Weinberg principle, and is natures way of ensuring variation. To create a "new" stable homozygous strain/
variety, a stable "line" would have to be created from the population by selecting individual plants, or back crossing to an already stable line. One is not supposed to stabilize an F1 hybrid but their parents. A proper F1 hybrid has no pheno variation. In reality, in the cannabis industry, they often do, because people don't start out with homozygous IBLs.
To create true F1 hybrids (that actually express heterosis aka hybrid vigor) two different stable homozygous lines are needed. To create a stable homozygous line through sibling mating you need to lock in traits, as in breed them true, as in make them homozygous. To have all homozygous offspring (making it a truebred) you need to cross homozygous with homozygous. This means having to select two plants with genotype AA instead of one or two Aa for example, which in case of complete dominance will both be the same pheno type. This requires really getting to know your plants as in how their traits inherit. This in turn allows you to predict the outcome of crossing two plants.
Obviously that is a lot of work, it requires what is called 'intense' selection and usually involves test crosses just to see how certain traits inherit. Trying to breed one or more traits true every generation. Cubing 'can' be a sort of brute force method to achieve similar results.
When you start breeding it's a bit of a chicken and egg thing, whether you create an F1 hybrid or go for backcrossing a new variety, homozygous IBLs are sort of mandatory. You can either use someone else's, or create your own. Cubing is a feasible way to create one with a relatively small amount of plants. Create another one and you can create a proper F1 hybrid.
I've had similar discussions in which this came up before so I will add up front: no, this is not just some academic theory, in fact, it's what many breeders who never even heard of Mendel already intuitively partly do. Although there are genes and traits that don't follow Mendel's "simple rules of inheritance", as in there are always exceptions, modern plant breeding is as tried and true as it gets.
To sum it up without punnet examples:
Back crossing to save genes in seed form.
- Backcross to a recurrent parent to end up with seeds that have 94% or more of the GENES of the recurrent parent, but not necessarily the GENOTYPES. For example to save the genes of a clone-only strain in seed form. The main (valid) reason for a clone-only strain however is that it's heterozygous and you don't have the parents to recreate it. This method of back crossing will not automatically lead to breeding traits true, as in homozygous, and it will still have genotype and some phenotype variety for the heterozygous genotypes that are in the recurrent parent (similar to the result of selfing an F1, but per trait). No matter how often you cross back. This method is common in the cannabis industry. It can lead to stable for growing (not a lot of pheno variation) but not necessarily stable for breeding (that goes for "best to best and cull the rest too). Some images to illustrate what happens:
http://www.extension.org/pages/3244...pulations-and-backcross-breeding#.VE5xr4t4pbw
Back crossing a specific trait into an existing line
- Backcross to a recurrent parent to end up with seeds that look most like that recurrent parent but with a trait of the donor parent. In this case you backcross one or more traits from a clone-only (or special pheno you found) to an IBL / stable line. So you backcross the clone-only trait to the IBL instead of backcrossing
to the clone-only. This is most applicable when the clone only has just one or few special traits of interest, (similar to the example in the link in my first post above). Auto genes is a good example, but also if you find a special pheno in a bag of land race seeds this is more effective than trying to acclimatize and beef up the land race (for which large population breeding would be more suitable). I have no idea how many cannabis breeders use this method, but it's the common method of backcrossing outside the cannabis industry. If you want to "borrow" a trait from another strain (PM resistance for example), you don't actually want mix the two strains as typical in cross and sibling mating but breed that trait and as little of the rest of the genes into the other strain.
Back crossing as in cubing a population.
- In short, when you back cross a population (in this context simply meaning more than a couple of plants) to a recurrent parent, the offspring (from those plants combined) will become more homozygous for all traits that are homozygous in the recurring parent. That's why it's important to back cross to a homozygous plant, instead of for example an F1 clone, but the key part is to use multiple plants from the offspring to cross back to the recurrent parent. For example, if the recurrent parent is a male, grow out 30 females, pollinate a small bud on all, get the seeds from the best ones (obviously those that show the desired traits most prominently), let's say 10 plants, pick 30 females from that, and pollinate with the same male again. The more plants (and the more seeds you pick from each), the less generations it will take to reach that point of being 90+% homozygous. Instead of using 1 male and many females, you could use 1 female and many males (and simply mix the pollen before pollinating that recurrent female). Downside is you get less bud, upside is you can cramp a LOT of males in a small space.
Essentially both the first and third method are about manipulating the gene frequency, but unlike the first, the third method (cross multiple plants from the offspring back to a homozygous plant) will increase the desired genotype frequency. The third still "also" increases the gene frequency (especially if you select out the once with the undesired traits, like the easy to spot recessive aa when AA is desired) which in turn increases the chances of ending up with homozygous alleles.
--------------------------------