I have a theory on how to do this with at least a humane budget (because there are much, much more expensive methods).
I take an original, untouched landrace and compare two things:
- I let the original tropical landrace grow for one generation in my non-authentic environment (e.g., in California). During this process, I select different phenotypes and take notes on them—e.g., I select the tropical phenotype, also the indica-type phenotype, the one with large roots, etc.
- I keep all the F1 phenotypes separately bagged (e.g., tropical pheno separately), and in the next season, I grow the tropical phenotype again in my non-authentic environment, side by side with the filial generation.
If I do this in California, the tropical F1 pheno will probably struggle a lot—because it has even less cold tolerance than the tropical phenos of the filial generation, which I now grow side by side.
So that means, for every phenotype I selected, I will see whether it will struggle more—for example, the tropical pheno will struggle more in California because it gets cold, because it hits a hard limit.
The idea is: if you move a Thai variety slightly within Thailand (e.g., to southern China), it will only struggle slightly in all aspects, because it still feels pretty “at home” there. So moving it in any direction would only cause minor changes and mild stress.
Vice versa, if you expose it to a very different environmental factor (e.g., cold), that phenotype will struggle much more severely.
So, if you’ve noted the degree of abrupt decline (ideally 2–3 plants per pheno), you’ll know that whenever you select toward that phenotype, you’ll get such-and-such crappy phenos.
From this, you can conclude whether the environment, the selection, or the phenotype is to blame.
Unfortunately, that still doesn’t tell us whether we should keep the phenotype or not…
That changes once we create a second grow environment where we change as many parameters (temperature, soil, light) as possible.
Because then, if the F1 phenotype is bad, it will perform poorly in all environments. But if the environment was the problem, the second environment will show that in one place the F1 develops poorly, and in the other it doesn’t—meaning the pheno is, in principle, okay.
So in the environment where the F1 pheno grows well, we can directly select the best specimens and continue working with them.
That’s my technique.
This way, we bypass the limitations of an environment that would suddenly cause certain phenotypes to weaken, and it’s almost as if we’re still selecting in Thailand.
Again, the difference in the degree of decline from the F1 to the filial generation indicates that the environment might be to blame (if you see a sudden increase in weakness in one particular environment).
