The science of seeds

I've always had a fascination with the idea of trying to grow all of a person's food in a lab under very controlled and optimized conditions.  Think space station or nuclear holocaust if you need a reason.  Obviously this is not the most efficient way to go about growing food.. or is it?  In a sterile, controlled environment you could eliminate a lot of variables that harm production and tune things to actually increase desirable outcomes.  I was thinking that this would only go so far though.  

To really improve productivity, yes you need to tune the environment to the plant, but also you need to selectively breed plants to better produce in an environment.  After generations of this process I doubt that the plants could survive at all in outside conditions.  After thinking about this, I've realized that I have very little knowledge on this topic, but would like to know more.  How do seed companies go about selecting and creating seeds year after year with good yields?  I know hybridization is part of it, but that its benefits also wear off over time.  There has to be some kind of uphill struggle to breed good plants that I am unaware of.  There have to be standardized ways to measure desirable properties of plants and log them for selection.  Can anyone point me in the right direction?  Have any of you had good experiences with saving seeds even in a regular garden?

I save seed every year. Sorghum seeds, millet seeds, squash seeds, bean seeds, and the list goes on. You just keep seeds from the best fruit, that's all.

The whole point is that they're NOT grown in a controlled environment. They're open pollinated, so the gene pool keeps diversifying even as the human involved retains the genes he likes.

My question is: what are you trying to optimize? Grow local, adapted seeds that have been selectively kept for millennia and you will find out that everything you need us already out there. Beware of chronological snobbery

Nathanael, that's cool that you've had good success!  I do get that people select seeds for the environment that they live in, I guess I'm proposing using an specific environment that can be replicated anywhere indoors.  This would allow anyone in the world who is able to maintain this environment to work on progressively improving crops with others in collaboration.

I would select for traits that would be ideal for growing indoors, such as:

Short maturity time
Use least amount of energy for lighting
Waste least amount of water
Works well with automation
Best nutrition, greatest caloric yields, etc.

Surely something like this is already being done?

NASA did a lot of work on this very idea and they started it back in the 1970's. At the end of the 1980's they determined that it was not working and they stopped the research.

Their problem was that they were trying to create too many environments and in a space station with very little room, that becomes a logistic nightmare.
It is far more efficient to set a single environment and select plants that fit well into that environment than it is to try for best fit environment to each plant you want to grow.
What works is to make each plant a best fit to the environment, it allows for maximum growing space in any size area.

When we breed plants for a specific location, we are tuning the plant to that environment, this is what Joseph Lofthouse does, it is the process of creating landrace species.

Seed companies do two types of breeding, they do closed pollination (gather the pollen and apply by brush then cover the female  flower to prevent external pollination) or they do open pollination where there is a large distance between subspecies so cross pollination doesn't take place.
Closed pollination is usually done when they are trying to improve a species (this is how we have so many different tomatoes and corn now) then for producing the seeds you buy they do open pollination as above.
The same protocols can be used by any gardener.

Redhawk

Daniel Richardson wrote:I've always had a fascination with the idea of trying to grow all of a person's food in a lab under very controlled and optimized conditions.

Iirc it needs about 1000 m² to feed one person, without meat. With meat much more. You can hardly control/optimize every condition, since many are not known completely.

As for seeds, I go on par with another reply. The best is to make your own seeds, as your plants adopt to your situation pretty fast. You just need to supply them with good humus, irrigation as needed, keep predators out and so on. Nature will do the rest for you. Planted a few onions today and it was great to see dozens of earthworms. Picked them up with matured cow manure, we used to get a better soil. Our soil is pretty acid and loamy. ;-( Now the soil seems to get slowly better.