Constructing a viable closed ecosystem is a hugely complex problem. A system with few elements is actually more difficult than one with many elements. The ability of an ecosystem to withstand stress increases with its diversity. A blight affecting one type of tree in a forest containing dozens of types of trees can be easily accommodated. A blight in a system of one species--for example, Irish potatoes--can be a disaster. Modern agriculture is vulnerable to this, and smart farmers vary their plantings and varieties from year to year to avoid being dependent on a single species.
The Biosphere, for all the grief it's gotten from some quarters, was a good try, and I think quite a bit was learned from it.
All of the long-term space systems I've designed use engineering and supply systems instead of "ecological" recycling. A LOT fewer variables and failure points to worry about. No doubt someday it will be worked out, but possibly by devices that mimic plant processes, not actual plants.
My way to get to Mars is not with an exotic small ship with plants and a closed ecological system. It's with a flotilla of redundant vehicles and a lot of storage space.
Power is a key: you can do a LOT of recycling by brute force with a nuclear reactor or massive solar arrays somewhere on your ship. This also helps with the large amount of propulsive capacity needed to fly a fleet instead of a capsule.
Space/Science » in reply to Euhyperamoeba biospherica
Closed ecosystem
