I’ve just read this article in the NYTimes about contained (though not sealed) greenhouse systems that grow fish and use their wastes to fertilize plants. Aquaponics, which I confess is new to me, ticks a lot of boxes: farming fish circumvents the problems of overfishing, but doing it indoors and recycling the wastes avoids a lot of the pollution inherent to aquaculture systems around the world. And – if the testimonials given in the article and comments here are true – plants absolutely thrive on fish poo, and – unlike hydroponics – the produce is tastier than you’d expect from greenhouse-grown winter tomatoes and other vegetables.
As some of the commenters have pointed out, these systems do require some inputs, in particular food for the fish (though this can be scrounged from, for example, restaurant and table scraps, as fish like tilapia will eat anything), and energy for water pumps and for grow-lights during times of the year when daylight is limited. But although you can buy these systems as kits, they can also be made from scavenged materials.
The biggest concern I can see here is the possibility of fish escaping into local ecosystems. If I think about my new backyard in Cambridge, for example, we’re only about 100 yards from the Cam. What if my tilapia fry somehow got out, into the sewer, and thus into the river? Southeastern England is just not that cold (well, except this winter); they could conceivably survive and – of course – breed. So, I’d probably choose a local species of trout, which tastes better to me anyway (though I’m not so sure about how it would work in fish and chips).
Here’s a bunch of case studies from a UK website dedicated to aquaponics; this group collaborates with the University of Stirling, in Scotland. I’ve put in a request to be added to their network.