T he idea of creating materially closed microbiospheres, including humans, to study ecological processes had its roots in several branches of research. One was the sealed microcosms and open, but boundary-defined, mesocosms that ecologists developed to study ecosystem processes. Another source was the experimental life-support systems designed for use in spacecraft and as prototypes for space habitations. During the 1960s, H. T. Odum began advocating life-support research in sealed greenhouses that would rely on the ecological self-organizing properties of the enclosed soils, plants, and animals (Odum 1963). In 1971, Dennis Cooke wrote, "The fact that we are not now able to engineer a completely closed ecosystem that would be reliable for a long existence in space...is striking evidence of our ignorance of, contempt for, and lack of