Comparisons of Early Life Support System Experiments
Bioregenerative unit Fungi, algae, rat
Gas balance, waste recycle
Algae, bacteria, mouse Gas balance, waste recycle, water recycle
Gas balance, waste recycle
Algae, bacteria, man
Gas balance, waste cycle, water recycle
Source: Adapted from Taub, F. B. 1974. Annual Review of Ecology and Systematics, 5:139.
lists a few of the systems tested. Most of these early systems relied on the green algae Chlorella pyrenoidosa, whose laboratory culture was well known. Chlorella cultures could absorb CO2 while producing oxygen and food under artificial lighting conditions. Much research was done sealing in various animals, such as mice, rats, and monkeys, with one or a few microbial cultures, such as Chlorella, to test for self-sufficiency, with the test lasting several days or weeks. This work was funded by the U.S. Air Force and the National Aeronautics and Space Administration (NASA) with millions of dollars spent. Eventually, these algal-based systems were abandoned due to unknowns of reliability, high weight of water required for culturing algae, and difficulties converting algae into an acceptable human food (Taub, 1974). Emphasis in the 1970s shifted to higher plants as the basis for bioregenerative life support systems and work in this direction continued both with NASA in the form of their Controlled Ecological Life Support System (CELSS), which was formally started in 1978 (Galston, 1992), and by the Soviet Union's similar approach, termed BIOS (Salisbury et al., 1997).
A small group of ecologists, including primarily H. T. and E. P. Odum, R. Beyers, G. D. Cooke, and F. Taub, became interested in the challenge of life support system design in the 1960s. The ecologists suggested that a multispecies ecosystem was required to support a human during space flight in order to ensure reliability or stability. Their basic argument relied on the diversity-stability relationship discussed earlier and on experience with microcosm experiments. In particular, microcosm experiments of ecological succession demonstrated that self-organization results in a stable ecosystem with balanced primary production and community respiration, which implies balanced gas exchange between oxygen and carbon dioxide. H. T. Odum (1963) described the process needed to design such a life support system for humans with the same approach of multiple seeding used for microcosm setup. He calculated that the multispecies life support system would require about 2 acres per human (0.8 ha) based on considerations of the energy transformation of sunlight through primary production. Such a relatively large area was required due to the low efficiency of photosynthesis in converting light energy to chemical energy in organic matter. The large area with many species was also needed to provide the "homeostatic mechanisms" (Cooke et al., 1968) required to maintain system stability. Multiple seeding and self-organization at the scale of 2 acres per astronaut did not meet with the approval of the engineers and physiologists working with single-species cultures. Such a system appeared to be too large and too heavy to be practical, which of course is a valid criticism.
The fundamental issue is long-term reliability. Once the spaceship leaves earth the astronauts must be able to rely on the life support system or risk death from lack of an atmosphere or food or buildup of waste products. The question is which approach best meets the requirement of reliability. The conventional engineering approach tried to develop and optimize subsystems for different functions, such as oxygen production, waste recycle, and food production, and then to connect the subsystems back together. The subsystems consisted of single-species cultures and mechanical components in this case. The overall design concept was to keep the system simple and well understood as noted by Brown (1966):
What we are trying to do is ... to make this closed system as simple as we can. The virtue of simplicity is that you can understand all of the regulatory factors that you have to worry about; you can put in the proper manual or automatic controls; and the fewer the components, the better.
This is essentially the well-tested and established engineering method of design. In comparison, the ecologists' approach was quite the opposite. It involved mixing together hundreds or thousands of species, most of which are little known, and through self-design a stable system would emerge. This approach is based on a faith in the ecosystem and its long evolutionary history rather than on the conventional engineering method. Several quotes from ecologists on this dichotomy reveal the extent of the disagreement:
When I read of schemes to create living spaces from scratch upon which human lives will be dependent for the air they breathe, for extrinsic protection from pathogens and for biopurification of wastes and food culture, I begin to visualize a titantic-like folly born of an engineering world view. At this point we don't know enough, being totally reliant on knowledge as well as physical subsidies from nature to survive on earth. In space there are no doors to open or neighbouring ecosystems to help correct our mistakes. (Todd, 1977)
... these simple systems are inherently unstable and depend upon a very large investment in power for control. The probability of failure of the two-species (man and microorganism) linkage during a long space voyage due to successional processes or to stress is high. It would seem obvious that these simple ecosystems pose a serious risk to the astronaut, and further work on them, as the basis for a life-support system, should be abandoned. Above all, the narrow engineering approach cannot be applied to bioregeneration. Organisms cannot be "designed" and "tested" like transistors or batteries to perform "one function" or to solve "one problem"; they have evolved with other organisms as a unit and they carry out a variety of functions which must dovetail with other activities of the ecosystem. The "minimum ecosystem" for man must clearly be a multispecies one. (Cooke, 1971)
In appraising the potential costs of closed system designs one has the alternative of paying for a complex ecosystem with self maintenance, respiration, and controls in the form of multiple species as ecological engineering, or in restricting the production to some reduced system like an artificial algal turbidistat and supplying the structure, maintenance, controls, and the rest of the functions as metallic-hardware engineering. Where the natural combinations of circuits and "biohardware" have already been selected for power and miniaturization for million of years probably at thermodynamic limits, it is exceeding questionable that better utilization of energy can be arranged for maintenance and control purposes with bulky, nonreproducing, nonself maintaining robot engineering. (H. T. Odum, 1963)
There is a kind of frustration in these quotes in opposition to conventional engineering. This sense can also be felt by reading the discussions that occurred at the symposia on bioregenerative systems where a single ecologist tried to defend the multispecies approach to a group of engineers and physiologists (Cooke in Cooke et al., 1968 and E. P. Odum in Brown, 1966).
This dialogue provides perspective on the gulf between ecologists and engineers not just in terms of design of life support systems but in more general terms. There are differences between the ecological engineering method and the conventional engineering method, and this text is an effort at bringing the differences to the forefront for consideration. In terms of life support systems design, the engineers and physiologists have temporarily won the funding battle over the ecologists, as can be seen by the heritage of NASA's CELSS program which involves essentially little or no ecology (see, for example, Brechignac and MacElroy, 1997). Before the advent of CELSS, H. T. Odum (1971) called this situation "something of a national fiasco" because NASA had "refused to recognize that multispecies designs are required for stability and that this energy cost is unavoidable." No successful long-term bioregenerative life support system has yet been designed and only time will tell if NASA made the correct choice between alternative life support system designs.
As a side note, NASA has supported a group of ecologists who published a set of papers on the rhetoric of closed systems (Botkin et al., 1979; Maguire et al., 1980; Slobodkin et al., 1980). They suggest that closed systems can be a valuable tool for developing ecological theory. Curiously, however, this group does not cite any of the work mentioned above and does not offer any discussion of the life support system design question. NASA also provided some support to Folsome's microcosm work mentioned earlier. Neither of these token efforts of support for ecological research seems to have had an influence on NASA's approach to life support.
As another side note, it is interesting that the case for the multispecies microcosm was again made here in terms of life support system design as it was made earlier in terms of ecotoxicology testing. In both cases ecologists were pitted against a rival group, and in both cases they lost the argument for funding support.
Against this backdrop of government funded research on life support systems, the Biosphere 2 project started in the desert north of Tucson, AZ. Biosphere 2 is an aircraft carrier-sized (3.0 acres/1.25 ha) mesocosm originally designed and constructed to test human closure in a bioregenerative life support system (Figure 4.23). It was an impressive project from several perspectives. First, it was a privately funded project that was run by a corporation (Space Biosphere Ventures) as a for-profit
enterprise. Second, it was much bigger than any other closed system ever attempted. Third, it utilized to a large extent the ecological engineering approach to life support system design, previously ignored by government sponsored research. The latter point makes Biosphere 2 an important case study for ecological engineering.
H. T. Odum (1971) had anticipated the Biosphere 2 project in discussing a large life support system (Figure 4.24), and he had written several grants in the 1960s for a similar though smaller-scale project that were not funded. Biosphere 2 realized this concept but with huge technological couplings. The project was started in 1984 and it followed a tumultuous series of events (Table 4.5). It still exists today but with a very different focus. The original purposes of the project were "... to develop bioregenerative and ecologically-upgrading technologies; to conduct basic scientific and ecological research; and to educate the public in ecosystem and biospheric
TABLE 4.5 Milestones in the
Was this article helpful?