technologies in local climate modification; for example, planting trees may save 200 billion kilowatt-hours annually in the United States by reducing the need for air conditioning (Committee on Science, Engineering, and Public Policy, 1992). Avise (1994) calculated the cost of Biosphere 2, which partially regulated the life-support systems for eight humans over a two-year period with an electricity subsidy from outside the sphere. The cost was $150 million (U.S.), or $9 million per person per year. The complications of calculating per capita ecosystem services at vast global or continental scales were definitely not present in this mesocosm experiment. Nevertheless, there is little doubt that practices tolerated in Biosphere 1 (the earth), such as human population growth, overexploitation of ecological capital, and massive destruction of habitats and species, could not be tolerated in Biosphere 2 even for a relatively short time (Avise, 1994). Clearly, some essential functions of natural ecosystems would be difficult or expensive to replace through technological systems.
Ecosystem health is a complex concept, but a consensus definition has been derived (Haskell et al., 1992; see also Karr, in this volume):
An ecological system is healthy and free from ''distress syndrome" if it is stable and sustainable—that is, if it is active and maintains its organization and autonomy over time and is resilient to stress.
It seems reasonable that a close correlation should exist between health and performance at any level of biological organization. Healthy plants capture more solar energy. Growing forests store more carbon. Indeed, one definition of stress demands that captured energy be diverted from growth into coping mechanisms (Calow, 1991). However, the relationship between ecosystem health and ecosystem services is not as well defined.
A number of interrelated hypotheses on the relationship between ecosystem health and ecosystem services deserve serious attention (Cairns and Pratt, 1995).