International Encyclopedia of Systems and Cybernetics

A community of interconnected and interacting systems, depending of specific energy inputs and evolving as a whole.

The original definition as introduced by the british plant ecologist A.G. TANSLEY, was: "A community of organisms and their physical environment interacting as an ecological unit"(as quoted by F. CAPRA, 1997, p. 33)

R. MARGALEF describes the ecosystem most generally "… as the material path followed by energy that generally enters in the form of photons and goes out into an energy sink" (1968, p.14). In the mean-time, energy flow may construct less or more complex ecosystems, with however a tendency to maximize complexity. The more organized parts tend to have a selective advantage on the less organized ones, due to their better possibility to use and increase their information. According to MARGALEF: "This is an extraordinaryly valuable principle not only in ecology but also in the general theory of systems" (p.28). See "St. Matthew Principle".

H. ODUM, in turn, describes the ecosystem as: "… a network of food and mineral flows in which the major pathways are populations of animals, plants and microorganisms, each specialized to live in a different way, doing a different job for the energy flows of the system" (1971, p.60).

He adds: "… the purpose of having many species is… the increased system effectiveness possible with specialization, division of labor, and more kinds of controls and regulatory circuits"(Ibid). A. LOTKA "world engine" model is a closely related formulation of the same global concept (1956, p.334).

ODUM's effectiveness is somehow related to ecosystems tendency toward global stabilized complexity (climax) at the highest level of energy flows use. Such a closely limited dynamic equilibrium means that they are stabilized by countervailing feedbacks. However stabilized ecosystems seem to lose resilience, unless they are submitted to periodical environmental disturbances. Their very complexity helps them and simultaneously hinders them in this process. S. REICE contends that: "… equilibrium is an unusual state for natural ecosystems", and that "… the normal state of communities and ecosystems is to be recovering from the last disturbance" (1994, p427).

This is coherent with St. KAUFFMAN's hypothesis according to which "Ecosystems co-evolve to the edge of chaos" (1993, p.261).

Living systems are particularly significant in many ecosystems, because they modified and go on modifying the interactions between the lithosphere, the hydrosphere and the atmosphere.

The planet is our global ecosystem (see Gaia hypothesis) and it is basically fuelled by the sun's energy, which has been and still is the prime mover of its evolution.

Ecosystems are in some sense hierarchized (but not truly hierarchical, as there are no specific regulators or controllers, with the dubious exception of man himself, a disturber more than a regulator), i.e. simultaneously organized from more global to more local and specific ones "top down", but also conversely from the base levels to the upper ones. This explains that a vast ecosystem can be deeply transformed by the intrusion or elimination of some apparently unimportant class of elements.

Some specific smaller ecosystems can be more or less isolated within more global ones, at different levels, as for example: the "tropical rain forest", some specific tropical rain forest, some small specific patch of the same, or even, part of the canopy corresponding to this patch or the bark of one specific tree. Of course, the environment should never be to tally forgotten.

G. BATESON, as well as many other authors, emphasizes that: "We are rapidly, of course, destroying all the natural systems in the world, the balanced natural systems. We simply make them unbalanced – but still natural" (1973, p.406).

We should never forget that any change of one part necessarily implies and triggers changes – frequently unpredictable – in the other components.

The introduction of chaotic models seems to be an urgent need in ecosystemics.