Systems Ecology @
MIT & SFI
Understanding the ecological limits of complex living systems
Our work has been centered on understanding the emergent context-dependent behavior of complex living systems (biotic communities and the nonliving environment) using formal procedures of systems thinking, synthesis, and mathematical modeling. Our work has been rooted on the notion of structural stability (the capacity of a system to display a particular behavior despite small perturbations to its dynamics) and uses tools as varied as: feasibility analysis, population dynamics, statistical mechanics, information theory, metabolic scaling theory, homology, matrix theory, network theory, geometry, causal inference, and empirical dynamic modeling. We have formalized and corroborated the feasibility principle in ecology explaining the limits in the diversity of biotic communities in nature (see below). We are now looking to understand the feasible bioenergetic limits shaping the development, evolvability, and adaptability of complex living systems under changing environments—the limits of life itself.
In the 1930's, G. F. Gause corroborated the competitive exclusion principle showing the impossible systems to be realized:
(i) Two species competing for the same limited resource cannot coexist.
(ii) Natural selection will favor the species having the larger growth rate.
In the 1970's, R. M. May formalized the stability principle showing the possible systems but cannot be realized:
(i) Living systems have a complexity threshold beyond which they will lose their capacity to return to their current state (stability)
(ii) Stable living systems will tend to be either small or characterized by weak interactions
Over the last 10 years, we have formalized and corroborated the feasibility principle showing the possibilities for a system to be realized:
(i) Every possible biological solution (e.g., a community of interacting populations) is feasible for a given collection of environmental challenges.
(ii) Self-organization leads to solutions that are feasible for the largest set of local environmental challenges.
[Notice that this is different from convergent evolution, which postulates that for every environmental challenge there is a finite number of feasible solutions (the possibility space of solutions).]