Ecosystem Linearization: An Evolutionary Design Problem

The hypothesis that ecosystems are nominally linear in their large-scale dynamic characteristics, and that linearization is achieved in evolution, is explored in four principal respects. (i) Ecosystems are well behaved due to mechanisms and organizational attributes which promote or reflect superposition: complexity and largeness, homeostasis, hierarchical structure, environmental rather than state causation of many discontinuities, linear frequency response, single free equilibrium, global stability, and single forced steady state. (ii) Ultimate restriction on ecosystem development by abiotic resources suggests linear donor control as a system-level metaorganization superimposed upon detailed, nonlinear mechanisms. (iii) Ecosystem behavior in the large can be viewed, because of the small amount of the physical world which is actually biotic, as local dynamics about the equilibrium, ground, or zero state. (iv) Ecosystems, rich in feedback interconnections, are linearized by feedback elements that may be linear, nonlinear, or stochastic, with increasing effect. Linearization is posed as an evolutionary design problem, with natural selection operating to rid ecosystems of undesirable nonlinear characteristics or nullify their effects. The linearization hypothesis is in no sense definitively framed at this point, but there appear to be enough reasonable considerations to qualify it for treatment by ecologists as a serious ecosystem theory.