I contend that focusing on the ''absence of health,'' or what I will c
all ''extreme cases,'' provides a basis for agreement about ecosystem
health in ecosystem management. An extreme case is defined where there
exists (1) scientific agreement about models, (2) scientific agreemen
t about scale, (3) policy agreement that the risks will be great and u
nnecessarily so. scientific agreement requires that a link be establis
hed between biodiversity and ecosystem processes, goods, and services.
I discuss recent work concerning ecosystem engineers that demonstrate
s such a link. Agreement about scale requires consensus about the scal
e of management decisions, both spatial and temporal,;md reconciliatio
n Of management goals and objectives of the various stakeholders in an
ecosystem. Finally, policy agreement requires that ecological risks b
e great and unnecessarily so Where these extreme cases do not apply, I
suggest that it will be useful to appeal to stakeholder negotiations
to seek agreement concerning ecosystem health.