The problems posed by adaptive management for improved ecosystem health are
reviewed. Other kinds of science-informed ecosystem management are needed
for those regions of conflict between rapid human population growth, increa
sed resource extraction. and the rising demand for better environmental ame
nities, where large-scale experiments are not feasible. One new framework i
s threshold-based resource management. Threshold-based resource management
guides management choices among four major science and engineering approach
es to achieve healthier ecosystems: self-sustaining ecosystem management, a
daptive management, case-by-case resource management, and high-reliability
management. As resource conflicts increase over a landscape (i.e., as the e
cosystems in the landscape move through different thresholds), management o
ptions change for the environmental decision-maker in terms of what can and
cannot be attained by way of ecosystem health. The major policy and manage
ment implication of the framework is that the exclusive use or recommendati
on of any one management regime, be it seif-sustaining, adaptive, case-by-c
ase, or high-reliability management, across all categories of ecosystems wi
thin a heterogeneous landscape that is variably populated and extractively
used is not only inappropriate, it is fatal to the goals of improved ecosys
tem health. The article concludes with detailed proposals for environmental
decision-makers to undertake "bandwidth management" in ways that blend the
best of adaptive management and high-reliability management for improved e
cosystem health while at the same time maintaining highly reliable flows of
ecosystem services, such as water.