Worldwide, local-scale anthropogenic stress combined with global climate change is driving shifts in the state of reef benthic communities from coral-rich to micro- or macroalgal-dominated (Knowlton, 1992; Done, 1999). Such phase shifts in reef benthic communities may be either abrupt or gradual, and case studies from diverse ocean basins demonstrate that recovery, while uncertain (Hughes, 1994), typically involves progression through successional stages (Done, 1992). These transitions in benthic community structure involve changes in community metabolism, and accordingly, the holistic evaluation of associated biogeochemical variables is of great intrinsic value (Done, 1992).

Effective reef management requires advance prediction of coral reef alteration in the face of anthropogenic stress and change in the global environment (Hatcher, 1997a). In practice, this goal requires techniques that can rapidly discern, at an early stage, sublethal effects that may cause long-term increases in mortality (brown, 1988; Grigg and Dollar, 1990). Such methods would improve our understanding of the differences in the population, community, and ecosystem structure, as well as function, between pristine and degraded reefs. This knowledge base could then support scientifically based management strategies (Done, 1992).

Brown (1988) noted the general lack of rigor in the assessment of stress on coral reefs and suggested that more quantitative approaches than currently exist are needed to allow objective understanding of coral reef dynamics. Sensitive techniques for the timely appraisal of pollution effects or generalized endemic stress in coral reefs are sorely lacking (Grigg and Dollar, 1990; Wilkinsin, 1992). Moreover, monitoring methods based on population inventories, sclerochronology, or reproductive biology tend to myopic and may give inconsistent results. Ideally, an improved means of evaluating reef stress would discriminate mortality due to natural causes from morality to anthropogenic causes (Brown, 1988).

Models of coral reef ecosystems, parameterized by process measurements and scaled in time-space using remote sensing, have the potential to address pressing research questions that are central to devising valid management strategies (Grigg el al., 1984; Hatcher, 1997b). To attain this goal, ecosystem-level models that integrate studies of physical and chemical forcing with observed biological and geological responses are required. This interdisciplinary approach to understanding reef biogeochemical dynamics can allow investigations that integrate the scales of time and space (Hatcher, 1997a), thereby enabling prediction of coral reef change (Andréfouët and Payri, 2001). In turn, prediction of holistic ecosystem function within various environmental focusing scenarios has substantial promise in mitigating future disturbance. Indeed, management of coral reefs at the ecosystem level has been suggested as the only meaningful approach to preserving coral reefs (Bohnsack and Ault, 1996; Christensen et al., 1996).