Background/Question/Methods    The purpose of this session is to elucidate how new perspectives on ecosystem response to disturbance, primarily in terms of resiliency, are related to biogeochemical properties of the system.  My objective is to contribute some philosophical perspectives, definitions of terms, and operational suggestions for addressing this goal.
     Biogeochemistry is a large, interdisciplinary field demanding definition for this specific purpose.  What aspects of biogeochemistry are significant in this context?  Ecosystems themselves are abstractions requiring characterization, in this case, both for biogeochemical properties and for responses to disturbances.  Major kinds of ecosystems (aquatic, marine, and terrestrial) are fundamentally different in both respects.  At the same time, disturbance has many dimensions; some intrinsic, some extrinsic. What are these dimensions and how are they correlated?  In particular, what meaning of resilience should be adopted and should it be the primary property of interest?  
     This session seeks to link largely abstract entities and phenomena, specifications of which are highly contingent on local system circumstances. This kind of goal and these realities are characteristic of much of ecology.  What conceptualizing approaches may be employed for fruitfully linking all of these abstractions in meaningful terms?

Results/ConclusionsResults/Conclusions       Ecosystem properties relevant for biogeochemical processes (e.g. element choice or pool size) only partially overlap with properties relevant for characterizing resilience (e.g. life histories of constituent species).  Which properties overlap and under what circumstances?  Further, the importance of properties varies across ecosystem types, e.g. terrestrial ecosystems vastly differ from streams in the relative importance of input and output fluxes with respect to internal cycling. 
     Disturbances themselves vary in kind (e.g. epidemics versus fires), temporal frequency and novelty (occurrence relative to lifetimes of affected species), intensity (degree of affliction), and scale (spatial extent). In turn, different metrics (e.g. phytomass, dead organic matter, trophic level composition, food chain relationships) are appropriate for estimating resilience for different ecosystems and disturbances.
    Generalizations are intellectually interesting and heuristically valuable, but in ecology they are typically incapable of predicting outcomes for real cases.  Because ecological systems are enormously complex, ecologists must concede that specifics matter. To get beyond mere collections of case studies, we must organize experience into predictive frameworks that take contingencies into account. Conflating the challenging topics addressed in this session will require extensive literature review, broad involvement of diverse researchers, and deployment of organizational devices such as frameworks, perhaps reorganized into diagnostic systems.