A non-value based framework for assessing ecosystem integrity

Indicators of ecosystem state change are currently selected based on the hu man value for some resource (e.g., biodiversity, water quality, etc.) or th e human desired end-point of a system (e.g., existence of late successional forests, lack of insect or pathogen outbreaks, etc.). Examples are used fr om ecosystem studies and forest certification protocols to highlight the pr oblems of selecting indicators based on our derived values. Selection of va lue-based indicators may misidentify the important constraints controlling the functioning of a system and may result in poor predictions of system su stainability. Several examples are presented on how to develop a non-value based framework of analysis to select parameters capable of predicting how a system functions, or to identify the dominant constraints for each site. The value of using large data sets that can be analyzed to select non-human value-based constraining variables at different scales is discussed. This process of selecting the dominant constraints also highlights why the same suite of indicators cannot be used automatically for all sites.