Qualitative models on stream ecosystem recovery: exploratory studies

Most of water ecosystems are endangered by human actions, in spite of their importance for all living systems. Qualitative models and simulations may be useful for stream ecosystem recovery in many ways: for understanding such systems, to predict values of variables and to combine such understanding with restoration and proactive actions of management. This study describes the development of qualitative simulation models about the effects of pollution by organic matter and its consequences on the amount of dissolved oxygen and mortality of fishes. We present simulations of good and bad management practices, that may result in increasing and reducing fish stocks. Problems we found and solutions we implemented during the modelling effort are discussed, including the explicit representation of assumptions and the role of ambiguities in the outcomes of the models. We argue that new ontologies and qualitative domain theories will be required in order to tackle processes are related to both energy flow and nutrient cycling and to represent complex interactions between physical, chemical and biological phenomena.