Quantitative versus qualitative modeling: a complementary approach in ecosystem study.

Natural disturbance or human perturbation act upon ecosystems by changing some dynamical parameters of one or more species. Foreseeing these modifications is necessary before embarking on an intervention: predictions may help to assess management options and define hypothesis for interventions. Models become valuable tools for studying and making predictions only when they capture types of interactions and their magnitude. Quantitative models are more precise and specific about a system, but require a large effort in model construction. Because of this very often ecological systems remain only partially specified and one possible approach to their description and analysis comes from qualitative modelling. Qualitative models yield predictions as directions of change in species abundance but in complex systems these predictions are often ambiguous, being the result of opposite actions exerted on the same species by way of multiple pathways of interactions. Again, to avoid such ambiguities one needs to know the intensity of all links in the system. One way to make link magnitude explicit in a way that can be used in qualitative analysis is described in this paper and takes advantage of another type of ecosystem representation: ecological flow networks. These flow diagrams contain the structure, the relative position and the connections between the components of a system, and the quantity of matter flowing along every connection. In this paper it is shown how these ecological flow networks can be used to produce a quantitative model similar to the qualitative counterpart. Analyzed through the apparatus of loop analysis this quantitative model yields predictions that are by no means ambiguous, solving in an elegant way the basic problem of qualitative analysis. The approach adopted in this work is still preliminary and we must be careful in its application.