This work presents the creation of causal maps derived from individual-based models (IBMs) of a forest to help the comprehension of how a devastated area recovers and in a further moment help ecological management decisions that enable speeding up such recovery.

A causal map is a directed graph that represents influence relations about variables of a given domain according to an expert in that domain, and is commonly used in management practice and decision support systems as a way of understanding a given system or problem.

Often in ecological management, the ecosystem onto one wished to actuate exhibits complex dynamics that is not entirely known and is not amenable to be easily decomposed into a causal map based solely on previous knowledge of the domain experts.

The individual-based modelling approach is advantageous in such situations because it allows incremental understanding of the studied ecosystem by modelling parts of it at a time. It also allows complex properties of the ecosystem to arise in an emergent way in the model, without the need of explicitly quantitying or modelling them.

We use causal bayes networks as a mean for obtaining causal maps.
Causal bayes networks are a formalism that is able to encode cause and effect relations among variables and that probabilistic reasoning about some domain of expertise to be conducted. It is possible to elaborate causal bayes networks relating a set of variables when experimental control about a phenomenon that relates such variables is available.


This work uses an individual-based model of sucession in the recovery of devastated forest areas to be able to obtain a level of experimental control that allows the ellaboration of causal bayes networks and simplifies them into causal maps that can assist management decisions related to the kinds of species reintroduced in those areas and the success and the speed the recovery.