Use of timed automata and model-checking to explore scenarios on ecosystem models

The interest to build ecosystem models is well acknowledged in order to improve the understanding of the sophisticated linkages between humans and natural species embedded within variable local and global environmental contexts. It is especially true when a complex temporal evolution intervenes as in population regulations. Ecological modellers usually resort to numerical models supported by accurate data and extensive knowledge on biological processes. Unfortunately, the task becomes more difficult to model ecosystems with limited data and knowledge. Qualitative models may be more suitable for designing data-poor systems in a decision-aid context. We propose a new qualitative approach for ecosystem modelling based on timed automata (TA) formalism combined with a high-level query language for exploring scenarios. TA rely on a discrete-event system formalism to reproduce the temporal dynamics of a system. Combined with model-checking techniques, TA enable the exploration of system properties in response to a wide range of scenarios based on a temporal logic. Our applicative case concerns the evolution of different fish biomass along time according to fishing policies, especially when exogenous environmental issues may also be considered. We have developed this approach to model a simplified marine ecosystem subject to different fishing policies. Using predefined query patterns, we show that TA and model-checking are relevant tools to query timed properties of a fishery system in response to different management options. This modelling approach may be especially useful for fostering better discussion among all stakeholders involved in fisheries management.

► We propose a new qualitative ecosystem modelling based on timed automata.
► Combined with model-checking, we can explore system properties and scenarios.
► We apply this model to a simplified fisheries system with various fishing efforts.
► We think this approach may help stakeholders for better ecosystem management.