Design of multi-paradigm integrating modelling tools for ecological research

Integrating modelling tools allow different modelling paradigms to coexist and cooperate in the same simulation model. The need for such tools in ecological modelling is due to the high level of complexity of ecological and environmental decision-making problems, their multiple scales of description, the diversity of the available approaches, and the size and heterogeneity of the available datasets. This article discusses problems and perspectives in developing integrating modelling tools and introduces the Simulation Network Interface (SNI), a software package for easy coordination of different existing simulation models. The interface allows the coordination of independent simulation models residing on different machines into higher-level, multi-paradigm, distributed simulation, with minimal recoding efforts of existing models. The interface can also be used to easily provide a remote interface to simulation or data retrieval services running on different architectures. As examples of its application, we describe three ongoing projects using the SNI: (1) the integration of Swarm, an agent-based simulation toolkit, with the Spatial Modelling Environment (SME), a process-based spatial simulation toolkit; (2) the straightforward implementation of a GIS-based spatial data repository for network-based data retrieval and manipulation; and (3) a network-based calibration service for complex simulation models.