Ecological models are used to formalize concepts of ecological processes and as such explore existing dynamics and characteristics. Ecological models can also be predictive or used to compare proposed management plans and explore their effects on other processes. Modeling tools are especially useful in communicating complex processes, spatial patterns, and data in a visual format that is clear and compelling and, when appropriately applied, can empower stakeholders to move forward with concerted efforts to address an ecological problem. It has been recognized that during the modeling process the modeler usually gains much understanding about the system workings, about what is most essential, and what controls the system behavior. It is this rich and exciting experience that comes from the modeling process that led to the idea of designing the whole decision-making process around the modeling process. The modeling process itself becomes the decision-making tool, and the decision making becomes part of the modeling process.
Both monitoring and modeling are scientific tools that can support good decision making in ecosystem-based management and are often most powerful when used together. Monitoring data collected at varying scales can be used as inputs to models, to calibrate and validate the accuracy of a model, or to address specific research questions using statistical models. Development of ecological models often indicates the types of information which are important to understand dynamics but for which no data are available. Whereas selective monitoring can give a good description of patterns and linkages within a system, it may be more difficult and expensive to determine the driving forces of these patterns. Simulation models help determine the mechanisms and underlying driving forces of patterns otherwise described statistically. In many cases the monitoring efforts that go along with modeling can serve as a good vehicle to engage the local stakeholders in the process. When stakeholders see how samples are taken, or, ideally, take part in some of the monitoring programs, they create bonding with the researchers and become better partners in the future decision-support efforts.
The modeling of physical, biological, and socioeconomic dynamics in an ecosystem requires attention to both temporal dynamics and spatial relationships. There are many modeling tools that focus on one or the other. To be useful in a participatory framework models need to be transparent and flexible enough to change in response to the needs of the group. Simulation (process) models may be formalized in software such as Stella (isee™ systems), Simile (Simulistics Ltd.), or Madonna (Berkeley Madonna) as a system of stocks, flows, and converters that are connected to form a series of partial differential equations. These and other software packages have user-friendly graphic user interfaces (GUIs), which make them especially helpful when models are demonstrated to stakeholders or when they are formulated in their presence and with their input. In this context, programming directly in C++ or other languages may be less effective. Additional efforts are essential to build interfaces or wrappers that would allow these models to be presented to the stakeholders, or embedded into other models (modularity). In general, process models may be very helpful to explain and understand the systems to be analyzed; however, they are not practical for exploring the role of the spatial structure of an ecosystem. Alternatively, Geographic Information Systems (GIS) explicitly model the spatial connectivity and landscape patterns present in a watershed, but are weak in their ability to simulate a system's behavior over time. Ecosystem-based management demands the coupling of these approaches such that spatial relationships, linkages, and temporal dynamics can be captured simultaneously. There are many specific models developed to analyze the spatiotemporal dynamics of specific systems or processes. So far there are not many generic tools that combine temporal and spatial modeling. One is the Spatial Modeling Environment (SME). There are also modules programmed as components of a GIS, say using the scripting language or Avenue in ArcINFO. The latest implementations of Simile promise some powerful tools to integrate with a GIS; however, most of these are yet to be tested and debugged.
Agent-based models are yet another modeling technique that is useful in participatory workshops. They offer some powerful techniques to engage the stakeholders in a dialog, with some role-playing games leading to more clearly defined rules of behavior for agents. Again for the participatory context, a GUI is essential. NetLogo or StarLogo are the two modeling frameworks that are very promising.
Was this article helpful?