NOAA HABITAT RECOVERY MODELS USING CELLULAR AUTOMATA TECHNIQUES1

Abstract

ABSTRACT Models that predict characteristics of habitat recovery following disturbance are useful tools for ecological risk assessment as well as natural resource damage assessment. NOAA is evaluating the suitability of cellular automata as a single modeling framework for simulating several habitat types (seagrass, coral, rocky intertidal and marsh). The primary challenge is creating rules that reflect biological relationships and complexity. The desire for biological realism in the model is balanced by the need for simplicity in the defined relationships. Once the rules are defined, cellular automata models are straightforward to construct and modify during development. The cellular automata technique divides a domain into discrete cells with a finite number of states. For habitat models dominated or characterized by a single species (e.g. seagrass, coral or marsh), the cells are defined with two states: empty and occupied. For the multiple species models (e.g. rocky intertidal), the cells are either empty (bare rock) or occupied by one of four biological covers: Fucus, barnacles, mussels, or “other” (e.g., coralline algae or red algal turf). Probabilistic rules are generated for cell transitions from one state (or dominant species) to another. In cases where there is not spatial influence, the model can be collapsed into a simple spreadsheet calculation. Though some ecosystems recover in timescales shorter than decades, some of these ecosystems recover over time periods longer than a human lifespan. Thus simulation modeling is the only feasible means to provide reasonable guidance for their final recovery. Within this modeling system, the user can create an initial distribution of species, and then choose either to run the simulation from the initial beach, or to selectively empty cells to create a specific type of disturbance before running the simulation. The goal for this work is to both intuitively (graphically) (1) display tradeoffs from the use of different cleanup and restoration techniques and (2) to provide a defensible and quantitative recovery forecast for computing monetary damages against responsible parties.