A FUZZY-LOGIC MODEL TO PREDICT CORAL-REEF DEVELOPMENT UNDER NUTRIENT AND SEDIMENT STRESS

Coral reefs are highly complex systems characterized by mostly nonline ar relationships between biotic and abiotic components. Traditional mo dels of reef dynamics often require unavailable data and precision, wh ich limits their success and usefulness. We tested a new approach in c oral reef modeling with fuzzy logic. Fuzzy logic has been applied succ essfully in modeling highly nonlinear systems in engineering, decision support systems, and ecology. As part of an integrated coastal zone m anagement model, we constructed a coral reef model that predicts chang es in coral cover and diversity under anthropogenic stress, namely nut rient enrichment and increased sedimentation. The model reflects our c urrent knowledge of the fringing reefs of Curacao, Netherlands Antille s. The seven input variables used were dissolved inorganic nitrogen an d phosphate, suspended particulate matter, maximum colony size, substr atum available for colonization, coral cover, and coral diversity. Eac h variable was divided into three triangular fuzzy sets reflecting low , medium, and high values. For each of the 2187 possible input combina tions we estimated cover and diversity after 10 years. We consulted ex perts with a thorough knowledge of the local reef system and have auto matically accounted for interactions between the variables described a bove. The model clearly shows how increases in nutrient and sediment i nputs affect coral cover and diversity. Although the model can be refi ned continuously, it appears to reflect accurately the current knowled ge of reef dynamics, making a beneficial contribution to education, ma nagement, and science.