Simulation, Multivariate Statistics, and the Degradation of Coastal and Estaurine Ecosystems

Methods are presented by which ecosystem modeling and multivariate statistics are blended to provide coherent and useful assessments of ecosystem degradation. Certain concepts of systems ecology are directly applicable to monitoring coastal and estuarine systems. These are related to ecological hypothoses dealing with resistance to environmental perturbations. The hypotheses lead to definitions of measurable variables. Multivariate tests of hypothesis and pattern recognition techniques can be used to validate model performance and to detect disturbances. The overall approach takes advantage of inherent properties of certain types of ecosystems and can be used as a basis of monitoring programs.