Modeling water resources: have we got it right?

Aquatic scientists generally recognize that controlled experiments are required to establish cause-effect relationships (e.g., Havens and Aumen, 2000), and understanding ecological processes is key to accurately predicting complex ecosystem responses. However, resource managers may have at their disposal only a limited amount of observational data when faced with management decisions. Hence, there may be a tendency to use simple empirical models for decision making. An example of eutrophication management in lakes illustrates a pitfall of this approach when used independently of other scientific information.