Classifying coastal waters: Current necessity and historical perspective

Coastal ecosystems are ecologically and commercially valuable, productive habitats that are experiencing escalating compromises of their structural and functional integrity. The Clean Water Act (USC 1972) requires identification of impaired water bodies and determination of the causes of impairment. Classification simplifies these determinations, because estuaries within a class are more likely to respond similarly to particular stressors. We reviewed existing classification systems for their applicability to grouping coastal marine and Great Lakes water bodies based on their responses to aquatic stressors, including nutrients, toxic substances, suspended sediments, habitat alteration, and combinations of stressors. Classification research historically addressed terrestrial and freshwater habitats rather than coastal habitats. Few efforts focused on stressor response, although many well-researched classification frameworks provide information pertinent to stressor response. Early coastal classifications relied on physical and hydrological properties, including geomorphology, general circulation patterns, and salinity. More recent classifications sort ecosystems into a few broad types and may integrate physical and biological factors. Among current efforts are those designed for conservation of sensitive habitats based on ecological processes that support patterns of biological diversity. Physical factors, including freshwater inflow, residence time, and flushing rates, affect sensitivity to stressors. Biological factors, such as primary production, grazing rates, and mineral cycling, also need to be considered in classification. We evaluate each existing classification system with respect to objectives, defining factors, extent of spatial and temporal applicability, existing sources of data, and relevance to aquatic stressors. We also consider classification methods in a generic sense and discuss their strengths and weaknesses for our purposes. Although few existing classifications are based on responses to stressors, may well-researched paradigms provide important information for improving our capabilities for classification, as an investigative and predictive management tool.