Correlating Enhanced National Wetlands Inventory Data with Wetland Functions for Watershed Assessment
The U.S. Fish and Wildlife Service has been conducting the National Wetlands Inventory for over 25 years. The NWI Program has produced wetland maps for 91% (78% final) of the lower 48 states, all of Hawaii, and 35% of Alaska. Wetlands are classified according to the Service's official wetland classification system (Cowardin et al. 1979). This classification describes
wetlands by ecological system (Marine, Estuarine, Lacustrine, Riverine, and Palustrine), by subsystem (e.g., water depth, exposure to tides), class (vegetative life form or substrate type), subclass, water regimes (hydrology), water chemistry (pH and salinity), and special modifiers (e.g., alterations by humans). The maps have been converted to digital data for 47% of the lower
48 states and 18% of Alaska. The availability of digital data and geographic information system (GIS) technology make it possible to use NWI data for various geospatial analyses. In the 1990s, the NWI Program for the Northeast Region recognized the potential application of NWI data for watershed assessments, but realized that other attributes would have to be added to
the data to facilitate functional analysis. Dr. Mark Brinson had recently developed a hydrogeomorphic (hgm) approach to wetland functional assessment (Brinson 1993a). This approach provided the impetus for developing other attributes to expand the NWI database and make it more useful for functional assessment. In the mid-1990s, a set of hgm-type descriptors were developed to describe a wetland's landscape position, landform, and water flow path (Tiner 1995, 1996a,b). Use of the initial set of keys for pilot watershed projects lead to a refinement and expansion of the keys in subsequent years (Tiner 1997a, 2000, 2002, 2003). These projects were watershed characterizations that included a preliminary assessment of wetland functions as a main component or the prime component of the study. The reports addressed the following watersheds: Casco Bay (Maine; Tiner et al. 1999), Nanticoke River (Maryland and Delaware; Tiner et al. 2000, 2001), Coastal Bays (Maryland; Tiner et al. 2000), and Cannonsville and Neversink Reservoirs (New York; Tiner et al. 2002), as well as the Pennsylvania Coastal Zone (Tiner and DeAlessio 2002).
In conducting these studies, we worked with local and regional wetland experts to develop correlations between wetland characteristics recorded in the database and wetland functions (see Acknowledgments for listing). The correlations reflect our best approximation of what types of wetlands are likely to perform certain functions at significant levels based on the characteristics
we have in the wetland database. Conducting wetland assessments in other areas, especially in arid, semiarid, and tropical regions, may identify other wetlands that need to be added to the significance list for various functions.