The Data

Wetlands, Lakes, Reservoirs, and Ponds
Data Source: Data for wetlands are from Dahl (2000). Data used here are from Figure 35 (p. 56) and Figure 42 A–C (p. 62). Data for lakes, reservoirs, and ponds come from Dahl (2000) and also from Frayer et al. (1983); Dahl and Johnson (1991); and unpublished data from the U.S. Fish and Wildlife Service.

Data Collection Methodology/Definitions: The data shown here are derived from the U.S. Fish and Wildlife Service’s National Wetlands Inventory (NWI), which produces periodic reports of changes in wetland area. For this report, decadal estimates are presented as the midpoint of the decade. For example, “1980s” data are presented as “1985.” The historic estimate for 1780 is based on the estimate of 221 million acres of coastal and freshwater wetlands at that time (see Dahl 1990) minus an estimate of 10 million acres of coastal wetlands in 1922, which should approximate the historical area of coastal wetlands because most of these were converted to other land cover types after World War II (see Gosselink and Baumann 1980). Estimates of wetland extent in the 1780s are based on colonial or state historical records plus land use records, drainage statistics, and information on the extent of hydric soils (i.e., drained and undrained).

NWI counts all wetlands, lakes, reservoirs, and ponds, regardless of land ownership, but recognizes only wetlands that are at least 3 acres, and ponds that are at least 1 acre. A permanent study design is used, based initially on stratification of the 48 conterminous states by state boundaries and 35 physiographic subdivisions. Within these subdivisions are 4375 randomly selected 4-mi2 (2,560-acre) sample plots. These plots were examined with the use of aerial imagery of varying scale and type; most images were 1:40,000-scale, color infrared, from the National Aerial Photography Program.

The wetland types selected for reporting here were recommended as the most relevant and most reliable for long-term reporting by the NWI (see Dahl 2000, p. 62). For wetlands, they include forested, shrub, and emergent wetlands. Ponds include the category of open-water ponds and non-vegetated palustrine wetlands (i.e., palustrine unconsolidated shore, which are mud flats and the shorelines of ponds); ponds are generally less than 6 feet (2 m) deep and less than 20 acres in size. Lakes and reservoirs are generally larger than 20 acres and deeper than 6 feet, although smaller bodies are included if they are deeper than 6 feet or have a wave-formed or bedrock shoreline.

Data Quality/Caveats: Field verification was conducted to address questions of image interpretation, land use coding, and attribution of wetland gains or losses, and plot delineations were completed. For example, for the 1980s-to-1990s analysis, 21% of the sample plots were verified.

Ephemeral wetlands and effectively drained palustrine wetlands observed in farm production are not recognized as a wetland type and are not included. Wetlands that are farmed during dry years but that normally support hydrophytic vegetation were classified as freshwater emergent wetlands.

The U.S. Geological Survey’s (USGS) National Hydrography Dataset (NHD) also has information on lake, reservoir, and pond area (at least 6 acres in size). Considerably higher total acreage (26.8 million acres) is found using this resource. NWI was used because time trends are possible; the cause of the disparity between datasets is not known.

Data Access: The Status and Trend of Wetlands in the Conterminous United States 1986 to 1997 is available on the Web at http://wetlands.fws.gov/bha/SandT/SandTReport.html.

Riparian Areas Note: This indicator uses a distance of roughly 100 feet from the edge of a stream to define its “riparian” area. This is based on the availability of remote-sensing data, as described below. We are cognizant that the definition of riparian areas is a complex one, and that no single value for the width of this feature will be appropriate in all situations.

Data Source: Data reported here for the classification of riparian areas along streams and rivers were provided by the U.S. Environmental Protection Agency’s National Exposure Research Laboratory, Environmental Sciences Division, and are based on the NHD. The NHD is a comprehensive set of digital spatial data that encodes information about naturally occurring and constructed bodies of water (see http://nhd.usgs.gov/). The NHD was developed based on EPA’s River Reach File 3 (RF3), which itself was based on digitization of streams from USGS topographic quadrangle maps. The dataset does not provide information on very small streams, and the lower limit of stream size that is reported in the database is unclear. Data on the vegetation cover within 100 feet of streams and rivers were produced by EPA from remote-sensing imagery and the NHD. The remote-sensing imagery is from the National Land Cover Dataset (NLCD; see the technical note for the national extent indicator for further details.

Data Manipulation: For this study, EPA combined these datasets to identify the land cover along streams and rivers (and the shores of ponds, lakes, and reservoirs—see the altered freshwater ecosystems indicator). For each stream reach described in the NHD, land cover was characterized, using the NLCD, in a band approximately 100 feet wide on either side of the stream. NLCD land cover classes were aggregated to produce four general categories (forested; agricultural; urban; and grasslands, shrublands, and woody and emergent wetlands). In one instance, the text describes this latter category as “other natural vegetation,” despite the fact that some of these land cover types may not be the historical (i.e., natural) vegetation for that site, or may have been altered in other ways. This terminology is used to highlight the contrast with the highly altered land covers (urban, agricultural). Estimates of the riparian area in each of these different land cover classifications were derived by overlaying stream reaches and land cover. Data Caveats/Limitations: The NLCD and the NHD are currently the most comprehensive datasets available for land cover and freshwater resources, respectively. However, both of these contain inaccuracies that could affect the calculations presented here. The NLCD is known to contain approximately 20% error in land cover classification; some of the known misclassifications that occur randomly in the dataset include suburban areas or tree farms classified as forest; grasslands classified as agriculture, or vice versa; and fallow agricultural fields classified as barren lands. The NHD is a relatively new dataset and is known to contain numerous errors and inconsistencies. Strahler first- and second-order streams (a method for ranking stream order, which is related to size) are poorly represented in the NHD as well as in the RF3 that serve as the base data. It appears that dry lake beds in the west may have occasionally been included as lakes in the NHD. Additionally, the architecture of the NHD results in some lakes being represented by numerous polygons with different identifications, thus being counted as separate lakes in this analysis. Numerous inconsistencies exist in the NHD attribute data. The designation of stream segments as perennial or intermittent is particularly problematic; in at least one case, this designation can be shown to follow USGS topographic quadrangle boundaries. In addition, many errors can be found in the attribution of ponds, lakes, and reservoirs. Although these inconsistencies were noted, it was not possible given the scope and scale of this analysis to provide across-the-board corrections, nor was it possible to coregister the datasets for all locations. Therefore, the most current versions of both datasets were used as is.

The Data Gap

Information on the number of small, medium, and large streams and rivers is not available. In general, the number of stream miles can be derived from sources such as the NHD; however, there is no universally accepted approach for categorizing streams and rivers based on size (i.e., small, medium, and large). Potential approaches include basing categories on flow rate, drainage area size, or stream order. USGS will soon incorporate a tool within the NHD dataset to allow determination of stream order, which can be determined from maps. Flow rate is a much more difficult parameter to determine.

In addition, there is concern that use of the NHD may understate the extent of small streams. Since the NHD is based upon historic mapping conducted for the USGS, there may be inconsistencies in the degree to which small streams were mapped. Since the rate of conversion and alteration of small streams is believed to be higher than for larger streams, it is important to ensure as great a coverage of small streams as is feasible.

For a discussion of the effects of human activities on small streams, see Meyer and Wallace (2001).

References

Dahl, T.E. 1990. Wetland losses in the United States 1780’s to 1980’s. Washington, DC: U.S. Department of the Interior, Fish and Wildlife Service.

Dahl, T.E. 2000. Status and trends of wetlands in the conterminous United States 1986 to 1997. Washington, DC: U.S. Department of the Interior, Fish and Wildlife Service.

Dahl, T.E., and C.E. Johnson. 1991. Status and trends of wetlands in the conterminous United States, mid-1970’s to mid- 1980’s. Washington, DC: U.S. Department of the Interior, Fish and Wildlife Service.

Frayer, W.E., T.J. Monahan, D.C. Bowden, and F.A. Graybill. 1983. Status and trends of wetlands and deepwater habitats in the conterminous United States, 1950’s to 1970’s. Ft. Collins, CO: Dept. of Forest and Wood Sciences, Colorado State University.

Gosselink, J.G., and R.H. Baumann. 1980. Wetland inventories: Wetland loss along the United States Coast. Zoological Geomorphology NF Supplement 34:173-187. Meyer, J.L., and Wallace, J.B. 2001. Lost linkage and lotic ecology: Rediscovering small streams. In M.C. Press, N.J.

Huntley, and S. Levin (eds), Ecology: Achievement and challenge (41st Symposium of the British Ecological Society). Blackwell Science.