You are here:
MID-ATLANTIC COASTAL STREAMS STUDY: STATISTICAL DESIGN FOR REGIONAL ASSESSMENT AND LANDSCAPE MODEL DEVELOPMENT
Citation:
Pitchford, A M., S. W. Ator, J. M. Denver, A R. Olsen, AND A C. Neale. MID-ATLANTIC COASTAL STREAMS STUDY: STATISTICAL DESIGN FOR REGIONAL ASSESSMENT AND LANDSCAPE MODEL DEVELOPMENT. Presented at Environmental Monitoring and Assessment Program (EMAP) Symposium 2002, Kansas City, MO, May 7-9, 2002.
Impact/Purpose:
The overarching objective is to develop and test landscape indicator statistical models for condition of streams and aquatic biota in relation to pesticides, nutrients, sediments and toxic substances, nationwide. The indicator models will serve as tools for managers who want consistent methods to compare potential impacts on streams within a biophysical region for differing landscape patterns. To accomplish this objective, several sub objectives will be important:
Develop landscape indicator statistical models of stream vulnerability for selected regions of the U.S., beginning with the Mid-Atlantic Coastal Streams; studies in the Midwest, California, and the Southeast will follow.
Demonstrate the application of the landscape indicator models for the ranking of watersheds, the identification of "hot spots," and the evaluation of management options using projected future scenarios of land use for the study areas selected above.
Develop statistical distributions for physical characteristics of small water bodies for use in OPP modeling.
Leverage resources for this research by incorporating existing data into the model development process, and by sharing field study costs with other projects in the same geographic areas.
This task represents a topic area within the Landscape Sciences research program which is described in A National Assessment of Landscape Change and Impacts to Aquatic Resources. A 10-Year Research Strategy for the Landscape Sciences Program, EPA/600/R-00/001. It also supports the Regional Vulnerability Assessment (ReVA) Program.
Besides being responsive to the Office of Pesticide Programs, this research directly supports long-term goals established in ORD's multi-year research plans related to GPRA Goal 2 (Water Quality) and GPRA Goal 8.1.1 (Sound Science/Ecological Research). Relative to the GPRA Goal 2 multi-year plan, this research will "provide tools to assess and diagnose impairment in aquatic systems and the sources of associated stressors" and "provide the tools to restore and protect aquatic ecosystems and to forecast the ecological, economic, and human health outcomes of alternative solutions" (Long Term Research Goals 2 and 3). Relative to the Goal 8 multi-year plan, this research will develop and demonstrate methods to provide states, tribes, and federal, state and local managers with abilities to: (1) assess the condition of waterbodies in a scientifically-defensible and representative way, while allowing for aggregation and assessment of trends at multiple scales; (2) diagnose cause and forecast future condition in a scientifically defensible fashion to more effectively protect and restore valued ecosystems; and (3) assess current and future ecological conditions, probable causes of impairments, and management alternatives.
Description:
A network of stream-sampling sites was developed for the Mid-Atlantic Coastal Plain (New Jersey through North Carolina) as part of collaborative research between the U.S. Environmental Protection Agency and the U.S. Geological Survey. A stratified random sampling with unequal weighting was used to select 175 first-order nontidal streams for synoptic sampling of base-flow water quality and benthic and riparian ecology during early 2000. Site replacement during reconnaissance yielded a final network of 174 streams. Additional streams were selected for sampling to allow comparison of data from this base network to downstream conditions and to assess seasonal and temporal variability. We designed the base network to provide data for two distinct but complementary objectives: a probabilistic assessment of regional conditions, and a gradient study over a range of land cover and hydrogeologic conditions. The base network includes twenty-five sites within each of seven hydrogeologic subregions that were delineated on the basis of physiography and the bulk texture of surficial sediments. Within each subregion, selection probabilities were adjusted to provide an approximately even distribution of network sites along a gradient of forested-to-developed (urban or agricultural) land in the contributing watershed. Alternative sites were selected in the same manner for use if the primary base-network sites were unsuitable for sampling. With these data, we developed landscape indicator models using stepwise regression to estimate in-stream concentrations of herbicides, nitrate + nitrite, and chloride. By applying weights to the data to compensate for land-use bias incorporated in the network design, we estimated concentrations of pesticides and nutrients for the entire population of first-order streams in the study area.