Watershed-Estuary-Species Nutrient Susceptibility

Objectives

A major objective of the project is the refinement of three existing watershed source-transport models and use of these models to assess the environmental and economic effects of changes in land-use, nutrient management, and climate on nutrient flux to estuaries. Comparisons of the flux predictions from the three models will provide information about the accuracy of model specifications of sources and transport processes and model predictions. The USGS has responsibility for the refinement and application of one of the models, namely the SPARROW (SPAtially Referenced Regressions On Watershed attributes; Smith et al., 1997) watershed model. The model applications will include the development of detailed watershed and stream monitoring data bases for the Atlantic and Gulf watersheds at a temporal resolution needed to assess the effects of climate change and land use on nutrient delivery to selected estuaries in these regions. The watershed models will be used in an economic analysis of policy and management scenarios that relate potential load targets (i.e., sources and inland watersheds) to their costs. Moreover, the model predictions of riverine flux will serve as input to trophic-response models of estuarine conditions (developed as part of the broader project goals) to support evaluations of the ecological impacts of changing land use and climate.

Statement of Problem

The USGS research objectives described in this Agreement are part of a larger collaborative project that has the objective of providing an improved predictive understanding of the potential causes and consequences of nutrient pollution on estuarine ecosystems, including the effects of estuarine hydrology and morphology on the upper trophic level responses in estuaries. The project researchers will develop a predictive framework that links models of watershed nutrient sources and riverine transport with models that describe the trophic functioning of coastal estuarine ecosystems. Collaborators include D. Scavia (Project Principle Investigator-U. of Michigan), R. Howarth (Cornell U.), D. Breitburg (Smithsonian Environmental Research Center), and G. Helfand (U. of Michigan). The project was recently awarded $2.5 million ($500,000 per year) from the NOAA Coastal Hypoxia Research Program for research extending over five years. The USGS will receive 8% of the total funding, amounting to $200K over five years. See the attached project proposal for details of the project objectives, methods, and timelines.

The project addresses information needs highlighted in a recent National Research Council report (NRC 2000), including the identification of coastal ecosystems that are most sensitive to nutrient enrichment, improved policies and technologies for reducing nutrient loads from land, and the quantification of nutrient-pollution effects in estuaries. The NRC report underscored the importance of these needs in view of the recent increases in nutrient inputs to coastal waters that have led to substantial changes to U.S. coastal ecosystems (NRC 2000; Howarth et al. 2000) with an estimated degradation of two-thirds of these systems (Bricker et al. 1999). Moreover, projected population growth, land development, and agricultural intensification suggest that nitrogen loadings to estuaries are likely to continue to increase (NRC 2000; Howarth et al. 2002).

Strategy and Approach

The USGS SPARROW water-quality model applied in this project uses a hybrid statistical and process-based approach to estimate the sources and transport of nutrients and suspended sediment in watersheds and surface waters. SPARROW nutrient models have been previously used to characterize loadings and sources to major U.S. estuaries (Bricker et al. 1999) and in a more detailed study of 40 major estuarine drainages (Alexander et al. 2001). Regional SPARROW nutrient models have also been developed for the Chesapeake Bay (Preston and Brakebill, 1999); major drainages to the Pamlico Sound (McMahon et al. 2003), and New England coastal waters (Moore et al. 2004). Additionally, Federal and State managers are using SPARROW to assess the sources of nutrient loadings for targeting reduction strategies for the Chesapeake Bay and the Connecticut River Basin (Moore et al. 2004).

A unique feature of the model is its spatially detailed infrastructure, which describes hydrologic flow paths and contaminant-loss processes separately in terrestrial and aquatic ecosystems. Land and water components in the model provide estimates of the rates of pollutant delivery from point and diffuse sources to stream reaches and downstream waters, including reservoirs and estuaries. All model parameters are determined statistically using nonlinear parameter estimation techniques and mass-balance constraints on model inputs (sources) and outputs (riverine nutrient export). Parameter estimation methods ensure that the calibrated model will not be more complex than can be supported by the data. The estimation techniques also have the advantage of providing measures of uncertainty in both model coefficients and predictions of contaminant loads. The SPARROW model has been shown to appreciably improve the accuracy and interpretability of model parameters and the predictions of pollutant loadings and sources in streams and rivers in comparison to those estimated using conventional statistical watershed modeling approaches (Smith et al. 1997; Alexander et al. 2000; Alexander et al. 2002).

In this project, new SPARROW models will be developed that can be used to describe inter-annual and seasonal changes in nutrient and suspended sediment loads in response to changes in precipitation, contaminant sources, and land use. The models will reflect mean conditions during selected periods from the 1970s through 2002 for fluvial drainages along the Atlantic and Gulf of Mexico coasts. The models will be used to assess the climatic and humanrelated variations in nutrient and sediment loads in coastal streams. The model predictions of loads will also serve as input to the trophic-response estuarine models developed in this project.

USGS activities in years one and two include the compilation and processing of watershed data to support the development of the new SPARROW models. Detailed historical databases of climate, contaminant sources, land use, and stream water-quality measurements will be developed for the Atlantic and Gulf coastal drainages for the specified time periods to support the model applications. The SPARROW models will use an existing spatial infrastructure based on a national 1:500,000-scale river reach network with 30-m land use and 1-km elevation data. All watershed data developed for SPARROW and the other watershed models will be jointly coordinated and shared among project collaborators.