APPLICATION OF A LINKED WATER QUALITY RESPONSE MODEL FOR ESTABLISHING NUTRIENT CRITERIA FOR THE CALIFORNIA OAK AND CHAPARRAL SUB-ECOREGION

Aquatic systems require some amount of nutrients to maintain their health. However, an excess of these nutrients can lead to impairments such as excessive algal growth, eutrophication, and alteration of the biological community. A current regulatory challenge is defining appropriate nutrient criteria which will protect the designated uses of a waterbody consistent with ecoregional characteristics. EPA Region 9 has initiated a pilot study focusing on Sub-ecoregion 6 (California oak and chaparral). The overall strategy is to develop three lines of investigation to support the development of criteria recommendations: 1) review of long-term monitoring studies; 2) empirical data analysis of existing water quality data; and 3) watershed modeling assessments of natural background loading and response. The modeling component of the pilot project will be used to provide a method for controlling confounding factors, such as habitat alteration, which may skew conclusions derived solely from monitoring data.Determining criteria requires information on both loads and responses. Criteria should not be less than background; neither should they be set at levels lower than those associated with a risk of degradation. In many cases, target nutrient levels may be “inherited” from downstream waterbodies that are likely to be more stringent than the concentration target needed to prevent impairment within the stream itself. A set of linked loading, transport, and response models were used to translate the potentially more stringent receiving water standards to upstream watershed export.Response models for lakes/reservoirs are needed to estimate loading targets consistent with protection of beneficial uses from excessive algal growth. The BATHTUB model provides a useful tool for estimating lake response to nutrients (Walker, 1996). The model simulates phosphorus, nitrogen, phytoplankton, chlorophyll a, Secchi depth, and hypolimnetic oxygen consumption. A combination of the United States Geological Survey (USGS) SPAtially Referenced Regressions On Watershed Attributes (SPARROW) model (Smith et al., 1997) and the USDA Soil and Water Assessment Tool (SWAT) model (Arnold et al., 1990) were used to predict the watershed loading and stream transport used to generate the inputs to the BATHTUB models. SWAT is a long-term, continuous time watershed model that simulates watershed nutrient export based on weather, soil, topography, and vegetation data. Significant losses of nutrients can occur during instream transport due to plant uptake, loss to sedimentation, or, in the case of nitrogen, volatilization. SPARROW provides a simple empirical approach to estimation of transport losses for application to multiple sites within an ecoregion.The result of the modeling analysis is an estimate of watershed loading rates and instream nutrient concentrations which is predicted to meet instream and receiving water designated uses and is consistent with ecoregional characteristics. The estimates provide a baseline from which to evaluate conditions where designated uses should be fully realized and allow decision-makers to discriminate water quality impacts that are due to nutrient over-enrichment.