1. Merced River at River Road Bridge near Newman, CA (11273500)
2. Orestimba Creek at River Road near Crows Landing, CA (11274538)
3. San Joaquin River near Vernalis, CA (11303500)
4. Cosumnes River at Michigan Bar, CA (11335000)

1. Salt Slough at Highway 165 near Stevinson, CA (11261100)
2. Merced River at River Road Bridge near Newman, CA (11273500)
3. Merced River at McConnell State Park near Livingston, CA (372450120423300)
4. San Joaquin River at Patterson Bridge near Patterson, CA (11274570)
5. Tuolumne River at Hickman Bridge near Waterford, CA (11289800)
6. San Joaquin River near Vernalis, CA (11303500)
7. Stanislaus River at Riverbank, CA (374419120570701)
8. Cosumnes River at Michigan Bar, CA (11335000)

Regional Synthesis Studies

The San Joaquin-Tulare Basins NAWQA has been involved in two regional synthesis studies:

Macroinvertebrate community responses to landscape and hydrologic alterations

This study includes the interpretation of data for 1993-2004 for three Major River Basins (MRB) in the Western United States: (1) Rio Grande, Colorado River and Great Basins (MRB6), (2) Pacific Northwest River Basins (MRB7), and (3) California River Basins (MRB8). The objectives of the study were:

1. Formulate measures of landscape and hydrologic alteration appropriate for western river systems.  Determination of land-use condition will be based on general measures of disturbance associated with human activities (e.g. agriculture, logging, mining, urban development, grazing), and will include mixtures of land-cover types.   Hydrologic alteration will represent deviations from natural flow regimes owing to impoundments, diversions, or other water-management practices.


2. Describe spatial patterns in macroinvertebrate community attributes (indices/metrics) in response to hydrologic and associated landscape conditions.


3. Determine whether stream habitat conditions are changing in response to landscape disturbance and hydrologic alteration.


4. Develop biological tolerance and environmental optima relations for western macroinvertebrates, and use optima standards to identify environmental thresholds of chemical and physical conditions associated with landscape disturbance.

Results from this NAWQA study are published in Konrad and others, 2008 (see publications).

(2) Developing predictive disturbance models for biology

This study is a follow-up study to (1) for the same area and time period. The objectives of this study are:

Trends in biological conditions of the nation’s waters are important to both federal and state regulatory agencies. These agencies usually address biological conditions using various metrics (e.g., species richness) or indices (e.g., IBIs) of community structure. Another approach for assessing biological condition is the development of predictive models. These models allow the user to determine how the observed macroinvertebrate taxa sampled from a “test” site relates to the expected taxa based on the reference sites used to construct the model. Degradation of biological condition is indicated when expected (based on reference conditions) species are not found at a test site. These types of models are also called O/E or “O over E models (i.e., ratio of observed to expected taxa).” This type of model like IBIs is very useful as a bioassessment tool for inferring impairment of sampled water bodies but is not a predictive model in the sense meant in this proposal; it does not predict the condition of unsampled stream sites.

We plan to develop predictive disturbance models that use measures of watershed disturbance including urban and agricultural land use, land cover, flow regime, and hydrologic infrastructure (e.g., number of dams, number of canals) as the predictors of biological condition at unsampled sites. In particular, we will be using data for benthic macroinvertebrate communities. We will be using various metrics of invertebrate community structure (e.g. number of species belonging to Ephemeroptera, Plecoptera, and Trichoptera) as our measure of biological condition. Methodologies for developing predictive disturbance models for biological data have not been well developed in the literature, consequently, the overall objective of this study is to test approaches for developing predictive disturbance biological models. In particular, we will:

1. Construct spatially dense data sets for benthic macroinvertebrate surveys from NAWQA, EPA, and state programs for Oregon, California, and Washington. EPA data will include samples from the Western EMAP and Wadeable Streams Assessment programs. California, Oregon, and Washington have state bioassessment programs that utilize sampling of benthic macroinvertebrate assemblages. A variety of standard invertebrate metrics will then be calculated from the data set using IDAS.


2. Construct companion data sets of landscape disturbance variables to be used as independent variables in the development of models. As part of the IR6 project we have already been evaluating generalized measures of disturbance based on land use and hydrologic infrastructure. These generalized measures will be used in the proposed analysis but we will also consider other national and regional features that we did not utilize in IR6. Pesticide National Synthesis and Nutrient and Trace Element National synthesis coverages might be particularly appropriate.


3. Hydrologic variability has been identified as a likely mechanism of disturbance that will produce high uncertainty in response predictions unless it is incorporated in the model. Consequently, we will: 1) continue and expand analyses addressing the relationship between flow regime and invertebrate communities started as part of the IR6 project (Konrad and others, 2008), 2) identify surrogate variables, such as basin area, precipitation, geology, elevation, and land use, that can be used to predict a few broad hydrologic characteristics (e.g., variability categories, frequent or infrequent flooding, high or low baseflow) based on readily available GIS coverages. and 3) integrate the surrogate variables into the predictive disturbance models.


4. Survey the literature and collaborate with practitioners to determine a variety of modeling approaches to apply to the data, then build and evaluate the best models.

Final products from this study will not be available until 2010.