California Bay – Delta Authority Fish Mercury Project: The Relationship between Landscape Features and Sport Fish Mercury in the Sacramento-San Joaquin Delta Watershed

The goal of this study was to quantify the relationship between landscape features and mercury in sport fish. We hypothesized that spatial variation in sport fish mercury would most strongly relate to major point sources associated with mining activity, habitat features (e.g., abundance of wetlands), and watershed-scale patterns in mercury deposition from the atmosphere.

Geographic information system (GIS) techniques were used to evaluate the influence of these features on the bioaccumulation of mercury in sport fish. In addition, we attempted to determine the most appropriate spatial scale for analysis of these influences. Fish mercury data were selected from a state-wide historical bioaccumulation database, which also included data collected by the Fish Mercury Project (FMP) in 2005. Based on the geographic distribution of the sampling locations, three species common across the study area were chosen for analysis: largemouth bass, channel catfish, and white catfish. For the time interval examined (1990 – 2005), the mean wet weight concentration for each species at a given location was calculated.

By superimposing hydrologic flow paths with watershed delineations, we were able to determine flow direction and pathways from upstream sources to downstream habitats. “Fish watersheds” (watershed areas upstream of fish sampling locations) were created for a subset of locations in the CalFed Ecosystem Restoration Program Geographic Scope. Regions heavily influenced by mining within each watershed were also delineated. We examined landscape features in fish watersheds and within close proximity (circular buffers) of fish locations. Using a GIS database, several statistics were derived for each fish watershed area and buffer, including the percent covered by habitat features (vegetated vs. non-vegetated wetlands, permanent vs. temporary inundation), number of gold and mercury mines, total land area (km2), and total length (km) of all streams and rivers.

The work presented in this report is the first attempt the authors are aware of to relate landscape features to fish mercury concentrations using GIS on a large regional scale. Fish mercury concentrations appeared to be influenced by many complex and interacting factors that obscured any obvious correlations with atmospheric deposition, historic mining districts, and wetlands at the watershed scale. Some features seemed to have an influence at the proximal-scale (buffer) analyses, namely vegetated wetland and temporarily inundated aquatic habitat. Though these tentative conclusions were based on small sample sizes, they contradicted the prevailing notion that wetlands generally increase methylmercury accumulation in the food web.

The lack of clear relationships in the dataset analyzed is probably due to shortcomings in the available data, rather than a true absence of influence of atmospheric deposition, mining, or wetlands on spatial patterns in mercury bioaccumulation. Future attempts to relate landscape features with mercury in the food web at the watershed scale should focus on watersheds with a stronger linkage between landscape features and mercury in fish and fewer confounding factors. Statewide monitoring of sport fish is currently being planned that should provide a better basis for conducting this type of analysis in the future.