Prediction of fish and sediment mercury in streams using landscape variables and historical mining

• Published in: The Science of the total environment Vol. 571; pp. 364 - 379
• Main Authors: Alpers, Charles N., Yee, Julie L., Ackerman, Joshua T., Orlando, James L., Slotton, Darrel G., Marvin-DiPasquale, Mark C.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.11.2016
• Subjects: Akaike Information Criteria; California; Fish; Geospatial modeling; Gold mining; Methylmercury; Oncorhynchus mykiss; Salmo trutta; Streambed sediment; USA, California, Sierra Nevada Mts; INE, USA, California, San Francisco Bay; USA, California, Sacramento; Methylmercury; Streambed sediment; Fish; Geospatial modeling; California; Akaike Information Criteria; Gold mining
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2016.05.088

Widespread mercury (Hg) contamination of aquatic systems in the Sierra Nevada of California, U.S., is associated with historical use to enhance gold (Au) recovery by amalgamation. In areas affected by historical Au mining operations, including the western slope of the Sierra Nevada and downstream areas in northern California, such as San Francisco Bay and the Sacramento River–San Joaquin River Delta, microbial conversion of Hg to methylmercury (MeHg) leads to bioaccumulation of MeHg in food webs, and increased risks to humans and wildlife. This study focused on developing a predictive model for THg in stream fish tissue based on geospatial data, including land use/land cover data, and the distribution of legacy Au mines. Data on total mercury (THg) and MeHg concentrations in fish tissue and streambed sediment collected during 1980–2012 from stream sites in the Sierra Nevada, California were combined with geospatial data to estimate fish THg concentrations across the landscape. THg concentrations of five fish species (Brown Trout, Rainbow Trout, Sacramento Pikeminnow, Sacramento Sucker, and Smallmouth Bass) within stream sections were predicted using multi-model inference based on Akaike Information Criteria, using geospatial data for mining history and landscape characteristics as well as fish species and length (r2=0.61, p<0.001). Including THg concentrations in streambed sediment did not improve the model's fit, however including MeHg concentrations in streambed sediment, organic content (loss on ignition), and sediment grain size resulted in an improved fit (r2=0.63, p<0.001). These models can be used to estimate THg concentrations in stream fish based on landscape variables in the Sierra Nevada in areas where direct measurements of THg concentration in fish are unavailable. [Display omitted] •Stream fish Hg positively influenced by Au mines, deciduous forests, urban areas.•Fish Hg negatively influenced by evergreen forests and wetlands in Sierra Nevada.•Sediment Hg and MeHg affected by LOI, grain size, land use and cover, Au mines.•Geospatial model predicts Hg in 5 fish species as a function of length and location.•Model requiring sediment MeHg slightly better than geospatial-only model.