Mercury in fish, bed sediment, and water from streams across the United States, 1998-2005

Abstract

Mercury (Hg) was examined in top-predator fish, bed sediment, and water from streams that spanned regional and national gradients of Hg source strength and other factors thought to influence methylmercury (MeHg) bioaccumulation. Sampled settings include stream basins that were agricultural, urbanized, undeveloped (forested, grassland, shrubland, and wetland land cover), and mined (for gold and Hg). Each site was sampled one time during seasonal low flow. Predator fish were targeted for collection, and composited samples of fish (primarily skin-off fillets) were analyzed for total Hg (THg), as most of the Hg found in fish tissue (95-99 percent) is MeHg. Samples of bed sediment and stream water were analyzed for THg, MeHg, and characteristics thought to affect Hg methylation, such as loss-on-ignition (LOI, a measure of organic matter content) and acid-volatile sulfide in bed sediment, and pH, dissolved organic carbon (DOC), and dissolved sulfate in water. Fish-Hg concentrations at 27 percent of sampled sites exceeded the U.S. Environmental Protection Agency human-health criterion of 0.3 micrograms per gram wet weight. Exceedances were geographically widespread, although the study design targeted specific sites and fish species and sizes, so results do not represent a true nationwide percentage of exceedances. The highest THg concentrations in fish were from blackwater coastal-plain streams draining forests or wetlands in the eastern and southeastern United States, as well as from streams draining gold- or Hg-mined basins in the western United States (1.80 and 1.95 micrograms THg per gram wet weight, respectively). For unmined basins, length-normalized Hg concentrations in largemouth bass were significantly higher in fish from predominantly undeveloped or mixed-land-use basins compared to urban basins. Hg concentrations in largemouth bass from unmined basins were correlated positively with basin percentages of evergreen forest and also woody wetland, especially with increasing proximity of these two land-cover types to the sampling site; this underscores the greater likelihood for Hg bioaccumulation to occur in these types of settings. Increasing concentrations of MeHg in unfiltered stream water, and of bed-sediment MeHg normalized by LOI, and decreasing pH and dissolved sulfate were also important in explaining increasing Hg concentrations in largemouth bass. MeHg concentrations in bed sediment correlated positively with THg, LOI, and acid-volatile sulfide. Concentrations of MeHg in water correlated positively with DOC, ultraviolet absorbance, and THg in water, the percentage of MeHg in bed sediment, and the percentage of wetland in the basin.