Geochemical assessment of metals and dioxin in sediment from the San Carlos Reservoir and the Gila, San Carlos, and San Francisco Rivers, Arizona

Abstract

In October 2004, we sampled stream-bed sediment, terrace sediment, and sediment from the San Carlos Reservoir to determine the spatial and chronological variation of six potentially toxic metals-Cu, Pb, Zn, Cd, As, and Hg. Water levels in the San Carlos Reservoir were at a 20-year low at an elevation of 2,409 ft (734.3 m). Four cores were taken from the reservoir: one from the San Carlos River arm, one from the Gila River arm, and two from the San Carlos Reservoir just west of the Pinal County line. Radioisotope chronometry (7Be, 137Cs, and 210Pb) conducted on sediment from the reservoir cores provides a good chronological record back to 1959. Chronology prior to that, during the 1950s, is based on our interpretation of the 137Cs anomaly in reservoir cores. During and prior to the 1950s, the reservoir was dry and sediment-accumulation rates were irregular; age control based on radioisotope data was not possible. We recovered sediment at the base of one 4-m-long core that may date back to the late 1930s. The sedimentological record contains two discrete events, one about 1978-83 and one about 1957, where the Cu concentration in reservoir sediment exceeded recommended sediment quality guidelines and should have had an effect on sensitive aquatic and benthic organisms. Concentrations of Zn determined in sediment deposited during the 1957(?) event also exceeded recommended sediment quality guidelines. Concentration data for Cu from the four cores clearly indicate that the source of this material was upstream on the Gila River. Lead isotope data, coupled with the geochemical data from a 2M HCl-1 percent H2O2 leach of selected sediment samples, show two discrete populations of data. One represents the dominant sediment load derived from the Safford Valley, and a second reflects sediment derived from the San Francisco River. The Cu concentration spikes in the reservoir cores have chemical and Pb isotope signatures that indicate that deposits in a porphyry copper deposit from the Morenci district is the likely source of these Cu-rich sedimentary deposits. Copper concentrations and Pb isotope data in premining terrace-sediment deposits indicate that the Cu peaks could not have resulted from erosion of premining sediment from terrace deposits downstream on the Gila River. The chemical and Pb isotope data also indicate that agricultural practices in the Safford Valley have resulted in an increased sediment load to the Gila River since large-scale farming began, prior to the time when the San Carlos Reservoir was built. Analyses of dioxin, which is an impurity in one of the herbicides used in the late 1960s and early 1970s, were completed in sediment from one of the cores in the reservoir to determine whether any of these pesticide residues have accumulated in the reservoir sediment. Dioxin concentration is expressed in terms of its toxicity (toxic equivalent concentration or TEQ). Concentrations of dioxin in the sediment ranged from 0.68 to 1.37 pg/g and are less than any of the benchmark concentrations recommended as threshold values for adverse effects of dioxin in sediment (> 2.5-10 pg/g).