Streamwater and solid samples (rock, unconsolidated sediment, stream sediment, and efflorescent material) in the Toll Gate Creek watershed, Colorado, were collected and analyzed for major and trace elements to determine trace-element concentrations and stream loads from December 2003 through March 2004, a period of seasonally low flow. Special emphasis was given to selenium (Se) concentrations because historic Se concentrations exceeded current (2004) stream standards. The goal of the project was to assess the distribution of Se concentration and loads in Toll Gate Creek and to determine the potential for rock and unconsolidated sediment in the basin to be sources of Se to the streamwater.
Streamwater samples and discharge measurements were collected during December 2003 and March 2004 along Toll Gate Creek and its two primary tributaries - West Toll Gate Creek and East Toll Gate Creek. During both sampling periods, discharge ranged from 2.5 liters per second to 138 liters per second in the watershed. Discharge was greater in March 2004 than December 2003, but both periods represent low flow in Toll Gate Creek, and results of this study should not be extended to periods of higher flow. Discharge decreased moving downstream in East Toll Gate Creek but increased moving downstream along West Toll Gate Creek and the main stem of Toll Gate Creek, indicating that these two streams gain flow from ground water. Se concentrations in streamwater samples ranged from 7 to 70 micrograms per liter, were elevated in the upstream-most samples, and were greater than the State stream standard of 4.6 micrograms per liter. Se loads ranged from 6 grams per day to 250 grams per day, decreased in a downstream direction along East Toll Gate Creek, and increased in a downstream direction along West Toll Gate Creek and Toll Gate Creek. The largest Se-load increases occurred between two sampling locations on West Toll Gate Creek during both sampling periods and between the two sampling locations on the main stem of Toll Gate Creek during the December 2003 sampling. These load increases may indicate that sources of Se exist between these two locations; however, Se loading along West Toll Gate Creek and Toll Gate Creek primarily was characterized by gradual downstream increases in load. Linear regressions between Se load and discharge for both sampling periods had large, significant values of r2 (r2 > 0.96, p < 0.0001) because increases in Se load (per unit of flow increase) were generally constant. This relation is evidence for a constant addition of water having a relatively constant Se concentration over much of the length of Toll Gate Creek, a result which is consistent with a ground-water source for the Se loads.
Rock outcroppings along the stream were highly weathered, and Se concentrations in rock and other solid samples ranged from below detection (1 part per million) to 25 parts per million. One sample of efflorescence (a surface encrustation produced by evaporation) had the greatest selenium concentration of all solid samples, was composed of thenardite (sodium sulfate), gypsum (calcium sulfate) and minor halite (sodium chloride), and released all of its Se during a 30-minute water-leaching procedure. Calculations indicate there was an insufficient amount of this material present throughout the watershed to account for the observed Se load in the stream. However, this material likely indicates zones of ground-water discharge that contain Se.
This report did not identify an unequivocal source of Se in Toll Gate Creek. However, multiple lines of evidence indicate that ground-water discharge supplies Se to Toll Gate Creek: (1) the occurrence of elevated Se concentrations in the stream throughout the watershed and in the headwater regions, upstream from industrial sources; (2) the progressive increase in Se loads moving downstream, which indicates a continuous input of Se along the stream rather than input from point sources; (3) the occurr