Concentrations and loads of suspended sediment and trace element pollutants in a small semi-arid urban tributary, San Francisco Bay, California.

Abstract

Water-quality policy documents throughout the world often identify urban stormwater as a large and controllable impact to sensitive ecosystems, yet there is often limited data to characterize concentrations and loads especially for rare and more difficult to quantify pollutants. In response, concentrations of suspended sediments and silver, mercury and selenium including speciation, and other trace elements were measured in dry and wet weather stormwater flow from a 100% urban watershed near San Francisco. Suspended sediment concentrations ranged between 1.4 and 2700 mg/L and varied with storm intensity. Turbidity was shown to correlate strongly with suspended sediments and most trace elements and was used as a surrogate with regression to estimate concentrations during unsampled periods and to compute loads. Mean suspended sediment yield was 31.5 t/km(2)/year. Total mercury ranged between 1.4 and 150 ng/L and was, on average, 92% particulate, 0.9% methylated, and 1.2% acid labile. Total mercury yield averaged 5.7 μg/m(2)/year. Total selenium ranged between non-detect and 2.9 μg/L and, on average, the total load (0.027 μg/m(2)/year) was 61% transported in dissolved phase. Selenate (Se(VI)) was the dominant species. Silver concentrations ranged between non-detect and 0.11 μg/L. Concentrations and loads of other trace elements were also highly variable and were generally similar to other urban systems with the exceptions of Ag and As (seldom reported) and Cr and Zn which exhibited concentrations and loads in the upper range of those reported elsewhere. Consistent with the semi-arid climatic setting, >95% of suspended sediment, 94% of total Hg, and 85-95 % of all other trace element loads were transported during storm flows with the exception of selenium which showed an inverse relationship between concentration and flow. Treatment of loads is made more challenging in arid climate settings due to low proportions of annual loads and greater dissolved phase during low flow conditions. This dataset fills an important local data gap for highly urban watersheds of San Francisco Bay. The field and interpretative methods, the uniqueness of the analyte list, and resulting information have general applicability for managing pollutant concentrations and loads in urban watersheds in other parts of the world and may have particularly useful application in more arid climates.