Factors affecting reservoir and stream-water quality in the Cambridge, Massachusetts, drinking-water source area and implications for source-water protection

This report presents the results of a study conducted by the U.S. Geological Survey, in cooperation with the city of Cambridge, Massachusetts, Water Department, to assess reservoir and tributary-stream quality in the Cambridge drinking-water source area, and to use the information gained to help guide the design of a comprehensive water-quality monitoring program for the source area. Assessments of the quality and trophic state of the three primary storage reservoirs, Hobbs Brook Reservoir, Stony Brook Reservoir, and Fresh Pond, were conducted (September 1997-November 1998) to provide baseline information on the state of these resources and to determine the vulnerability of the reservoirs to increased loads of nutrients and other contaminants. The effects of land use, land cover, and other drainage-basin characteristics on sources, transport, and fate of fecal-indicator bacteria, highway deicing chemicals, nutrients, selected metals, and naturally occurring organic compounds in 11 subbasins that contribute water to the reservoirs also was investigated, and the data used to select sampling stations for incorporation into a water-quality monitoring network for the source area.

All three reservoirs exhibited thermal and chemical stratification, despite artificial mixing by air hoses in Stony Brook Reservoir and Fresh Pond. The stratification produced anoxic or hypoxic conditions in the deepest parts of the reservoirs and these conditions resulted in the release of ammonia nitrogen orthophosphate phosphorus, and dissolved iron and manganese from the reservoir bed sediments.
Concentrations of sodium and chloride in the reservoirs usually were higher than the amounts recommended by the U.S. Environmental Protection agency for drinking-water sources (20 milligrams per liter for sodium and 250 milligrams per liter for chloride). Maximum measured sodium concentrations were highest in Hobbs Brook Reservoir (113 milligrams per liter), intermediate in Stony Brook Reservoir (62 milligrams per liter), and lowest in Fresh Pond (54 milligrams per liter). Bed sediments in Hobbs Brook and Stony Brook Reservoirs were enriched in iron, manganese, and arsenic relative to those in the impounded lower Charles River in Boston, Massachusetts.

Trophic state indices, calculated for each reservoir based on nutrient concentrations, water-column transparency, and phytoplankton abundances, indicated that the upper and middle basins of Hobbs Brook Reservoir were moderately to highly productive and likely to produce algal blooms; the lower basin of Hobbs Brook Reservoir and Stony Brook Reservoir were similar and intermediate in productivity, and Fresh Pond was relatively unproductive and unlikely to produce algal blooms. This pattern is likely due to sedimentation of organic and inorganic particles in the three basins of Hobbs Brook Reservoir and in Stony Brook Reservoir. Molar ratios of nitrogen to phosphorus ranged from 55 in Stony Brook Reservoir to 120 in Hobbs Brook Reservoir, indicating that phytoplankton algae in these water bodies may be phosphorus limited and therefore sensitive to small increases in phosphorus loading from the drainage basin. Nitrogen loads were found to be less important than phosphorus to the trophic condition of the reservoirs.
Hobbs Brook and Stony Brook, the two principle streams draining the Cambridge drinking-water source area, differed in their relative contributions to many of the estimated constituent loads. The estimated load of fecal coliform bacteria was more than seven times larger for the mainly residential Stony Brook subbasin upstream from Kendal Green, Mass., than it was for the more commercial and industrial Hobbs Brook subbasin, though the drainage areas of the two subbasins differ only by about 20 percent. The State standard for fecal coliform bacteria in streams in the Cambridge drinking-water source area (20 colony forming units per 100 milliliters) was exceeded at all sampling stations.

Estimated s