The U.S. Geological Survey, in cooperation with the Pioneer Valley Planning Commission, the U.S. Environmental Protection Agency, and the Massachusetts Department of Environmental Protection Senator William X. Wall Experiment Station, assessed the presence of 14 commonly used human-health pharmaceutical compounds, fecal indicator bacteria, and other man-made compounds indicative of the presence of human sewage in the lower reach of the Mill River near its confluence with the Connecticut River in Springfield, Massachusetts. The study was part of the Tri-State Connecticut River Targeted Watershed Initiative and involved the collection and analysis of raw river water at three sites along the reach, extending from Watershops Pond to the mouth, over the course of a low-flow period, July through November 2010. Previous studies in the region indicated that nonpoint or undocumented sources of wastewater contributed a variety of organic contaminants and potentially harmful bacteria to rivers under both high- and low-flow conditions. Additional samples, including a raw sewage sample collected near a Mill River combined sewer overflow during a non-overflow period, were collected in March 2011.

The study was designed to determine if city sewage or other domestic sources of wastewater were entering the river within this reach during low-flow conditions. No definitive evidence of sewage was measured in Mill River water samples collected during the study period. Fecal indicator bacteria, including Escherichia coli (E. coli) and enterococci bacteria, were detected in all Mill River water samples. In the DNA analysis of enterococci cultures from the Mill River, samples generally tested negative for the Enterococcus faecium (esp) human-specific genetic marker, whereas the raw sewage sample tested positive. Samples also generally tested negative in the human-specific rDNA marker assay for the anaerobic bacterium Bacteroidetes. Samples tested negative in 2010 for two Bacteroidetes human-specific genetic markers, HF134 and HF183, except samples from near the mouth of the Mill River, which tested positive. Samples collected in March 2011 from all three measurement sites tested positive for both markers. The results of bacterial analyses suggest that the fecal bacteria in summer and fall months are most likely of animal origin rather than human. Despite the urban setting, long history of development, and many potential sources of man-made contamination in the Mill River, none of the 12 water samples collected during the study contained targeted pharmaceutical compounds at concentrations greater than the analytical reporting levels. Other man-made compounds, like fluorescent whitening agents, were measured and detected in samples at low concentrations 4 out of 5 times the samples were collected; however, the other lines of evidence do not support a sewer source but rather other nonpoint sources upstream in the watershed.

The results of this study do not support the hypothesis that aging sewer lines or combined sewer overflow infrastructure leak into the Mill River as tested during the low-flow conditions during sampling for this study. None of the results from Mill River samples offer conclusive evidence of the presence of sewage. Some low-level detections of pharmaceutical compounds, other man-made chemicals, and bacteria suggest an upstream, nonpoint source.

A single raw sewage sample was collected, diluted, and examined for comparison with Mill River water samples and to ensure that the analytical methods could detect typical wastewater constituents. High levels of bacteria were measured, and low levels of three anthropogenic pharmaceutical compounds were detected, confirming the effectiveness of the sub-part-per-million method. The concentration of fluorescent whitening agent-1 in the sewage sample was 90,000 times greater than the median concentration in the Mill River samples.