Fecal coliform bacteria in stream sediments: Sampling methodology, measurement and modeling the impact on stream water quality.

摘要

Three studies were conducted related to the measurement and impact of stream sediment fecal coliform (FC) bacteria on stream water quality. In part one an enumeration technique for sediment FC was defined and statistically characterized. This characterization necessitated the development of a sample splitting mechanism, which was found to split samples with no significant bias. The proportion of variation (PV) due to measurement error of the technique was found in the lab to be 3% for two sediment types tested. Varying agitation method used in the enumeration technique significantly affected counts in one of the sediment types tested.; This technique was utilized in the second study, where sediment and water samples were collected from a reach in the Willamette River to determine whether or not sediments underlying slow-moving water areas in streams ('alcoves') can harbor more FC than main channel sediments. Alcove sediments were found to harbor significantly more FC than sediments in the main channel. FC concentrations measured in one alcove's sediments over the course of a year showed a correlation between alcove sediment and alcove water FC concentrations during non-storm periods (r = 0.80), and no correlation (r = .03) during storm periods.; The third study was an analysis of the potential impact of FC-laden alcove sediments on a stream's water quality. A process-based model was constructed in order to identify the conditions under which alcoves might impact main channel water quality, and to determine which of the processes are the primary drivers of the system. Sensitivity analyses were run to identify process drivers. Alcoves with high sediment FC concentration were found to significantly impact main channel FC concentration during periods of high main channel flow and low upstream FC inputs. High FC concentrations in the main channel masked the alcoves' effect. The impact of the alcove on the main channel was found to be sensitive to the hydraulic exchange parameters and the FC die-off rate in water, and not sensitive to the in-alcove processes of resuspension, settling, and FC die-off in sediment. The model was applied to a reach of the Willamette River using 14 months of flow and FC concentration data. The model predicted that a single alcove with typical FC contamination in the sediment would not significantly impact main channel water quality. The correlation (r) between predicted alcove sediment and water FC concentrations was found to be 0.66 during non-storm periods.