Monitoring and Modeling Temporal Variations of Pathogens and Indicators in an Impacted Watershed Used for Drinking Water Supply
监测和模拟饮用水供应受影响流域病原体和指标的时间变化
Results are presented from a study examining the temporal changes in pathogen and
indicator numbers during storm events in April and May of 2003 in a watershed in
Southwestern Ontario, Canada. Water samples were collected for total coliforms, fecal
coliforms, Escherichia coli, Escherichia coli O157:H7, and Campylobacter spp. Samples
were not analyzed for Cryptosporidium spp. and Giardia spp. at that time; however,
subsequent storm event samples were collected and analyzed for the two protozoa. Event
samples were collected from within an intensely farmed subwatershed of the Grand
River.
All samples were positive for total coliforms, fecal coliforms and Escherichia coli.
During precipitation events, coliform indicators increased to peak numbers very rapidly
from their baseline concentrations, and a tailing of their numbers was also observed
following an event, as expected. Peak concentrations of coliform bacteria were more
than 3 orders of magnitude greater than baseline concentrations for the events sampled.
E. coli O157:H7 was not detected in any event samples. Campylobacter spp. was
observed in water samples at the beginning of one of the events; however, the numbers of
Campylobacter spp. dropped off to undetectable levels as the event progressed. Using
the Spearman rank correlation test, a positive correlation was observed between E. coli
and turbidity (Rs=0.89). Although E. coli bacteria originate in fecal matter, they are
widespread in the environment and are therefore rarely flushed out completely following
a precipitation event.
In order to predict the levels of indicator bacteria and potentially pathogenic
microorganisms in a stream, an existing watershed-scale hydrologic model,
WATFLOOD/SPL, was augmented for pathogen fate and transport from point and nonpoint
sources. The pathogen model considers transport as a result of overland flow,
subsurface flow to tile drainage systems, as well as in-stream routing. Model results were
reliable in estimating E. coli concentrations (at an order-of-magnitude level). However,
for pathogenic microorganisms which are less widespread in the environment with highly
localized sources, more research in estimating actual environmental loadings is needed.
It is essential that drinking water treatment plants be capable of handling peak
concentrations of pathogens. Information on baseline levels and peak concentrations of
major pathogens of concern should be available to ensure robust treatment plant design.
This study offers additional insight into temporal fluctuations of potential pathogen
concentrations arriving at a treatment plant intake. Includes 25 references, figures.