U.S. Geological Survey, Water Resources of Pennsylvania

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2005-5039

Occurrence of Fecal-Indicator Bacteria and Protocols for Identification of Fecal-Contamination Sources in Selected Reaches of the West Branch Brandywine Creek, Chester County, Pennsylvania

By: Peter J. Cinotto

This report is available online in Portable Document Format (PDF). If you do not have the Adobe Acrobat PDF Reader, it is available for free download from Adobe Systems Incorporated.

View the full report in PDF 5.38 MB

ABSTRACT

The presence of fecal-indicator bacteria indicates the potential presence of pathogens originating from the fecal matter of warm-blooded animals. These pathogens are responsible for numerous human diseases ranging from common diarrhea to meningitis and polio. The detection of fecal-indicator bacteria and interpretation of the resultant data are, therefore, of great importance to water-resource managers. Current (2005) techniques used to assess fecal contamination within the fluvial environment primarily assess samples collected from the water column, either as grab samples or as depth- and (or) width-integrated samples. However, current research indicates approximately 99 percent of all bacteria within nature exist as attached, or sessile, bacteria. Because of this condition, most current techniques for the detection of fecal contamination, which utilize bacteria, assess only about 1 percent of the total bacteria within the fluvial system and are, therefore, problematic. Evaluation of the environmental factors affecting the occurrence and distribution of bacteria within the fluvial system, as well as the evaluation and modification of alternative approaches that effectively quantify the larger population of sessile bacteria within fluvial sediments, will present water-resource managers with more effective tools to assess, prevent, and (or) eliminate sources of fecal contamination within pristine and impaired watersheds.

Two stream reaches on the West Branch Brandywine Creek in the Coatesville, Pa., region were studied between September 2002 and August 2003. The effects of sediment particle size, climatic conditions, aquatic growth, environmental chemistry, impervious surfaces, sediment and soil filtration, and dams on observed bacteria concentrations were evaluated. Alternative approaches were assessed to better detect geographic sources of fecal contamination including the use of turbidity as a surrogate for bacteria, the modification and implementation of sandbag bacteria samplers, and the use of optical brighteners. For the purposes of this report, sources of bacteria were defined as geographic locations where elevated concentrations of bacteria are observed within, or expected to enter, the main branch of the West Branch Brandywine Creek. Biologic sources (for example, waterfowl) were noted where applicable; however, no specific study of biologic sources (such as bacterial source tracking) was conducted.

Data indicated that specific bacterial populations within fluvial sediments could be related to specific particle-size ranges. This relation is likely the result of the reduced porosity and permeability associated with finer sediments and the ability of specific bacteria to tolerate particular environments. Escherichia coli (E. coli) showed a higher median concentration (2,160 colonies per gram of saturated sediment) in the 0.125 to 0.5-millimeter size range of natural sediments than in other ranges, and enterococcus bacteria showed a higher median concentration (61,830 colonies per gram of saturated sediment) in the 0.062 to 0.25-millimeter size range of natural sediments than in other ranges. There were insufficient data to assess the particle-size relation to fecal coliform bacteria and (or) fecal streptococcus bacteria.

Climatic conditions were shown to affect bacteria concentrations in both the water column and fluvial sediments. Drought conditions in 2002 resulted in lower overall bacteria concentrations than the more typically wet year of 2003. E. coli concentrations in fluvial sediment along the Coatesville study reach in 2002 had a median concentration of 92 colonies per gram of saturated sediment; in 2003, the median concentration had risen to 4,752 colonies per gram of saturated sediment.

Symbiotic relations between bacteria and aquatic growth were likely responsible for increased bacteria concentrations observed within an impoundment area on the Coatesville study reach. This reach showed evidence of elevated aquatic growth and sharp increases in E. coli concentrations from upstream to downstream through the impoundment area in both 2002 and 2003. In 2003, E. coli concentrations within the waters column increased from 940 colonies per 100 milliliters upstream to 6,000 colonies per 100 milliliters at the dam crest. Given that these bacteria likely resulted from natural bacterial regrowth, the use of E. coli as an indicator of fecal contamination was severely impaired.

Variable environmental conditions along the West Branch Brandywine Creek made the common field-chemical parameters of specific conductance, temperature, pH, and dissolved oxygen ineffective and (or) impossible to use for the determination of inputs of fecal contamination. Extreme variations in chemical gradients commonly were related to the urban/industrial signature of the watershed. For example, during base-flow sampling in 2002, specific-conductance values exceeding 1,000 microsiemens per centimeter observed in effluent from a local steel mill. This effluent raised the specific conductance within the West Branch Brandywine from just above 200 microsiemens per centimeter upstream from the outfall to just below 500 microsiemens per centimeter downstream from the outfall. These chemical gradients also, likely, had an effect on the initial colonization of bacteria, the formation of biofilms, and the persistence of certain types of bacteria along the study reach.

Data collected in 2003 indicated that nutrients increased during both base-flow and stormflow conditions along the Coatesville study reach. For example, during base-flow sampling in 2003, 20 pounds of phosphorus was shown to enter the West Branch Brandywine Creek along the Coatesville study reach. The largest contributors to this base-flow nutrient load were likely two wastewater-treatment facilities adjacent to the study reach. During stormflow sampling in 2003, 480 pounds of phosphorus was shown to enter the West Branch Brandywine Creek along the Coatesville study reach. Data, along with other research, indicated the largest contributor to this stormflow nutrient load was likely remobilized sediment originating from a large dam impoundment. These elevated nutrient concentrations were considered sufficient to promote accelerated aquatic growth along the reach.

Data collected in 2003 showed that wastewater constituents entered the West Branch Brandywine Creek largely from urban storm-sewer systems. Samples from the primary storm sewer for the city of Coatesville had detections for 20 of 69 wastewater constituents. These constituents included both strong and weak indicators of fecal contamination and generally indicated the storm-sewer system along the Coatesville study reach was a likely source of fecal-indicator bacteria and fecal contamination under base-flow conditions. By comparison, 5 constituents were detected in samples from the upstream end of the reach, and 10 constituents were detected in samples from the downstream end of the reach. During stormflow, numbers of detections were similar along the entire length of the study reach-five in samples from the upstream end, eight in samples from the center of the reach, and seven in samples from the downstream end of the reach. These data indicate that point sources (such as culverts and pipes, septic systems, and wastewater-treatment facilities) are not likely the origin of bacteria contamination during stormflow. The bacteria concentrations observed during stormflow events probably result from remobilized sessile bacteria stored within fluvial sediments. In this case, these bacteria should not be considered indicators of current fecal contamination.

Impervious surfaces were found to increase bacteria concentrations along the West Branch Brandywine Creek because contaminated runoff from impervious areas generally flows into, and is concentrated within, the confines of the local storm-sewer system. During 2002, storm-sewer outfalls draining impervious areas were associated with all major locations of elevated bacterial concentrations (greater than 1,200 colonies per gram of saturated sediment) in fluvial sediments. During 2003, wetter conditions and overall bacteria concentrations higher than in 2002 resulted in point sources of bacterial contamination becoming less pronounced; however, the storm-sewer system, draining adjacent impervious areas, was still observed to be the primary source of bacteria along the reach. Where stormwater and (or) other runoff from these areas was allowed to infiltrate and (or) flow through wetland and riparian buffers, bacteria concentrations were not observed to be elevated above background levels commonly observed throughout similar areas of the same reach.

Two run-of-the-river dams along the Coatesville study reach were evaluated for their effects on observed bacterial concentrations. These dams were shown to have greater or lesser effects on bacterial concentrations depending on the size of the structure and the capacity of the structure to impede flows. The smaller upstream dam had an approximate height of 3 feet and showed little observed effect on measured turbidity values; these data indicated that the dam did not effectively impede the flow of water or sediment within the West Branch Brandywine Creek. Consequently, this small dam did not show any observed effect on bacterial concentrations either upstream or downstream of the structure. The larger dam, near the middle of the reach, had an approximate height of 20 feet and showed greater effects on both turbidity and bacteria concentrations. The capacity of the larger dam to impede flows, combined with nutrients entering the reach, resulted in increased biologic activity throughout the impoundment area. Within this larger impoundment, enterococcus bacteria populations were observed to decrease sharply and E. coli bacteria populations were observed to increase sharply as flow approached the dam crest. All bacteria levels were then observed to drop to background levels, in both the water column and fluvial sediment, immediately downstream from the dam crest. Additional study is required to determine the cause for this rapid die off.

Turbidity was assessed as a potential surrogate for E. coli bacteria. Regression analysis indicated higher turbidity levels usually can indicate higher concentrations of bacteria (R2 = 0.67), but the relation was too sporadic on the West Branch Brandywine Creek to use turbidity as a surrogate for estimated bacteria concentrations. Evaluation of data from individual base-flow and stormflow events resulted in variable and generally poor statistical relations between E. coli bacteria and turbidity (R2 values ranged from 0.02 to 0.94).

Sandbag samplers were used in 2003 to determine their suitability for the assessment of fecal contamination. Sandbag samplers rely on the ability of bacteria to attach to surfaces and use the larger sessile bacteria populations instead of the more commonly used planktonic bacteria populations. E. coli bacteria concentrations observed in the sandbag samplers, after 1 week in place, were similar to those found within natural sediments collected concurrently. Enterococcus bacteria concentrations within the same sandbag samplers were not similar, and were generally lower, than those observed within the natural sediments. This discrepancy was likely because sand within the samplers was sieved to a size that was likely too coarse for enterococcus bacteria to persist.

Optical-brightener samplers were installed along with each sandbag sampler. Optical brighteners are additives used in common household detergents; therefore, detection of optical brighteners, along with elevated fecal-indicator bacteria concentrations, strongly indicates a link to humans. Positive results for optical brighteners were detected only at the outfalls of two sewage-treatment facilities; because of treatment of the effluent from these facilities, these samples did not have elevated bacteria concentrations. The lack of additional positive results was largely because this method is not sensitive to low concentrations of optical brighteners.

Table of Contents

Abstract
Introduction
     Purpose and Scope
     Description of Study Area
          Coatesville Study Reach
          Wagontown Study Reach
     Previous Studies
Methods of Sample Collection and Analysis
     Collection and Analysis of Bacteria Within the Water Column
     Collection and Analysis of Bacteria Within Fluvial and Artificial Sediments
     Field Chemistry     
     Water-Quality Samples
     Quality Assurance / Quality Control
Relation of Bacteria to the Fluvial Environment in the West Branch Brandywine Creek and
Its Tributaries     
     Human and Environmental Factors Potentially Affecting Concentrations of Bacteria in the
          West Branch Brandywine Creek and Its Tributaries
          Particle Distribution of Sediments
          Climatic Conditions
          Aquatic Growth in the Water Column and Sediments
          Field Chemistry     
          Nutrients and Wastewater Constituents
          Impervious Surfaces
          Sediment and Soil Filtration
          Dams
          Turbidity
     Potential Sources of Bacteria on Selected Reaches of the West Branch Brandywine Creek
          Sediment
          Storm Sewers
          Bacterial Regrowth
Protocols for Assessment of Fecal Contamination Using Sandbag Samplers and Optical-Brightener
Monitoring
     Description of Method and Equipment
     Results From Trial Installation of Sandbag and Optical Brightener Samplers
     Application of Method for Assessment of Escherichia Coli and (or) Enterococci Bacteria
          Construction of Samplers
          Installation of Samplers
          Analysis
               Natural Sediment and Sandbag Samplers
               Optical-Brightener Samplers
          Equipment List for Construction and Installation of Samplers
     Limitations of Samplers
Limitations of the Investigation
Summary and Conclusions
Acknowledgments
References Cited
Appendixes
Glossary

Figures

1. Map showing study area on West Branch Brandywine Creek, Chester County, Pennsylvania
2. Photograph showing headwall of 1929 bridge that previously spanned Gibbons Run, Coatesville study reach
3-6.      Maps showing:
     3. Base-flow sampling sites on the Coatesville study reach
     4. Stormflow sampling sites on the Coatesville study reach
     5. Sandbag sampling sites on the Coatesville study reach
     6. Base-flow sampling sites on the Wagontown study reach
7-10.      Photographs showing:
     7. U.S. Geological Survey personnel collecting bacteria sample from the water column in the West Branch Brandywine Creek, Chester County, Pennsylvania
     8. U.S. Geological Survey personnel collecting bacteria sample from fluvial sediments in the West Branch Brandywine Creek, Chester County, Pennsylvania
     9. U.S. Geological Survey personnel recording field-chemical measurements in Coatesville study reach, 2002, West Branch Brandywine Creek, Chester County, Pennsylvania
     10. Sand-gauge card used for visual estimation of sediment particle size
11-12.      Boxplots of:
     Distribution of Escherichia coli in fluvial sediment, by particle-size range
     Distribution of enterococci in fluvial sediment, by particle-size range
13-19.      Graphs showing:
     13. Escherichia coli in fluvial sediment, Coatesville study reach, 2002 and 2003
     14. Escherichia coli in the water column, Coatesville study reach, 2002 and 2003
     15. Escherichia coli and turbidity in the water column during base-flow conditions, Coatesville study reach, 2002
     16. Escherichia coli and turbidity in the water column during base-flow conditions, Coatesville study reach, 2003
     17. Enterococci and turbidity in the water column during base-flow conditions, Coatesville study reach, 2003
     18. Specific conductance, Coatesville study reach, September 9-12, 2002
     19. Specific conductance, Coatesville study reach, July 7-9, 2003
20.Map showing nutrient and wastewater-constituent sampling sites, Coatesville study reach West Branch Brandywine Creek, Chester County, Pennsylvania
21-26.      Graphs showing:
     21. Escherichia coli in fluvial sediment during base-flow conditions, Coatesville study
reach, 2002
     22. Escherichia coli in fluvial sediment during base-flow conditions, Coatesville study reach, 2003
     23. Escherichia coli and enterococci concentrations in the water column during base-flow conditions, Wagontown study reach, 2003
     24. Escherichia coli and enterococci concentrations in the water column during base-flow
conditions, Coatesville study reach, 2003
     25. Fecal coliform and fecal streptococci concentrations in fluvial sediment during base-flow conditions, Wagontown study reach, 2002
     26. Escherichia coli and enterococci concentrations in fluvial sediment during base-flow
conditions, Wagontown study reach, 2003
27-29.      Photographs showing:
     27. Dam at station 0 on Coatesville study reach, West Branch Brandywine Creek
     28. Dam at station 2,500 on Coatesville study reach, West Branch Brandywine Creek
     29. Dam at station 10,905 (dam #4) on Coatesville study reach, West Branch Brandywine Creek
30-38.      Graphs showing:
     30. Escherichia coli and turbidity in the water column during base-flow conditions, Coatesville study reach, 2002
     31. Escherichia coli and turbidity in the water column during base-flow conditions, Coatesville study reach, 2003
     32. Enterococci and turbidity in the water column during base-flow conditions, Coatesville study reach, 2003
     33. Relation between base flow turbidity and Escherichia coli bacteria within Coatesville study reach, 2002
     34. Relation between base flow turbidity and Escherichia coli bacteria within Coatesville study reach, 2003
     35. Relation between stormflow turbidity and Escherichia coli bacteria within Coatesville study reach, 2002
     36. Relation between stormflow turbidity and Escherichia coli bacteria within Coatesville study reach, 2003
     37. Relation between turbidity and Escherichia coli bacteria within Coatesville and
Wagontown study reaches, 2002-03
     38. Relation between stormflow turbidity and Escherichia coli bacteria at USGS streamflow-gaging station 01480617
39. Photograph showing sandbag and optical-brightener samplers
40-41.     Graphs showing:
     40. Escherichia coli concentrations in sandbags and natural fluvial sediment, Coatesville study reach, 2003
     41. Enterococci concentrations in sandbags and natural fluvial sediment, Coatesville study reach, 2003
42. Photograph showing mounting hardware for sandbag and optical-brightener samplers

Tables

1. Sample-processing protocols for collected bacteria samples from the West Branch
Brandywine Creek, Chester County, Pennsylvania


View the full report in PDF 5.38 MB

This report is available online in Portable Document Format (PDF). If you do not have the Adobe Acrobat PDF Reader, it is available for free download from Adobe Systems Incorporated.

For more information about USGS activities in Pennsylvania contact:
Director
USGS Pennsylvania Water Science Center
215 Limekiln Road
New Cumberland, Pennsylvania 17070
Telephone: (717) 730-6960
Fax: (717) 730-6997
or access the USGS Water Resources of Pennsylvania home page at:
http://pa.water.usgs.gov/.


U.S. Department of the Interior, U.S. Geological Survey
Persistent URL:
Page Contact Information: Contact USGS
Last modified: Thursday, December 01 2016, 06:39:55 PM
FirstGov button  Take Pride in America button