United States Office of Research EPA/600/R-08/110
Environmental Protection and Development September, 2008
Agency Washington DC 20460
Evaluation of Receiving
Water Improvements from
Stream Restoration
(Accotink Creek, Fairfax
City, VA)
*2tt
^ £\"*-r U.S. Environmental Protection Agency
1 Office of Research and Development
" National Risk Management Research Laboratory
-------
THIS PAGE INTENTIONALLY LEFT BLANK
-------
EPA/600/R-08/110
September, 2008
Evaluation of Receiving Water
Improvements from Stream Restoration
(Accotink Creek, Fairfax City, VA)
By
Ariamalar Selvakumar
Thomas P. O'Connor
Urban Watershed Management Branch
Water Supply and Water Resources Division
National Risk Management Research Laboratory
Edison, NJ 08837
Scott Struck
Former Federal Post-doc with U.S. EPA
(Currently with Tetra Tech, Golden, CO 80401)
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
Notice
The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development
performed and managed the research described in this report. It has been subjected to the Agency's peer
and administrative review and has been approved for publication as an EPA document. Any opinions
expressed in this report are those of the author and do not, necessarily, reflect the official positions and
policies of the EPA. Any mention of products or trade names does not constitute recommendation for use
by the EPA.
-------
Abstract
Installation of best management practices (BMPs) in watersheds or streams is widely used as a
means of reducing, eliminating, or controlling the input of human-based physical, chemical, or
hydrologic stressors to those systems. Although BMPs may be effective in managing a particular
stressor, installation of stream bank and channel restoration alone may not fully restore nor fully
protect the biological condition of the receiving waterbody since multiple stressors are known to
affect aquatic biota.
The National Risk Management Research Laboratory (NRMRL), part of U.S. Environmental
Protection Agency's (U.S. EPA) Office of Research and Development (ORD) evaluated the
effectiveness of stream bank and channel restoration as a means of improving in-stream water
quality and biological habitat in Accotink Creek, Fairfax City, Virginia using discrete sampling
and continuous monitoring techniques before and after stream restoration. Continuous water
quality monitoring showed that temperature of the creek changed with season and wet weather
flow events with the highest temperature observed in summer (e.g., July). Specific conductivity
was higher in winter due to street salting while pH stayed close to neutral year around. There were
no statistically significant differences in other chemical constituents and bacteriological indicator
organisms before and after restoration as well as upstream and downstream of the restoration.
Macroinvertebrate indices such as Virginia Stream Condition Index (VASCI) and Hilsenhoff
Biotic Index (HBI) and Emphemeroptera, Plecoptera, Trichoptera (EPT) taxa showed a general
improvement in biological quality between pre- and post-restoration. The differences were
statistically significant for VASCI, HBI, and EPT taxa. However, they were all below the
impairment level, indicating poor water quality conditions. The United States Geological Survey
(USGS) also performed continuous monitoring and discrete sampling under an Interagency
Agreement (IAG No. DW-14-922064010) to U.S. EPA. Their monitoring and predictive
equations showed a stronger relationship between turbidity and suspended sediment concentration
than turbidity and E. coli. There was no change in the results derived from the predictive
equations before and after restoration, likely because improved conditions have yet to be realized
or stream restoration did not reduce sediment transport. A pebble count analysis also suggested
that very little has changed in the restoration reach.
These results indicate that stream restoration alone may have little effect on improving the
conditions of in-stream water quality and biological habitat. It should be recognized that
improvement may not be reflected in a two year post-restoration period and that additional
monitoring is needed. Also, reduction of stormwater runoff volumes and associated pollutants of
concern should be addressed in the watershed through source control and stormwater retrofits to
achieve desired biological outcomes.
in
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between
human activities and the ability of natural systems to support and nurture life. To meet this
mandate, EPA's research program is providing data and technical support for solving
environmental problems today and building a science knowledge base necessary to manage our
ecological resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks from
pollution that threaten human health and the environment. The focus of the Laboratory's research
program is on methods and their cost-effectiveness for prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water systems;
remediation of contaminated sites, sediments and ground water; prevention and control of indoor
air pollution; and restoration of ecosystems. NRMRL collaborates with both public and private
sector partners to foster technologies that reduce the cost of compliance and to anticipate emerging
problems. NRMRL's research provides solutions to environmental problems by: developing and
promoting technologies that protect and improve the environment; advancing scientific and
engineering information to support regulatory and policy decisions; and providing the technical
support and information transfer to ensure implementation of environmental regulations and
strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
Sally C. Gutierrez, Director
National Risk Management Research Laboratory
IV
-------
Contents
Notice ii
Abstract Hi
Foreword iv
Contents v
List of Figures vii
List of Tables viii
Acronyms and Abbreviations ix
Acknowledgements x
Executive Summary xi
Chapter 1 Introduction 1-1
Background 1-1
Objectives 1-4
Project Partners 1-5
Chapter 2 Project Description 2-1
Site Location and Background 2-1
Sampling and Monitoring 2-3
Chapter 3 U.S. EPA Sampling and Monitoring 3-1
Sampling and Monitoring 3-1
Continuous Water Quality Monitoring 3-1
Discrete Water Quality Sampling 3-2
Macroinvertebrate Sampling 3-3
Results 3-5
Continuous Water Quality Monitoring 3-5
Discrete Water Quality Sampling 3-11
Macroinvertebrate Sampling 3-15
Stream Channel Cross Sections 3-21
Pebble Count 3-22
Chapter 4 USGS Sampling and Monitoring 4-1
Background 4-1
Continuous Water Quality Monitoring 4-3
Discrete Water Quality Sampling 4-4
Results 4-4
Continuous Water Quality Monitoring 4-4
Discrete Water Quality Sampling 4-5
Utility of the Continuous Data for the Prediction of SSC and Bacteria Concentrations 4-10
Patterns in Turbidity Concentrations Before, During, and After Restoration 4-10
Chapter 5 Conclusions 5-1
Summary 5-2
-------
Recommendations for Further Action 5-3
Chapter 6 References 6-1
VI
-------
List of Figures
Figure 2-1. Accotink Creek areal map showing major highways and lOOyr flood plain 2-3
Figure 2-2. Three photos of stream restoration 2-4
Figure 3 -1. Water quality sampling and continuous monitoring stations (indicated by blue marker) 3-2
Figure 3-2. U.S. EPA's macroinvertebrate sampling locations (indicated by orange marker) 3-4
Figure 3-3. Continuous water quality monitoring at Site WQ1 for pH 3-6
Figure 3 -4. Continuous water quality monitoring at Site WQ1 for temperature and depth 3-6
Figure 3 -5. Continuous water quality monitoring at Site WQ 1 for conductivity and turbidity 3-7
Figure 3-6. Continuous water quality monitoring at Site WQ2 for pH 3-8
Figure 3 -7. Continuous water quality monitoring at Site WQ2 for temperature and conductivity 3-8
Figure 3 -8. Continuous water quality monitoring at Site WQ2 for level and estimated flow 3-9
Figure 3-9. Continuous water quality monitoring at Site WQ3 for pH 3-9
Figure 3-10. Continuous water quality monitoring at Site WQ3 for temperature and conductivity 3-10
Figure 3-11. Continuous water quality monitoring at Site WQ3 for level and estimated flow 3-10
Figure 3-12. Relationship between level and rain at Site WQ3 3-11
Figure 3-13. Average of wet weather concentrations of COD and SS before and after restoration 3-13
Figure 3-14. Average of wet weather concentrations of TKN, NH3, and TPO43" before and after restoration
3-13
Figure 3-15. Summary of indicator organism concentrations before and after restoration 3-15
Figure 3-16. Virginia Stream Condition Index (VASCI) scores for before and after the Accotink Creek
stream restoration 3-18
Figure 3-17. Hilsenhoff Biotic Index (HBI) scores for before and after the Accotink Creek stream
restoration 3-18
Figure 3-18. Differences in VASCI between post and pre-restoration 3-19
Figure 3-19. Differences in HBI between post and pre-restoration 3-19
Figure 3 -20. Representative depth profiles before and after restoration at an upstream location 3-21
Figure 3-21. Representative depth profiles before and after restoration at a restored location 3-22
Figure 3 -22. Pebble count results at Site 5 - Ranger Road (upstream of restoration) 3 -24
Figure 3-23. Pebble count results at Site D - downstream of Old Lee Highway (downstream of restoration)
3-24
Figure 3-24. Pebble count results at Site A - upstream of Lee Highway (upstream of restoration) 3-25
Figure 3-25. Pebble count results at Site C - upstream of Old Lee Highway (restoration reach) 3-25
Figure 3-26. Pebble count results at Site B - below Lee Highway at Harley Dealer (restoration reach) ..3-26
Figure 4-1. Example of continuous water quality data determined by sensor technology 4-2
Figure 4-2. Example of correlation between turbidity and fecal coliform concentration 4-2
Figure 4-3. Continuous water quality monitoring by USGS at Old Lee Highway 4-5
Figure 4-4. Graph of predicted vs. observed for E. coll 4-6
Figure 4-5. Graph of observed E. colivs. residual 4-6
Figure 4-6. Graph of predicted vs. observed for suspended sediment 4-7
Figure 4-7. Graph of observed suspended sediments vs. residual 4-7
Figure 4-8. Graph of observed SSC vs. predicted SSC (based on Turbidity Only) 4-8
vn
-------
Figure 4-9. Graph of SSC showing pre- and post-restoration samples 4-9
Figure 4-10. Graph of E. coll showing pre- and post-restoration samples 4-9
Figure 4-11. Distribution of Accotink Creek turbidity values before, during, and after restoration 4-11
List of Tables
Table 1 -1. Maj or pollutants (stressors) in stormwater runoff and their effects on streams 1-2
Table 3-1. Results of water quality analysis (physical and chemical constituents) 3-12
Table 3-2. Results of water quality analysis (indicator organisms) 3-14
Table 3-3. Results of macroinvertebrate data 3-16
Table 3-4. Average macroinvertebrate indices and EPT taxa families before and after restoration 3-17
Table 3-5. Total number of macroinvertebrates 3-20
Table 3-6. Results of pebble counts 3-23
Vlll
-------
Acronyms and Abbreviations
ANOVA Analysis of Variance
APHA American Public Health Association
BMP Best Management Practice
COD Chemical Oxygen Demand
CPU Coliform Forming Unit
CWA Clean Water Act
CWP Center for Watershed Protection
DNA Deoxyribonucleic Acid
DPWES Department of Public Works and Environmental Service
EC E. coli
EN Enterococci
EPA U.S. Environmental Protection Agency
EPT Emphemeroptera, Plecoptera, Trichoptera
FC Fecal Coliforms
HBI HilsenhoffBiotic Index
IAG Interagency Agreement
LogTurb Log of Turbidity
MS4 Municipal Separate Storm Sewer System
MSL Mean Sea Level
NPDES National Pollutant Discharge Elimination System
NRMRL National Risk Management Research Laboratory
NTU Nephelometric Turbidity Unit
ORD Office of Research and Development
SOP Standard Operating Procedures
SS Suspended Solids
SSC Suspended Sediment Concentration
TKN Total Kjeldahl Nitrogen
TMDL Total Maximum Daily Loads
U.S. EPA U.S. Environmental Protection Agency
USGS United States Geological Survey
UWMB Urban Watershed Management Branch
UWRF Urban Watershed Research Facility
VADEQ Virginia Department of Environmental Quality
VASCI Virginia Stream Condition Index
WT Water Temperature
YSI Yellow Springs Instruments
-------
Acknowledgements
An undertaking of this type requires the dedication and cooperation of many individuals. The
technical direction and coordination for this project was provided by the technical project team of
the Urban Watershed Management Branch. Many members of the Branch assisted in making this
product available to the public. Anthony Tafuri and Michael Borst performed reviews of this
report. Carolyn Esposito reviewed the quality assurance project plan and this report.
Jim Keating of U.S. EPA's Office of Water, Ron Landy of U.S. EPA Region 3, and Wayne Green
of Green Management Ltd. peer reviewed this report. Paula Estornell, Charlie App, Ann Carkoff
from U.S. EPA Region 3 gave support throughout the project. Greg Pond of U.S. EPA Region 3's
Wheeling Laboratory conducted the macroinvertebrate analysis and reviewed the section
pertaining to macroinvertebrates. Jeanne Classen of Virginia Department of Environmental
Quality assisted with continuous operation of field equipment. Christian Jones of George Mason
University and his graduate student, lyad Aburdeineh, assisted with wet weather sampling. Adrian
Fremont, Engineer for the City of Fairfax was instrumental in finding the location and helped us in
many other ways. Ken Hyer of USGS performed continuous monitoring and discrete sampling
under an Interagency Agreement (TAG No. DW-14-922064010) to U.S. EPA.
Most monitoring, sampling and laboratory analysis was performed by personnel from PARS
Environmental, Inc., the operating contractor of the Urban Watershed Research Facility (UWRF)
under U.S. EPA contract number EP-C-04-064. Most notably, Christa Casciolini, Michael
Cerrato, John Lapinski, Yogesh Parikh, and Clarence Smith should be recognized for their
contribution to the completion of the sampling and analyses.
-------
Executive Summary
Increased urbanization results in a larger percentage of connected impervious areas and can
contribute large quantities of stormwater runoff and significant quantities of debris (litter and
floatables) and pollutants (oils, microorganisms, sediments, nutrients, organic matter, and heavy
metals) to receiving waters. Land-use practices directly impact urban streams. Stream flows in
urbanized watersheds increase during wet weather events as a function of impervious area and
can result in degradation of the natural stream channel morphology affecting the physical,
chemical, and biological integrity of the stream. Stream bank erosion, which also increases with
increased stream flow, can lead to bank instability, property loss, infrastructure damage, and
increased sediment loading to the stream. Increased sediment loads may lead to water quality
degradation downstream and have negative impacts on fish, benthic invertebrates, and other
aquatic life. To improve water quality in urban and suburban areas, watershed managers often
incorporate best management practices (BMPs) to reduce the quantity of runoff and to minimize
pollutants and other stressors contained in stormwater runoff.
This study addresses the effectiveness of stream bank and channel restoration techniques on
improving benthic macroinvertebrate indices and in-stream water quality within an urban
watershed. The project monitored the effects of restoring 1,800 linear ft (550 m) of degraded
stream channel in the North Fork of Accotink Creek in the City of Fairfax, Virginia.
Restoration, which was completed in June 2006, included planting native plant materials along
the stream and installation of bioengineering structures to stabilize the stream channel and bank.
These actions were intended to restore the stream channel to a stable condition, thereby reducing
stream bank erosion and sediment loads in the stream. Monitoring was performed before and
after the restoration by both U.S. Environmental Protection Agency (U.S. EPA) and United
States Geological Survey (USGS).
Chapter 1 provides an introduction and justification of the project. Chapter 2 describes the
project background and site location. Chapters 3 and 4 detail water quality monitoring, sampling
and analysis conducted by U.S. EPA and USGS, respectively. Chapter 3 also summarizes the
results of macroinvertebrates sampling conducted by U.S. EPA. Finally, Chapter 5 provides the
conclusions drawn from this project.
XI
-------
Chapter 1 Introduction
Background
Since the inception of the Clean Water Act (CWA) in 1972, the United States has made great
efforts in restoring and preserving the physical, chemical, and biological integrity of nation's
waters. However, nearly half of the nation's assessed surface waters remain incapable of
maintaining water quality adequate for supporting one or more designated uses, i.e., recreational
swimming or drinking water supply (U.S. EPA, 2007). One of the top causes of river and stream
impairment is sediment or siltation. The National Water Quality Inventory 2000 Report (U.S.
EPA, 2002a) estimated that about 30% of identified cases of water quality impairment are
attributable to stormwater runoff. Since its formation, the U.S. EPA has established several
regulatory programs to address the various point and non-point sources; however, the laws the
Agency implements have led to less regulatory emphasis on non-point source pollution. In 1987,
CWA was amended to establish National Pollutant Discharge Elimination System (NPDES)
stormwater discharge requirements. To implement these requirements, the U.S. EPA published
the "Phase I" stormwater permit program on November 16, 1990 to address certain stormwater
discharge categories associated with 10 categories of industrial activity, construction and
development activities disturbing more than five acres, and medium and large municipal separate
storm sewer systems (MS4s) with populations greater than 100,000. In December 8, 1999, the
U.S. EPA promulgated "Phase II" stormwater regulations expanding the list to include small
MS4s located in "urbanized areas" as defined by the Bureau of Census, and those small MS4s
located outside of a urban area that are designated by NPDES permitting authorities with
populations fewer than 100,000 and construction sites disturbing more than one acre and less
than five acres of land (U.S. EPA, 2000).
Land development and urbanization impact receiving streams by adversely altering watershed
hydrology in several ways. The conversion of natural forested or grassland areas to impervious
surfaces results in an increased volume of surface runoff because less water is able to infiltrate
into the ground, i.e., more water enters the receiving water by surface runoff than via
groundwater pathways. Examples of impervious surfaces in an urban area include roadway
surfaces, parking lots, and rooftops. Surface runoff is also directed to the stream channel more
quickly than water that infiltrates the soil. The routing to the receiving stream is expedited by
curbs, gutters, and stormwater pipes, which convey water rapidly from impervious surfaces to
1-1
-------
the stream. Consequently, stream flows in urbanized watersheds increase in magnitude during
wet weather flows as a function of impervious area (Schueler, 1995).
Natural streams follow predictable meandering patterns, which dissipates energy and minimize
scouring of the streambed and banks. Increased stream flows during wet weather impact the
natural stream channel morphology, which affects the physical, chemical, and biological
integrity of the stream (Natural Resources Conservation Service, 1998). The increased flows
alter the stream channels by widening the cross-sectional area between stream banks and down-
cutting of the stream bed. This, in turn, triggers a cycle of streambank erosion and habitat
degradation (Schueler, 1994). Streambank erosion can lead to bank instability and increased
sediment loading to the stream. The increased sediment load may cause channel instability and
water quality degradation and have negative impacts on fish, benthic invertebrates, and other
aquatic life in the stream. Channel instability leads to the loss of in-stream habitat structures,
such as the loss of pool and riffle sequences. Klein (1979) noted that macroinvertebrate diversity
dropped sharply in Maryland urban streams as a result of an increase in imperviousness in the
catchment areas of streams. Sensitive aquatic insects such as stoneflies, mayflies, and caddisflies
are replaced by species tolerant to pollution and hydrologic stress such as chironomids, tubificid
worms, amphipods, and snails. In addition to the physical damage done to the streams,
stormwater runoff may bring many types of pollutants which have the potential to significantly
impact the biological community. Table 1-1 lists major pollutants in stormwater runoff and their
effects on streams.
Table 1-1. Major pollutants (stressors) in stormwater runoff and their effects on streams
Stressor
Sediment
Nutrients (Nitrogen
and Phosphorus)
Organic Matter
Bacteria
Oil and Grease
Heavy Metals
Temperature
Potential Sources
Construction sites, winter road sand, in-stream
erosion, unvegetated lots (bare soils)
Landscaping practices (application of
fertilizers), animal wastes, leaks from sanitary
sewers and septic systems, air deposition
Leaks from sanitary sewers and septic
systems, garbage, yard waste
Leaking sanitary systems, garbage, pet waste,
homeless populations, animals
Vehicle traffic, maintenance and fueling
activities, leaks and spills, runoff from areas
with industrial land use
Automobiles, paints, preservatives, motor oil
Runoff from hot impervious surfaces, water
stored in shallow unshaded ponds and
impoundments, removal of natural vegetation
(tree canopy), industrial discharges
Environmental Effect
Increases turbidity, disturbs aquatic
and benthic habitat, embeds substrate
Stimulates algae growth, depletes
dissolved oxygen, accelerates
eutrophication
Depletes dissolved oxygen, impacts
life in the surface waters
Affects recreational uses and aquatic
life, imposes health risks
Creates slicks and disrupts water/air
exchanges, stresses stream biota
Toxic to some aquatic life,
accumulates in aquatic animals
Impacts water body's ability to
support certain fish and aquatic
organisms, reduces capacity for
dissolved oxygen
The use of effective best management practices (BMPs), such as stream restoration is one way to
mitigate these impacts of urbanization and increased impervious cover. Stream restoration
projects are popular in the United States as a result of public awareness concerning the
connection between stream health and community health. Many projects aim to protect
infrastructure or to improve aesthetics. Communities spend millions of dollars annually on
watershed restoration and stream habitat improvement, yet little is known about the effects of
1-2
-------
stream restoration as post-restoration evaluation and monitoring are not that common. Even if
they are monitored, they are not monitored for a long time (i.e., longer than 1-year). This may be
due to lack of funding or other reasons. Laeser and Stanley (2004) have noted no detectable
changes in nutrient concentrations in association with streambank improvement programs. It is
important to conduct post-restoration evaluation to understand the effects of restoration, which
will lead to better planning of restoration projects.
Bohn and Kershner (2002) pointed out that aquatic habitat restoration must be implemented at a
watershed scale to be effective. Unless larger scale watershed issues are addressed in restoration
planning, the current practice of direct structural modification of channels at the site level is
unlikely to reverse aquatic population densities. Many habitats result from a change in
environment; attempts to fix them at a particular point in space or time fail to recognize that
stream channels are dynamic and that high quality habitats are a product of this dynamism
e^a/., 1996).
In the past decade, hundreds of millions of dollars have been invested in the implementation of
BMPs in watersheds or streams as a means of reducing, eliminating, or otherwise controlling the
input of human-based physical, chemical, or hydrological stressors to these systems. In Virginia
alone, over $60 million have been spent on agricultural BMP implementation activities from
2000 to 2006 with the explicit goal of improving water quality (Virginia Department of
Conservation and Recreation, 2007). Earlier research conducted by the U.S. EPA and the data in
the International BMP Database (www.bmpdatabase.org) have demonstrated that BMPs can be
effective at the pilot-scale and the field-scale (Strecker et a/., 2002; Struck et a/., 2006).
However, less information is available to document the effectiveness of these BMPs at the
subwatershed to watershed scale, which is precisely the scale at which water quality compliance
and water quality improvements are typically judged. Because of the costs associated with the
implementation of these BMPs, federal, state, and local agencies are asking:
• Are the implementation activities working?
• How long will it take for the BMPs to work?
• Are there more time-efficient, cost-effective methods for detecting these improvements?
Answers are needed to these questions to support the development of watershed implementation
plans, to motivate stakeholders to implement BMPs, and to ensure the vitality of the cost-share
programs that have supplemented the cost of implementing these BMPs.
Unfortunately, few studies have been able to provide rigorous evidence of improvements in
water quality following implementation activities. This inability to detect statistically significant
improvement in water quality conditions occurs for several reasons.
First, numerous individual samples are often needed every year to provide sufficient data,
depending on parameters, with which to statistically determine trends in water quality
parameters. The cost associated with the collection of these samples is significant and often
limits monitoring after BMP implementation.
1-3
-------
Second, environmental factors cause extensive variability (noise) in concentrations of monitored
parameters and confound attempts to quantify improvements (signal) that are related to BMP
activities. Flow is the single greatest of these confounding environmental factors because even
small changes in flow typically are associated with measurable changes in nutrient, sediment and
bacterial concentrations. Additional confounding variables include rainfall rate, rainfall amount,
and seasonality.
Lastly, lag times between implementation of BMPs and corresponding improvement in water
quality may be considerable, depending on the sensitivity of the parameter, scale of management
compared to runoff area, and the current and future condition of the watershed (current level of
degradation and future changes to the watershed). For example, when using a biological
indicator such as fish populations, 10 years or more of monitoring is often required to detect a
response to restoration because of the large inter-annual variability in abundance of juvenile and
adult salmonids (Bisson et a/., 1992; Reeves et a/., 1997). Roni et al. (2002) asserted that
biological response to various restoration techniques is the ultimate measure of restoration
effectiveness, but drawing conclusions about the biological effectiveness of various techniques
has been difficult and has hampered efforts to provide scientific guidance on restoration
activities.
Objectives
The objective of this project was to investigate the effectiveness of BMPs, specifically, stream
restoration techniques, to improve biological and in-stream water quality in an impaired stream
in an urban watershed. This objective was achieved by continuous monitoring of water quality
and by collecting physical, chemical, and biological data in the receiving stream before and after
stream restoration.
This project tested the following three hypotheses at the 90% level of confidence:
Hypothesis #1: The quality of the water, as measured by physical, chemical, and
biological parameters, in the stream before and after stream restoration
will be different.
Hypothesis #2: The type and number of macroinvertebrate community in the stream
before and after stream restoration will be different.
Hypothesis #3: The physical habitat parameters in the stream before and after stream
restoration will be different.
Accotink Creek in Fairfax City, Virginia was selected as the project site mainly because the City
of Fairfax was proceeding with restoration of 1,800 linear ft (550 m) of degraded stream channel
in the North Fork of Accotink Creek from Lee Highway to Old Lee Highway. In-stream samples
were collected and analyzed for physical, chemical, and biological (macroinvertebrates, bacterial
indicators) parameters before and after restoration to document the changes in-stream quality as
a result of the restoration.
1-4
-------
Project Partners
This project was a joint effort between U.S. EPA ORD and U.S. EPA Region 3. Additional
cooperators were the Center for Watershed Protection (CWP) under a cooperative agreement
with U.S. EPA Region 3 and the USGS under an interagency agreement with U.S. EPA ORD.
1-5
-------
Chapter 2 Project Description
Site Location and Background
The Accotink Creek watershed covers about 3,400 acres (5.3 square miles) of drainage area
within the Fairfax City limits. There are about 22,000 people living in the city. The majority of
the soils in the city is well-drained with moderately coarse-texture and moderate infiltration
rates. Percentage of imperviousness is about 35% (DPWES, 2001). Elevation in the city
watershed ranged from 425 ft (130 m) above mean sea level (MSL) at its highest point to 285 ft
(87 m) above MSL at the point Accotink Creek flows out of the city. Recent land development
and redevelopment projects have included provisions for stormwater management practices that
effectively slow and distribute high stormwater flows over a period of time, thereby reducing
erosion in the streams (The Louis Berger Group, 2005).
The Accotink Creek headwater watershed has uncontrolled urban runoff that has resulted in the
deepening and widening of the creek's channel, sediment removal from the stream reach and
deposition downstream, and streambank instability. The Creek and its tributaries within the city
are important natural features that provide recreational and aesthetic values that enhance the
quality of life in the city. The headwaters of Accotink Creek originate within the City of Fairfax
and flow southeast through Fairfax County to its confluence with Potomac River at Gunston
Cove, which flows into the Chesapeake Bay.
According to Virginia Water Quality Standards, "all state waters are designated for the following
uses: recreational uses; the propagation and growth of a balanced indigenous population of
aquatic life, including game fish, which might be reasonably expected to inhabit them; wildlife;
and the production of edible and marketable natural sources (e.g., fish and shellfish)." Many of
the fish and other aquatic life, which are important for the Creek's viability, began to disappear
when the open areas were developed and paved (Fairfax, 2005). Overall stream health measured
by the physical, biological, and habitat assessment is fair to poor in the majority of the city,
erosion potential remains at a very high level, sedimentation is a problem, and down-cutting
streams threaten city utilities and surrounding property.
2-1
-------
High runoff volume from impervious surfaces is the primary cause of stream degradation in the
Accotink Creek watershed. The amount of stormwater runoff generated under existing
conditions is almost double the runoff that would be generated under 100% forested conditions
(The Louis Berger Group, 2005). The Fairfax County Stream Protection Strategy Baseline Study
conducted by Department of Public Works and Environmental Services (DPWES) concluded
that the benthic macroinvertebrate community health in the Accotink Creek were poor; habitat
conditions were very poor; and fish taxa richness is low (DPWES, 2001).
Point sources do not appear to be an important factor in water quality impairment in the Accotink
Creek watershed. The Creek was listed as impaired on Virginia's 1998 303(d) Total Maximum
Daily Load (TMDL) priority list due to violation of the State's water quality standard for fecal
coliform (VADEQ, 1998). As part of the TMDL study, the U.S. Geological Survey (USGS)
Virginia District conducted DNA fingerprinting, called ribotyping, on fecal coliform samples
from Accotink Creek (downstream of the restoration reach). The dominant bacterial sources
were found to be geese (24%), humans (20%), and dogs (13%). Other sources identified
included ducks, cats, raccoons, sea gulls, cattle, and deer (USGS, 2003).
Impacts of stormwater runoff are common in highly urbanized areas. Changes in land use in the
City of Fairfax, which has grown and developed over the years, affected the stream conditions in
many parts of the city. The city is characterized by commercial and, high- and low-density
residential development that accounts for greater than 60% of land use. Consequently, the city
proactively developed a Watershed Management Plan when faced with major water quantity and
quality problems (The Louis Berger Group, 2005). One cause of poor water quality and stream
degradation, as reported by the plan, was elevated volumes of uncontrolled stormwater runoff
due to directly connected impervious surfaces.
More than 75% of the overall stream health condition assessments (calculated using the physical,
habitat, and biological conditions) performed for the management plan indicated a fair to poor
result. Along with other BMPs, the management plan called for streambank restoration as an
important facet to improve stream conditions. Fairfax chose to focus on areas which stood to
gain the most benefit from the use of BMPs and have attempted to coordinate improvements
with an overall watershed strategy by utilizing regional and holistic approaches where possible.
Restoration of the stream channel of Accotink Creek was necessary to reduce loss of property,
restore public safety, protect infrastructure, stop the destruction of downstream habitat, and
improve aquatic life in Accotink Creek. The city determined that stream restoration was the
most cost effective way to minimize channel erosion (Personal Communication with Adrian
Fremont, City Engineer, City of Fairfax, 2006). Since 1994, the city has been conducting
systematic stream restoration in the Accotink Creek watershed. More than three miles of stream, just
over half of the city's total stream miles, have been restored or stabilized to date. In the spring of
2002, the city completed stream restoration improvements on the North Fork of Accotink Creek
from Stafford Drive to Lee Highway.
The subject of this monitoring project was a more recent stream restoration of a segment of
1,800 linear ft (550 m) of the North Fork of Accotink Creek from Lee Highway to Old Lee
Highway in the City of Fairfax, Fairfax County, Virginia (Figure 2-1). The stream restoration
2-2
-------
liCn ^H^^^ir^frl
Figure 2-1. Accotink Creek areal map showing major highways and lOOyr flood plain
included placing bioengineering structures (coir fiber logs, erosion control fabrics, and live
willow stakes) to prevent erosion and establish deeper rooted vegetation to stabilize the bank.
Rocks were individually placed to divert stream flow from the edge of the channel to the center
of the stream. Rock veins were constructed to reduce slope and form step pools to slow water
velocity. Dense planting and seeding of native vegetation along the stream was done to protect
exposed soils from erosion and sedimentation during heavy rainfall and high flows completed
the channel restoration (Figure 2-2). These actions were intended to restore the stream channel
to a stable condition and reduce streambank erosion thereby reducing sediment loads in the
stream. The construction started in March of 2006 and was completed in June of 2006.
Sampling and Monitoring
The U.S. EPA and USGS carried out continuous monitoring and discrete water quality
monitoring and sampling beginning in December 2005 following standard sampling protocols.
At the designated locations, electronic water quality and quantity monitoring equipment was
installed to monitor pH, temperature, turbidity, conductivity, water depth, and water velocity
continuously. In addition, discrete samples were collected during storm events, with the
objective of attempting to obtain samples at least twice per season. Additionally, dry weather
samples were collected. These samples were analyzed for physical and chemical [i.e., pH,
dissolved oxygen, turbidity, conductivity, temperature, suspended solids (SS), suspended
2-3
-------
Figure 2-2. Three photos of stream restoration
2-4
-------
sediment concentrations (SSC), particle size distribution, and nutrients], and bacteriological (i.e.,
fecal coliform, enterococci, and E. coli) parameters. The results of both U.S. EPA and USGS
monitoring and sampling and analysis are described in detail in Chapters 3 and 4, respectively.
Physical habitat monitoring and biological sampling (including macroinvertebrate sampling)
were conducted three times before restoration to establish the pre-existing condition. Biological
sampling was conducted five times following restoration. U.S. EPA Region 3's Wheeling
Laboratory in West Virginia performed the macroinvertebrate identification, classification, and
enumeration.
2-5
-------
Chapter 3 U.S. EPA Sampling and Monitoring
Sampling and Monitoring
Continuous Water Quality Monitoring
Standard water quality parameters (pH, conductivity, temperature, turbidity) were measured both
upstream and downstream of the restoration from December 2005 to March 2008. This water
quality monitoring enabled the quantification of physical and chemical changes in the receiving
water. Four continuous water quality monitoring stations were deployed (Figure 3-1) - three by
U.S. EPA (WQ1 to WQ3) and one by USGS (WQ4). Water quality monitoring was conducted
continuously except during the restoration period. Area-velocity flow meters combined with
other monitoring probes (American Sigma, Loveland, CO) installed at two selected locations
recorded average flow depth, velocity, water temperature, conductivity, and pH at 15-min
intervals (Figure 3-1; Sampling Stations WQ2 and WQ3). Depth was measured using
differential pressure (bubbler) or pressure transducers. Twin 1 MHz piezoelectric crystals were
used to measure Doppler-based velocity. Internal electronics combined the measured values
using the stream cross-section and computed an associated flow rate. In addition, a YSI (Yellow
Springs Instruments, Yellow Springs, OH) probe placed at the upstream border of the restoration
reach was used to measure water temperature, specific conductivity, turbidity, and pH also at 15-
min intervals (Figure 3-1; Monitoring Station WQ1). All field instrumentations were battery
powered. The instruments were connected to data logging and telemetry equipment that
transferred all data to the U.S. EPA office in Edison, NJ. Following the initial deployment,
approximately monthly maintenance visits were performed on the continuous water quality
monitoring equipment to clean and check the calibration of the sensors. In-field recalibration
was performed during these monthly maintenance visits, as necessary. The sampling and
monitoring locations are shown in Figure 3-1.
3-1
-------
Figure 3-1. Water quality sampling and continuous monitoring stations (indicated by blue marker)
Discrete Water Quality Sampling
Both dry and storm event discrete samples were collected following standard U.S. EPA protocols
from the middle of the water column in approximately the center of the stream flow.
Dry weather conditions were defined as time that was proceeded by at least 72 hours of no or
only trace amounts of precipitation as per NPDES protocol (U.S. EPA, 1992).
Discrete samples were collected in duplicate, at WQ2 and WQ4, to represent water quality above
and below the restored area, respectively. Samples were collected in pre-cleaned, sterile two-
liter bottles by lowering the bottles from the bridge during significant wet weather events or by
hand grab during dry weather or lesser wet weather events. Samples were either shipped by
courier or brought back to the laboratory for analysis at the UWRF in Edison, New Jersey.
The samples were analyzed for SS, chemical oxygen demand (COD), nutrients (total phosphate
(TPO43"), orthophosphate (OPO43"), total Kjeldahl nitrogen (TKN), ammonia (NH3), nitrate
(NCV), and nitrite (N(V)) and bacteriological indicator organisms (fecal coliform, enterococci,
and E. coif). The samples were analyzed following Standard Methods (APHA et al, 1998). All
the analysis was conducted in triplicate.
3-2
-------
For indicator organisms, samples were sequentially diluted with sterile buffered water using
three dilution factors based on previous analyses of similar samples. Dilution factors were
estimated to obtain the method recommended colony count on at least one dilution set.
Sequential dilutions usually used at least 10 mL aliquots and always used at least 5 mL. All
results were volume-normalized to give concentrations in colony forming units (CPU) per 100
mL. Each analytical batch included laboratory blanks and positive controls. Blanks were run
before and after each analytical set. Verification was performed on 10 colonies for each
organism following Standard Methods (APHA et al, 1998). After incubation, the plates were
manually enumerated.
Macroinvertebrate Sampling
Biological integrity above, within, and below the restored area before and after restoration were
evaluated using benthic macroinvertebrate data. Water quality monitoring programs use
macroinvertebrates as indicators of water quality. For example, Collins et al. (2008) reported
that invertebrate community index (ICI) developed by Bennett et al. (2004), which comprised of
10 metrics based on the structure, function, and condition of the taxa collected, is capable of
predicting the biological integrity of the urban streams in Choctawhatchee and Pea River
watersheds in Alabama. Benthic macroinvertebrates are a major component of healthy stream
systems and are an important link in any aquatic food web, forming the core diet of many fish.
Macroinvertebrates play an important role in the nutrient processing and organic energy cycling
in lotic environments. Most of the organic matter that enters a stream is ingested and excreted by
macroinvertebrates many times along the length of a stream. Benthic macroinvertebrates, as the
name implies, are insects generally visible to the naked eye (though identification typically
requires a dissecting microscope (10 x magnification)) that often inhabit areas of streams,
especially under rocks and near the sediment water interface, for at least part of their life cycle.
They can include larval forms of many common insects such as mayflies, caddisflies,
damselflies, and craneflies; or crustaceans like crayfish and scuds. They make good indicators of
watershed integrity because they:
• live in the water for all or most of their life,
• inhabit areas suitable for their survival,
• are relatively easy to collect,
• differ in their tolerance to amount and types of pollution,
• are relatively easy to identify in a laboratory,
• have limited mobility (in the larval forms),
• are among the first organisms to recruit disturbed areas, and
• are indicators of environmental condition.
r\
Individual macroinvertebrate kick-net samples covering 2 m of each riffle were collected using
modifications of the established protocols of the U.S. EPA's Rapid Bioassessment protocol for
Use in Wadeable Streams and Rivers (Barbour et a/., 1999) which the Virginia Department of
Environmental Quality (VDEQ) employs for bioassessments. An area of 0.5 m x 0.5 m (0.25
meter square) upstream of the net was sampled using the 0.5 m-wide kick-net. Using the toe or
heel of the boot, the upper layer of cobble or gravel was dislodged and the underlying bed was
3-3
-------
scraped. Larger substrate particles were picked up and rubbed by hand to remove attached
organisms prior to kicking and allowing the detached macroinvertebrates to float downstream
into the net. A total of 8 kick-net collections were composited into one sample for a total of 2 m2
within each riffle during dry weather flow conditions. All organisms caught in the net were
transferred to a two liter sampling container. Samples were preserved with 70% ethanol before
sending to EPA Region 3's Wheeling Laboratory for analysis. Five locations, selected for
macroinvertebrate sampling, are shown in Figure 3-2. Site C (just above the bridge) had to be
moved about 50 ft (15.2 m) below the bridge after restoration due to the relocation of the riffle
following the restoration. The sixth location, not shown, was in the previously restored (2003)
upstream riparian park. Macroinvertebrate collections were initiated at the downstream location
and proceeded upstream. Samples were collected in riffle and run habitats. Macroinvertebrates
retained on a No. 35 mesh dip net (500 |im) were randomly subsampled to 110±20 organisms
and identified using macroinvertebrate identification keys of Merritt and Cummins (1996),
Pennak (1989), Peckarsky etal. (1990), and Thorp and Covich (1991). After identification and
enumeration of macroinvertebrates, the Virginia Stream Condition Index (VASCI), total taxa,
total taxa family, EPT (Emphemeroptera, Plecoptera, and Trichoptera) taxa, EPT family,
Hilsenhoff biotic index (FIBI), percent of scrapers, and percent of most dominant taxon were
calculated.
Figure 3-2. U.S. EPA's macroinvertebrate sampling locations (indicated by orange marker)
3-4
-------
Although HBI (Hilsenhoff, 1987) was originally developed to assess low dissolved oxygen
caused by organic loading, a purpose for which it works best, the HBI is also considered to be
sensitive to the effects of impoundment, thermal pollution, and some types of chemical pollution
(Hilsenhoff, 1998; Hooper, 1993). This index has also been used to detect nutrient enrichment,
high sediment loads, and thermal impacts. In addition, since originally developed, this index has
been modified to accommodate comparisons of samples collected throughout the year. The HBI
ranges from 0-10 with 10 being the worst and 0 being the best; there is no defined impairment
threshold value. Samples with HBI values of 0-2 are considered clean, 2-4 are slightly enriched,
4-7 are enriched, and 7-10 are polluted (Hilsenhoff, 1988).
The VASCI is a multi-metric biological index developed using recent advances in bioassessment
methods and is calibrated from Virginia data for use in the assessment of Virginia's nontidal,
upland streams. This index was used to compare with regional and local reference datasets. The
VASCI ranges from 0-100 (100 is the best possible), with 60 being the impairment threshold in
VA. VASCI and HBI are inversely related with respect to water quality.
EPT family richness is also commonly used to assess water and habitat quality and is defined as
<2=poor water quality; 2-5=fair; 6-10=good; and >10=excellent quality.
Results
Continuous Water Quality Monitoring
Daily averages of the continuous monitoring 15-minute data collected for pH, conductivity,
turbidity, temperature, and depth recorded by YSI are shown in Figures 3-3 through 3-5. The
gap in the data from June 2006 to August 2006 was due to the equipment being damaged after a
large storm event, which came right after the restoration was completed. Three inches of rainfall
fell in 2 hours on June 9th. During the period of June 23-26, 2006, there was major flooding in
the area. On June 25, 2006, in Fairfax County, VA, two stream flow gages recorded peaks near
the 50-year recurrence interval and one stream flow gage recorded a peak near the 100-year
recurrence interval.
As expected, pH stayed close to neutral ranging between 6.5 and 8. Temperature changed
seasonally, but also had an event-related effect, where the daily average temperature decreased
with increasing depth due to increased flow during wet weather events and likely due to the
difference in temperature between rain water and stream water. Turbidity and conductivity
appear to be event-related with spikes occurring during wet weather events. The conductivity
also was seasonally dependent as it peaked during winter, likely due to runoff from salt during
snow melt. Salting is a regular snow and ice management practice in the City of Fairfax.
3-5
-------
8
6
iS S3 ?= (3
Date
Figure 3-3. Continuous water quality monitoring at Site WQ1 for pH
Date
Figure 3-4. Continuous water quality monitoring at Site WQ1 for temperature and depth
3-6
-------
conductivity
Turbidity
800
+ 700
Date
Figure 3-5. Continuous water quality monitoring at Site WQ1 for conductivity and turbidity
Continuous monitoring data for pH, conductivity, temperature, level, and velocity recorded by
the area-velocity flow meters for WQ2 and WQ3 are shown in Figures 3-6 through 3-11. Again,
it can be seen that the conductivity was higher in February due to events involving street salting.
Temperature of the creek water changed with the season and the wet weather flow events. The
highest temperatures were observed in July. pH ranged between 5 and 10. Flow data are not
very reliable as can be seen in the figures. The flow rates were calculated by the American
Sigma unit based on the flow level (a pressure transducer or bubbler), a velocity measuring
device (sonar) and a specified area. Negative or vey low velocity values were recorded by the
velocity probe. Open channel conditions in the field are not ideal conditions for this type of
velocity measurement. Increasing flow can lead to turbulent conditions, and eddies can trigger
the velocity sensor to record negative values, resulting in negative flow calculations. The level
sensor was calibrated in the field, and the velocity sensor was factory calibrated. The in-situ
flow values presented are not considered calibrated flow values in the sense that standard stream
gauges which typically use weirs or flumes. They are presented for demonstrative purposes only
to show changes in the flow regime in the stream being monitored.
At monitoring station WQ3, the base flow level changed from approximately 85 cm to 28 cm as
the location changed slightly after restoration placing the probe in shallower water.
3-7
-------
Date
Figure 3-6. Continuous water quality monitoring at Site WQ2 for pH
Temperature
• Conductivity
4 ^
.o
55
3 E
Date
Figure 3-7. Continuous water quality monitoring at Site WQ2 for temperature and conductivity
3-S
-------
100
80
E 60
o
40
20
4000
3500
3000
2500 ^
d
O
o
d
o
o
d
o
o
d
if)
d
(35
CO
O
O
d
C2
co
o
o
Date
Figure 3-9. Continuous water quality monitoring at Site WQ3 for pH
3-9
-------
40
Date
Figure 3-10. Continuous water quality monitoring at Site WQ3 for temperature and conductivity
120
Date
Figure 3-11. Continuous water quality monitoring at Site WQ3 for level and estimated flow
Relationship between the level and the rainfall recorded by the nearby station is plotted in Figure
3-12. It can be seen that the level increased with rainfall and has a direct relationship as
3-10
-------
expected. For example, on June 25, 2006, the rainfall was about 3.07 in. and the level jumped
from 36.46 cm on June 24, 2006 to 86.35 cm on June 25, 2006.
120
Date
Figure 3-12. Relationship between level and rain at Site WQ3
Discrete Water Quality Sampling
Results of the discrete samples collected before and after restoration in both upstream (Lee
Highway) and downstream (Old Lee Highway) locations and analyzed for physical and chemical
constituents are shown in Table 3-1. Seven wet weather (two before restoration and five after
restoration) and seven dry weather (two before restoration and five after restoration) sampling
events were conducted with the full suite of analytes. Data in Table 3-1 indicate that wet
weather concentrations of TPO43", NHs, TKN, SS, and COD were higher than the dry weather
concentrations typically. SS concentrations ranged between 0.20 - 20 mg/L and 89 - 291 mg/L
respectively for dry and wet weather samples. COD concentrations ranged between 0.4 - 15
mg/L and 11 - 73 mg/L for dry and wet weather samples, respectively.
Concentrations of wet weather SS increased significantly after restoration. This may be because
restoration work disturbed the stream channel and liberated sediments. Also, it takes time to
stabilize the stream banks as plants require time to grow before being effective. Concentrations
of SS ranged between 3-13 mg/L and 97 - 291 mg/L for before and after restoration,
respectively at the downstream location. Concentrations of COD did not change and ranged
between 12 - 67 mg/L. TPO43", NHa, and TKN concentrations increased slightly after
restoration. Concentrations ranged between 0.07 - 0.35 mg/L, 0.5 - 1.3 mg/L, and <0.01 - 0.29
mg/L for TPO43", TKN, and NHs, respectively after restoration. However, these changes are not
3-11
-------
Table 3-1. Results of water quality analysis (physical and chemical constituents)
Pre -Restoration
Post -Restoration
Date
3/1/06
4/5/06
5/2/06
5/9/06
6/20/06
9/21/06
10/13/06
11/16/06
12/14/06
4/4/07
4/15/07
7/11/07
9/18/07
1/16/08
Flow
Condition
Dry
Wet
Dry
Wet
Wet
Dry
Wet
Wet
Dry
Dry
Wet
Wet
Dry
Dry
Concentrations in mg/L
Upstream (Lee Highway)
SS
0.67
6.67
1.40
26.67
4.40
(0.28)
0.40
89.10
(1.84)
252.10
(10.04)
0.20
30.32
127.50
171.30
(0.99)
2.40
(1.41)
0.30
(0.14)
COD
0.37
(0.44)
14.92
(0.39)
7.61
(0.34)
61.66
(0.88)
28.86
(1.21)
15.24
(0.92)
11.08
(0.67)
72.52
(0.94)
7.69
(1.27)
30.11
(1.21)
21.96
(1.60)
23.27
(2.38)
3.36
(1.08)
1.05
(1.14)
TPO43
0.03
(0.01)
0.02
(0.01)
0.04
(0.06)
0.35
(0.04)
0.06
(0.01)
0.06
(O.01)
0.28
(0.01)
0.24
(O.01)
0.04
(O.01)
0.03
(0.02)
0.13
(0.01)
0.29
(0.06)
0.04
(O.01)
0.03
(O.01)
TKN
0.07
(0.01)
0.40
(0.02)
0.20
(0.02)
0.58
(0.04)
0.61
(0.01)
0.32
(0.03)
0.49
(0.02)
0.83
(0.04)
0.18
(0.01)
1.35
(0.02)
0.99
(0.07)
1.35
(0.05)
1.14
(0.06)
0.64
(0.02)
NH3
0.03
(0.01)
0.08
(0.01)
0.01
(O.01)
0.19
(0.01)
0.06
(0.01)
0.08
(O.01)
0.01
(0.01)
0.01
(O.01)
O.01
(O.01)
0.73
(O.01)
0.29
(O.01)
0.14
(O.01)
ND
0.03
(O.01)
Downstream (Old Lee Highway)
SS
2.00
3.33
25.07
(1.04)
13.25
(0.35)
6.30
0.22
96.80
290.60
1.20
19.60
(2.26)
120.20
(0.85)
204.40
ND
ND
COD
1.88
(0.34)
19.44
(2.31)
10.33
(1.11)
68.01
(2.04)
22.25
(0.88)
28.45
(0.35)
12.48
(0.85)
67.08
(2.31)
6.03
(0.57)
20.35
(2.42)
27.03
(1.40)
29.24
(1.02)
3.81
(1.21)
1.81
(0.56)
TPO43
0.02
(0.01)
0.03
(0.01)
0.04
(O.01)
0.13
(0.01)
0.07
(0.01)
0.01
(O.01)
0.32
(0.01)
0.22
(O.01)
0.04
(0.01)
0.04
(0.02)
0.12
(0.01)
0.35
(0.02)
0.03
(O.01)
0.02
(0.02)
TKN
0.09
(0.01)
0.39
(0.02)
0.54
(0.05)
0.44
(0.02)
0.65
(0.01)
0.37
(0.01)
0.50
(0.03)
0.95
(0.05)
0.21
(0.02)
1.28
(0.01)
0.63
(0.02)
1.30
(0.07)
0.51
(0.07)
0.49
(0.04)
NH3
0.03
(0.01)
0.11
(0.01)
0.10
(O.01)
0.07
(0.01)
0.06
(0.01)
0.03
(0.01)
0.01
(0.01)
0.01
(O.01)
0.01
(0.02)
0.76
(0.02)
0.29
(0.01)
0.14
(0.01)
ND
0.04
(O.01)
Note: Restoration was completed on June 6,
Brackets indicate standard deviation
2006
3-12
-------
great enough to associate with restoration activities. The One-way ANOVA statistical analysis
indicates that there is no statistically significant difference between before and after restoration
and as well as upstream and downstream of the restoration except for wet weather SS. These
concentrations are well below Virginia Water Quality Standards (State Water Control Board,
2007). Concentrations of SS, COD, TPO43", NH3, and TKN in wet weather samples before and
after restoration in both upstream and downstream locations are shown in Figures 3-13 and 3-14.
1 \J \J -
140 -
—. 12° "
j"
"3>
£ 100 -
w
W 80 -
•D
C
n
Q 60 -
O
O
40 -
20 -
n _
D Lee-Before Restoration
• Lee-After Restoration
D Old Lee-Before Restoration
D Old Lee-After Restoration
1
T
I
T
1
,—
T
1
COD SS
Figure 3-13. Average of wet weather concentrations of COD and SS before and after restoration
1.4
I 1
O 0.8
co" 0.6
z- 0.4
H 0.2 H
0
Figure 3-14.
I
D Lee-Before Restoration
• Lee-After Restoration
D Old Lee-Before Restoration
D Old Lee-After Restoration
jfl
i
T _L
TKN NH3 TPO4
Average of wet weather concentrations of TKN, NH3, and TPO43 before and after restoration
3-13
-------
Results of the discrete samples collected before and after restoration in both upstream (Lee
Highway) and downstream (Old Lee Highway) locations and analyzed for bacteriological
constituents are shown in Table 3-2. Data in Table 3-2 indicate that wet weather concentrations
of fecal coliform, enterococci, and E. coli are much larger than the dry weather conditions as
expected. Except for the November 6, 2006 wet weather sampling event, both upstream and
downstream samples had concentrations in the same order of magnitude. Concentrations of all
three indicator organisms in the November 6, 2006 samples were much higher in the downstream
samples compared to upstream samples. The November 2006 sampling event may be
anomalous. Concentrations of organisms vary with seasons and summer concentrations were
significantly higher compared to other seasons as expected. The One-way ANOVA statistical
analysis indicates that there is no statistically significant difference between before and after
restoration as well as upstream and downstream of the restoration. Concentrations of fecal
coliform, enterococci, and E. coli in wet weather samples before and after restoration in both
upstream and downstream locations are shown in Figure 3-15.
Table 3-2. Results of water quality analysis (indicator organisms)
Pre-Restoration
Post -Restoration
Date
3/1/06
4/5/06
5/2/06
5/9/06
6/20/06
9/21/06
10/13/06
11/16/06
4/4/07
4/15/07
7/11/07
9/18/07
1/16/08
Flow
Condition
Dry
Wet
Dry
Wet
Wet
Dry
Wet
Wet
Wet
Wet
Wet
Dry
Dry
Concentrations in CFU/100 mL
Upstream (Lee Highway)
FC
31±9
1867±719
160±36
77000±
6557
33667±
6807
643±133
3887±271
777±25
5117±776
5033±513
67500±
4950
1000±586
72±71
EN
7±3
119±39
66±26
4900±608
1163±162
70±13
3933±457
3453±334
6183±
2646
ND
148±53
15±5
ND
EC
23±7
365±55
101±30
7733±252
8700±
1082
436±76
11267±
1106
30±26
3900±
1905
2300±361
18000±4000
104±23
200±141
Downstream (Old Lee Highway)
FC
74±46
1865±212
224±85
74333±6028
26667±5033
285±157
7033±252
143500±
12021
4567±208
6167±351
52333±
10693
233±48
37±28
EN
20±10
214±94
338±84
7100±361
1665±398
78±1
6333±551
6600±600
2883±
1089
ND
163±90
19±8
ND
EC
31±12
520±204
179±42
3533±551
11167±2021
302±12
13667±666
99000±8485
2600±385
2700±458
9267±513
93±11
250±71
These physical, chemical, and bacteriological data suggest that local restoration in and around
streams may not be sufficient enough to see significant measurable effects on the water quality
of the stream.
3-14
-------
_o
+J
+: 5
c
8
O 4
0 4
(A
'= 3
(0
S1
o
5 2
"re
o
I 1
0)
o
-1 n
1
f [ ^
.1
^
1
D Lee-Before Restoration
• Lee-After Restoration
D Old Lee-Before Restoration
D Old Lee-After Restoration
M
E. co/i
Fecal Coliform
Enterococci
Figure 3-15. Summary of indicator organism concentrations before and after restoration
Macroinvertebrate Sampling
The results for VASCI and HBI indices, number of EPT taxa families, and number of total taxa
families for all sampling events are summarized in Table 3-3. Total number of taxa families
between sampling locations ranged between 3 and 10 and typically had more than 5 families
represented. EPT taxa families ranged between 0 and 3 between sampling locations and
typically are 1 and 2 indicating poor water and habitat quality. All of the sites, including the
control sites, received VASCI scores less than 60, the impairment threshold in Virginia,
indicating impaired macroinvertebrate conditions. The scores of the HBI index for all the sites
are within the "enriched" category (4-7) as defined by Hilsenhoff (1988) which indicates that
most species identified are moderately tolerant to polluted water with high organic content or
excessive nutrient.
Seasonal variation in-stream biota makes it difficult to compare data over time unless
comparisons are made only with data from a single season. Three and five sampling events were
conducted before and after restoration, respectively. Longer duration of pre-restoration sampling
was not possible as the city had the project design, funding, and implementation plan in place.
Comparison of data within a season was only possible for the fall season. The fall season data
were collected in 2005 (2 events) before restoration and 2006 (2 events) and 2007 (2 events),
which were collected after the restoration. Paired Mest, which examines the changes that occur
before and after a treatment to determine whether or not the treatment had any effect, indicated
the changes that occurred are not great enough to exclude the possibility that the difference are
due to chance for both indices in all locations (P>0.05).
3-15
-------
Table 3-3. Results of macroinvertebrate data
Pre- and
Post-
Restoration
Pre- Restoration
Post- Restoration
Date
11/03-
04/2005
12/07-
08/2005
3/13-
14/2006
9/21/2006
11/15/2006
5/9/2007
9/18-19/07
11/14-15/07
Species
VASCI
HBI
#ofEPTTaxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
VASCI
HBI
# of EPT Taxa Families
# of Total Taxa Families
Site A
(-120 m North of
Lee Hwy)
Upstream
21.2
6.86
1
5
21.5
5.91
1
5
25.2
6.03
2
5
36.8
6.02
3
5
29.6
5.35
2
6
27.9
6.09
3
7
32
5.9
3
6
27.1
6.47
1
6
SiteB
(-100 m South of
Lee Hwy)
Restoration Area
29.1
5.87
2
6
25.1
6.17
1
5
23.9
6.82
1
5
28.2
5.9
2
4
26.6
6.09
1
5
22.8
6.59
1
5
30.5
5.93
2
7
28.5
6.02
1
7
SiteC
(-10 m North of
Old Lee Hwy)
Restoration Area
24.3
5.94
1
5
30.7
6.03
1
9
26.3
6.03
1
6
33.5
5.75
2
7
28.4
6.03
2
7
12.3
6.02
0
3
22.5
6
2
8
30.4
6.13
1
8
SiteD
(-200 m South of
Old Lee Hwy)
Downstream
25.9
6.06
1
5
25.6
6.13
1
6
27.2
6.59
1
6
32.2
5.71
2
6
24.8
5.98
1
5
22.2
6.79
2
5
31.7
5.86
2
7
29.2
5.97
1
6
Site RUP
(-50 m West of
Bridge at River
Road) Upstream
28.5
5.95
1
6
24.2
6.13
1
8
38.6
5.28
3
4
33.3
5.79
2
10
26
6.08
2
6
32.2
5.84
2
7
28.8
6.16
1
9
3-16
-------
Table 3-4 summarizes the average values for the parameters before and after restoration.
Benthic invertebrate data collected to date indicate areas within the restoration reach have
VASCI scores that are not significantly different than before the restoration. Controls show
substantial variability before and after restoration. The VASCI score at control site A was much
smaller than expected in the pre-restoration sampling event. This may be due to seasonal
variability and related to the velocities experienced in this stream that remain unchanged with
this management strategy. Upstream control site VASCI scores following restoration were
intended to provide an attainable goal for sites B and C within the current restoration reach.
Both sites B and C in the restored section were moved owing to the fact that the restoration
altered the riffle locations that the original riffle did not exist in the same location. The HBI
average was 6.05 in the restored area, 6.06 downstream, and 5.89 upstream sites after restoration.
All were ranked as enriched per Hilsenhoff (1988) and there was no significant difference
between indices.
Macroinvertebrate data completed for VASCI, HBI, and EPT taxa families showed a slight
improvement trend in conditions between pre- and post- restoration for all sites up to two years
after the restoration (Table 3-4). Paired t-test indicated a statistically significant change in
VASCI (P=0.014) and HBI indices (P=0.012) and total number of EPT Taxa families (P=0.017)
between before and after restoration as the change occurred was greater than would be expected
by chance.
Table 3-4. Average macroinvertebrate indices and EPT taxa families before and after restoration
VASCI
HBI
EPT Taxa
Families
Site RUP*
Pre
26.4
(3.0)
6.04
(0.13)
1.00
(0.0)
Post
31.8
(4.8)
5.83
(0.35)
2.00
(0.71)
Site A
Pre
22.6
(2.2)
6.27
(0.52)
1.33
(0.58)
Post
30.7
(3.9)
5.96
(0.41)
2.40
(0.89)
Upstream Controls
SiteB
Pre
26.0
(2.7)
6.29
(0.49)
1.33
(0.58)
Post
27.3
(2.9)
6.11
(0.28)
1.40
(0.55)
SiteC
Pre
27.1
(3.3)
6.17
(0.32)
1.00
(0.0)
Post
28.7
(4.6)
5.99
(0.14)
1.40
(0.89)
Restoration Reach
SiteD
Pre
26.2
(0.9)
6.26
(0.29)
1.00
(0.0)
Post
28.0
(4.4)
6.06
(0.42)
1.60
(0.55)
Downstream
Affects
*RUP: Restored Upstream Park
Parentheses indicate standard deviation
An important factor influencing the slow recovery of benthic invertebrates in this system may be
the substantial increase in wet weather flow velocities. The stream restoration likely created
more habitats through the added pool-riffle structure incorporated in the restoration, but little or
no volume control management was done in the watershed to attenuate wet weather flow
volumes during this phase of watershed enhancement. Volume control to reduce flow velocities
(e.g., stream bed scouring) from directly connected impervious areas and continuation of
invertebrate collection sensitive to timing may improve recorded macroinvertebrate conditions in
the restored reach. Moreover, macroinvertebrate communities may be limited by water quality
since many of the taxa collected were considered tolerant of adverse chemical conditions. The
3-17
-------
results for VASCI and HBI indices are shown in Figures 3-16 and 3-17, respectively.
Differences in invertebrate indices are shown in Figures 3-18 and 3-19.
38
o
O
O
36 -
34 -
32
28
26
24
22
20
1
1
n Pre-Restoration
D Post-Restoration
_L
1
SiteRUP Site A Site B Site C Site D
Figure 3-16. Virginia Stream Condition Index (VASCI) scores for before and after the Accotink Creek
stream restoration
7.0
6.8
6.6 H
6.4
£ 6.2
O
o
(/) 6.0
X 5.8
5.6
5.4
5.2 1
5.0
D Pre-Restoration
D Post-Restoration
n
Site RUP
Site A
SiteB
SiteC
SiteD
Figure 3-17. Hilsenhoff Biotic Index (HBI) scores for before and after the Accotink Creek stream restoration
3-18
-------
*
•o
C
o
•f
2
3
£
3
o
Q.
Site RUP Site A Site B Site C Site D
Figure 3-18. Differences in VASCI between post and pre-restoration
0.00
x
| -0.05
I "°-1°
IS -0.15
2
$ -0.20
£ -0.25
o.
•JJi -0.30
Site RUP Site A Site B Site C Site D
-0.35
Figure 3-19. Differences in HBI between post and pre-restoration
Table 3-5 summarizes the type and number of dominant species for all sampling events. Except
for March 2006, all the sites had a similar total number of macroinvertebrates (i.e., taxa) and
there was no statistically significant difference in total macroinvertebrate relative abundance
over all the sampling dates. There was also no significant difference in the total number of
macroinvertebrates between the upstream, downstream, and restored sites. The upstream,
downstream, and restored areas have similar percent dominance values. The most dominant taxa
at these sites were Chironomidae, Hydropsychidae, Naididae, and Lumbriculidae representing
87% of the upstream site samples, 93% of the restored area samples, and 92% of the downstream
samples composed of these four families. All other families were relatively rare, with most
composing less than 1% of represented taxa (i.e., 1-2 taxa).
While there were no differences in the total number of families, there were more Chironomidae
than Hydropsychidae at all sites before restoration. This was reversed after the restoration as
3-19
-------
Table 3-5. Total number of macroinvertebrates
Pre- and
Post-
Restoration
Pre- Restoration
Post- Restoration
Date
11/03-
04/2005
12/07-
08/2005
03/13-
14/2006
09/21/2006
11/15/2006
05/9/2007
09/18-
19/07
11/14-
15/07
Species
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Chironomidae
Hydropsychidae
Naididae
Lumbriculidae
Site A
(-120 m North
of Lee Hwy)
Upstream
79
31
3
98
6
3
27
1
10
12
30
52
1
1
25
67
7
4
100
1
19
2
30
87
6
60
35
SiteB
(-100 m South of
Lee Hwy)
Restoration Area
45
59
1
55
14
1
9
8
1
2
5
7
102
15
93
1
1
80
42
2
42
126
2
24
93
1
3
SiteC
(-10 m North of
Old Lee Hwy)
Restoration Area
72
33
65
29
2
1
72
3
22
5
15
80
1
4
90
4
1
104
1
132
9
2
25
22
3
SiteD
(-200 m South of
Old Lee Hwy)
Downstream
40
77
1
67
30
3
4
23
1
10
4
13
106
1
7
106
1
1
68
2
52
1
50
80
2
46
58
10
Site RUP
(-50 m West of
Bridge at River
Road) Upstream
76
32
4
69
9
31
46
16
71
3
90
9
30
91
1
45
70
12
3-20
-------
there were more Hydropsychidae than Chironomidae except for the 5/19/2007 sampling event. It
is plausible that restoration created more stable substrates which are required for attachment by
net-spinning Hydropsychids. Many Chironomidae are silt and sand tolerant and early colonizers
following streambed scouring. The May 2007 sampling event may be anomalous, or other water
quality factors may be responsible for higher Chironomidae on this sampling occasion.
Overall, the poor VASCI scores and relatively high HBI indicate that water quality may be
limiting macroinvertebrate recovery following restoration activities. The dominant taxa found in
Accotink Creek (pre- and post restoration) suggest a variety of pollutants (e.g., nutrients, metals,
other trace toxicants) could be responsible for structuring the observed communities. Moreover,
additional monitoring is needed to detect changes in macroinvertebrate communities over time.
Improvement may not be realized in two years post-restoration, a finding common to many
stream restoration projects (Personal communication with Wheeling biologist Gregory Pond,
2008).
Stream Channel Cross Sections
Stream channel cross sectional measurements were taken using a folding ruler and a flexible tape
measure stretched perpendicular to the direction of stream flow. Measurements were taken from
bank to bank at 0.5 ft (0.15 m) increments close to the banks and at 1 ft (0.3 m) interval else
where at four different locations (one upstream, one downstream, and two in restored area).
Figures 3-20 and 3-21 show the channel profiles at two different locations. In the upstream
location, the bottom contours did not change much after restoration. In the restored area, the
depth profile showed deeper and more sharply defined bottom contour after restoration compared
to before restoration. Bottom depth changed from approximately 7 ft (2.13 m) to 11 ft (3.35 m).
Substrate was mostly gravel and cobble comprising 90-95% of the streambed of the creek in the
restored area, whereas gravel and cobble comprised 77-84% of the streambed upstream of the
restoration.
Distance Across Stream Channel (m)
6 9
12
15
before restoration
after restoration
Figure 3-20. Representative depth profiles before and after restoration at an upstream location
3-21
-------
Distance Across Stream Channel (m)
4 6 8 10 12
14
16
before restoration
after restoration
Figure 3-21. Representative depth profiles before and after restoration at a restored location
Pebble Count
The pebble count was conducted 2 times before restoration (November 3, 2005 and March 1,
2006) and once after restoration (October 3, 2006). The pebble count was conducted at 5
different riffle locations (Ranger Road - upstream of restoration; Site A - Upstream of Old Lee
Highway - upstream of restoration; Site B - Below Lee Highway at Harley Dealer - restoration
reach; Site C - upstream of Old Lee Highway - restoration reach; and Site D - downstream of
Old Lee Highway - downstream of restoration) to evaluate streambed particle-size distributions.
Counts were performed in a manner similar to that described by Wolman (1954); minor
modifications to the methods were needed to accommodate site characteristics. The pebble
count technique developed by Wolman in 1954 has long been used to document the surface
particle size distribution of coarse riverbed material. Because Accotink Creek is a relatively
narrow stream, an entire stream riffle with multiple transects were needed for the pebble count to
be more representative, rather than just an individual transect within a riffle. On average, the
sampled riffles were about 25 ft (7.62 m) long and approximately 18 ft (5.5 m) wide. Pebbles
were selected for size determination from within the wetted perimeter of the stream, and were
chosen for size determination using the first-blind-touch approach. Particle size was determined
using a pebble count template (which provided a standard classification system). Particles that
were smaller than 2 mm were compared to a sand gauge card to determine size. A total of 100
pebbles were selected from within each riffle section. By classifying particles using the template
and sand card, the particles could be grouped into sieve size classes according to the Wentworth
scale. Following size classification, the data were plotted to summarize the relative size classes
identified in each riffle.
The pebble count data are summarized below in Table 3-6 and Figures 3-22 through 3-26, and
these data indicate that as of this study period, very little has changed in this stream reach. A
more quantitative statistical analysis is limited by the number of samples collected at each site.
By evaluating the pebble count data at each site with time, there is a slight increase in the post-
restoration (October 2006) sampling at both the most upstream and downstream cross sections.
3-22
-------
However, the most upstream site (at Ranger Road) is a control that is above the restoration, it
cannot be concluded that the slight increase in particle size at the most downstream site is caused
by the restoration. The other three intermediate sites demonstrate very little changes in the size
distributions over time. This lack of change in the stream bed size classes is most likely due to
the restoration not changing the rate of water courses down the stream.
Table 3-6. Results of pebble counts
Date
11/03/2005
Pre-
Restoration
3/1/2006
Pre-
Restoration
10/03/2006
Post-
Restoration
Pebble
Count Data
% Silt/Clay
% Sand
% Gravel
% Cobble
Particle Size
(mm)
% Silt/Clay
% Sand
% Gravel
% Cobble
Particle Size
(mm)
% Silt/Clay
% Sand
% Gravel
% Cobble
Particle Size
(mm)
Site A -
Above Lee
Hwy (Above
Restoration)
3
13
84
0
7.9±3.9
3
20
76
1
6.3±4.5
1
17
79
3
8.0±4.5
Site B -
Below Lee
Hwy
(Within
Restoration)
0
11
76
13
19.1±3
0
5
87
8
20.5±2.6
0
5
82
13
24.2±2.4
Site C -
Above Old
Lee Hwy
(Within
Restoration)
0
9
61
30
34.0±2.5
0
9
60
31
29.5±3.0
0
2
77
18
32.2±2.4
Site D -
below Old
Lee Hwy
(Below
Restoration)
2
4
90
4
19.9±2.0
2
33
65
0
4.3±8.0
4
4
76
16
29.2±2.2
Site 5 -
Ranger
Road
(Above
Restoration)
2
14
68
16
8.2±6.1
0
21
73
6
24.8±2.8
3-23
-------
0%
0.01
1 10 100
Particle size (mm)
1000
10000
Figure 3-22. Pebble count results at Site 5 - Ranger Road (upstream of restoration)
100%
90%
80%
70%
C 60%
re
"£ 50%
0)
c
'f 40%
•4-1
C
30%
0)
20%
10%
0%
. Nov. 3, 2005
-Mar. 1, 2006
- Oct. 3, 2006
0.01
0.1
1 10 100
Particle size (mm)
1000
10000
Figure 3-23. Pebble count results at Site D - downstream of Old Lee Highway (downstream of restoration)
3-24
-------
100%
90%
80%
re
0.01
1000
1 10 100
Particle size (mm)
Figure 3-24. Pebble count results at Site A - upstream of Lee Highway (upstream of restoration)
10000
0.01
1 10 100
Particle size (mm)
1000
10000
Figure 3-25. Pebble count results at Site C - upstream of Old Lee Highway (restoration reach)
3-25
-------
0%
0.01
0.1
100
1000
10000
Particle size (mm)
Figure 3-26. Pebble count results at Site B - below Lee Highway at Harley Dealer (restoration reach)
3-26
-------
Chapter 4 USGS Sampling and Monitoring
Background
The USGS conducted continuous water quality monitoring and collected grab samples from
December 2005 to August 2007 under an interagency agreement (TAG No. DW-14-922064010)
with U.S. EPA ORD. A YSI sensor was used to monitor turbidity, specific conductance, pH, and
water temperature. The sensor, an YSI extended deployment sonde, was installed just
downstream of the restoration area hanging from the bridge at Old Lee Highway. In general,
continuously monitored data can provide detailed records of water quality (Figure 4-1), and
allow scientists and watershed managers to better understand their systems. As part of the IAG,
the USGS also collected approximately 21 water quality samples over a wide range of flow
conditions and analyzed them for E. coli and suspended sediment concentrations and performed
pebble counts (results were presented in Chapter 3) at designated sites before and after the
restoration.
Relationships often exist between the water quality parameters that can be measured with sensors
and other contaminants of interest; these relationships make the technology even more powerful.
For example: turbidity values typically correlate well with both suspended sediment and bacteria
concentrations. When discrete water quality samples are collected manually during both low
flow and storm flow periods, in conjunction with the continuously monitored data, regression
equations can be developed to relate a target water quality constituent in the discrete samples
(e.g., suspended sediment or bacteria concentration) to the water quality parameters that are
monitored continuously (e.g., turbidity). This regression equation can then be used to estimate
continuous concentrations of the target water quality constituent. This approach is completely
analogous to the standard methods for developing continuous discharge records, in which stream
stage (water level) is recorded continuously and a regression equation (a rating curve) is
developed to relate continuous stage and discrete discharge measurements. Instead of
developing a stage-discharge relationship to calculate continuous discharge, this approach is used
to develop such models as turbidity-bacteria correlations to calculate continuous bacteria
concentrations. Figure 4-2 shows the relationship between turbidity and fecal coliform from a
previous study further downstream.
4-1
-------
James River at Cartersville, VA
17 18 19 20
Date In May
21
70,000
- 60,000
- 50,000 •sr
Turbidity
Conductance
Discharge
22
Figure 4-1. Example of continuous water quality data determined by sensor technology
(http://va.water.usgs.gov/ContinuousWaterQuality.pdf)
Accotink Creek near Annandale, VA
6.00
o 5.00 -
o
1 4.00 -
E
,0 3.00
"5
^ 2.00 -
ro
o
01
"- 1.00 H
0.00
0.25 0.75 1.25 1.75 2.25
Log Turbidity(NTU)
2.75
Figure 4-2. Example of correlation between turbidity and fecal coliform concentration
One novel application of this continuous water quality monitoring and development of regression
equations is for the detection of change in water quality that is related to BMP implementation
activities. Detection of measurable improvements in water quality can be achieved through
numerous univariate and multivariate statistical analyses of these data. These analyses can
include an evaluation of changes in the developed regression equations, the regression residuals,
and the overall distribution of continuously estimated constituents. The direct benefits of this
approach are that the data analysis is largely independent of confounding environmental factors,
the continuous data provide a better dataset with which to efficiently detect environmental
change, and over the long term this approach should be less costly than traditional monitoring
4-2
-------
plans.
Continuous Water Quality Monitoring
All continuous water quality monitoring operations in the USGS are performed according to the
USGS standard methods for the operation of this equipment. Published USGS standard
operating procedures (SOP) (Wagner etal., 2000) were followed during this study, therefore,
only a summary of these procedures are outlined below and an internet link to the full SOP is
provided in the references section.
The continuous water quality monitor (YSI Model 6920 multi-parameter monitor) was deployed
at the Old Lee Highway Bridge on December 14, 2005 and configured to measure water
temperature, turbidity, specific conductance, and pH at 15-minute intervals. The instrument was
connected to data logging and telemetry equipment that transferred all data to the USGS office in
Richmond, VA, where the data were displayed on the internet for access by all interested
individuals during the time of monitoring. Following initial deployment, monthly maintenance
visits were performed on the monitoring equipment to clean and check the calibration of the
sensors. In-field recalibration was performed during these maintenance visits as necessary
following the equipment tolerances as specified by the monitor manufacturer and the SOPs
(Wagner et a/., 2000).
Following the monthly maintenance visit, the maintenance data were used to determine whether
the monitoring equipment was subject to bio-fouling or calibration drift. If either of these
conditions was observed to be outside the SOP tolerances, the continuous water quality record
may be shifted to correct these data. At the conclusion of each water year, the data were
reviewed for accuracy, all shifts were checked, the quality of the data were rated (as excellent,
good, fair, or poor), a station analysis for the water year was prepared, and the finalized data
were published in the Annual Virginia Water Science Center Data Report. By following the
SOPs outlined by Wagner et al. (2000), these continuous data were of known quality and were
able to be compared to any other continuous water quality data that also were collected following
these guidelines.
Continuous water quality monitoring continued during most of stream restoration construction in
the Accotink Creek above Old Lee Highway through until early May 2006, when the contractors
needed the monitor removed so that they could clear sediment from underneath the bridge and do
minor restoration downstream of Old Lee Highway. The monitor was re-deployed on June 1,
2006, after the stream restoration around the Old Lee Highway Bridge was completed. The
restored stream channel around the Bridge caused considerable monitoring difficulty following
the June 1, 2006 re-deployment, because it was difficult to keep probes of the unit submerged.
The restored channel was considerably wider and shallower than it had originally been and the
creek drawdowns with the growing season. Following several storm events, a slightly deeper
channel had developed, allowing the water quality monitor to be fully submerged.
4-3
-------
Discrete Water Quality Sampling
During the 19-month monitoring period, approximately 21 discrete water quality samples were
collected and analyzed by the USGS from the bridge at Old Lee Highway over a wide range of
flow conditions, with special effort paid to the collection of water quality samples during storm
flow conditions. Approximately 13 samples were collected before restoration and 8 samples
were collected after restoration. These discrete water quality samples were collected and
analyzed following standard USGS protocols (USGS, 1998). Samples for analysis of suspended
sediment concentration (SSC) and E. coli were collected as grab samples from the approximate
center of stream flow, under varying hydrological conditions. Samples for analysis of SSC were
collected in clean, pre-weighed glass bottles, while samples for the analysis of E. coli were
collected in clean, sterilized glass bottles. For all samples collected, SSC and percent of the
sediments finer than sand size were determined. For as many of these samples as possible, E.
coli concentrations also were determined; the decision on which samples to analyze for E. coli
were based on their ability to process these samples within the prescribed 6-hour holding times.
Sediment samples were shipped to the USGS Eastern Region (Kentucky) Sediment Laboratory
for analysis following approved sediment analysis techniques (Sholar and Shreve, 1998; ASTM,
2007). Bacterial samples were processed using standard membrane filtration techniques (USGS,
1998; U.S. EPA, 2002b). As described in the USGS manuals for water quality sampling and
analysis, approximately 10% of the samples were made up of quality control samples, such as
blanks and duplicate samples.
Results
Continuous Water Quality Monitoring
A sample of continuous monitoring data for pH, conductivity, turbidity, and water temperature
recorded by the YSI Model 6920 multi-parameter monitor is shown in Figure 4-3. It can be seen
that the conductivity was higher in February due to snow and freezing events requiring street
salting. The temperature of the creek water changed with the season and the wet weather flow
events. The highest temperatures were observed in July and August. pH stayed close to neutral.
Similar results were observed with U.S. EPA monitoring equipment.
4-4
-------
30
25 -
O
E 20-
15 -
£
1
0)
Q.
I
•o
C
ra
Q.
10 -
8000
-Temperature
-pH
- Specific Conductance
-Turbidity
0)
ra
-------
The stage information was determined from the metadata that were recorded on field sheets
during the collection of stream samples. An interaction term between turbidity and the
categorical flow variables was found to be significant and justified on the basis of residual plots.
Predictive equations (with 1:1 lines) and residual plots are presented below for both E. coli and
SSC in Figures 4-4 through 4-7.
5.0
4.5
4.0
5 3.5
Uj
-I 3.0
2.5
2.0
1.5
1.0
1.0 1.5 2.0 2.5 3.0 3.5
Observed Log E. coli
Figure 4-4. Graph of predicted vs. observed for E. coli
4.0
4.5
5.0
1.5
1.0 -
?
9-
.r. o.o -
"ra
TJ
-0.5 -
-1.0 -
-1.5
1.5 2.0 2.5 3.0 3.5
Predicted Log E. coli
Figure 4-5. Graph of observed E. coli vs. residual
4.0
4.5
5.0
4-6
-------
4.0
3.5 -
3.0 -
IS
•a
§
w 2.0 -
1.5 -
•o
£
o
1.0 -
0.5 -
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Observed Log Suspended Sediment
Figure 4-6. Graph of predicted vs. observed for suspended sediment
3.5
4.0
0.6
0.4-
0.2 -
o
^
Q.
i
a
(0
o
OL
-0.2 -
-0.4-
-0.6
*
«
0.0 0.5 1.0 1.5 2.0 2.5
Predicted Log Suspensed Sediments
Figure 4-7. Graph of observed suspended sediments vs. residual
3.0
3.5
4-7
-------
Figure 4-8 is a plot of a simplified SSC equation prior to the addition of the stage interaction
terms (the equation isLogSSC = 0.9823(X0gTwr6) +0.1052). Before the addition of the stage
terms, there is a consistent over-prediction of SSC on the falling limb, and an under-prediction
on the rising limb and peak samples.
3.5
^ 3.0 -
H
3 2.5
o 2.0
c
o
1.5 -I
0.5 -
0.0 -
n
.Q
O
W
-------
3.5
3.0 -
2.5 -
O
CO
CO
u> 2.0
O
1.5 -
1.0 -
0.5 -
0.0
0.0
G D
D
•
G
0.5
1.0
1.5 2.0 2.5
Observed Log SSC
Figure 4-9. Graph of SSC showing pre- and post-restoration samples
• Pre-Restoration
a Post-Restoration
3.0
3.5
4.0
5.0
4.5 -
4.0 -
•5 3.5 -|
o
ui 3.0
O)
5 2.5 J
t> 2.0 -|
I-
1.0 -
0.5 -
0.0
a n
• Pre-Restoration
a Post-Restoration
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Observed Log E. coli
Figure 4-10. Graph of E. coli showing pre- and post-restoration samples
3.5
4.0
4.5
5.0
4-9
-------
Utility of the Continuous Data for the Prediction ofSSC and Bacteria Concentrations
The better equation for predicting SSC involved the determination of a relative flow condition
(rising stage, falling stage, or baseflow conditions); however, a simpler predictive equation exists
that does not require this determination (the simpler equation includes only SSC and turbidity).
Continuous estimates of E. coli and SSC are useful for several applications which are described
below.
One application of predictive equations for estimating continuous SSC andE. coli concentrations
is that the records could be used to evaluate the frequency with which concentrations of either
constituent exceeds a particular level. For example, a continuous record of estimated E. coli
concentrations can be analyzed to estimate how often a given bacterial water quality standard
might be exceeded.
Additionally, the continuous estimations of SSC and E. coli could be combined with a
continuous record of stream flow to produce loading estimates of these constituents for the
stream (load can be computed as the product of a concentrations term and a flow term). As there
was no stream gage at the Old Lee Highway site, the load computations cannot be performed
with the current data.
Another application of these continuous data is in the calibration and verification of watershed
models for SSC and E. coli. Continuous records of estimated SSC and E. coli could provide
more robust data sets with which to evaluate models.
Patterns in Turbidity Concentrations Before, During, and After Restoration
Another application of the continuous turbidity data is in the evaluation of the turbidity patterns
before, during, and following the stream restoration. The detail provided by 15-minute interval
data can be used to provide a robust image of the distribution of turbidity values that occurred at
and around the monitoring site. Figure 4-11 below presents the distribution of turbidity values
that were observed before restoration, during restoration, and after restoration. The dates used
for the different restoration periods are:
-pre-restoration: December 14, 2005 - April 2, 2006
-during restoration: April 3, 2006 - May 31, 2006
-post-restoration: June 1, 2006 - August 28, 2007
In this figure, the turbidity value corresponding to the 50% on the y-axis represents the median
turbidity value observed for a given period. These plots can be described as S-curves.
Unusually low turbidity values are on the lower left corner of the plot and unusually high
turbidity values are in the upper right corner of the plot.
4-10
-------
ACCOTINK CREEK TURBIDITY
Magenta Line = Pre Restoration Blue Line = During Restoration Black Line = Post Restoration
100
0.1
10
Turbidity (FNU)
1,000
Figure 4-11. Distribution of Accotink Creek turbidity values before, during, and after restoration
An interesting pattern is observed in these turbidity distribution lines, in that the distributions of
pre- and post-restoration data are similar, while the during-restoration data set indicates an
increase in turbidity concentrations at all frequencies. This increase in turbidity concentrations
during the restoration period is completely consistent with the in-stream patterns that were
observed by field crews during the restoration period; turbidity values were frequently elevated
(relative to the pre-restoration period), as the restoration work disturbed the stream channel and
liberated sediments. While this pattern of increased turbidity levels during the restoration effort
isn't a surprise it is interesting to observe that over the short term, the restoration appeared to
result in increased turbidity levels.
Perhaps more significant than the increase in turbidity observed during the restoration work is
the similarity of the distributions of turbidity values observed in the pre- and post-restoration
periods. The median concentrations observed during these two monitoring periods are
essentially identical, and the overall shape of the distribution curves is almost identical. This
seems to indicate, that overall, the restoration did not appear to result in major changes to the in-
stream turbidity levels. This observation is important because the same conclusion was reached
in the analysis of the discrete water-quality samples; the restoration activities had not had an
4-11
-------
impact on sediment transport within Accotink Creek. Reaching the same conclusion regarding
the effects of the restoration effort in Accotink Creek using these two different monitoring
approaches lends additional support to this particular conclusion for the study.
In the analysis of these distribution curves, it is important to acknowledge that the distribution
plots cannot take into account that the turbidity data were collected during time periods of
differing length, and over differing hydrological conditions. These differing lengths and
hydrological conditions could play a role in causing apparent differences in the distribution of
turbidity values, which makes it that much more interesting that the pre-restoration and post-
restoration turbidity data look so similar.
4-12
-------
Chapter 5 Conclusions
Data collected from Accotink Creek in Fairfax City, Virginia before and after stream bank and
channel restoration of 1800 linear ft (550 m) of degraded stream channel indicates the stream
restoration alone has little effect on improving the conditions of in-stream water quality, stream
bed, and biological habitat within a two year period of time.
Continuous water quality monitoring data showed that temperature of the creek water changed
with season and wet weather flow events as expected. Temperature decreased with increasing
level (i.e., increasing flow) because rain water temperature may be less than creek water
temperature. The highest temperature was observed in July. This indicates that the stream
temperature responds to atmospheric temperature. The rainfall temperature (particularly
associated with cold fronts) appeared to dominate and was not as affected by surface temperature
of impervious areas as one might expect. The other source for this cooling affect can be routing
of runoff through the sewer where the runoff cools to the surrounding temperature of the buried
pipe. The pH stayed close to neutral and ranged between 5 and 10. The pH was not affected by
flow or wet weather events. Turbidity and conductivity appear to be event related with spikes
occurring during wet weather events. Conductivity was higher in winter due to street salting
during frozen precipitation events. Otherwise, conductivity decreased with wet weather flow and
recovered quickly afterwards within 6 hours.
Analysis of discrete samples for chemical constituents such as SSC, SS, COD, total phosphate,
total nitrogen, and ammonia and indicator organisms such as fecal coliform, enterococci, andE.
coli indicated that wet weather concentrations were typically higher than dry weather
concentrations. However, there was neither statistically significant difference in concentrations
between before and after restoration, nor between upstream and downstream of the restoration.
Concentrations of organisms vary with seasons and summer concentrations were significantly
higher compared to other seasons.
Macroinvertebrate data such as for VASCI, HBI, and EPT taxa showed a general improvement
in conditions between pre- and post- restoration for all sites. The differences are statistically
significant for VASCI and HBI indices and EPT taxa. However, all sites are still below the
impairment level, indicating poor water quality conditions. Further because of the large standard
deviation of the invertebrate index score values, there was no statistical significant difference
5-1
-------
between the upstream control sites and the sites within the restoration reach for the VASCI and
HBI indices. This is further complicated by the inherent temporal variation in
macroinvertebrates studies which makes it difficult to determine the effectiveness of restoration
with the short sampling periods such as the ones used in this study. The system which includes
previously restored areas may not have achieved equilibrium, and might take a longer sampling
period to result in greater differences in index scores.
Except for one sampling event before restoration, all the sites had similar total numbers of
macroinvertebrates and there was no statistically significant difference in total macroinvertebrate
abundance over the sampling dates. Most of the species identified were moderately tolerant to
polluted water. There was also no significant difference in the total number of
macroinvertebrates between the upstream, downstream, and restored sites. The upstream,
downstream, and restored areas have similar percent dominance values. The most dominant
species at these sites were Chironomidae, Hydropsychidae, Naididae, and Lumbriculidae,
comprising 87% of the upstream site samples, 93% of the restored area samples, and 92% of the
downstream samples. All other families were relatively rare, most composing of less than 1-2
species.
Macroinvertebrates composition changed after restoration, but did not decrease in abundance.
There were more Chironomidae in all sites than Hydropsychidae before restoration. After the
restoration, it was reversed; i.e., there were more Hydropsychidae than Chironomidae. These
differences are probably due to the disturbance in the restored area caused by the restoration.
The regression equations developed by the USGS to relate a target water quality constituent in
the discrete samples (i.e., SSC or E. coif) to the water quality parameters that are monitored
continuously (i.e., turbidity) using discrete water quality samples collected manually during both
low flow and storm flow periods, in conjunction with the continuously monitored data showed a
stronger relationship between turbidity and suspended sediment than for turbidity and E. coli.
The same conclusion was reached with the U.S. EPA data, but the relationship was much
weaker. No detectable change occurred in the results of the predictive equations before and
after restoration. Also, the median turbidity concentrations observed during the before and after
monitoring periods were essentially identical, and the overall shape of the distribution curves
was almost identical. The lack of change is either because data have not been collected over a
long enough monitoring period, or the restoration activity did not impact a sufficiently
substantial portion of the watershed to reduce sediment transport. This latter conclusion is
supported by the pebble count data, which indicated that very little has changed in the restored
reach.
Summary
One of the three hypotheses tested (Hypothesis #2) in this project was satisfied. The differences
are statistically significant for VASCI and HBI indexes and EPT taxa between before and after
restoration at 90% level of confidence. However, all sites are still below the impairment level,
indicating poor water quality conditions in comparison to Virginia reference streams. Stream
restoration was successful in stabilizing stream banks, preventing bank sloughing and further
5-2
-------
incision. This was important to the infrastructure in the stream restoration area and property
owners of Fairfax City. If one of the goals of stream restoration is to restore habitat and
biological communities, stabilizing banks alone may not enough to bring back species that
depend on good water quality. Reduction of stormwater runoff volumes and associated
pollutants of concern must be addressed through pollution source control and stormwater
retrofits to achieve improved biological outcomes (e.g., detention ponds, swales, downspout
disconnection program, oil grit separators, etc.). Beechie et al. (1996) pointed out that traditional
approaches to aquatic habitat restoration concentrating on repairing or enhancing specific habitat
conditions rather than restoring the landscape processes that form and sustain high quality
aquatic habitats is not effective. Laeser and Stanley (2004) concluded that local restoration in
and around streams are insufficient for improving the water quality of the stream as there were
no changes in nutrient concentrations in association with restoration activities. Many habitats are
a result of change; attempts to fix them at a particular point in space or time fail to recognize that
stream channels are dynamic and that high quality habitats are a product of this dynamism.
Unless larger scale watershed issues are addressed in restoration planning, the current practice of
direct structural modification of channels at the site level is unlikely to reverse aquatic
population declines (Bohn and Kershner, 2002).
It should be noted that the current restoration was limited by the confined area of the stream
section; however, the previous restoration efforts were able to reconnect the stream flood plain
and therefore were able to provide some storage in the flood plain. This project would indicate
that neither the current or previous restoration measures were enough and that further volume
and flow controls are necessary for the runoff further up in the watershed, before it reaches the
stream channel and the modified flood plain to achieve greater habitat restoration.
Restoring healthy ecosystems that have been impacted over the years by human mismanagement
is not an easy task. Restoration requires understanding of factors that caused deterioration of the
ecosystem. Stream restoration alone rather than addressing the whole watershed may yield no
net improvement in the health of aquatic systems. However, stream restoration was successful in
protecting infrastructure and adjacent properties.
Recommendations for Further Action
The study results indicate that the stream restoration did not improve the water quality of this
particular or previously restored reaches. The indications are that the hydrology has not
significantly changed, though the restoration has lessened further degradation to the stream
banks in critical areas. Longer term monitoring may yet prove that streambank erosion is not the
source of continued sediment transport and that the sediment measured in the control and
restored reaches in this study are from inherent sources upstream.
Restoration by design transitioned the stream to a step pool function to accommodate current
flows, while the natural state before watershed development may have been a pool riffle
structure. Because of this change, there may be a shift in the biota type as the ecosystem is in a
continued state of change based on upstream watershed activities that result in the need for
continued adaptation or replacement of tolerant stream biota.
5-3
-------
A fish assemblage study should be performed and Rapid Bioassessment Survey and habitat
assessment should have been performed before restoration (U.S. EPA, 1999). This can still be
done in areas that need to be restored for comparative purposes.
Few run off volume water quantity controls have been implemented prior to, during or even after
restoration activities. A recommendation to improve water quality involves the institution of wet
weather flow controls, upstream in the watershed. Stormwater BMPs, strategically placed
throughout the watershed could reduce and delay discharge to the stream. This coupled with the
existing restoration, may ultimately lead to improved habitat and water quality conditions. A
failure to incorporate Stormwater BMPs and controls will result in continued high and flashy
flows to the Accotink creek. These continuing conditions will wear on the existing restoration,
and ultimately will once again begin to alter the stream channels in ways that will further
degrade the system or even negate the effects of the restoration.
As improvement may not be realized in the two years post-restoration, continued monitoring,
and particularly of macroinvertebrates may be warranted, as indications are that the indices are
potentially still trending to improve. Also, indices for the macroinvertebrates are based on scores
obtained from pristine conditions and these may be unachievable in these disturbed urban
systems. To date there is no index score or attainability level that has been mapped out or
charted for affected urban streams. The Accotink Creek and other restored streams like it may
be approaching the highest macroinvertebrate scores for the type of watershed that now shapes
the creek. Developing relevant index scores for urban and suburban areas may require a larger
study (i.e., at regional and national level) to index and catalog results.
5-4
-------
Chapter 6 References
American Public Health Association, American Water Works Association and Water
Environment Federation. 1998. Standard Methods for the Examination of Water and
Wastewater 20th edn. APHA, Washington, DC.
ASTM. 2007. ASTM D3977 - 97 - Standard Test Methods for Determining Sediment
Concentration in Water Samples. http://www.astm.com/Standards/D3977.htm
Barbour, M.T., Gerritson, J., Snyder, B.D., and Stribling, J.B. 1999. Rapid Bioassessment
Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates, and
Fish. Second Edition. EPA 841-B-99-002. U.S. Environmental Protection Agency, Office of
Water, Washington, D.C. http://www.epa.gov/OWOW/monitoring/techmon.html.
Bennett, H.H., Mullen, M.W., Stewart, P.M., Sawyer, J.A., and Webber, E.C. 2004.
Development of an Invertebrate Community Index for an Alabama Coastal Plain Watershed.
Journal of the American Water Resources Association, Vol. 40, No. 1, pp. 43-51.
Beechie, T., Beamer, E., Collins, B., and Benda, L. 1996. Restoration of habitat-forming
processes in Pacific Northwest watersheds: a locally adaptable approach to aquatic ecosystem
restoration. In The Role of Restoration in Ecosystem Management (D.L. Peterson and C.V.
Klimas, eds), pp. 48-67. Madison, WI: Society for Ecological Restoration.
Bisson, P.A., Quinn, T.P., Reeves, G.H., and Gregory, S.V. 1992. Best management practices,
cumulative effects, and long-term trends in fish abundance in Pacific Northwest river systems.
Pages 189-232 in R. J. Naiman, editor. Watershed Management.
Bohn, B.A. and Kershner, J.L. 2002. Establishing Aquatic Restoration Priorities Using a
Watershed Approach. Journal of Environmental Management, Vol. 64, pp. 355-363.
Collins, C.L., Mullen, M.W., Stewart, P.M., and Webber, E.C. 2008. Validation of an
Invertebrate Community Index for Urban Streams for an Alabama Coastal Plains Watershed.
Journal of the American Water Resources Association, Vol. 44, No. 3, pp. 663-669.
6-1
-------
Department of Public Works and Environmental Services (DPWES). 2001. Fairfax County
Stream Protection Strategy Baseline Study. Stormwater Management Branch, Stormwater
Planning Division, Fairfax County, Virginia.
Fairfax. 2005. http://www.fairfaxva.gov/environment/streams.asp, 10/25/2005.
Fremont, A. 2006. Fairfax City Engineer, City of Fairfax, Virginia. Personal communication.
Hilsenhoff, W.L. 1987. An improved biotic index of organic stream pollution. Great Lakes
Entomology, Vol. 20, pp. 31-39.
Hilsenhoff, W. L. 1988. Rapid field assessment of organic pollution with a family level biotic
index. The Journal of the North American Benthological Society, Vol. 7, pp. 65-68.
Hilsenhoff, W.L. 1998. A modification of the biotic index of organic stream pollution to
remedy problems and to permit its use throughout the year. Great Lakes Entomology, Vol. 31,
pp. 1-12.
Hooper, A. 1993. Effects of season, habitat, and an impoundment on twenty five benthic
community measures used to assess water quality. M.S. Thesis, College of Natural Resources,
University of Wisconsin-Stevens Point. 443 pp.
Klein, R. 1979. Urbanization and Stream Quality Impairment. American Water Resources
Association. Water Resources Bulletin, Vol. 15, No. 4.
Laeser, S.R. and Stanley, E.H. 2004. Evaluation of effects of riparian and streambank
restoration activities on water quality in a southern Wisconsin agricultural stream. Presented at
the NABS Annual Meeting, Vancouver, British Columbia.
The Louis Berger Group. 2005. City of Faifax Watershed Management Plan. July.
Merritt, R.W. and Cummins, K.W. (editors). 1996. An introduction to the aquatic insects of
North America, 3rd ed. Kendall/Hunt Publishing Company, Dubuque, Iowa.
Natural Resources Conservation Service. 1998. Stream Corridor Restoration, Principles,
Processes, and Practices. United States Department of Agriculture. NEH 653, Washington, D.C.
Peckarsky, B. L., Fraissinet, P., Penton, M.A., and Conklin, Jr., D.J. 1990. Freshwater
macroinvertebrates of Northeastern North America. Cornell University Press, Ithaca, NY. 442
pp.
Pennak, R.W. 1989. Freshwater invertebrates of the United States, 3rd ed. J. Wiley & Sons,
New York.
Pond, G. 2008. Personal communication. U.S. EPA Region 3 Wheeling Laboratory.
6-2
-------
Reeves, G.H., Hohler, D.B., Hansen, B.E., Everest, F.H., Sedell, J.R., Hickman, T.L., and
Shively, D. 1997. Fish habitat restoration in the Pacific Northwest: Fish Creek of Oregon. Pages
335-359 in Williams, I.E., Wood, C.A., Dombeck, M.P., editors. Watershed restoration:
principles and practices. American Fisheries Society, Bethesda, Maryland.
Roni, P., Beechie, T.J., Bilby, R.E., Leonetti, F.E., Pollock, M.M., Pess, G.R. 2002. A Review
of Stream Restoration Techniques and a Hierarchical Strategy for Prioritizing Restoration in
Pacific Northwest Watersheds. North Am. J. Fish. Manage., Vol. 22, pp. 1-20.
Schueler, T. 1994. The importance of imperviousness. Watershed Protection Techniques, Vol.
l,No. 3.
Schueler, T. 1995. The importance of imperviousness. Watershed Protection Techniques, Vol.
l,pp. 100-111.
Sholar, CJ. and Shreve, E.A. 1998. Quality-Assurance Plan for the analysis of fluvial sediment
by the northeastern region, Kentucky District Sediment Laboratory. U.S. Geological Survey
Open-File Report 98-384, 27 pages.
State Water Control Board. 2007. 9 VAC 25-260 Virginia Water Quality Standards.
Strecker, E., Urbonas, B., Quigley, M., Howell, J., and Hesse, T. 2002. "Urban Stormwater
BMP Performance Monitoring: A Guidance Manual for Meeting the National Stormwater BMP
Database Requirements" Prepared by GeoSyntec Consultants, Urban Drainage and Flood
Control District, and Urban Water Resources Research Council of ASCE, in cooperation with
Office of Water, US EPA. Report No. EPA-821-B-02-001.
Struck, S.D., Selvakumar, A., Borst, M. 2006. Performance of Stormwater Retention Ponds and
Constructed Wetlands in Reducing Microbial Concentrations. Office of Research and
Development, Washington, D.C. EPA-600-R-06-102.
Thorp, J.H. and Covich, A.P. (editors). 1991. Ecology and Classification of North American
Freshwater Invertebrates. Academic Press, New York, New York.
U.S. EPA. 1992. NPDES Storm Water Sampling Guidance Document. United States
Environmental Protection Agency, Office of Water, Washington, D.C. EPA-833-B-02-001.
U.S. EPA. 1999. Rapid Bioassessment Protocols for Use in Streams and Wadeable Rivers:
Periphyton, Benthic Macroinvertebrates, and Fish. Second Edition. United States
Environmental Protection Agency, Office of Wetlands, Oceans, and Watersheds, Washington,
D.C. EPA-841-B-99-002.
U.S. EPA. 2000. Storm Water Phase II Final Rule: An Overview. United States Environmental
Protection Agency, Office of Water, Washington, D.C. EPA 833-F-00-001.
6-3
-------
U.S. EPA. 2002a. National Water Quality Inventory - 2000 Report. United States
Environmental Protection Agency, Office of Water, Washington, D.C. EPA-841-R-02-001.
U.S. EPA. 2002 b. Method 1603: Escherichia coli (E. coli) in Water by MembraneFiltration
Using Modifiedmembrane-Thermotolerant Escherichia coli Agar (Modified mTEC). United
States Environmental Protection Agency. Office of Water, Washington, D.C. EPA 821-R-02-
023
U.S. EPA. 2007. National Water Quality Inventory: Report to Congress - 2002 Reporting
Cycle. United States Environmental Protection Agency. United States Environmental
Protection Agency, Office of Water, Washington, D.C. EPA-841-R-07-001.
U.S. Geological Survey (USGS). 1998. National Field Manual for the Collection of Water-
Quality Data: Techniques of Water-Resources Investigations, Book 9, variously paginated.
Available on the internet: http://water.usgs.gov/owq/FieldManual/index.html
USGS. 2003. UseoftheHydrological Simulation Program - FORTRAN and Bacterial Source
Tracking for Development of the Fecal Coliform Total Maximum Daily Load (TMDL) for
Accotink Creek, Fairfax County, Virginia. Water Resources Investigation Report 03-4160.
VADEQ. 1998. Virginia's 1998 303(d) Total Maximum Daily Load Priority List and Report.
Virginia Department of Environmental Quality, Richmond, VA.
Virginia Department of Conservation and Recreation. 2007. Virginia Agricultural BMP Cost-
Share Program History. Richmond, VA.
Wagner, R.J., Mattraw, H.C., Ritz, G.F., and Smith, B.A. 2000. Guidelines and Standard
Procedures For continuous Water-quality Monitors: Site Selection, Field Operation, Calibration,
Record Computation, and Reporting. U.S. Geological Survey Water-Resources Investigations
Report 00-4252. Available online at: http://pubs.usgs.gov/wri/wri004252.
Wolman, M.G. 1954. A method of sampling coarse river-bed material. Transactions of the
American Geophysical Union (EOS), Vol. 35, pp. 951-956.
6-4
------- |