United States
Environmental Protection
Agency
EPA/600/F-11/004 I March 2011 I www.epa.gov/ord
Experimental Stream Facility
Design and Research
Providing Research Solutions to
Manage Water Quality
Office of Research and Development
National Risk Management Research Laboratory
-------
Background
The Experimental Stream Facility (ESF) is a valuable research
tool for the U.S. Environmental Protection Agency's (EPA)
Office of Research and Development's (ORD) laboratories in
Cincinnati, Ohio. This brochure describes the ESF, which is
one of only a handful of research facilities in the United States
designed to conduct small stream research.
The Need for Small Stream Research
Small stream ecosystems comprise over 72% of the river miles
in the U. S.; yet, the role they play in managing watershed-
level water quality is uncertain. While the potential impact of
naturally occurring biogeochemical functions within an individual
small stream may seem minimal, the accumulated impact of the many small streams within a watershed may be
significant. Small streams are largely unmapped, unregulated, overlooked in watershed models, and generally not
considered in the total maximum daily load (TMDL) process or in land use regulations. As such, they are often
ignored in land development plans, and are often replaced with culverts, storm sewers, or lined ditches, which
eliminates any role they may play in maintaining water quality. Research is needed to increase our understanding
of the relative importance of small stream ecosystems and the services they provide for effective watershed
management.
In addition to watershed management research, there is a continual need to better characterize how streams and
rivers, in general, react to and process emerging contaminants and stressful mixtures. The numerous physical,
chemical, and biological variables interacting under natural conditions, as aquatic ecosystems are exposed
to contaminants of concern, confound the interpretation of effects on aquatic biology in receiving streams.
Mesoscale studies are needed to gain a better understanding of the impact of contaminant stressors on multiple
trophic levels and the levels which are the most sensitive to stress.
As a final point, biotic indicators form the basis for stream
condition assessment in this Nation, but current design criteria
for watershed best management practices (BMPs) are based
largely on peak flow control and sediment removal. Research
is needed to begin to bridge the gap between physical-based
engineering design and biological-based assessment criteria.
The Need for the Stream Facility
The facility has eight stream mesocosms; a mesocosm is an
experimental system designed to simulate natural conditions,
often by using naturally occurring organisms and artificial Experimental stream mesocosms
The Experimental Stream Facility 1
-------
The Experimental Stream Facility
structures that simulate nature as closely as possible. Mesocosm design
provides a balance between the benefits of a controlled laboratory study
(e.g. controlled flow rates, solar irradiance and streambed type) and the
benefits of a field study (e.g. continuous biota for colonization, and
natural variability of biota and water quality). This allows researchers
to study how pollutant loads interact with important characteristics of
stream habitat that may or may not change as a result of anthropogenic
stress. Changes to the stream ecosystem structure and function can be
measured and observed in ways that are not possible in field or laboratory
studies. Well designed and controlled experimental stream mesocosm
studies can represent a primary tool for providing the process-level
understanding necessary to move basic research and development from
the laboratory bench-top to field applications.
Research Approach
ESF studies are designed to obtain information on both watershed
management and the impact of contaminants of concern.
Watershed Perspective
The ESF sits at the downstream end of the watershed of the East Fork
of the Little Miami River (1295 km2 / 500 mi2), and experiments are
designed with watershed scale management as one focus of the studies.
Studies at the ESF afford EPA the opportunity to collect and analyze
data on in-stream process-level biological, chemical, and physical
interactions. Each ESF study is designed to provide information on
small stream ecosystem structure and function that can be used for the
development and testing of: new indicators of ecological stress; water
quality monitoring technologies; watershed monitoring strategies; and
water quality management methods and models. Characteristics of small
stream sub-watersheds in the East Fork watershed provide information
for the design of ESF experiments, and the results of ESF experiments are
subsequently incorporated into a multi-spatial scale watershed research
plan designed to characterize, track, and model changes to water quality
-------
associated with the National Pollutant Discharge Elimination System (NPDES) and the Safe Drinking Water Act
-associated BMP implementation projects. Hence, an overarching goal is to develop numerical descriptions of
these stressor-biotic linkages for watershed models used as decision support in TMDL and TMDL-like analyses of
watershed management.
Contaminant Perspective
The ESF mesocosms are especially useful for studying the impacts of contaminants of concern (e.g. endocrine
disrupters, pharmaceuticals, pesticides) on the natural stream biota. Contaminants can be added at
environmentally relevant levels; levels observed in natural ecosystems and reported in the literature. Exposure
at these levels lets researchers observe more subtle changes (e.g. gene and protein expression) in stream biota as
well as changes in biotic communities and ecosystem structure and function. This approach provides additional
information to laboratory studies which commonly use lethal effects (e.g. LC50) as the endpoint. The post-
dose recovery period built into each experiment provides additional information on the ability of the biota to
recover when the contaminant is removed from the ecosystem. The relative importance of specific changes in
stream habitat (e.g. sediment load, flow rates) on contaminant impact can be simulated and are often included
as interacting variables. Additional biotic indicators can be studied through the use of emergent traps, fish
deployments, in-stream invertebrate exposures, and real time biomonitoring. Interactions and effects observed in
the ESF setting help identify important assessment endpoints, and inclusion functions for field-scale models of the
fate and transport.
Experimental Framework
Natural river water flows through the ESF, which allows studies to be conducted when stream biota are most
abundant and active, approximately mid-March through late October. Typically two experiments are conducted
per year, with each experiment running for about 12-15 weeks: 4-6 weeks for colonization of the streams by
natural biota; 4-6 for applying a stressor(s) such as a chemical dose or flow manipulation; and 3-4 weeks for stream
recovery after the stressor period ends.
Base Analytical Endpoints
The baseline analytical endpoints for each
experiment include: sediment size fractions;
carbon, nitrogen, and phosphorus in
sediments; inter-gravel water nutrient species
concentrations; surface water temperature,
conductivity, pH, dissolved oxygen, oxidation/
reduction potential, and turbidity; stored
organic matter; invertebrate community
structure; periphyton (a gelatinous matrix,
also known as a biofilm, that may consist
of attached algae, bacteria, fungi, trapped
fine sediment and deterital material, and
The Experimental Stream Facility
-------
The Experimental Stream Facility
microinvertebrates that use it both as a food source and habitat); per capita invertebrate drift emigration rates;
periphyton chlorophyll-a concentrations; carbon, nitrogen, and phosphorus in periphyton; surface water nutrient
species concentrations; and algal community structure. These data are acquired and managed across all ESF
studies, and are likewise a part of the associated field research. They
serve as a basis for meta-analyses, scaling ESF results to the field, or
comparing against novel indicators of exposure and stress.
The Research Team
National Risk Management Research Laboratory (NRMRL) is
responsible for the operation, maintenance, and facilitation of the
experiments at the ESF. However, the experimental designs are
developed and implemented by a multi-disciplinary team of research
scientists and engineers, and technical support personnel from across
EPA's ORE) laboratories. The lead for each experiment can be a
researcher from any of the EPA laboratories. Using the combined
expertise of the multi-disciplinary team members optimizes
coordination, leverages resources, and encourages collegiality in the
integration of both risk assessment and risk management objectives
into novel experiments.
The Experimental Stream Facility
The facility, located in Milford, Ohio, was originally built by the
Proctor & Gamble Company, and was well known as a premier
laboratory in the area of mesocosm research in ecotoxicology and risk
assessment. The facility is currently owned by Clermont County,
and became available to EPA in 2005, through a lease with the
County (some lab and office space is used by the Clermont County
Sewer District Water Quality Testing Laboratory). Since taking
occupancy, EPA has made several modifications to the facility which
make it a unique small stream research laboratory. The capabilities of the facility have been extended for research
applications in the fields of ecohydrology/sedimentology, biogeochemistry and risk management.
Experimental Stream Configuration
Each stream mesocosm includes a head tank, an upper biotic colonization channel, a lower biotic colonization
channel, a tail tank, and a water recirculation system (see figure, page 4), and is 12 meters (m) long, or 39 feet (ft).
The upper channel typically has a tile streambed, and the lower channel usually has a gravel streambed, but the
channels can be interchanged. A small pool section can be added between the upper and lower channels. Plastic
or stainless steel mesocosms are available, depending upon the physiochemical properties of the contaminant
under study. A removable baffle to subdivide the channel sections into like units is optional for increasing
statistical replication.
Scientist and engineers
-------
Head Tank
All source water, chemical dose, and
recirculated water flows pass through an in-line
mixer into the bottom of the 126 liter (L), or
33 gallon (gal), head tank. Water flows over a
rectangular weir into the upper channel section.
The head tank is mounted on a jacking system
which can raise or lower the head tank, which
allows the channel slope to be varied between
0 and 4.3% to simulate different stream flow
conditions.
Head Works
1) Head Tank
2) Main River Water Influent
3) Alternate Water Source
Influent
4) Flow Meter-Transmitter
5) Actuator
6) Chemical Dose Influent
7) Scissor Jack-Slope Control
Tile Section
The upper channel is usually configured to simulate large, flat rocks in a
glide reach of a natural stream where periphyton attach and grow. The
channel is approximately 30 centimeters (cm), or 12 inches (in), wide; 9
cm (3.5 in) deep; and 4.5 m (14 ft) long. It is usually lined with 47 rows
of unglazed ceramic tiles, with 3 tiles per row. The tiles provide a defined
surface area, as well as a removable substrate for the growth of periphyton.
Water moves through this section in a more laminar flow at a depth of
2.5 cm (1 in). The tiles are sampled randomly during an experiment to
derive several measurements, such as algal biomass, stored nutrients, and
periphyton community structure.
1. Tiled sampled in low light channel
tile section
?. Tiled sampled in highlight channel 3. Cyanobacteria in
The Experimental Stream Facility
-------
The Experimental Stream Facility
Gravel Tray Section
The lower channel is usually configured to simulate a riffle in a natural stream where small rocks and gravel
provide both substrate for periphyton and shelter for aquatic invertebrates. The channel is approximately 50 cm
(21 in) wide, 17 cm (6.5 in) deep, and 4.5 m (14 ft) long. It contains an array of gravel-filled plastic trays in 15
rows, with 3 trays per row (mesh baskets can be used instead of plastic trays to allow interflow between sampling
units.). Gravel (small river rock) ranging in size from 1.5 to 4 cm (0.6 to 1.6 in) diameter has been used to date,
but substrates ranging from fine silt/clays to smaller gravel could be used depending on experimental objectives.
Each gravel tray provides a removable substrate of defined volume and area, which can be used to quantify many
of the base analytical endpoints.
Water moves through this section
in a more turbulent flow at a depth
of 5 cm (2 in), and water jets
can be configured in this section
to simulate peak flow velocities
during storm events. Gravel
trays are also sampled randomly
during an experiment. Similar
gravel trays are used at several field
sampling locations in the East
Fork Watershed, and provide the
same base analytical endpoint data.
Standard operating procedures
have been written for sampling,
processing, and analyzing all
variables obtained from the gravel
trays and the tiles.
1. Trays at the start of experiment 2. Trays with accumulation & recently replaced trays
3a-c. Gravel biota 4. Samplers are buried in gravel 5. Sampled tray
-------
Drift Measurements
Invertebrate drift responses can be measured by
positioning drift nets to isolate the gravel section.
The upstream net secludes influent drift, while the
downstream net catches all organisms drifting from
the gravel during the measurement period. Per capita
emigration is determined as the number of individuals
captured by the downstream net divided by their
respective density counted in the gravel section.
Tail Tank
A 222 L (59 gal) tail tank at the end of each mesocosm
simulates conditions of a pool in a natural stream
channel. The tail tanks house continuous clam-based
behavioral biomonitoring and real-time water quality
sensing equipment, and can be configured for exposing
larger stream biota (i.e., fish). Side-streams are plumbed
in-line with the discharge from each mesocosm to supply
flow to additional small tanks for other biotic exposure
options. The biomonitors allow behavioral responses to
be tracked in relation to the measured changes in stream
structure and function. Outflow from the tail tank can
be discharged under an NPDES permit or sent to the
neighboring wastewater treatment plant. A new version
of clam biomonitors is currently being tested.
Upstream Drift Net
1. Electronics
2. Plunger
3. Clam
4. Adhesive
5. Pedestal
6. Adjustable Height
Tail tank with exposure cage
Clam biomonitor housing in tail
tank
Clams in biomonitor; plunger moves
proportional to shell gape
The Experimental Stream Facility
-------
The Experimental Stream Facility
Recirculation System
Recirculation loops with pumps can return water at a constant pre-set flow rate from the tail tank to the head tank
for each mesocosm, which allows residence time to be manipulated while maintaining velocities and turbulent
features for the biotic colonization sections. This can be important for studies considering the fate and transport
of stream solutes. The recirculation flow can also be diverted to the head tank of an adjacent mesocosm so that the
effluent of one mesocosm can serve as influent to another, allowing up to eight streams to be studied in series.
1. Tail tanks 2. Recirculation line from tail tank to pump 3. Flow meter on recirculation loop
Unique ESF Features
There are several unique features that make the ESF a one-of-a-kind facility for conducting controlled, flow-
through, meso-scale simulation studies of stream ecosystems. The process flow diagram below depicts the key
components and linkages of the process-controlled systems.
Multiple Water Sources
Water is pumped into four ESF
supply tanks from several sources,
and flows by gravity from the
supply tanks to each mesocosm.
Each mesocosm has two sets of
piping, flow meters, and computer-
controlled valves to allow water
from two supply tanks to flow to the
head tank of each mesocosm. Flow
through one set can be controlled
within 0.38 liters/minute (L/min),
or 0.1 gallons per minute (gpm), in
the range of 19 to 190 L/min (5 to
50 gpm). Flow though the other set
can be controlled within the range of
-------
East Fork River intakes
East Fork River main channel
ESF water supply tanks
0 to 57 L/min (0 to 15 gpni). This flow control, along with the ability to adjust the mesocosm slopes, allows for a
range of controllable in-stream velocities.
Natural water flow can be supplied from two nearby sources: Heiserman Stream, a relatively unimpacted
headwater stream, or the East Fork, a 6th order river channel. The East Fork water has two to three times the
carbon and nutrient concentration of the Heiserman water. The availability of water from such different stream
sources expands the natural simulation horizon of ESF experiments.
Tertiary wastewater effluent (oxidation ditches followed by
clarification and sand filtration) from the adjacent Lower East Fork
Wastewater Treatment Plant (WWTP) is pumped to an ESF supply
tank which can be aerated to avoid stagnation. Excess wastewater is
routed back to the WWTP. The availability of wastewater allows the
effects of emerging contaminants in wastewater to be studied alone or
in combination with other variables. Finally, carbon-treated tap water
is also available for experiments.
Variable Simulated Solar Irradiance
High intensity grow lights provide simulated sunlight to fuel primary
production in ESF mesocosms. The ceiling mounted lights provide
daily-integrated irradiance of approximately 12% of unshaded (open-
field) photo synthetically active radiation (PAR), which is similar to
that experienced by a forested stream where leaves in the tree canopy
shade the channel. Additional lights can be configured to provide
daily-integrated irradiance of approximately 100% of unshaded
(open-field) PAR to simulate a stream flowing through an agricultural
field or developed area without a forested riparian buffer. The lights
are programmed to run on a 24-hour cycle based on the ESF's
geographical location.
Grow lights - low light channels
High light channel
The Experimental Stream Facility 9
-------
The Experimental Stream Facility
Continuous Water Quality and Climate Monitoring
Each mesocosm has sensors in the tail tank that measure the
surface water quality base analytical endpoints listed previously.
The sensors continually transmit the water quality data for each
mesocosm at five minute intervals to the data acquisition system.
A complementary set of these sensors monitors the influent waters
to the facility. The ESF has an outdoor weather station and
indoor sensors to continuously monitor PAR, air temperature, and
humidity. Precipitation is continuously tracked by the outdoor
station as well.
Chemical Dosing System
Chemical doses of stressors/pollutants can be metered into the
head of each mesocosm from the chemical delivery system. Doses
are continuously mixed in 570 L (150 gal) stainless steel tanks.
The chemical feed pumps are accurate to within 0.5 milliliters/
minute (ml/min), or 0.2 gallons per day (gpd), within the range of
25 to 150 ml/min (9.5 to 57 gpd), and are integrated with pressure
transducers to ensure even pumping and fail safe operation. They
work together with meters and valves on the river flow lines to
provide constant in-stream dose concentration over extended
dosing periods (e.g. 30 days, typically).
Continuous water quality monitoring equipment
Chemical dosing tanks
-------
Supervisory Control and Data Acquisition
A supervisory control and data acquisition (SCADA) system
performs several functions:
• monitors/controls natural river, wastewater, and
recirculation flows;
• activates/controls the timing of the simulated solar
irradiance;
• monitors/controls the chemical delivery system;
• acquires data from all the continuous water quality and
behavioral monitoring sensors; and
• initiates phone calls to the operators in case of any system
failure (e.g. low water flow).
ft
®
H
H
©
'' I I
0 © 1^] ©
1J
©• ®
Screenshot ofSCADA System
Data files generated within the SCADA system are transferred to a data storage device, in which data management
functions are housed. An effort is underway to make data accessible to all collaborators via email or the web.
Programmable Environmental Triggers
Logic control algorithms can be written for the SCADA system that trigger changes in flow rate, chemical dosing,
or flow distribution (river inflow vs. recirculation) in response to natural changes in climate or water quality.
Some examples include mesocosm flow regimes that can be programmed to change at the onset of a rain event, as
detected by the weather station, or a change in water quality, as detected by one of the real-time sensors in the tail
tank. The timing and proportion of flow that is recirculated can be used to manipulate sediment accumulation,
colonization rates, and/or discharge rate from each mesocosm. Contaminant dosing can be programmed to be
continuous or intermittent and triggered by a rain event.
Significance
Understanding the impact of contaminants on small streams, and reducing the loading of stressors to watersheds,
is of concern to environmental stakeholders, including local and state governments, utilities, developers, and
homeowners. Knowing how contaminants of concern affect aquatic biota will help determine which stressors
are most important for the consideration of environmental regulators. Adoption of the watershed approach and
best management practices to control both urban and rural sources of waterborne pollutants is helping to reduce
contaminants at the watershed level. The chief beneficiaries are the environmental decision makers who will use
the ESF data in watershed models to better characterize how streams react to, and process, emerging contaminants
and stressful mixtures, and to quantitatively link known stressors in stream flow with the structure and function of
stream ecosystems.
The Experimental Stream Facility
-------
The Experimental Stream Facility
Contacts
Technical Contact
Christopher Nietch PhD
Systems Ecologist
U.S. Environmental Protection Agency
Office of Research & Development
National Risk Management Research Laboratory
Water Suppy & Water Resources Division
(513) 569-7460 nietch.christopher@epa.gov
Communications
Michelle Latham
Biologist/Communications Specialist
U.S. Environmental Protection Agency
Office of Research & Development
National Risk Management Research Laboratory
Water Supply & Water Resources Division
(513) 569-7601 latham.michelle@epa.gov
EPA Scientist leading a tour
at the Experimental
Stream Facility.
Resources
Website
Experimental Stream Facility
http://www.epa/nrmrl/wswrd/wqm/esf/esf.html
Video
U.S. EPA (2009) Experimental Stream Facility -
Design and Research
http://www.youtube.com/user/USEPAgov#play/
uploads/5/jwO!Yv7heE4
Fact Sheet
Experimental Stream Facility
(EPA/600/F-08/006)
http://www.epa.gov/ord/NRMRL/pubs/
facilities/600f08006.pdf
-------
SEPA
United States
Environmental Protection
Agency
Office of Research and Development
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
Official Business
Penalty for Private Use
$300
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERM IT NO. G-35
Printed with vegetable-based ink on paper that
contains a minimum of 50% post-consumer fiber
content and processed chlorine free.
------- |