The Use of Wetlands
^ot-Controlling
Bj ^-^
lutio i
-------
The Use of Wetlands for
Controlling Stormwater Pollution
reprinted by:
U. S. E.P.A., Region 4, Wetlands Section, Water Division
61 Forsyth St., S. W.
Atlanta, Georgia, U. S. A. 30303-3415
April, 1997
PREPARED BY
Eric W. Strecker
Joan M. Kersnar
Eugene D. Driscoll
Woodward-Clyde Consultants
111 S.W. Columbia, Suite 990
Portland, Oregon 97201
AND
Dr. Richard R. Horner
University of Washington
Seattle, Washington
Technical Advisor
Thomas E. Davenport
U.S. EPA, Region V
77 West Jackson Boulevard
Chicago, Illinois 60604
DISTRIBUTED BY
The Terrene Institute
1700 K Street, NW, Suite 1005
Washington, DC 20006
April 1992
-------
ACKNOWLEDGMENTS
The authors of this revised and updated edition of this report were Mr. Eric W. Strecker, Ms. Joan M.
Kersnar, and Mr. Eugene D. Driscoll of Woodward-Clyde Consultants, and Dr. Richard R. Homer of the
University of Washington. Authors of an earlier edition also included Mr. Gary Palhegyi and Ms. Joan
Duffield of Woodward-Clyde Consultants.
Mr. Thomas E. Davenport (Region V Water Division, Wetlands and Watershed Section, Watershed Man-
agement Unit, U.S. Environmental Protection Agency, Chicago, IL) was the EPA Technical Advisor. His
support and guidance for this effort are gratefully acknowledged. Comments received from Ms. Nancy
Phillips of EPA Region V and Mr. Robert Good of EPA Headquarters on the draft report were sincerely ap-
preciated. Cover photo contributed by Dr. Harvey Olem, Olem Associates, Hemndon, Virginia.
Produced by The Terrene Institute under cooperative agreement X-995048 with the U.S.
Environmental Protection Agency Region V. Points of view expressed in this report do
not necessarily reflect the views or policies of The Terrene Institute or EPA, nor does any
mention of trade name and commercial products constitute endorsement of their use.
For copies of this publication, contact
The Terrene Institute
1700 1< Street, NW, Suite 1005
Washington, DC 20006
(202) 833-8317
Limited copies of a Technical Appendix are also available from the Institute. The ap-
pendix contains an annotated bibliography with summaries of the documents
selected for detailed review in this report.
Printed on recycled paper.
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1
1.1 Purpose of Study 1
1.2 Definitions of Wetlands 1
1.3 Sources of Information 3
1.4 Report Organization 3
2.0 REPORTED PERFORMANCE OF WETLANDS
FOR STORMWATER TREATMENT 7
2.1 Pollutant Removal Mechanisms 7
2.1.1 Sedimentation 7
2.1.2 Adsorption 8
2.1.3 Precipitation and Dissolution 9
2.1.4 Filtration 9
2.1.5 Biochemical Interactions 9
2.1.6 Volatilization and Aerosol Formation 10
2.1.7 Infiltration 10
2.2 Wetland Stormwater Pollutant Removal Efficiencies 10
2.3 Probable Causes of Variation and Dissimilarities with Wetland Performance. 15
2.4 Comparison of Factors Affecting Reported Treatment Efficiencies 19
2.5 Assessment of the Reliability of Wetland Data 29
2.6 Summaiy 31
3.0 NOTED IMPACTS OF STORMWATER RUNOFF ON WETLANDS BIOTA. 33
3.1 Introduction 33
3.2 Hydrologic Impacts 33
3.2.1 Impacts on Wetland Morphology 34
3.2.2 Impacts on Plants 34
3.2.3 Impacts on Animals 34
3.3 Accumulation of Toxins 35
3.3.1 Accumulation in Sediments 36
3.3.2 Accumulation in Plants 36
3.3.3 Accumulation in Animals 37
3.4 Need for Further Studies 37
4.0 COMPARISON OF WETLAND AND DETENTION BASIN PERFORMANCE.... 39
4.1 Introduction 39
4.2 Case Studies 39
4.2.1 The Orange County Treatment Facility 40
4.2.2 The Pittsfield-Ann Arbor and Swift Run Systems 41
4.2.3 The Lake Apopka Reservoir and Flooded Field Experiment 42
4.2.4 The McCarrons Treatment System 44
4.3 Summary 45
5.0 IMPROVING THE PERFORMANCE OF WETLANDS 46
1
-------
TABLE OF CONTENTS ( Continued)
Section Page
6.0 ISSUES RELATED TO THE USE OF WETLANDS
FOR STORMWATER POLLUTION CONTROL 50
6.1 Constructed versus Natural Wetlands 50
6.2 Additional Studies 50
6.3 Known Studies Currently Under Way 51
7.0 REFERENCES 56
8.0 ABBREVIATIONS 65
9.0 MEASUREMENT UNITS - ABBREVIATIONS
AND CONVERSION FACTORS 65
APPENDIX (separately bound)
LIST OF TABLES
Table Page
1 Literature Researched to Investigate Performance Characteristics of Wetlands 4
2 Average Removal Efficiencies for Total Suspended Solids and Nutrients in
Wetlands Reported in the Literature 11
3 Average Removal Efficiencies for Metals and Oil and Grease in
Wetlands Reported in the Literature 12
4 Wetland Geographic and Hydraulic Characteristics 16
5 Comparison of Reported Removal Rates for Constructed and Natural Stormwater
‘ Vetlands 20
6 Comparison of Reported Removal Rates for Constructed and Natural Stormwater
Wetlands Sized Less-Than and Greater-Than 2% of the Contributing Drainage Area 26
7 Comparison of Reported Removal Rates for Constructed Storm water Wetlands
Sized Less-Than and Greater-Than 2% of the Contributing Drainage Area 27
8 Sampling Characteristics from the Wetlands Reviewed 30
9 Suggested Reporting Information for Studies that Assess the Ability of Wetlands to Treat
Stormwater Pollution 52
11
-------
TABLE OF CONTENTS (Concluded )
LIST OF FIGURES
Figure Page
1 Location of Wetlands Researched for Their Ability to Treat Stormwater Runo . 5
2 PollutantRemovalRatesforTSSandTP 13
3 Pollutant Removal Rates for NH3 and Pb 14
4 Box Plot Percentiles Comparison of Site Average Pollutant Removals for Natural
and Constructed Wetland Systems 21
5 Comparison of Site Average Pollutant Removals for Natural and Constructed Wetland
Systems:
Connected Percentiles for (a) Total Suspended Solids and (b) Total Phosphorus 22
6 Comparison of Site Average Pollutant Removals for Natural and Constructed Wetland
Systems: Connected Percentiles for (a) Ammonia and (b) Total Lead 23
7 Average Site Pollutant Removal Comparisons for All Wetlands with Less Than 2% and
Greater Than 2% Wetland to Watershed Area Ratios (WWQR): (a) Scatter Plot for TSS
and TP and (b) Percentile Box Plots for TSS, TP, and TPb . 25
8 Average Site Pollutant Removal Comparisons for Constructed Wetlands With Less Than
2% and Greater Than 2% Wetland to Watershed Area Ratios (DAR): (a) Scatter Plot for
TSS and TP and (b) Box Percentiles 28
111
-------
1.0
INTRODUCTION
1.1 PURPOSE OF STUDY
Wetlands are receiving increasing attention as attractive systems for removing pollutants from
scormwater surface runoff before the runoff enters downstream lakes, streams, and other open
water bodies. Wetlands have long been employed for the treatment of wastewaters from
municipal, industrial (particularly acid mine drainage), and agricultural sources (Hammer 1989).
The Environmental Protection Agency (EPA) encourages the use of constructed wetlands for
water pollution control, through the innovative and alternative technology provisions of the
construction grants program (Bastian 1989). The use of natural wetlands for treatment of
wastewater and stormwater is regulated by EPA, the Corps of Engineers and various state
agencies. Although the use of wetlands for treatment of point source wastewater discharges is
fairly extensively documented, little information has been compiled on the use of wetlands for
treating pollution from stormwater runoff.
The purpose of this document is to provide information that will assist EPA, state, and local
technical personnel to assess the capabilities and limitations of the use of wetlands as a control
measure for reducing the environmental impacts of stormwater pollution in downstream water
bodies. The approach taken was to summarize and evaluate information developed through a
review of published literature and reports dealing with the aspects of wetland design, operation,
and performance. The reviews were focused on information concerning the ability of these
systems to serve as control measures for stormwater pollution. The report summarizes the
available information on the effect of stormwater pollutants on wetlands and presents a summary
of recommendations regarding the use of wetlands for stormwater pollution control. The
document focuses specifically on wetlands that receive stormwater runoff, rather than wetlands
that are used to treat wastewaters. It includes information from studies on both natural and
constructed wetlands. However, it is important to note that EPA Region V discourages the
use of natural wetlands for stormwater control; inclusion of information on natural
wetlands is for the purpose of providing information to the reader.
1.2 DEFINITIONS OF WETLANDS
Several legal and administrative definitions of the term “wetlands” are in existence. A definition is
included in the federal agency regulations for the Clean Water Act, the Food Security Act of 1985,
and the Emergency Wetlands Resources Act of 1986. This definition is given as follows (33
CFR 328.3(b); 1986):
“The term ‘wetlands’ means those areas that are inundated or saturated by surface
or ground water at a frequency and duration sufficient to support, and that under
normal circumstances do support, a prevalence of vegetation typically adapted for
life in saturated soil conditions. Wetlands generally include swamps, marshes,
bogs, and similar areas.”
1
-------
The U.S. Fish and Wildlife Service (USFWS), in cooperation with other federal and state
agencies, private organizations, and individuals, developed the following definition (Cowardin et
al. 1979):
“Wetlands are lands transitional between terrestrial and aquatic systems where the
water table is usually at or near the surface or the land is covered by shallow
water. For purposes of this classification, wetlands must have one or more of the
following three attributes: (1) at least periodically, the land supports
predominantly hydrophytes, (2) the substrate is predominantly undrained hydric
soil, and (3) the substrate is non-soil and is saturated with water or covered by
shallow water at some time during the growing season of each year.”
This definition was initially used by the USFWS as the basis for the National Wetlands
Inventory. According to this definition, an area must satisfy one or more of these conditions to be
considered a wetland. Thus, a tidal flat with wetland soil and hydrology but no vegetation would
meet this definition of a wetland. Other definitions differ from the USFWS defmition by
excluding unvegetated areas and requiring that wetland hydrology, hydric soils, and
predominantly hydrophytic vegetation would have to all exist to signify a wetland.
To resolve the conflict in definitioh, four federal agencies agreed on a common methodology in
early 1989 (USFWS, EPA, Corps of Engineers, and SCS). The inter-agency methodology,
which has been adopted by other government agencies and jurisdictions, states that wetlands
“possess three essential characteristics: (1) hydrophytic vegetation, (2) hydric soils, and (3)
wetland hydrology, which is the driving force creating all wetlands.” Under this approach, an
area must meet all three technical criteria in order to be considered a wetland. Even though the
Congress has since directed the Corps to revert to the 1987 wetlands manual, it still preserves this
three parameter approach.
In general terms, areas of shallow water and areas with saturated soil, which are dominated by
water-tolerant woody plants and tree are considered swamps; those dominated by soft-stemmed
plants are considered marshes; and those with major quantities of mosses are considered bogs.
These general categories of wetlands can be further described as follows:
Freshwater swamps contain a variety of woody plants and water-tolerant trees. Southern
swamps contain bald cypress, tupelo gum, willow, white oak, and birch. Northern
swamps contain alder, black ash, black gum, white cedar, red maple, and willow.
Coastal marshes are dominated by cordgrass, blackrush, and glasswort.
• Freshwater marshes can often include submerged and floating plants, but emergent plants
usually distinguish marshes from other aquatic environments. Common emergent plants
include cattail, bulrush, reed, grasses, and sedges.
• Bogs typically contain plants such as cranberry, tamarack, black spruce, leather leaf,
pitcher plants, and mosses. Commonly found in the northeastern and north central
2
-------
regions of the United States, bogs are dependent on stable water conditions and are
characterized by acidic and low-nutrient waters, and acid-tolerant mosses.
1.3 SOURCES OF INFORMATION
As part of the first edition of this report, an initial literature search was conducted and revealed
about 140 papers and reports available for possible inclusion in this study. Additional literature
sources were identified for possible inclusion in the second edition through an updated
computerized literature search, through direct contacts with authors of previously cited literature,
and as feedback to the first edition. The literature sources were screened to identify the papers
and reports that are related to the use of wetlands for reducing the environmental impacts of
stormwater pollution. The criteria for this screening addressed whether the reports included one
or more of the following: (1) storinwater treatment; (2) inlet and outlet conditions; (3) effects on
the wetland environment; (4) effectiveness and limitations of the system; and (5) how the system
was managed. It should be noted that the above criteria were used as guidelines and not as strict
rules.
This literature review and screening process reduced the potential list of candidates to the reports
that were subsequently examined in detail. A thorough review was completed on these selected
reports, and a 1- to 6-page summary was prepared to highlight pertinent information for each.
These detailed summaries are presented in the Appendix.
The detailed review revealed that 17 of the selected reports discussed the results of research on a
functioning wetland system. This literature included the appropriate data and discussion by the
authors to assess the capabilities and limitations of using wetlands to control stormwater source
pollutants. These data included influent and effluent water quality, the effectiveness of the
system, flows and volumes, wetland and watershed areas, and the biological characteristics of the
system. Reports without all of these data may be included in the set of selected reports because
the studies were believed to be thorough enough to warrant inclusion. Table 1 presents a list of
the selected reports and some general characteristics of the wetlands that have been studied.
Table 1 illustrates some of the large variations among the subject wetlands. These wetlands differ
widely in location and wetland type, ranging from Florida’s southern swamplands to
Minnesota’s northern peatlands to California’s brackish marshlands to Puget Sound’s palustrine
wetlands. Each of these geographical locations differs in climate, vegetation, and soil types.
Figure 1 is a map showing the geographic location of the selected studies.
1.4 REPORT ORGANIZATION
The principal pollutant removal mechanisms that have been studied and reported for wetlands
include: sedimentation, adsorption, precipitation and dissolution, filtration, biochemical
interactions, volatilization and aerosol formation, and infiltration. These mechanisms are
discussed briefly in Section 2.1. The pollutant removal efficiencies that have been observed for
stormwater routed through wetland systems are presented and discussed in Section 2.2.
3
-------
Table 1. LITERATURE RESEARCHED TO INVESTIGATE PERFORMANCE CHARACTERISTICS OF WETLANDS
Detention Pond Constructed Wetland
Study Location Name/I. D. /Wetland /Natural Classificaton
Martin and Smoot 1986 Orange County, Orange County detention pond constructed hardwood
Honda Treatment System wetland cypress dome
Harper ci al. 1986 Florida Hidden Lake wetland natural hardwood
swampland
Reddy ci al. 1982 Orange County. Lake Apopka wetland constructed cattail marsh
Honda
Blackburn et al. 1986 Palm Beach. Palm Beach PGA wetland constructed southern
Honda Treatment System and natural marshland
Esry and Cairns 1988 Tallahassee, Jackson Lake detention pond constructed southern
Honda wetland marshland
Brown. R. 1985 Twin Cities Metm Twin Cities Metro wetlands natural . northern
Area. and pealland
Minnesota constructed
Wotzka and Oberts 1988 Roseville. McCarrons detention pond constructed cattail marsh
Minnesota Treatment System wetland
Hickok ci aL 1977 Minnesota Wayzota wetland natural northern
peatland
Batten 1987 Waseca, Cl Lake wetland constructed cattail marsh
Minnesota
Meioria 1986 Fremont, DUST Marsh wetland constructed brackish marsh
California
Morris et al. 1981 Tahoe Basin. Tahoe Basin wetland natural high elevation
California Meadowland riverine
Scherger and Davis 1982 Ann Arbor. Pittsfield-Ann Arbor detention pond constructed northern
Michigan Swift Run wetland and pealland
natural
ABAG 1979 Palo Alto. Palo Alto Marsh wetland natural brackish marsh
California
Jolly 1990 St. Agatha, Long Lake Wetland-Pond detention pond constructed cattail marsh
Maine Treatment System wetland
Oberts ci aL 1989 Ramsey-Washington Tanners Lake, McKnight detention ponds constructed cattail marsh
Metro Area, Lake. Lake Ridge. and wetlands
Minnesota Carver Ravine
Reinelt ci a!. 1990 King County. B31 and PC12 wetlands natural palustrine
Washington
Rushton and Dye 1990 Tampa, Tampa Office Pond wetland constructed cattail marsh
Florida
Hey and Barren 1991 Wadsworth. Des Plaines River Wetland wetland constructed freshwater
Illinois Demonstration Project river ine
4
-------
Figure 1. LOCATION OF WE1’LANDS RESEARCHED FOR THEIR ABILiTY TO TREAT
STORMWATER RUNOFF
5
-------
Section 2.3 discusses probable causes of variations and dissimilarities in wetland polIutan
removal from one studied wetland to another, including a comparison of wetland characteristic
and reported performance. Section 2.4 presents a comparison of factors affecting reporteE
treatment efficiencies of wetlands. Section 2.5 discusses the differences in the monitoring
procedures and the methods of quantifying performance in the studied wetlands. Because 01
these dissimilarities, the results are site-specific, making it difficult to generalize the functioning
and effectiveness of wetlands in controlling stormwater runoff pollutants.
Hydrologic impacts due to increased stonnwater runoff to wetlands has been a major concern.
Also, the uptake of toxins by plants and animals has been identified as a concern whet
considering the long-term implications of using wetlands for pollution control. Section 3 presents
a brief discussion of the present state of knowledge regarding the hydrologic impacts and the
uptake of toxins in biota in wetlands receiving stormwater.
Some constructed wetlands can be considered to be detention basins that include aquatic
vegetation. The use of detention basins for controlling stormwater-borne pollutants has been
fairly well documented elsewhere. Section 4 presents a discussion of those studies where both
wetlands and detention basins were studied (in either parallel or series configurations) and
compared.
Many of the authors of the reports reviewed herein presented their own recommendations
regarding the use of wetlands for stormwater pollution control. Abstracts of these authors’
comments are provided in Section 5.
Finally, Section 6 briefly discusses some of the issues, regarding the use of wetlands for
stormwater pollution control and recommends additional studies that could be undertaken to better
define the issues. A brief description of suggested future wetland pollutant removal study
reporting data is presented.
References used in this report are given in Section 7, and abbreviations and unit conversions are
given in Sections 8 and 9, respectively. An annotated bibliography with summaries of the
documents reviewed in this study is separately bound in the appendix.
6
-------
2.0
REPORTED PERFORMANCE OF WETLANDS
FOR STORMWATER TREATMENT
2.1 POLLUTANT REMOVAL MECHANISMS
Pollutants in stormwater can be removed by wetlands through a combination of 1)incorporation
into or attachment to wetland sediments or biota, or 2)degradation, or 3)export to the atmosphere
or groundwater. Both physical and chemical pollutant removal mechanisms are thought to occur
in wetlands. These mechanisms include sedimentation, adsorption, precipitation and dissolution,
filtration, biochemical interactions, volatilization and aerosol formation, and infiltration. Because
of the many interactions between the physical, chemical and biological processes in wetlands,
these mechanisms are not independent. The large variation in wetland characteristics (e.g.,
hydrology, biota, etc.) causes the dominant removal mechanism to vary from wetland to wetland.
These variations can help explain why wetlands differ so much in their removal efficiencies. The
following subsections briefly describe the principal removal mechanisms.
2.1.1 Sedimentation
Sedimentation is a solid-liquid separation process utilizing gravitational settling to remove
suspended solids. It is considered the predominant mechanism for the removal of many
pollutants from the water column in wetland and detention basin systems. Sedimentation of
suspended material, along with pollutants that are highly adsorbed, has been documented as the
primary mechanism of removal in wetlands by several authors including Martin and Smoot (1986)
and Oberts (1982).
Gravity will settle particles that have settling velocities large enough to overcome upward
impelling forces caused by fluid motion. Settling velocities are a function of the particle diameter
and density. While gravity tends to settle particles, turbulence tends to resuspend particles.
Suspended solids in natural waters tend to range from 0.005 to 100 microns ( .t) in diameter
(Chan et al. 1982). Sartor and Boyd (1972) investigated street surface contaminants in
stormwater runoff, and found that about 6 percent of the total solids were less than 43 in
diameter, 37 percent ranged from 43 to 246 p. in diameter, and 57 percent were greater than 246 p.
in diameter. Scherger and Davis (1982) found that 100 percent of the sediments greater than 60 p.
in diameter were removed by settling.
The most significant factors affecting settling of suspended material pertain to the hydraulic
characteristics of the wetland system. More specifically, the removal of suspended solids is a
function of the detention time, inlet-outlet conditions, turbulence, and depth. For example, Martin
and Smoot (1986) reported that residence time and turbulence were the most important factors
affecting sedimentation. Morris et al. (1981) reported that sheet flow (spreading out flows), as
opposed to channelized flow, was the most important factor affecting settling.
7
-------
The opposite of sedimentation is flotation. Many of the same processes occur a
sedimentation, only in reverse. Such floatable pollutants as oil and grease, litter, and pollut
that accumulate in the micro layer (including metals) may be removed in wetlands by proce
discussed below.
2.1.2 Adsorption
Adsorption of pollutants onto the surfaces of suspended particulates, sediments, vegetation,
organic matter is a principal mechanism for removal of dissolved pollutants. The litera
suggests that pollutants such as phosphorus, dissolved metals, and other adsorbents (inclu
colloidal pollutants) are removed through these processes (Harper et al. 1986; ABAG P
Hickok et al. 1977). Adsorption occurs through three main processes: (1) electros
attractions; (2) physical attractions (e.g., Van der Waals forces and hydrogen bonding); am
chemical reactions. The rates by which these processes occur are thought to be inversely re]
to the particle size and directly related to the organic content of the particles in the wetland.
(Harperetal. 1986).
Adsorption processes have been shown to be enhanced by increasing the contact of stormy
with the underlying soils and organic matter. In addition, high residence times, shallow v
depths, and even distribution of influent enhance the interactions of water with soil and j
substances, thereby increasing the adsorption potential.
Gersberg et a!. (1984) stated that adsorption of metals onto particulates, sediments, and or
matter is the predominant mechanism for heavy metal removal, followed by the subseq
settling of these particles. They tested their hypothesis by adding dissolved copper, zinc,
cadmium in high concentrations to the influent waste stream of an constructed wetland.
results suggest that 97 percent of the added zinc and 99 percent of the added copper and cadn
were removed from the waste stream.
Wisseman and Cook (1977) reported that chromium and lead were highly retained by sedim
whereas zinc and cadmium showed some transport from the sediments. Their study investi
the concentration of heavy metals in sediments downstream from a culvert discharging u
stormwater runoff. The results showed that chromium and lead concentrations decreased ra:
with distance from the culvert, whereas zinc and cadmium had more gradual decreas
concentration with distance.
Hickok et al. (1977) have shown that phosphorus removal is a function of the incrc
interactions of runoff with the mineral soils in wetlands. They found that the Way
Minnesota, wetland organic soil contained about 5.5 times more phosphorus than the nc
holding capacity of soils.
8
-------
2.1.3 Precipitation and Dissolution
Many ionic species (e.g., metals) dissolve or precipitate in response to changes in the solution
chemistry of the wetland environment. Metals such as cadmium, copper, lead, mercury, silver,
and zinc form insoluble sulfides under reduced conditions commonly found in wetlands
(Benforado 1981). Fulvic and humic acids released by decaying organic matter can form
complexes with metal ions. The resultant decreased pH promotes the dissolution of metals,
thereby making them available for bonding to inorganic and organic molecules (Gersberg et al.
1984).
2.1.4 Filtration
Filtration occurs in most wetlands through the simple act of vegetation removing pollutants and
sediments from the water column in a sieve-like fashion. Brown (1985) has suggested that the
increased density of vegetation slows the velocity and wave action which increased settling of
suspended material. Wotzka and Oberts (1988) also have suggested the use of filtration by
vegetation to improve the effectiveness of the wetland system for the same reasons discussed
above. Dense vegetation can be very effective at removing floatables and litter from stormwater.
Filtration can also take place in the soil matrix when infiltration occurs.
2.1.5 Biochemical Interactions
Vegetative systems possess a variety of processes to remove nutrients and other material from the
water column. In general, these processes are (1) high plant productivity and nutrient uptake; (2)
decomposition of organic matter, (3) adsorption; and (4) aerobic or anaerobic bacterially mediated
processes. Through interactions with the soil, water, and air interfaces, plants can increase the
assimilation of pollutants within a wetland system. Plants provide surfaces for bacterial growth
and adsorption, filtration, nutrient assimilation, and the uptake of heavy metals (Chan et al.
1982).
Meiorin (1986) suggested that overland flow enhances the nutrient and bacterial removal due to
increased soil and plant interactions. The study suggested that increased contact with the plant
roots and the bacteria associated with the rhizosphere is more efficient at pollutant removal than
unvegetated lagoons or ponds.
Through sedimentation, heavy metals and phosphorus settle out into the upper layers of the
wetland soils. Plant uQtake of these pollutants provides temporary removal of metals and
phosphorus from the sediments, allowing renewed adsorption sites within the sediment for the
attraction of other ions. Banus et al. (1975) reported that 6 to 8 percent of lead and about 20
percent of zinc and cadmium in sediments are taken up by marsh grasses.
Hickok et al. (1977) and Reddy et al. (1982) reported on the processes of ammonification and
nitrification. In an aerobic environment, nitrifying bacteria convert ammonia ions into nitrate for
further uptake by plants, and in an anaerobic environment, nitrate is converted to nitrogen gas
9
-------
(denitrification). These processes occur most rapidly during periods of warm temperatures when
microbial activity is highest.
2.1.6 VolatilIzation and Aerosol Formation
Volatilization (or evaporation) can remove volatile pollutants from wetlands. Air and water
temperature, wind speed, subsurface agitation, and surface films can affect the rate of
volatilization. Surface films may act as a barrier for the volatilization of some substances.
whereas evaporation may be a key mechanism for export of some substances such as chlorinated
hydrocarbons or oils often found in the surface films of water bodies receiving urban storm water
runoff (Chan et aL 1982). Aerosol formation may play only a minor role in removing pollutants
in wetlands and will occur only during strong winds (Chan et aLl982).
2.1.7 Infiltration
For wetlands with underlying permeable soils, pollutants can be removed through infiltration.
Stormwater can percolate through the soil, eventually reaching groundwater. Passage through the
soil matrix can provide physical, chemical and biological treatment depending on the matrix
thickness, particle size, degree of saturation and organic content Infiltration is also dependent on
the groundwater level at a site. In some instances, seasonal fluctuations in groundwater levels
may cause some wetlands to discharge groundwater during part of the year and recharge to
groundwater during other times of the year. Pollutant migration to groundwater depends on the
type of pollutant and aquifer characteristics. Contamination of unconfined aquifers by stormwater
may be more significant from upland infiltration than recharge through wetlands (Stockdale
1991).
2.2 WETLAND STORMWATER POLLUTANT REMOVAL EFFICIENCIES
Studies investigating the effectiveness of wetlands to treat stormwater runoff have been limited
(Figure 1), and those that have been conducted are primarily in a few geographical locations (e.g.,
Florida, Minnesota, California). The studies that are summarized herein represent a wide
diversity of wetland types, ranging from southern cypress swamplands and northern peatlands, to
brackish marshlands and high-elevation meadowlands. This section presents a discussion of
wetland stormwater pollutant removal efficiencies found in the literature.
Tables 2 and 3 summarize reported removal efficiencies for total suspended solids (TSS) and
nutrients, and metals, respectively. These tables reveal that a wide variety of parameters have
been measured and reported in each of the studies. Because the wetlands vary in their hydraulic
conditions, climate, and vegetation, and because the studies employed various monitoring and
reporting procedures, the broad range of pollutant removal efficiencies was not unexpected.
Figures 2 and 3 present histograms of pollutant removal efficiencies reported for total suspended
solids, phosphorus (TP), ammonia (NH3), and lead (Pb). Note that if the wetland does not
indicate removal, then the constituent was not measured.
10
-------
Table 2. AVERAGE REMOVAL EFFICIENCIES FOR TOTAL SUSPENDED S(
D NUTR .NTS IN WETlANDS REPORTED IN THE LiTERATURE
Study
System Name
System Type
Martin and Smoot
Orange County
detention pond
1986
-Treatment System
wetlands
entire system
Harper ci aL
Hidden Lake
wetland
1986
Reddy et al.
Lake Apopka
reservoirs
1986
flooded fields
Blackburn et al.
Palm Beach POA
system
1986
Treatment System
Esry and Cairns
Jackson Lake
system
1988
Browii
Fish Lake
wetlandlpond
1985
Lake Elmo
Lake Riley
Spring Lake
wetland
wetland
wetland
Wotzka and Obert
McCarrons Wetland
detention pond
1988
Treatment System
wetland’
system
Hickok et al.
Wayzata Wetland
wetland
1977
Barten
Clear Lake
wetland
1987
Meionn
DUST Marsh
1986
Basin A
BasinB
BasinC
System
wetland’
wetland’
wetland’
wetland
Morris ci al.
Angora Creek
wetland
1981
Tallac Creek
wetland
Scherger and Davis
Pittsfield-Ann Arbor
detention pond’
1982
SwIft Run
wetland
ABAG
Palo Alto Marsh
wetland
1979
Jolly
Long Lake Wetland-Pond
entire system
1990
Treatment System
Oberts ci at.
Tanners Lake
detention pond’
1989
MclCnlght Lake
Lake Ridge
Carver Ravine
detention ponds’
wetland
wetland-pond system
Reinelt ci at.
B3 1
wetland
1990
PC I2
wetland
Rushton and Dye
Tampa Office Pond
wetland
1990
Hey and Barrett
Des Plaines River Wetland
1991
Median
EWA3
EWA 4
EWA S
EWA 6
pollutant efficiency ror wetlan
wetland
wetland
wetland
wetland
d systems (without ‘):
Negative (“ .“) removal efficiencies mdmcatc net export in pollutant loads
TSS
VSS
Th
P0
TEN
LTJ ANT
Ore. N
REMO
NH3
VAL EFFI
N03
CIENC?
IP
(PERCENT)
Ortho-P
Dls. P
COD
SOD
65
66
89
60
60
85
19
21
36
17
23
39
60
54
61
-17
40
9
33
17
43
57
2
28
76
-30
21
7
18
17
83
-16
-24
62
80
7
-109
81
50
4.8
-7.6
16
57.5
51.9
17
68.1
64.2
33
60.9
7.3
62
75 1
16.7
35
96
76
37
70
90
78
95
88
-20
-300
78
80
20
-20
-20
38
20
-14
36
-36
7
11
0
50
25
-86
37
27
-43
-7
28
25
-30
-10
91
87
94
95
87
94
85
24
83
88
26
85
60
fl
63
78
36
78
57
25
53
90
79
93
94
-44
78
76
25
55
54
52
40
63
40
51
76
23
-77
-1
-1
4
-5
18
16
32
2
12
29
46
-4
36
58
65
28
37
68
-25
-46
-18
-57
54
36
-20
-88
20
33
50
35
5
-120
39
76
14
20
23
49
87
85
37
.6
54
95
94
92
63
85
85
20
50
57
67
1
5
14
24
-6
7
15
28
-10
1
11
17
9
7
34
37
1
20
34
-5
-3
-14
12
8
1
14
56
4
20
-2
-2
64
-3.7
55
65
72
76
89
98
76
70
42
70
95
46
59
55
69
97
46
79
24
5
7
33
28
23
55
45
-------
AVERAGE REMOVAL EFFICIENCIES FOR METALS AND OIL AND GREASE IN
DS REPORTED IN THE LITERATURE
Study System Name System Tyne
Marlin and Smoot Orange County detention pond’
1986 Treatment System wetland’
entire system
Harper Ct at Hidden Lake wetland
1986
Reddy et al Lake Apopk* reservoirs
1986 flooded fields
Blackburn et at. Palm Beach PGA system
1986 Treatment System
Esry and Cairns Jackson Lake system
1988
Brown Fish Lake wetlandfpond
1985 Lake Elmo wetland
Lake Riley wetland
Spring Lake wetland
Wolzka and Obest McCarrons Wetland detention pond’
1988 Treatment System wetland’
system
Hickok et at Wayzata Wetland wetland
1977
Barten Clear Lake wetland
1987
Meionn DUST Marsh
1986 Bas inA wetland’
Basin B wetland’
BasinC wetland’
System wetland
Morris et at Angora Creek wetland
1981 Tallac Creek wetland
Scherger and Davis Pittsfield-Ann Arbor detention pond’
1982 Swift Run wetland
ABAO Palo Alto Marsh wetland
1979
Jolly Long Lake Wetland-Pond entire system
1990 Treatment System
Oberta et at Tanners Lake detention pond’
1989 Mc Knight Lake detention ponds’
Lake Ridge wetland
Carver Ravine wetland-pond system
Reinelt et at 133! wetland
1990 PCI2 wetland
Rushton and Dye Tampa Office Pond wetland
1990
hey and Barrett Des Plaines River Wetland
1991 EWA3 wetland
EWA 4 wetland
EWA 5 wetland
EWA6 wetland
Median pollutant efficiency for wetland systems (without ‘): 83
Negative (“-“) removal efficiencies indicate net export In pollutant loads.
Lead Zinc Copper Cadmium Nickel Dromlum ‘ d and
total dissolved total, dissolved total dissolved total dissolved total dissolved total dissolved Grease
39 29 15 -17
73 54 56 75
83 70 70 65
55 56 41 57 40 29 71 79 70 70 73 75
85
68
90
94
82
80
67
30
42
-20
36
55
32
27
24
-60
-12
47
-57
83
-29
17
Il
13
13
88
42
-19
26
66
-25
61
0
83
0
59
63
52
6
34
63 42
61 40 29 69
79 48
70 70 75 -13
-------
(b)
Measured TP Removal by Indicated Wetlands
C
E 100
0
•
80
I 9 HUflU o nMH H
-20
• -40
0
-60
-80
C
•
2 -100
0
Q.
(a)
Measured TSS Removal by Indicated Wetlands
C
•
g 100
0
V
60
0
40
. 20
(I) 0
-20
0
•i -40
0
E -60
-J
-a
Note: No bar indicates that the removal estimates were not reported for this parameter at the indicated wetland.
Figure 2. POLLUTANT REMOVAL RATES FOR (a) TSS AND (b) TP
13
-------
(a)
Measured NH3 Removal by Indicated Wetlands
C
E
100
80
60
40
20
0
-20
-40
-60
-80
-100
I
Note: No bar indicates that the removal estimates were not reported for this parameter at the indicated wetland.
Figure 3. POLLUTANT REMOVAL RATES FOR (a) NH3 AND (b) Pb
14
-------
Despite the variability observed in pollutant removal efficiencies, some similarities exist among
the wetlands. The following observations can be made:
• Suspended solids and total lead, followed by total zinc and chromium, show the greatest
consistency with pollutant removal efficiencies.
• Suspended solid removal efficiencies tend to be more consistent and larger in constructed
wetlands than in natural systems. This is likely due to the design and management of the
constructed systems.
• In some cases, concentrations of dissolved lead, zinc, and copper can be reduced
significantly.
• Nutrient removal efficiencies vary widely among wetlands. The variations appear to be a
function of the season, vegetation type, and management of the wetland systems.
• Total phosphorus and nitrate show the greatest consistency with nutrient removal
efficiencies. Total phosphorus removal efficiencies tend to be more variable for the
natural wetlands and less variable for detention basins and constructed wetlands.
2.3 PROBABLE CAUSES OF VARIATIONS AND DISSIMILARiTIES OF
REPORTED WETLAND POLLUTANT REMOVAL EFFECTIVENESS
In addition to the efficiencies tabulated by the authors, several reports presented conclusions to
help explain the effectiveness of wetland treatment and their variations. Hydrology was reported
to be the most critical parameter influencing wetland performance. Variations in local hydrology,
detention times, rates of runoff, water level fluctuations, and seasonality were all reported to
affect the function of wetlands and, thus, their effectiveness at removing pollutants (Benforado
1981). Table 4 presents geographic, hydrologic and hydraulic characteristics for each of the
wetlands reviewed.
The size and volume of a wetland system can greatly affect both the actual removal efficiencies
and one’s ability to estimate these efficiencies. Chan (1982) reported difficulties in estimating
pollutant removal efficiencies due to the volume of the wetland basin. The volume of the
Demonstration Urban Stormwater Treatment (DUST) marsh was sufficiently large that the
treatment cycle spans several storms; no one storm provided a complete picture of pollutant
efficiencies. The DUST marsh was found to accumulate storm water flows within the system and
discharge effluent slowly over periods of days or weeks, depending on the interval between
storms. Thus, the water collected at the discharge from the DUST marsh is probably a mixture of
water that entered from the previous storms.
The type of inlet structure and the flow patterns through wetland areas can significantly affect
pollutant removal efficiencies. Morris et al. (1981) found that sheet flow (as opposed to
15
-------
Table 4. WETLAND GEOGRAPHIC AND HYDRAULIC CHARACTERISTICS
Wetland Watershed Wetlandl Average Basin Detention
Watershed % System Constructed Size Size Watershed Flows Volume lime Depth Inlet
Land Use Land Use Type /Naiural (acres) (acres) Ratio (cfs) (acre-fl) (hours) (fi) Condition Comments
resIdential 33 detention pond constructed 02 41.6 0.5% 25 1.2-1.9 75 8- 11 discrete a
blgbway 27 wetland conatnicted 0.78 NA 1.9% NA 0.5-28 8 0-5 discrete
=
Study
System Name
Marlin and S moot
Orange County
1986
Treatment System
forest
40
system constructed
098
2.4%
Harper at al
Hidden Lake
residential
NA
wetland natural
2.5
552
4.5%
022
NA
NA
NA
diffuse
b
1986
Reddy et al
1982
Lake Apopka
agriculture
100
reservoirs constructed
0.9
NA
NA
0.56
2.6
9.4 days
3.3
diffuse
c
flooded fields constructed
09
NA
NA
023
06
48 days
07
diffuse
Blackburn at ai
1986
Palm Beach PGA
Treatment System
residential
gol(couzse
NA
NA
wetland constructed
89
2350
3.8%
NA
NA
NA
NA
diffuse
d
wetland ccnst + nat
296
2350
12.6%
NA
NA
NA
NA
diffuse
Ersy and Cairns
1988
Jackson Lake
urban
NA
detention pond constructed
20
2230
0.9%
NA
150
NA
7.5
diffuse
e
f
wetland constructed
9
2230
04%
NA
135
NA
1.5
diffuse
Brown
FishLake
residential
30
wetland natural
16
700
2.3%
0001-001
64
NA
4
discrete
g
1985
Lake Elmo
Lake Riley
commercial
agriculture
open
residential
commercial
agrIculture
open
residential
S
12
53
12
I
34
53
13
wetland natural
wetland natural
225
.
77
2060
2475
10.9%
3.1%
0.001-0.65
0.004-135
900
231
NA
NA
4
3
dIscrete
discrete
I i
I
Spring Lake
commercial
agriculture
open
residential
2
30
55
5
wetland constructed
64
5570
1.1%
0008-4
256
NA
4
discrete
commercial
agriculture
open
1
57
37
Wotzka and Obert
1988
McCarrons Wetland
Treatment System
urban
NA
detentlonpond constructed
29.7
600
5.0%
0.05-2
2.3-9.7
24 days
25
dimise
j
wetland constructed
6 2
600
1.0%
system constructed
35.9
60%
Hickoketal.
Wayzata Wetland
resIdential
NA
wetland natural
7.6
65.1
11.7%
0.08
NA
NA
NA
discrete
k
1977
commercial
Batten
Clear Lake
urban
NA
wetland constructed
52.9
1070
49%
1.5
10
3-5 days
0.5
dIffuse
1987
Meionn
1986
DUSTMarsh
urban
agriculture
93
7
wetlandA constructed
5
-
-
•
10-250
150
4-4odaya
47
diffuse
I
wetland B constructed
6
-
-
wetland C constructed
21
2960
0.7%
wetland(system) constructed
32
2960
1.1%
Morris at al.
Angora Creek
residential
NA
wetland natural
NA
2816
NA
8.46
NA
NA
NA
diffuse
in
1981
Scherger and Davis
1982
Taflac Creek
Pittsfield-Ann Arbor
Swift Run
forest
NA
residential
commercIal
agriculture
open
NA
NA
45
19
13
23
wetland natural
detentIon pond Constructed
NA
25.3
2781
4872
NA
0.5%
868
0-29 16
NA
21-176
NA
4-105
NA
0-6
diffuse
discrete
ii
wetland natural
255
1207
2.1%
0-166
15-60
12-82
0.3
discrete
-------
Table 4. WETLAND GEOGRAPHIC AND HYDRAULIC CHARACTERISTICS (concluded)
Study
Watershed
System Name Land Use
%
Land Use
Wetland
System Constructed Size
Type !Natural (acres)
Watershed
Size
(acres)
Wetland!
Watershed
Ratio
Average
Flows
(c(s)
Basin
Volume
(acre-fl)
Detention
Thue
(hours)
Depth
(ft)
Inlet
Condition Comments
ABAG
Palo Alto Marsh residential
62
wetland natural 613
17600
35%
150-320
400-750
30
1-6
discrete o
1979
commercIal
12
p
open
26
Jolly
Long Lake Wetland-Pond agriculture
100
wetland-pond constructed 1.5
18
83%
0.01
1.5
NA
0.5-8
diffUse q
1990
Treatment System
Oberts et a!.
Tanners Lake resIdential
NA
pond constructed 0.07
1134
neglIgible
NA
(11
NA
3.0
discrete r
1989
McKnight Lake residential
Lake Ridge residential
Carver Ravine residential
NA
NA
NA
pond constructed 5.53
wetland constructed 0.94
wetland-pond constructed 0.37
5217
531
170
0.1%
0.2%
(12%
NA
NA
NA
13.2
2.0
1.0
NA
NA
NA
4.9
48
2.0
discrete
discrete
discrete
Reinelt et a!.
B3 1 urbanized
NA
wetland natural 4.9
461.7
1.1%
15
0.03-0.43
3.3
NA
discrete a
1990
PC I2 rural
NA
wetland natural 3.7
2148
1.7%
0.7
005-0.60
2.0
NA
discrete
Rushton and Dye
Tampa Office Pond commercIal
100
wetland constructed 0.35
6.3
5.6%
NA
0.32
NA
0-1.5
discrete u
1990
Hey and Barreu
1991
Des Plaines River Wetland agriculture
Demonstration Project urban
80
20
EWA 3 constructed 5.6
-
-
5
NA
NA
I
discrete v
EWA 4 constructed 5.6
-
-
06
NA
NA
1
discrete
EWA 5 constructed 45
-
-
4
NA
NA
1
discrete
EWA 6 constructed 83
-
-
1
NA
NA
I
discrete
TABLE NOTES:
- Si NA = Not available
Comments: a Short circuiting was observed during several storms.
b The wetland is not a basin, bit similar to a grassy swale.
c Design configuration suggests little short circuiting occurred.
d Design configuration suggests little short circuiting occurred.
e Generally sheet flow exists within the artificial wetland.
(Design configuration suggests little short circuiting occurred.
g The major influent to these natural wetlands Is discrete channelized flow.
h The schematic suggests large areas of dead storage.
I Short circuiting was ant discussed by the author.
Three discrete Inlets help to minimize short circuiting and dissipate surface water energy.
lc Design configuration suggests minimal short circuiting existed regardless of a single discrete Inlet
I Design configuration suggests little short circuiting occurred due to long and narrow wetland basins.
m Flow occurs as channclized flow until the storm volume is large enough to force sheet flow through the meadowland,.
n The schematic suggests large areas of dead storage exlsL
o Waxer level and volume are controlled by the tidal cycle.
p Coannelazed flow exist until the tide Increases causing the surrounding marsh to become Inundated.
q Entire system consists of a sedimentation basin, grass filter strip, constructed wetland, and deep pond.
r Monitoring occurred during a dry penod.
sStorm flows reduce detention times.
Channelizatlon reduced effective area In wetland.
u Overflow from adjacent wetlands occurred during extremely high water; leak and breach problems occurred during study.
v Water is pumped to the system from the river (drainage area of 210 square miles) for 20 hours per week.
-------
channelized flow) was the most critical factor in the effectiveness of meadowland treatment. This
finding is consistent with the theory that shallow, vegetative overland flow decreases velocities
and increases sedimentation. In addition, close contact with the soil matrix was found to increase
assimilation of nutrients and bacteria. Brown (1985) found that an undefined inflow (multiple
input locations) to the wetland, which results in better dispersion of incoming loads, was critical
in the effectiveness of the wetland. An undefined inflow reduced short-circuiting and increased
mixing and contact of the stormwater with the soil and plant substrates.
The change in seasons has been considered another important factor in the effectiveness of
wetland treatment of storm runoff. High evapotranspiration rates and seasonal productivity and
decay of plant and animal life are considered typical factors of seasonality. Removal efficiencies
in wetlands located in areas with strong seasonal variation were found to vary significantly
between seasons. Meiorin (1986) reported that high summer evapotranspiration rates caused a
200 to 300 percent increase in the total dissolved solids concentrations within the DUST Marsh.
Furthermore, high productivity during wann periods leads to decreases in nutrients, and increases
in SOD and suspended solids. Morris et al. (1981) reported that flushing and leaching effects of
spring snowmelt caused an increase in total Kjeldahl nitrogen and organic carbon in flows leaving
the Tahoe Basin meadowlands. The available organic material and nutrients leaving the
meadowlands were found to be a function of decay processes, which increase during the winter
season. Harper et al. (1986) reported that d&ention times greater than 2 days caused an increase
in the export of ortho-phosphorus from the Hidden Lake wetland.
Hickok et al. (1977) described microbial activity as the most important factor affecting
phosphorus removal. Microbial activity was found to be highest during the warmer temperatures
of summer. However, the microbial activity decreased when soils were submerged and became
anaerobic. They also reported that an acclimation period was required for microbes to adjust to
the surge of stormwater into the Wayzata wetland system.
Other factors are also known to cause variat.i ns in the reported pollutant removal effectiveness of
wetlands which receive storm runoff. For example, Chan (1982) described how the maturity of
the recently-constructed DUST Marsh affected its treatment potential. The newly exposed soils of
the wetland exhibited a leaching of salts. As the wetland became inundated, the new marsh
system caused the leaching of salts and some pollutants that were associated with past land uses.
Inundation also resuspended particulate matter. In addition, a gradual transition from salt-tolerant
plant species to less tolerant species occurred early in the study.
Benforado (1981) suggested that the buildup of nutrients and heavy metals in a wetland system
may reduce its effectiveness. Over time, the adsorption sites available on sediment and plant
substrates may become saturated, thus reducing the capacity for the assimilation of added
pollutants.
Scherger and Davis (1982) explained how particle-size distribution affects the settling of
suspended sediments. A particle whose settling velocity exceeds that of the surrounding fluid
motion will eventually settle out. However, particles whose settling velocities are less than the
18
-------
motion of fluid will remain suspended. As a result, the gradation of the localized sediment
particle sizes becomes an important factor in the removal of suspended sediments and their
associated contaminants.
Maintenance practices performed at a wetland also can influence its performance. Accumulated
sediments and debris can be removed to maintain storage volumes and reduce the potential for
resuspension and conveyance of these materials downstream. For some systems such as Jackson
Lake (Esry and Cairns 1988), sediment removal need only be performed at a sedimentation basin
located upstream of a wetland and not within the wetland itself. Plants can be harvested
periodically to remove vegetation and excess nutrients from the wetland (Harper et al. 1986;
Barten et a!. 1977) or to control undesirable species at newly constructed wetlands (Blackburn et
a!. 1986).
2.4 COMPARISON OF FACTORS AFFECTING REPORTED TREATMENT
EFFICIENCIES
There were 26 different wetland systems which presented data on removal efficiencies. Factors
that were evaluated by this study with regards to their effect on wetlands pollutant removal
performance included whether the wetland was a constructed or natural system, vegetation types
found in the wetland, land-uses types draining to the wetland, area of the wetland system as
compared to the contributing watershed, estimated average storm-flow quantities draining to the
wetland, and inlet types. Of these, few meaningful direct relationships were found. This was
probably due to the limited amount of data available to determine these relationships, and that
there are multiple factors which effect performance, including those above and many others. With
the lack of a large data base, a meaningful multiple regression analysis was not possible.
However, several trends were noted. First, constructed systems were generally found to have a
higher average removal performance than natural systems, with less variability; and second,
larger wetlands as compared to watershed size also showed the same trend, a higher average
removal performance, with less variability.
Table 5 presents a comparison of reported removal rates for constructed and and natural wetlands
systems. Figure 4 displays a scatter plot for total suspended solids (TSS) and total phosphorus
(TP) and a percentile Box Plot for the constructed and natural systems. Note that in all cases for
the pollutants summarized constructed systems had a higher average and median performance
level. More significant however is the variability differences of the 2 types of wetlands, with
constructed sites being much less variable. This is not a surprising finding. Given that
constructed systems have generally been designed to handle expected incoming flows and to
minimize short-circuiting, it is expected that they should generally show a higher performance
level with more consistency. Figure 5 display percentile plots for total suspended solids and total
phosphorus and Figure 6 for ammonia (NH3) and total lead (TPb). These plots demonstrate the
larger variability by the generally higher slope and spread of the natural system percentile curves.
19
-------
Coiistnded Wetlands
and Natural Stormwater Wetlands
Wetland Sites
WWAR DAR
TSS NH3
TP Pb
LakeRidge
0.18%
565
85
Carver Ravine
0.22%
459
20
DUST Marsh
1.10%
91
76
Jackson Lake
130%
77
96
Orange County
2.40%
42
89
Clear Lake
4.90%
20
76
Tampa Officc
5.60%
18
64
McCanons
6.00%
17
94
Long Lake
8.30%
12
95
Paint Beach
12.60%
8
50
LakeApoka
-
-
-
EWA3
-
fl
EWA4
-
76
EWA5
-
89
EWA6
-
98
Wetland Site
WWAR
DAR
TSS
NH3
1?
Pb
ZN
1131
1.10%
91
14
-
-2
-
-
PC1Z
1.70%
59
56
-
-2
-
-
Swift Run
2.10%
48
76
-
49
83
-
Fish Lake
2.30%
43
95
0
37
-
-
Lake Riley
Palo Alto
3.10%
3.50%
32
29
-20
87
25
-
-43
-6
-
-
-
-
Hidden Lake
4.50%
22
83
62
7
55
41
Lake Elmo
10.90%
9
88
50
27
-
-
Wayzata
AngoraCreek
Tallac Creek
11.70%
-
-
9
-
-
94
54
36
-44
20
33
78
5
-120
94
5
-
82
-
-
Median
3.10%
32
76.0
25.0
5.0
69.0
61.5
CV
87.2%
68.6%
61.7%
167.8%
1900.6%
67.0%
47.1%
Average
4.54%
38
60.3
20.9
2.7
59.3
61.5
N
9
9
11
7
11
4
2
ZN
All Wetlands
WWAR
DAR TSS NH3
1?
Pb
Median
3.10%
32
33.0
46.0
83.0
42.0
CV
88.5%
176.4%
99.7%
139.2%
56.9%
38.9%
Average
4.39%
87
29.5
34.1
61.8
53.8
N
19
19
13
26
9
5
76.0
42.9%
69.7
25
I
Table Notes
WWAR= Ratio of wetland system to wgershcd area (expressed as a percent)
DAR= Drainage Area Ratio (ratio of Watershed to Wetland Areas)
CV= Coeffiecienc of Variation
TSS = Total Suspended Solids
NH3 = Ammonia
TP = Total Phosphorus
1?b = Total Lead
ZN
16
37
61
55
17
52
37
1
58
90
43
54
55
78
92
62
7.3
59
55
69
97
42
70
34
52
6
88
83
90
Natural Wetlands
20
-------
(a
E
a,
cc
a,
0,
(a
a,
a,
0 )
(a
C
a,
a,
Pollutant
Percentiles
90
50
25
110
Figure 4. Box Plot Percentiles Comparison of Site Average Pollutant Removals for Natural and
Constructed Wetland Systems
TSS = Total Suspended Solids TP
NH3 = Athmonia TPb
N
= Total Phosphorus
= Total Lead
= Number of Wetland Sites
Natural Constr Natural Consir Natural Constr Natural Constr
21
-------
Percentile
(a)
Total Phosphorus
A Natural Wetlands
• Constructed Wetlands
• Natural Wetlands
C) Constructed Wetlands
0 20 40 60 80 100
Percentile
(b)
Figure 5. Comparison of Site Average Pollutant Removals for Natural and Constructed
Wetland Systems:
Connected Percentiles for (a) Total Suspended Solids and (b) Total Phosphorus
E
0
0
0
0
C
0
0
0
0 20 40 60 80 100
1
vu
75
50
25
a
0
E
0
0
0
0
0
0
—
C
0
0
0
-25
-50
-75
-100
/
22
-------
Ammonia
(a)
Total Lead
‘ Z r
( U
E
0
cc
( U
a ’
(0
I-
( U
( U
a’
( U
( U
2
( U
0 -
- 90
I :
0- 20
10
f ’ s
Figure 6. Comparison of Site Avenge Pollutant Removals for Natural and Constructed Wetland
Systems: Connected Percentiles for (a) Ammonia and (b) Total Lead
Percentile
(b)
• Natural Wetlands
D Constructed Wetlands
• Natural Wetlands
O Constructed Wetlands
o 20 40 60 80 100
Percentile
0 20 40 60 80 100
23
-------
Note that for phosphorus removal at constructed site, above the 10th percentile of sites removals
were all greater then 50 percent, while for total phosphorus above the 20th percentile showed 40
percent removals. For ammonia both constructed and natural systems showed above 20 percent
removals at the 30 percentile for constructed sites. For total lead removal, constructed and natural
sites showed above 50 percent removal at about the 30th percentile.
The size of the wetland system as compared to the size of the contributing watershed was also
investigated. Regression of the wetland to watershed area ratio (DAR) to pollutant removal
performance did not reveal good direct relationships as Figure 7 shows. However, grouping of
sites with a greater than or less than 2 percent of wetland to watershed ratio did result in some
general trends.
Table 6 and Figure 7 present performance results for all wetland systems with reported tributary
watershed areas. In general, the larger DAR wetlands had higher performance levels with less
variability. This analysis includes all wetland sites, natural and constructed.
To separate out the effects of natural versus constructed systems, Table 7 and Figure 8 present the
same analyses for constructed sites only. ,Generally, for constructed sites the trends are the same,
although the differences in performance levels and variability in performance is much less. The
data indicates that if systems are carefully constructed, the DAR is probably not as an important
factor in determining performance. Therefore, at this time we are not suggesting that minimum of
a 2 percent DAR is the proper design criteria for constructed wetlands.
The Jackson Lake Wetland is an example of one with a small DAR which still achieved excellent
performance (85 percent TSS removal). The DUST marsh and the Lake Ridge Wetlands also
showed high performance levels (76 and 85 percent TSS removals, respectively). One factor for
the Dust Marsh performance is that it is an off-line device; it only receives flow volumes up to a
certain flow rate and then by-passes flows. This type of design is particularly appropriate for
wetlands receiving stormwater from larger catchments relative to wetland size.
We believe that a better measure of wetland capacity to treat runoff from a given watershed would
be to evaluate runoff volumes as compared to storage volumes and contact surface area.
However, the data from the studies did not consistently include data on rainfall statistics, percent
impervious for land-uses, specific percentages for land-uses in a catchment, flow volumes to the
wetland, capacity of the wetland system, and surface areas for contact with stormwater (including
soils and plants). Therefore, we were not able to analyze the wetland systems with this approach.
Section 6 contains some recommendations regarding reporting information for future studies, so
that such analyses can be completed in the future.
Finally, maintenance activities in wetlands that are treating stormwater have not been well
documented or studied. These activities could impact performance characteristics of wetlands,
24
-------
I (“
80
60
40
20
0
.
•
G
0
*
0
0
0
-20
-40
-60
.o
0
.
10 12 14
• Tss
0 TP
Percentiles
90
50
25
110
Figure 7. Average Site Pollutant Removal Comparisons for All Wetlands With Less Than 2% and
Greater Than 2% Wetland to Watershed Area Ratios (WWAR): (a) Scatter Plot for TSS
and TP and (b) Percentile Box Plots for TSS, TP, and TPb
LT2% =lessthan2%WWAR
GT 2% = greater than 2% WWAR
N = Number of Wetland Sites
TSS
TP
TPb
= Total Suspended Solids
= Total Phosphorus
= Total Lead
E
0
0,
(U
I-
0
0
0 )
(U
C
2
a,
(U
E
a,
0:
a,
0)
(U
a,
a,
C)
CU
C
a,
C)
a,
-2 0 2 4 6 8
WWAR
(a)
LT2% GT2% LT2% GT2% LT2% GT2%
(b)
25
-------
Comparison of Reported Removal Rates for Constructed
and Natural Stormwater Wetlands Sized Less-Than and
Greater-Than 2% of the Contributing Drainage Area
Wetland Systems Smaller Than 2% of Watershed Area
Wetland Site WWAR DAR TSS NIB
1?
Pb
W
LakeRidge 0 . 18% 565 85 -
37
52
CaiverRavine 0.22% 459 20 -
1
6
DIJSTMarsh 1.10% 91 76 16
58
88
42
831 1.10% 91 14 -
-2
-
-
Jackson Lake 1.30% 77 96 37
90
.
PC I2 1.70% 59 56 -
-2
-
-
Median 1.10% 9 1 66.0 26.5
19.0
52.0
42.0
CV 0.4% 181 .2% 56 .8% 783%
120.8%
92 .6%
-
Average 0.93% 224 57.8 26.5
303
48.7
42.0
N 6 6 6 2
6
3
1
Wetland Systems Greater Than 2% of Watershed Area
Wetland Site WWAR DAR TSS N W
1?
Pb
ZR
SwiftRun 2.10% 48 76 -
49
83
FishLake 2.30% 43 95 0
37
-
-
OrangeCowity 2.40% 42’ 89 61
43
83
70
LakeRiley 3.10% 32 -20 25
-43
-
-
PaloAlto 3.50% 29 87 -
-6
-
-
HiddeaLake 430% 22 83 62
7
55
41
Clearlaire 4S0% 20 76 55
54
-
-
Tampa office 5.60% 18 64 -
55
-
34
McCarrons 6.00% 17 94 -
78
90
.
LongLake 8.30% 12 95 -
92
-
-
LakeElmo 10.90% 9 88 50
27
-
Wayzata 11.70% 9 94 -44
78
94
82
Paim keach 12.60% 8 50 17
62
-
-
Median 4.90% 20 87.0 37.5
49.0
83.0
55.5
CV 62.0% 58 4% 42.1% 130.8%
91.4%
18.9%
404%
Average 5.99% 24 74.7 28.3
41.0
81.0
56.8
N 13 13 13 8
13
5
4
Wetland Systems With Reported Watershed Areas (all sites above)
1?
Pb
7 14
WWAR DAR TSS NIB
Table Notes
WWAR= Ratio of wetland system to watershed area (expaeased as a percent)
DAR= Drainage Area Ratio (ratio of Watershed to Wetland Areas)
CV= Coeffiecient of Vatiation
155 = Total Suspended Solids
14 1 13 = Ammonia
= Total Phosphorus
Wb = Total Lead
Table 6.
26
-------
Table 7. Comparison of Reported Removal Rates for Constructed
Stormwater Wetlands Sized Less-Than and
Greater-Than 2% of the Contributing Drainage Area
Wetland Systems Smaller Than 2% of Watershed Area
Wetland Site WWAR DAR TSS 1?
LakeRidge 0.18% 565 85 37
Carver Ravine 0.22% 459 20 1
DUST Marsh 1.10% 91 76 58
Jackson Lake 1.30% 77 96 90
Median
0.66%
275
80.5
47.5
CV
83.7%
84.2%
48.9%
80.3%
Average
0.70%
298
69.3
46.5
N
4
4
4
4
Wetland Systems Greater Than 2% o(Waterehed Area
Wetland Site WWAR DAR TSS
Orange County 2.40% 42 89 43
Clear Lake 4.90% 20 76 54
Tampa Office 5.60% 18 64 55
McCazrons 6.00% 17 94 78
Long Lake 8.30% 12 95 92
Palm Beach 12.60% 8 50 62
Median
5.80%
17
82.5
58.5
CV
52.5%
60.5%
23.3%
28.0%
Average
6.63%
19
78.0
64.0
N
6
6
6
6
Constructed Wet
land Systems
With Renorted
Watershed A
rena (all sites aborel
WWAR
DAR
TSS
1?
Median
3.65%
31
80.5
56.5
CV
94.6%
156.2%
32.5%
47.2%
Average
4.26%
131
74.5
57.0
N
10
10
10
10
Table Notes
WWAR= Ratio of wetland system to watershed area (expressed as a percent) TSS = Total Suspended Solids
DAR= Drainage Area Ratio (ratio of Watershed to Wetland Areas) T1 = Total Phosphorus
CV= Coeffiecient of Vanation
27
-------
irir
V
V yV
V
- 10 12
WWAR
(a)
Tss
YTP
Percentiles
90
50
25
110
Figure 8. Average Site Pollutant Removal Comparisons for Constructed Wetlands With Less Than
2% and Greater Than 2% Wetland to Watershed Area Ratios (WWAR): (a) Scatter Plot
for TSS and TP and (b) Box Percentiles
LT2%=lessthan2%WWAR
GT 2% = greater than 2% WWAR
N = Number of Wetland Sites
TSS = Total Suspended Solids
TP = Total Phosphorus
E
0
a-
0
0
0
—
C
0
H
0
0
90
80
70
60
50
40
3C
20
10
0
V
r
V
2
14
TSS
I
TP
90
• 80
70
0 60
‘U
50
40
30
C
0
20
0
N=4
N=4
LT2%
G12%
LT 2%
(b)
GT2%
28
-------
particularly over the long-term. In addition, the need for maintenance and level of maintenance is
not well understood or documented.
2.5 ASSESSMENT OF THE RELIABILITY OF WETLAND DATA
This section discusses some of the difficulties in comparing one wetland study to another.
Specifically discussed are the length of the study, the number of storms monitored or samples
taken, and the methods used to compute the reported pollutant removal efficiencies. Table 8
presents a list of the selected literature and the respective information on sampling characteristics
that were employed.
From the table, it can be seen that the studies identified were generally about a year or less. There
was also quite a variation in the number of samples collected (from 3 to about 150) as well as the
sampling methods (i.e., grab sample or samples versus composite sample for an event). These
factors all contribute to the difficulty of comparing results from the various studies.
All data collection, however well performed, is subject to random variations that cannot be
completely eliminated. These variations, or errors, are defined as either “chance variations” or
“assignable variations.” Chance variations are due to the random nature of the parameters
measured ; increased testing efforts and accuracies cannot eliminate these variations. Although
assignable variations cannot be eliminated altogether, these variations can be reduced and the
reliability of the data increased. Assignable variations are those errors that result from
measurement error, faulty machine settings, dirty containers, etc. Increasing both the length of a
study and/or the number of storms sampled can reduce the assignable variations and increase the
reliability of the data. Table 8 shows that in most cases, the studies reviewed have lasted no more
than one year, and in many cases had relatively small sample sizes.
Another complication in comparing the performance of wetlands is the method of quantifying their
effectiveness. Martin and Smoot (1986) discussed the following three types of methods to
compute efficiencies:
• The first method employs the efficiency ratio (ER), which is defined in terms of the
average event mean concentration (EMC) of pollutants, thus:
ER = 1 - average outlet EMC
average inlet EMC
• The second method is based on the summation of loads (SOL) of pollutants removed
during the monitored storms, thus:
SOL — 1 sum of outlet loads
— sum of inlet loads
29
-------
Table & SAMPLJ1 G CHARACrERIS7ICS FROM THE WETLANDS REVIEWED
Study
Location
Time
of Study
Length of
Study
Type of
Sample
Number
of Sterms
Monitored
Method of Computing
Efficiencies
1982-1984
2 years
7 multi grab
6 conipomte
13
ROL
1984-1985
1 year
composite
18
1977-1979
2 years
single grab
—ISO
MC
1985
1 year
single grab
36
1985
NA
NA
1
SOL
1982
lycar
composite
5-7
Martin and Smoot
1986
Harpor ci aL
1986
ReddyetaL
1982
Blackbirn et aL
1986
Eary and Cairns
1988
Brown
198 5
Wotalca a d Oberts
1988
Hickok ci aL
1971
Barten
1987
Metoris
1986
Morris et aL
1981
Scherger and Davis
1982
ABAG
1979
Jolly
1990
Obetis etal
1989
ReinehetaJ.
1990
R ishton and Dye
1990
Hey and Barrett
1991
Orange County,
Florida
Florida
Orange County.
Flotida
Palm Beach,
Plotida
Tallahassee.
Florida
Twin Cities
Metro Area,
Minnesota
Roseville ,
Minnesota
Minnesota
Waseca.
Minnesota
Coyote fliEs,
Fremont, Ca.
Tahoe Basin.
California
Ann Arbor,
Michigan
Palo Alto,
California
St. Agatha.
MaIne
Ramsey-Washington
Metro Area,
Minnesota
King County.
Washington
Tampa,
Floiida
Wadsworth .
1984-1988
2 years
composite
25
SOL
1974-1975
10 months
NA
NA
ER
1982-1985
3 years
composite
27
SOL
1984-1986
2 years
composite
11
MC
19174978
1 year
single grab
—75
SOL
1979-1980
8 months
composite
7
1979
3 months
composite
8
1989
5mosths
cowpo&te
I i
SOL
1988-1990
2 years
composite
13
SOL
ER
1989-1990
12 months
composite
3-8
Table Notes
ER = Event menu concentration
SOL = Sum of event loads
ROL = Regression of event loads
MC = Mean concentration
NA = Not available
30
-------
The third method of determining efficiency was developed by Martin and Smoot
(1986). l’his method defines the ratio as the slope of a simple linear regression of inlet
loads and outlet loads of pollutants. The equation for the regression of loads (ROL)
efficiency is thus:
Loads in = 13 • Loads out
where 13 equals the slope of the regression line, with the intercept constrained at zero.
The ER and SOL methods assume that the monitored storms are a representative sample of all
storms that occur. The ROL method assumes that the treatment efficiency is the same for all
storms.
Yet another method used in the literature is defmed in terms of the average sample or mean
concentration (MC), thus:
MC — 1 average outlet concentration
— - average inlet concentration
In some studies, the pollutant removal efficiencies are computed using one of the above methods
for separate storms or seasons, and the average of all of these efficiencies is reported as a single
value. These different methods of computation can produce differences in reported removal
efficiencies.
The available literature indicates the need for additional long-term research on the performance of
wetlands for the treatment of stormwater runoff. These studies should include investigating a
number of different types of wetlands in different geographical areas and climates. Section 6.2
discusses known current and on-going studies.
2.6 SUMMARY
Wetlands have been shown to have a good capability for removal of pollutants from stormwater
runoff. There are a number of factors which contribute to and influence removal efficiencies,
including sedimentation, adsorption, precipitation and dissolution, filtration, biochemical
interactions, volatilization and aerosol formation, and infiltration. The reported removal
efficiencies are, as expected, quite variable. For the wetlands systems reviewed, removal
efficiencies for total suspended solids (TSS) had a median of 76 percent. TSS removal is a good
indicator of pollutant removal potential for heavy metals and phosphorus, as well as other
pollutants which are associated with fine particulate matter. Constructed wetlands tended to be
more consistent than natural wetlands in their removal of TSS and other parameters that were
analyzed. Wetlands have also shown the ability to remove dissolved metals. Nutrient removal in
wetlands is variable, depending on both wetlands characteristics as well as seasonal effects.
31
-------
The fact that there are many dissimilarities between the wetlands that have been studied
contributes greatly to the variability seen in wetlands stormwater pollutant removal efficiencies. It
is evident, however, that properly designed, constructed, and maintained wetlands can be
effective pollution control measures. There is a definite need to look at additional wetlands in a
variety of geographical areas, and to look at long-term pollutant removal efficiencies.
32
-------
3.0
NOTED IMPACTS OF STORMWATER RUNOFF ON WETLAND BIOTA
3.1 INTRODUCTION
Many researchers have expressed concern over the impact of the quantity and quality of
stormwater runoff on wetland biota, especially in natural wetlands (Newton 1989; Stockdale
1991). The quantity of stormwater runoff determines the hydrologic characteristics of a wetland
including the average and extreme water levels and duration and frequency of flooding.
Stormwater runoff also contains pollutants that can adversely affect wetland biota if accumulated
in high concentrations. The following sections document the impacts of stormwater runoff on
wetland biota as noted in the literature.
3.2 HYDROLOGIC IMPACTS
The hydrology of a wetland is considered to be one of the most important factors in establishing
and maintaining specific types of wetlands and wetland processes (Mitsch and Gosselink 1986).
Hydrology is recognized as a key factor in wetland productivity, vegetation composition, nutrient
imports, salinity balance, organic accumulation, sedimentation transport, and soil anaerobiosis.
The hydrology of wetlands is reflected in its hydroperiod, the seasonal pattern of water level
fluctuations described by the timing, duration, frequency, and depth of inundation. The
hydroperiod of a wetland is a result of the balance between the inflows and outflows of water
(water budget) and the storage capacity of the wetland as defined by its morphology. Tidal
patterns can have a large influence on the hydroperiod for coastal wetlands. The residence time of
a wetland is also a product of the wetland hydrology. Modification of inlet and outlet conditions
at existing wetlands or the design of inlets and outlets at newly constructed wetlands will affect
the hydrology of the wetland.
Watershed hydrology upstream of the wetland also plays an important role in the hydrology of a
wetland. The quantity and quality of surface and subsurface inflows to the wetland are dependent
on conditions in the upland areas. Urbanization of upland areas increases the percentage of
impervious area (roof tops and paved surfaces) in the watershed, which results in increased
runoff volumes and peak flow rates and decreased infiltration and base stream flow. The removal
of trees and other vegetation cover and the installation of piped or channelized storm drainage
systems magnify these hydrologic changes. In general, the hydroperiod of a wetland will have
higher storm event peaks and longer periods of low water levels as its watershed area becomes
more urbanized.
Through the hydraulic forces and water level fluctuations associated with wetland hydrology,
stormwater runoff has the potential for changing wetland morphology and impacting the plant and
animal life of a wetland.
33
-------
3.2.1 Impacts on Wetland Morphology
The hydrology of a wetland influences its morphology. Inflows to the wetland transport
sediments to the wetland from the surrounding watershed. The particle size of the sediment and
the flow velocities and water depths within the wetland determine where this sediment is
deposited. Decomposing organic matter also accumulates on the bottom of the wetlands.
Sediments and bottom deposits within the wetland may be eroded or scoured by increased
velocities, turbulence or wave action and redeposited in a new location, either within the wetland
or at a downstream location. Depending on the pattern of sediment scour and deposition, a
wetland may have a veiy uniform surface that may lead to homogeneous stands of single species
or a surface with many niches that allows diverse habitats to develop.
Stormwater runoff containing high loads of sediment can cause excessive siltation, reduce light
penetration, cover fish spawning substrate, reduce dissolved oxygen content, clog fish gills, bury
benthic organisms, and decrease storage and channel capacities (Stockdale 1991). Although these
impacts are reduced in downstream receiving waters, the impacts may be significant for the
wetland. Significant changes to wetland morphology can impact the plant and animal life of the
wetland.
3.2.2 Impacts on Plants
The plant life in a wetland is highly dependent on the hydrology of the wetland. Changes to the
hydroperiod characteristics of the wetland (the Lime of year, frequency, duration, and depth of
inundation) can lead to decreased growth or mortality in some species. Plant species diversity,
measured by richness, evenness, and dominance, has been observed to decrease with high water
level fluctuations (Azous 1991). Flooding can cause soils to become anaerobic and lower the
oxidation-reduction potential of the soil, which can reduce the availability of nuthents in the soil
and affect normal root functions.
Extensive literature sources exist on the flood tolerance of woody plants (Stockdale 1991). The
flood tolerance and sensitivity of plants vary with species and the maturity of the stand. Seedlings
are generally more sensitive to flooding, although some species are capable of germination under
water. Some species of woody plants are able to survive deep, prolonged flooding for more than
one year, whereas others are unable to survive more than a few days of flooding during the
growing season (Whitlow and Harris 1979). Submerged vascular plant communities have been
shown to experience few impacts due to water level fluctuations (Davis and Brinson 1980).
Additional data are needed on the tolerance of submerged and emergent, non-woody plant species
to temporary and long-term flooding.
3.2.3 Impacts on Animals
Although animal life in an existing wetland is adapted to flooding to some extent, direct and
indirect impacts due to changes in the timing. frequency. duration and depth of flooding
associated with stormwater treatment can be identified (Lloyd-Evans 1989). Reduced animal
34
-------
populations or local extinction may occur from flooding deaths, loss of eggs or larvae, reduced
fertility, shortened reproductive seasons, shorter life spans, or slowed growth. flooding can also
reduce suitable habitat for nesting, rearing and cover; eliminate sources of available food; increase
the risk to predators; and increase disease and parasites. Extreme variations in water temperatures
from increased or decreased water depths can also impact wetland wildlife, particularly for non-
motile life forms. Increased areas of warm, shallow water can also create favorable habitat for
disease organisms and parasites. No studies have documented these impacts at stormwater
wetlands. Azous (1991) found no association between the species richness of aquatic
amphibians, mammals and birds and the range of observed water level fluctuations.
Increased inflows at existing wetlands can have positive affects on wildlife. Wetlands enlarged or
deepened by the increased inflows may have greater carrying capacity for wildlife. Desirable new
habitat areas such as productive edges, islands, or permanent pools may also be created. Large
wetland areas are thought to be more desirable than smaller areas as wetland bird habitats (Azous
1991). Wetlands constructed for stormwater treatment can also provide new habitats for animal
life in urbanized areas. Increases in wildlife populations have been observed at constructed
wetland sites (Hickman and Mosca 1991; Kadlec 1987 as cited in Niering 1989).
3.3 ACCUMULATION OF TOXINS
This review generally explored the uptake and accumulation of potential toxins by wetland
sediments, plants, and animals as reported in the literature. It is important to note that few of the
reports indicated concern regarding the fate or effects of contaminants in urban stormwater. Many
of the reports reviewed in this section referenced studies performed in wetlands receiving sewage
effluents or industrial discharges of some type. Urban runoff, especially from residential
watersheds, frequently has much lower concentrations of pollutants than sewage effluents or
industrial discharges.
Wetlands can serve as intermediaries in the hydrologic circulation of toxic materials from both
natural and human-caused sources. These toxins include metals and organic compounds that can
be deleterious to living organisms. Metals of significant concern include lead (Pb), copper (Cu),
zinc (Zn), chromium (Cr), cadmium (Cd), nickel (Ni), mercury (Hg), arsenic (As), and selenium
(Se) (Homer 1986). In sufficient concentrations, aluminum (Al) and manganese (Mn) may also
be toxic (Homer 1986; Small 1971).
Organic toxins include fulvic and humic compounds that may result from sources of human-
related processes and natural sources. Sources of organics from humans include petroleum-based
substances such as combustion products (polycyclic aromatic hydrocarbons), industrial chemicals
(PCBs and pentachlorophenol), and pesticides (chlorinated hydrocarbons, organophosphate
insecticides, and phenoxy herbicides). These compounds include many carcinogens and
mutagens (Homer 1986). The body of research on the behavior of organic toxic materials in
wetlands is less complete than that of metals, due to the expense of collecting samples and
performing organic chemical analysis.
35
-------
3.3.1 Accumulation in Sediments
Sediments typically constitute the most significant store of toxic substances available to organisms
in a wetland (Mitsch and Gosselink 1986). Metals and toxic organic compounds can be taken up
by plants from the sediments and can be introduced into the food web (Kadlec and Kadlec 1979;
Kreiger et al. 1986; Kadlec and Tilton 1979).
Both metals and organics tend to be adsorbed to finely-divided solids, depending on such
conditions as pH, oxidation-reduction potential, and salinity (Homer 1986). Hart (1982) stated
that metals are most likely to be adsorbed to clay- and silt-sized particles, because they have the
largest collective surface areas. Solid organic matter has an adsorption capacity intermediate
between clays and metal oxides (such as A1203, FeO2H, and Mn02). Metals may occur in
sediments as insoluble sulfides or complexed with fulvic and humic acids (Boto and Patrick
1979). The way a metal is complexed determines its availability to plants (Homer 1986).
The relatively longer residence time of water in wetlands, as compared to more swiftly moving
waters, is due to their flatness and the filtering action of the vegetation, and allows suspended
solids to drop out and be retained (Kadlec and Tilton 1979; Homer 1986). Woodward-Clyde
Consultants (1991) found that the greatest concentration of metals in sediments occuned at the
location nearest the stormwater inlet. The sediment concentration and bioavailability of copper,
lead and zinc were found to be at or near background levels in the downstream marsh area. The
long-term storage of metals depends on their burial in deep sediment (Mitsch and Ciosselink
1986). Sediments can be resuspended by storms and tides (Homer 1986) or remobilized by
vegetation (Windom 1977; Teal et al. 1982).
No wetland is a closed system, but some types retain more toxins than others. Wetlands that are
located high within a watershed may encounter less frequent flooding, which allows greater
amounts of peat to be permanently deposited (Homer 1986). If a wetland is saturated with
metals, the concentration of metals in the outflow increases (Mitsch and Gosselink 1986). In salt
marshes and estuaries, saltwater may cause some metals to pass through wetlands more rapidly.
Teal et al. (1982) found that, in a jalt marsh treated with sewage sludge, most iron (Fe), mercuiy
(Hg), and lead (Pb) were retained in sediments, while significant portions cadmium (Cd), zinc
(Zn), and chromium (Cr) formed soluble complexes and were flushed Out.
3.3.2 Accumulation in Plants
Plants take more metals from the sediment than from the water column. However, phytoplankton
can remove metals directly from the water, releasing them upon death to the sediments or to the
water (Hart 1982). In general, far greater amounts of metals remain in the sediment than are taken
up by plants (Dubinski et al. 1986; Banüs et al. 1975; Teal et al. 1982; Homer 1988). Metals
often do not circulate freely in the wetland but remain close to the source of contamination (Teal et
al. 1982; Simpson et al. 1983; Homer 1988).
36
-------
Some plants are apparently able to exclude toxic metals selectively. For example, in an
ombrotrophic (acidic) bog, aluminum (Al) and manganese (Mn) levels in plants were found to be
comparable to the levels in plants in other habitats, despite the potentially toxic levels of these
metals in the bog (Small 1972). The ability of some plants to regulate the amounts of metals they
assimilate may explain why productivity is not diminished by polluted conditions. Valiela et al.
(1975) did not observe detrimental effects on plant productivity from metals or chlorinated
hydrocarbons during an experiment in which sewage sludge was applied to a salt marsh. The
decomposition of plant parts can allow toxins to be released into the sediments and detrital food
chain, however (Mitsch and Gosselink 1986; Kadlec and Tilton 1979).
Organic compounds undergo many of the same processes in wetlands as metals, including
adsorption to sediments and plant uptake. In addition, they can be biodegraded. For example, a
Scirpus lacustris (bullrush) marsh was found to reduce organic compounds such as phenol, p-
cresol, pyridine, and aniline in 7 to 52 days (Seidel 1966). Petroleum hydrocarbons can be
decomposed by microbes, if loadings are not excessive. Pesticides present a special case due to
their potential toxicity to plants and varying degradability (Kadlec and Kadlec 1979).
3.3.3 Accumulation in Animals
The uptake of toxic materials by plants can introduce them into the grazing and detrital food
chains, with potentially deleterious effect. Metals from sewage effluents introduced to wetlands
tend to accumulate in the food chain (Kadlec and Tilton 1979). Renfro (1972) found that
radioactive zinc ( 65 Zn) from the Hanford Nuclear Reservation in Washington was taken up by
perch and amphipods in a Columbia River wetland. Several hundred swan deaths per year in the
lower Coeur d’Alene River Valley in Idaho have been attributed to high concentrations of lead
(Pb) from mine spoils in wetland sediments and in the horsetail ferns eaten by the swans
(Kreiger, personal communication; Kreiger et al. 1986).
Teal et al. (1982) found that the concentrations of toxic metals in salt marsh animals did not
correspond directly with the level of contamination from sewage sludge but varied with metal
type. In 10 years of sludge additions to a salt marsh, two cases of organic compound
contamination, one by aidrin and the other by PCBs and other halogenated hydrocarbons, were
found to have reduced populations of fiddler crabs and Tabinid larvae, respectively. In
conclusion, the relative responses of plants and animals to toxic metals and organic compounds
indicate that these contaminants are more likely to affect animals negatively.
3.4 NEED FOR FURTHER STUDIES
Hydrologic impacts to wetland biota depends on changes to the established hydroperiod. The
magnitude of change that results in significant impacts is not known. Impacts will also vary with
the type of wetland and the species present. Additional studies are needed to assess impacts to
submerged and emergent, non-woody plant species to temporary and long-term flooding. Use of
indicator species to identify stormwater impacts also should be considered (Lloyd-Evans 1989).
37
-------
There has been relatively little study of the cycling of organic toxic materials in wetlands due to
the expense of analyses. Most studies have focused on metals, most typically from sewage
effluent. More research is needed on the cycling of organic and metal toxins from other effluents,
such as urban runoff, highway runoff, and industrial discharges. Studies should concentrate
especially on accumulation in animals, because these biota have been studied less than plants and
because there is evidence of relatively greater uptake by animals. The geographic focus of many.
studies has been on such wetlands as salt marshes, southern hardwood swamps, and northern
peat bogs. More work needs to be done on northern inland freshwater wetlands and lacustrine
and riverine systems.
38
-------
4.0
COMPARISON OF WETLAND AND DETENTION BASIN PERFORMANCE
4.1 INTRODUCTION
Detention facilities have traditionally been constructed to control stormwater runoff quantities.
These facilities temporarily store stormwater runoff and later release the water at a lower flow
rate. Detention basins and ponds can be designed for water quality enhancement by including a
permanent pool of water and designing inlet and outlet structures to maximize detention.
Quiescent velocities within the basins allow sediments to settle out of the stormwater and chemical
and biological removal processes to occur. Detention basins usually do not have vegetation
within the permanent pool, but the banks may be planted with grasses for erosion control.
Detention basin/constructed wetland treatment systems have been recommended for stormwater
treatment (e.g., Meyer 1985; Martin and Smoot 1986; Wotska and Oberts 1988). Typically in
these systems, stormwater runoff discharges to the detention basin which then releases the water
to the wetland for additional treatment. The detention basin can provide pre-treatment for the
wetland, reducing the sediment and pollutant loads to the wetlands.
In other instances, detention basins and constructed wetlands are competing alternatives under
consideration for stormwater treatment. The designer or planner requires knowledge of the
relative.pollutant removal efficiencies, environmental impacts, maintenance requirements and
costs for the two alternatives.
This section discusses the results from four case studies which have compared the performance of
wetlands to detention basins through simultaneous monitoring of both systems. These studies
include combined detention basin/wetland systems and independent detention basins and wetlands
within the same watershed.
4.2 CASE STUDIES
Of the stormwater studies reviewed for this report, the following were selected as case studies:
• The Orange County Treatment System in Florida (Martin and Smoot 1986).
• The Lake Apopka Reservoir and Flooded Field Experiment (Reddy et al 1982).
• The Pittsfield-Ann Arbor and Swift Run System (Scherger and Davis 1982).
• The McCarrons Treatment System in Minnesota (Wotzka and Oberts 1988).
39
-------
4.2.1 The Orange County Treatment Facility
The Orange County Treatment Facility consists of a detention basin which operates in series with
a natural wetland. The detention basin was the first treatment unit in the two-unit system and
received untreated stormwater, whereas the wetland received already treated effluent from the
detention basin. The watershed which contributes runoff to this system is approximately 41.6
acres of forest, highway, and high- and low-density residential land uses. The drainage system
transports runoff through a network of curbs, gutters and storm drains to the detention basin and
then to the wetland. The basin’s surface area during dry weather is 8600 square feet (0.2 acres),
with water depths ranging from 8 to 11 feet. The wetland area during dry weather is 32,000
square feet (0.73 acres), with water depths ranging from 3 to 5 feet. Vegetation in the wetland
consists of cypress trees with a dense growth of hyacinths, duckweed, and cattails. In addition,
there are thick berry vines (such as blackberry and wild grape) growing in portions of the
wetland.
Martin and Smoot (1986) reported that the detention basin did allow settling of some suspended
material from the storrnwater runoff, but did not perform as well as other detention basins. The
following list presents the removal efficiencies for the detention basin and the wetland for five
pollutants: total suspended solid (TSS), total lead (TPb), total zinc (TZn), total nitrogen (TN),
and total phosphorus (TP). The detention basin’s reported TSS removal of 65 percent would
only be considered a moderate performance for a detention basin. The list below also indicates
that the wetland performed better than the detention basin for three out of the five pollutants
studied.
PERCENT REMOVAL
PARAMEIER Detention Basin Wetland
TSS 65 66
TPb 41 75
TZn 37 50
TN 17 30
TP 21 19
Martin and Smoot (1986) also discussed the effectiveness of the detention and the wetland in
reducing dissolved constituents. For example, they found that the detention basin had no effect
on major ions (Cl-, Ca 2 , Mg 2 , etc.), whereas the wetland removed about 20 percent of the Ca+
and Mg . Furthermore, Martin and Smoot (1986) found that the efficiencies in the detention
basin for dissolved lead and nitrogen were 29 and 24 percent, respectively. The reduction
efficiencies of the wetland were found to be 54 and 13 percent, respectively. Dissolved
phosphorus and dissolved ortho-phosphorus in the detention basin were each reduced by 70
percent. Dissolved phosphorus and dissolved ortho-phosphorus showed no reduction in the
wetlands. Also the efficiency for dissolved zinc was reported to be 17 percent for the detention
basin and 75 percent for the wetland.
Residence time and turbulence were indicated as the key factors affecting the settling of suspended
particles. Residence time will vary with flow rate, the mixing with water in dead storage, and the
40
-------
volume of live storage in the detention basin. When the storage capacity is exceeded, flow moves
directly from the detention basin inlet to the outlet. This condition, known as “short-circuiting”,
was observed during several monitored storms. Short-circuiting can change the residence time
from a hypothetical period of several hours to an actual period of just a few minutes. The result
of short-circuiting is a decrease in the effectiveness of settling suspended material. Turbulence
was also observed during their investigation for at least one storm. Turbulence, which is
dependent on rainfall intensity and detention basin inlet flow rate, can scour the bottom material,
causing an increase in loads in detention basin effluent. Wind can also induce turbulence in some
systems.
The observed short-circuiting and turbulence may help explain why the detention basin did not
show higher efficiencies than the wetland and had only moderate performance for a detention
basin system. Typically, detention basins can remove up to 90 percent of the suspended material
under more favorable cbnditions. Based on the observations of Martin and Smoot (1986), the
detention basin and wetland were equally effective at removing suspended material, even though
the wetland was acting on already treated stormwater from the detention basin.
4.2.2 The Pittsfield-Ann Arbor and Swift Run Systems
Scherger and Davis (1982) presented a paper assessing the treatment performance of a man-made
detention basin (Pittsfield-Ann Arbor) and a naturally-occurring wetland (Swift Run). These
units received runoff from different watersheds. The watershed areas for the Pittsfield-Ann Arbor
and Swift Run systems are approximately 6300 and 3100 acres, respectively. The watershed
areas consist of open, high-, medium- and low-density residential, commercial, industrial, park,
and agriculture land uses. The drainage systems transport runoff through a network of curbs,
gutters and storm drains. The Pittsfield-Ann Arbor detention basin’s surface area ranges from
226,510 square feet to 1,122,400 square feet (5.2 to 25.8 acres), with water depths ranging from
4 to 10 feet. The Swift Run wetland area ranges from 419,900 square feet to 1,077,000 square
feet (9.6 to 24.7 acres), with water depths ranging from 1.5 to 4.5 feet.
The list below presents the removal efficiencies in the detention basin for TSS, TP, Kjeldahl
nitrogen (TKN), total iron (TFe) and total lead (TPb). In addition, the natural wetland removal
efficiencies are presented for the same pollutants. Overall, the effectiveness of the wetland was
greater than that of the detention basin.
PERCENT REMOVAL
PARA1 ’U TER Detention Basin Wetland
TSS 39 76
TP 23 49
TKN 14 20
TFe 17 62
TPb 61 83
41
-------
The performance of the detention basin was lower than what would be expected from a theoretical
analysis. Scherger and Davis (1982) discussed several factors which were thought to influence
the removal efficiencies. These included detention time, short-circuiting, peak flow rates, and
particle sizes. They determined that detention time was the principal factor that affected the
performance. They reported average detention times ranging from 3.7 to 36 hours, depending on
the event. The wetland had detention times ranging from 12 to 82 hours. In general, the wetland
had larger detention times than the detention basin, which could partially explain why the wetland
performance was superior in removing suspended material.
The longer detention time of the wetland also helped to reduce the effects of peak flow rates. The
removal of suspended material was found to be dependent on the peak flow rates entering the
system. Peak flow conditions were reported to have a significant influence on the detention basin
performance, and scouring was thought to have occurred at the outlet of the detention basin on a
regular basis. Scherger and Davis (1986) suggested that these factors were the principal cause of
the differences between the performance of the detention basin and wetland for removal of
suspended material.
Scherger and Davis (1982) reported that nitrogen and phosphorus removals by both the detention
basin and wetland were low. They suggested that nitrogen removal in the detention basin was
low because more than 50 percent of the Kjeldahl nitrogen was in the soluble form. In the
wetland, however, the removal was increased by the assimilation of nitrogen in plants. However,
the response of the wetland was reported to depend on the season of the year. Warmer months
showed good removal of nutrients probably because of nutrient uptake for plant growth. Winter
months showed increases in the export of nutrients from the wetland (probably due to decaying
vegetation). The only explanation given as to why the phosphorus removal was low in the
detention basin was that most of the phosphorus was in the insoluble form.
In conclusion, the wetland performed better at removing the suspended material (including TSS,
TPb, and TFe) due in part to greater detention times with less influence from peak discharge. The
wetland was also more efficient at removing both phosphorus and nitrogen due to the nutrient
assimilation in aquatic plants.
4.2.3 The Lake Apopka Reservoir and Flooded Field Experiment
The Lake Apopka Reservoir and Flooded Field Experiment investigated the treatment efficiency of
reservoirs and flooded fields, some of which were stocked with aquatic plants to reduce nutrients
(Reddy et al. 1982). This experiment included three configurations of reservoirs and three
configurations of flooded fields. The first consisted of a single reservoir or a single flooded field
with no vegetation designated as “control”. The second consisted of a single reservoir or flooded
field with vegetation. And the third consisted of reservoirs in series or flooded fields in series
with vegetation. Each configuration had the same total area of 0.92 acres. The inlet water
consisted of agricultural runoff pumped for a period a 6 hours per day and 6 days a week at 251
gallons/minute and 114 gallons/minute for the reservoirs and flooded fields, respectively. The
depths of water in the reservoirs and flooded fields were approximately 1 and 0.2 feet,
42
-------
respectively. The vegetation that was used in the analysis included water hyacinth, elodea, and
cattails.
The following results were indicated for the reservoirs:
___ PERCENT REMOVAL
PARAMEF ER Control Single Series
Nitrate 55 54 68
Ammonium (NH ) 34 42 58
Ortho-phosphorus 21 63 75
Total phosphorus (TP) 26 44 61
For the flooded fields, the following results were indicated:
PERCENT REMOVAL
PARAMFJER Control Single Series
Nitrate 48 51 64
Ammonium (NIL ) 39 44 52
Ortho-phosphorus 28 24 17
Total phosphorus (TP) 11 16 7
Reddy et al. (1982) discussed some possible processes which may have led to the removal of
nutrients from the drainage water. For example, they suggested that uptake by aquatic plants,
denitrification, and diffusion into the underlying anaerobic soils were responsible for nitrate
reductions. These processes probably occur in both the reservoirs and the flooded fields, and
may explain why both the reservoirs and the flooded fields perfonned equally. Ammonia was
thought to be removed by assimilation into aquatic plants and algae and through nitrification, and,
again, no appreciable difference was found between the reservoirs and the flooded fields for the
removal of ammonia. The reservoir system significantly reduced both ortho-phosphorus and total
phosphorus, whereas the flooded fields did not perform effectively in the removal of these
constituents. The process of removal for phosphorus was thought to be assimilation by aquatic
plants and algae, adsorption, and precipitation. No indication was given to suggest why the
flooded fields functioned poorly in the removal of phosphorus.
The results of this study suggest that detention basins (control reservoir or flooded fields without
aquatic vegetation) and wetlands (single or series of reservoirs or flooded field with aquatic
vegetation) perform equally in the removal of nitrate and ammonium. Furthermore, the reservoir-
type wetlands were observed to be more effective in the removal of phosphorus than the flooded
field wetland systems. Reddy (1982) indicated that the flooded fields were unable to handle the
same hydraulic loading as the reservoirs. They conclude that six times more drainage water can
be pumped into the reservoirs while obtaining the same removal efficiency as the flooded fields.
43
-------
4.2.4 The McCarrons Treatment Facility System
Wotzka and Oberts (1988) presented a paper discussing a combined detention-wetland stormwater
treatment facility. The McCarrons Treatment Facility consisted of a 30-acre detention basin with
an average depth of 1.2 feet and a 6.2-acre constructed wetland with an average depth of 2.5 feet.
The detention basin received stormwater and then discharged to the wetland. The contributing
watershed consists of 600 acres of primarily urban land use. The predominant vegetation in the
wetland consists of cattails with other emergent plant species.
Overall, they found very good results for the system. The detention basin proved to be more
effective than the wetland in reducing several pollutants. For example, the following removal
efficiencies were given for the detention basin and wetland:
PERCENT REMOVAL
PARAMETER Detention Basin Wetland
TSS 91 87
TP 78 36
TN 85 24
TPb 85 68
Wotzka and Oberts (1988) discussed some of the possible explanations for the good results of the
detention basin, and the differences between it and the wetland. In general, they believe the
treatment efficiencies are lower in the wetland due to pre-treatment by the detention basin. They
stated that the inflows into the detention basin are spread equally around the perimeter of the
detention basin, thus dissipating the entry velocities of the storm runoff. Dissipation of inflow
energy was thought to promote settling and minimize short-circuiting.
Wotzka and Oberts (1988) also suggested that the percent of phosphorus in the dissolved and
particulate phases affected the reduction potential. They found that more than 80 percent of the
phosphorus was in the particulate form, thus resulting in high removal efficiencies due to settling.
Apparently, the wetland did not perform as well as the detention basin because of the periodic
release of nutrients from decaying vegetation and the fact that significant pre-treatment had
occurred.
Wotzka and Oberts (1988) suggested that the high removal of phosphorus was due in part to the
newly-exposed soils on the bottom of the detention basins. They explained that the newly-
exposed soils were likely to have more sorption capacity available than the soils in the wetland
further downstream. They also suggested that once saturated soil conditions occur, the
phosphorus removal may become greatly reduced.
In conclusion, this study indicates that the detention basin performed better than the wetland
system. However, this may be misleading since the wetland receives pre-treated waters from the
detention basin. The detention basin removed the fraction of pollutants that are more readily
settled and treated, leaving the wetland with the finer, more difficult to treat pollutants.
44
-------
An extension of the study on the McCarrons Treatment System is presented in Oberts et al.
(1989). This study included data from four sites located in the Ramsey-Washington Metro
watershed in Minnesota in addition to the McCarrons Treatment System. These additional sites
included two detention basins (Lake Ridge and McKnight Basin), a wetland (Tanners Lake) and a
wetland/detention system (Carver Ravine). The detention basins located at Lake Ridge and
McKnight Basin performed well for most pollutants. At these sites, removal of particulate-
associated pollutants was comparable to the McCarrons Treatment System, but removal of soluble
nutrients was not as great. The authors indicated that this may be from the “polishing” provided
by the vegetation at McCarrons. The lack of permanent storage was reported as the cause for the
noted poor performance of the Tanners Lake wetland and the Carver Ravine wetland/detention
system.
4.3 SUMMARY
Due to the physical differences and variability between the treatment systems, it is not reasonable
to compare specific performance; however, in general, the detention basins and wetlands appear
to function equally well for the parameters reported. When considering that two of the studies
(i.e., Orange County and McCarrons Treatment Systems) included systems with a detention basin
upstream of the wetland, wetlands appear to be very effective treatment devices.
Removal rates of floatable materials such as oils and greases have not been studied in wetland and
detention basin systems. We would expect wetland vegetation to provide significantly more
surface area for oils and greases to adsorb to under storm conditions. These materials would then
become exposed as the water levels decline during normal and low flows. This capture and
exposure process would facilitate biodegradation of these types of materials and perhaps lead to
overall high removal efficiencies.
Relatively extensive information is available on the performance of detention basins as settling
devices. Ideally designed settling devices can remove up to 90 percent or more of the settleable
material suspended in water. Detention basins discussed in this review did not function as well as
might be expected. In many cases, short-circuiting and turbulent flows were thought to have
caused a decrease in the detention times and performance of the basins. Measures to reduce short-
circuiting and turbulent flows should be included in designs and plans. Properly designed
wetlands can have the advantage of less short-circuiting and turbulence.
The Orange County and McCarron’s treatment systems are examples of what we consider to be
good designs. There are many issues surrounding the use of wetlands for stormwater pollution
control. These include both concern about hydrologic impacts and pollutant uptake (Section 3)
and maintenance issues. Creating a detention area (open ponded water) upstream of a wetland
allows for the pre-treatment of stormwater before it reaches biota and provides an area where
maintenance dredging can be performed without disturbing vegetation, If designed properly, the
detention basin area will tend to settle out the heavier particulate matter thereby significantly
reducing the need for maintenance in the wetland and minimizing the disturbance of wetland soils
and biota.
45
-------
5.0
IMPROVING THE PERFORMANCE OF WETLANDS
This section presents a summary of suggestions from selected studies on ways to improve the
effectiveness of wetland systems to treat stormwater. This summary is a compilation of the
authors’ thoughts based on their observations of the specific wetland systems that they studied.
Hickok (1977 )
• Improve physical entrapment by increasing settling time.
• Microbial utilization can be enhanced by providing a substrate for growth (i.e., plants and
soil) and fostering aerated conditions.
• Plant uptake can be optimized by increasing the density of vegetation.
• Maximize soils and organic matter available for adsorption.
• Water level management increases surface microbial activity by allowing the soil to
become aerated during thy periods.
Reddy. Graetz. Campbell. and Sinclair ( 1982 )
• Reservoirs in series had higher removal efficiencies than a single reservoir with nearly the
same volume.
• Both the single reservoir and the reservoirs in series functioned better as treatment devices
than the control reservoir which contained n vegetation.
• Flooded fields in series had higher performance than a single field for nitrogen removal
but not for phosphorus.
• Three to 6 days of detention time was found to be required to remove 50 percent of
nitrogen in the water hyacinth reservoir.
• A detention time of 7.2 days was found to be required to remove 60 percent of ortho-
phosphorus in reservoirs containing water hyacinth and elodea.
46
-------
Brown (1985 )
• Create a detention basin just before the inlet to a wetland to improve the removal of
suspended malerial and preserve the life of the wetland.
• The detention basin can be used to spread inflows evenly over the wetland, minimizing
short-circuiting.
• Dense vegetation slows velocity and wave action, which increase sedimentation.
• Increase contact of stormwaters with vegetation, such as cattails, to increase the
assimilation of nutrients (especially phosphorus).
Canning (1985 )
• Maintain a high detention thne. An average detention time of 18 to 24 hours is generally
considered adequate.
• Most pollutants run off the landscape during smaller, more frequent storms. Therefore, a
detention basin must control the small storms as well as the large ones.
• Re-suspension of sediments can be minimized by fostering vegetation on the wetland
banks and bottoms.
• Grassy swales in combination with other freatment systems can improve performance.
The use of a grassy swale in lieu of a curb and gutter system showed 98 to 99.8 percent
removal of BOD, TSS, TKN, N0 3 -N, TP and TFe.
• Key elements in performance are vegetation density and species selection. Dense growing
species are preferred.
• The most effective plant species found for stormwater treatment were bulrush ( Scirpus
lacuatris) , common reed ( Phragmites commonis) , and the common cattail ( Tvpha
latifolia) .
• Vegetation should be harvested periodically to improve long-term performance.
• Use of a detention basin or sedimentation basin upstream of a wetland system can
preserve the life of the wetland.
• Water should move through the wetland as sheet flow (not channelized).
• Peat can be used to line the bottom of a detention basin to maximize the removal by
adsorption (e.g., heavy metals and organics).
47
-------
Meyer ( 1985 )
• Increased detention times allow greater sedimentation and adsorption.
• Anaerobic wetland soils are ideal for removing a variety of pollutants.
• Dissolved metals are adsorbed onto wetland soils having a higher percentage of organic
matter.
Blackburn. PimentaL. and French (1986 )
• Design the system to retain the first runoff of rainfall from the watershed (the first inch of
rainfall was used in their design).
• Broad planting shelves on the fringe of the wetlands proved to be most effective and
maintenance free (6.5 to 33ft in width).
• Vegetation such as spatterdock, arrowhead, and water lily proved to be the most desirable
for aesthetics and the most manageable in terms of maintenance.
• Wetlands should be designed to minimize management problems.
• Increased density of wetland plants and increased width of wetland shelves can increase
pollutant removal efficiencies.
• As wetlands matured, efficiency of treatment and wildlife utilization increased.
Harper. Wanielista. Fries, and Baker (1986 )
• Runoff into a wetland system should be released slowly to reduce erosion and increase the
adsorption potential.
• Runoff inputs discharging into the wetland system should be distributed evenly by use of
a swale or a diffuse-inlet structure to minimize short-circuiting.
• Flow through the system should not exceed 48 hours. Detention times greater than 48
hours reduced efficiencies for ortho-phosphorus due to releases from vegetation.
• Mean flow velocities in wetlands should not exceed 3.2 feet/hour.
48
-------
I artin and Smoot ( 1986 )
• Maintain high detention time and low turbulence.
• Minimize dead storage and short-circuiting.
• Maintain a high surface area to volume ratio.
• Minimize build-up of organic matter.
otzlca and Oberts (1988 )
• Use diffuse inlets to minimize short-circuiting.
• Develop an adequate maintenance program.
• Quiescence between events is important.
• Maintain a high detention time.
• Optimize the use of filtration by vegetation.
King County Resource Planning Section (1991 )
• Wetlands should be managed in coordination with the associated watershed.
• Protect wetlands from non-stormwater related disturbances such as human intrusion.
• Determine baseline conditions of natural systems, including at least the wetland
hydroperiod and heavy metal accumulation, from which to measure and mange changes in
the wetland.
• For natural systems, large deviations from the predeveloped hydroperiod should be
avoided until more is known about plant succession.
• On-site best management practices should be used to control stormwater quantity and
quality prior to release to wetlands.
• Use presettling ponds upstream of the wetland to remove the largest particles and
minimize the need to disturb the wetland by dredging.
49
-------
6.0
ISSUES REGARDING WETLANDS USE AND EFFECTIVENESS OF
WETLAND SYSTEMS FOR STORMWATER POLLUTION CONTROL
6.1 USE OF NATURAL WETLANDS FOR STORMWATER POLLUTION
CONTROL
A significant issue is whether natural wetland systems should be used as stormwater control
measures. It is important to note that EPA Region 5 discourages the use of natural wetlands for
stormwater control. In general, natural wetlands have been found to be somewhat less predictable
than constructed wetlands in terms of pollutant removal efficiency. This difference may be due to
the fact that constructed wetlands have generally been engineered to provide favorable flow
capacity and routing patterns. As a result, they tend to detain inflows for longer periods of time
and have less short-circuiting than many natural systems.
People often question whether it is appropriate to use a natural, healthy wetland for such
purposes. The concern is whether the modified flow regime and the accumulation of pollutants
will result in undesirable environmental effects. There are many situations where natural wetlands
have been receiving urban runoff for years. Some have shown significant degradation due to a
number of factors, including urban runoff, whereas others have been less affected. A general
consensus from the literature is that the use of a healthy natural wetland for stormwater pollution
control should be discouraged. In the case of rehabilitating a natural but degraded wetland,
careful attention should be given to the design of modifications so that the applied runoff receives
sufficient pre-treatment. One pre-treatment technique would be to use pond areas to provide an
opportunity for suspended materials to settle out before the flows enter the wetland. Other
possible options include routing inflows to the wetlands through upstream grass swales, oil/water
separators, heavily vegetated areas (e.g., thick, shallow cattail area), and overland flow areas.
Under current federal regulations, stormwater discharges from industries and larger municipalities
(over 100,000 population) to natural wetlands considered to be waters of the United States require
permits through the Clean Water Act NPDES permit program. Restrictive conditions on the
permits are determined on a case-by-case basis. Also, filling of natural wetlands requires a permit
from the U.S. Army Corps of Engineers under section 404 of the Clean Water Act and review
under the National Environmental Policy Act.
6.2 NEED FOR ADDITIONAL STUDIES
Although there is little evidence of problems in wetlands that have been receiving stormwater
runoff, the available.data are quite limited, and there is a critical need to develop additional
information on impacts. It is recommended that additional studies on the impacts to biota be
undertaken.
Another important aspect of wetlands use for storm water treatment that has not been well studied
is the maintenance needs for wetland systems. Such maintenance activities could include
50
-------
sediment removal and plant harvesting. Further studies should be conducted to address the need
for, frequency of, and appropriateness of maintenance.
The gathering of more information on wetland effectiveness would be of benefit to developing
design procedures for sizing wetland treatment facilities. There is currently not enough
information fl the existing literature to develop design guidelines for constructed wetland
treatment systems. Additional studies are needed to broaden the type of wetland systems studied,
develop information, long-term performance and evaluate seasonal characteristics of wetland
performance.
In completing the analysis of the current data available on stormwater treatment wetland
effectiveness, it was found that most studied did not contain enough information on study and
wetland characteristics to perform a detailed analyses on factors affecting wetlands treatment
performance among different wetlands. Table 9 presents a summary of the information that we
feel would provide a better means to compare wetland designs and treatment effectiveness from
different wetland systems. This list is probably not inclusive.
An example of how this type of information could be useful, involves comparing watershed to
wetland characteristics in affecting performance. In section 2.4, we compared watershed to
wetland size ratios. We feel that a comparison of average storm volume to wetland volume would
have made a better analysis of the effect of wetland “sizes” for treatment abilities. The currently
available data, which predominantly presents areas of wetlands and watersheds, did not allow for
this kind of comparison. Percent impervious factors and therefore runoff volumes could be very
different in different watersheds. Data such as percent imperviousness, land-use information, and
rainfall statistics, along with wetland volume information would have allowed us to compare
average runoff volumes, wetland volumes, and resulting performance characteristics.
6.3 KNOWN STUDIES CURRENTLY UNDER WAY
We are aware of several short- and long-term studies that are now underway or will begin shortly
and will add significantly to the information currently available. Brief descriptions of these
studies follow.
Demonstration Urban Storrnwater Treatment (DUSD Marsh
The DUST Marsh is a constructed wetland system on the eastern shore of San Francisco Bay. It
was monitored for about a year during its initial operating period (Chan et al. 1982). In a recent
study, concentrations and potential bioavailability of selected metals were measured within the
sediment and water column of the Crandell Creek - DUST Marsh system (Woodward-Clyde
1991). This study will be continued in the winter and spring of 1991-92 to investigate the uptake
of copper, lead, and zinc by emergent macrophytes and to determine the spatial variability of soil
metal concentraijous downstream of various stormwater inlets.
51
-------
Table 9. Suggested Reporting Infonnation for Studies that Assess the Ability of Wetlands to Treat Stormwater Pollution
Classificarion
Constructed or Natural or Combination Wetland?
Vegetatron Species
Groundwater Interaction?
Total Flow from Average Storm
Wetland Volume
Vegetation Density
(percentage open and vegetated)
Vegetation Types
(submerged, emergent, floating)
Wefland Area
Aspect
Side Slopes
Soil Type and Defllm
Watershed Area
(acres)
Land Use
(percent residential, industrial, agricultural, undeveloped, etc.)
Percent Impervious
(percent impervious area)
Rainfall Data/Statistics :
Average Rainfall During Study (in/year)
Average Number of Storms per year
Average Storm Intensity (in/br)
Average Storm Duration (hours)
Avg Time Between Storms (days)
Low flow inflow rate(sl
(maximum storage volume)
Average Detention Time for Average Storm
Water Depths
Inflow Condition
(hours)
4rr5 ment of Inflow
(settling forebays, overland flow, detention basin, grassed swales, etc.)
Maintenance Practices ( iacl frequencyl :
Plant Harvesting?
Flushing?
Sediment Removal?
Chemical Treatment?
Other Maintenance?
Provide Hydrolocy and WateuOuality Data for all Storms Monitored
Type of Samples
(grab or oumposite)
Number of Storms Monitored
Method Used to Compute Pollutant Removal Efficiencies
Dominant Removal Mechnnisms
(crtl iiiir ritatit ,n. .i,Ic ‘t;’t i,ni flit. t it’ll i’..’ hr nu. il C l •
(length.to-width ratio)
(.11
t ’J
(minimum, maximum, average)
(discrete or diffuse inlets)
-------
Des Plaines River Wetlands Demonstration Project
This research project involves 450 acres of riparian land along 3 miles of the upper Des Plaines
River, approximately 35 miles north of Chicago. The project involves creation of eight
experimental wetland cells by rebuilding and revegetating areas previously drained or devoid of
wetlands. The purpose of the study is to assess the effectiveness of the wetlands to treat runoff
from an urban and agricultural watershed, and to investigate the effectiveness of different wetland
management techniques including the type of vegetation used, variations in flow rates and water
depths, and soil conditions. Construction of the experimental wetland began in April 1986 and is
now complete at four sites. Construction at the remaining sites is anticipated to be completed in
1991, and the research work will be completed by 1996. Research has begun on the completed
wetlands and preliminary results for the project are available (e.g., Hey and Barrett 1991). The
restored wetlands are estimated to be trapping more than 80 percent of the incoming sediment and
nutrients from the river. Native plant communities are displacing weeds, and shore bird
populations have increased since creation of the wetland (Hickman and Mosca 1991). A
description of the research plan is summarized in Hey (1987), and project publications are listed
in the project bibliography (Wetlands Research, Inc. 1991).
Emerald Square Mall. North Attleborough. Massachusetts
A series of detention basins and riverine wetlands were constructed as part of this shopping center
development near Boston (Daukas et al. 1989). To date, water quality data has been collected
only at the outlet of the system. Research plans for more complete monitoring are to be
implemented in the near future.
Fort Collins. Colorado
The City of Fort Collins has constructed a number of wetlands within its jurisdiction (Horak
1988). Projects include creation of a new wetland in an existing detention basin and protection of
existing wetlands. Funding for proposed research programs on these wetlands is currently being
sought.
Orange County Treatment System
The U.S. Geological Survey, Altamonte Springs, Florida, Field Office is continuing research on
the Orange County Treatment System. Initial work on this wetland system was conducted in
1982 through 1984 and is documented in Martin and Smoot (1986). In 1988, the system was
modified by relocating the inlet further from the outlet to increase residence time. Preliminary
assessments indicate that relocation of the inlet has resulted in flushing of sediments and other
pollutants from the wetland. Water quality monitoring has been conducted during the past two
years, and the results are now being documented.
53
-------
Peconic Bay Constructed Wetlands
Town of Southhold, Long Island, New York, has constructed a wetland to collect stormwater
runoff, and is in process of designing a second wetland. The wetlands are expected to improve
the water quality of the Peconic Bay System which has been plagued by brown tide algal blooms
since 1985. The Cornell University Cooperative Extension will assist the Town in the design and
construction of the second constructed wetland and monitor the effectiveness of both constructed
wetlands over a two-year period. The effectiveness of the constructed wetlands will be
determined by analysis of water samples and flow measurements taken upstream and downstream
of the wetlands as well as through analysis of ambient water quality conditions in the Peconic Bay
System. Plant tissue and wetland basin sediments will also be sampled. Plant survival will also
evaluated in this project
Pu get Sound Wetlands and Stormwater ManagementResearch Program
This research program is designed to examine the effects of urban stormwater runoff on the
wetlands of the Puget Sound region, Washington (King County Resource Planning 1987).
Implementation of the research plan began in 1987. The initial research effort was a broad survey
to define the characteristics of wetlands that had and had not been affected by urbanization of their
watershed. In 1988, a long-term field study was initiated to follow the urbanization process by
monitoring hydrology and water quality before, during and after urbanization. Wetland
sediments, vegetation, and animal communities are also being monitored. To spread out the
resources allocated to this long-term program, a total of five years of data will be collected during
a longer period of time with gaps of years having no data collected. Stonnwater management
guidelines have been developed in draft form and will be extended and refined as additional
research data become available. Initial results and program status are presented by King County
Resource Planning Section (1991).
Shop Creek Drainage Outfall System
The Shop Creek Drainage Outfall System is a water quality enhancement project located in a 640-
acre urban watershed draining into Cherry Creek Reservoir in southwest Aurora near Denver,
Colorado. Design recommendations for the system are presented in Wulliman et al. (1988).
Construction of the system’s detention pond, drop structures and wetland areas was completed in
July 1989. Water quality monitoring of the system began in the spring of 1990. Although the
primary focus of the monitoring program is on phosphorus removal during storm events,
analyses will be conducted to determine suspended sediment, nitrogen, alkalinity, COD, and
metal concentrations. Preliminary results for stormwater monitoring from May through
September 1990 indicate total phosphorus removal efficiencies of 51 percent in the detention pond
and 12 percent in the wetland areas (Wulliman, personal communication, 1991). Subsequent
papers will report on the results of the monitoring program.
54
-------
uthwestF1orida Water Management District Stormwater Research Program
This comprehensive research program developed from a 1988 Southwest Florida Water
Management District stormwater management initiative (Southwest Florida Water Management
District 1990). The research focuses on wet detention and wetland treatment ponds to reduce
stormwater pollution in Florida State waters. Three types of research projects are included in the
program: (1) broad surveys of 24 detention ponds and 16 wetland treatment sites; (2) detailed
studies of four systems; and (3) detailed investigations of individual parameters at a pair of
constructed wet detention ponds. The broad survey of the wetland treatment ponds is in
progress. For at least one year, the ponds will be monitored for temperature, pH, conductivity,
water depth , dissolved oxygen, total suspended solids, and metals. The research program also
includes a work plan for a two-year study of stormwater pollution reduction at a native
herbaceous marsh and the effect of stormwater on the sediments and vegetation in the marsh.
Initial results from these studies are presented by Southwest Florida Water Management District
(1990) and other related publications.
US EPA Database on Constructed Wetlands for Water Oualitv Treatment
A database is currently being developed for information concerning the use of North American
wetlands for water quality treatmenL The first phase of the project is near completion and
involved selection and development of the database software, input of data from 96 wetland
treatment sites with 127 separate systems, and preparation of a brief report to summarize how to
use the database and its contents. Of the systems included in Phase 1, 70 percent receive
municipal wastewater and 66 percent are constructed wetlands. The database is composed of
seven individual databases: (1) site information, (2) system information, (3) permit information,
(4) design information for individual cells, (5) operational data for flows and water quality, (6) a
literature summary for wetland sites in the database, and (7) a catalog of contacts have relevant
knowledge about each wetland treatment system. The second phase will include a final updating
of the database with all readily available wetland information. The third phase will provide a
detailed analysis of the information in the database with guidelines for permitting and design of
new wetland systems for waste water treatment.
55
-------
7.0
REFERENCES
Association of Bay Area Governments (ABAG). 1979. Treatment of Stormwater Runoff by a
Marsh/Flood Basin. Interim Report
Azous, A. 1991. An Analysis of Urbanization Effects on Wetland Biological Communities.
M.S. Thesis. Department of Civil Engineering, University of Washington, Seattle,
Washington.
Banus, M.D., I. Valiela, and J.M. Teal. 1975. Lead, Zinc and Cadmium Budgets in
Experimentally Enriched Salt Marsh Ecosystems. Estuarine Coastal Mar. Science 3:42 1-
430.
Barten, J.M. 1987. Stormwater Runoff Treatment in a Wetland Filter: Effects on the Water
Quality of Clear Lake. Lake and Reservoir Management: Volume 3.
Barten, J. 1987. Nutrient Removal from Urban Stormwater by Wetland Filtration. Lake Line,
North American Lake Management Society 3(3): 6-7, pp. 10-11.
Benforado, J. 1981. Ecological Considerations in Wetland Treatment of Wastewater. In
Selected Proceedings of the Midwest Conference on Wetland Values and Management,
June 17-19, 1981. Edited by B. Richardson. pp. 307-323. St. Paul, Minnesota.
Bastian, R.K., P.E. Shanaghan, and B.P. Thompson. 1989. Constructed Wetlands for
Wastewater Treatment, Municipal, Industrial, and Agricultural. Lewis Publishers, Inc.
ISBN 0-87371-184-X.
Blackburn, R.D., P.L. Pimental, and G.E. French. 1985. Treatment of Stormwater Runoff
Using Aquatic Plants. Northern Palm Beach County Water Control District West Palm
Beach, Florida.
Boto, K.G., and W.H. Patrick, Jr. 1979. Role of Wetlands in the Removal of Suspended
Sediments. In Wetland Functions and Values: The State of Our Understanding. Edited
by P.E. Greeson, J.R. Clark, and J.E. Clark. pp. 479-489. American Water Resources
Association, Minneapolis, Minnesota.
Brown, R.G. 1985. Effects of Wetlands on Quality of Runoff Entering Lakes in the Twin Cities
Metropolitan Area, Minnesota. U.S. Geological Survey Water Resources Investigation
Report 85-4170.
Brown, R.G. 1984. Effects of an Urban Wetland on Sediment and Nutrient Loads in Runoff.
Wetlands, Volume 4, pp. 147-158.
56
-------
Canning, D.J. 1988. Urban Runoff Water Quality: Effects and Management Options.
Shorelands Technical Advisory Paper Number 4. Second Edition. Washington
Department of Ecology, Shorelands and Coastal Zone Management Program, Olympia,
Washington.
Chan, E., T.A. Bunsztynsky, N. Hantzsche, and Y.J. Litwin. 1981. The Use of Wetlands for
Water Pollution Control. EPA-600/S2-82-086. Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio.
Cooke, S.S. 199 La. The Effects of Urban Stormwater on Wetland Vegetation and Soils: A
Long-Term Ecosystem Monitoring Study. Puget Sound Research ‘91 Proceedings,
Washington State Convention Center, Seattle, Washington, January 4-5, 1991. Edited by
T.W. Ransom. Puget Sound Water Quality Authonty, Olympia Washington.
Cooke, S.S. 1991b. The Vegetation and Soils Components of the Puget Sound Wetlands and
Stormwater Management Research Program. In Development of Guidance for Managing
Urban Wetlands and Stormwater. Final Report May 1991. Report to Washington State
Department of Ecology, Coastal Zone Management Program, by King County Resource
Planning Section, Seattle, Washington.
Cooke, S.S., R.R. Homer, C. Conolly, 0. Edwards, M. Wilkinson, and M. Emers. 1989.
Effects of Urban Stormwater Runoff on Palustrine Wetland Vegetation Communities —
Baseline Investigation (1988). Report to U.S. Environmental Protection Agency, Region
10, by King County Resource Planning Section, Seattle, Washington.
Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of Wetlands and
Deepwater Habitats of the United States. U.S. Fish and Wildlife Service Publication
FWS/OBS-79/3 1, U.S. Fish and Wildlife Service, Biological Services Program,
Washington, DC.
Daukas, P., D. Lowry, and W.W. Walker, Jr. 1989. Design of Wet Detention Basins and
Constructed Wetlands for Treatment of Stormwater Runoff from a Regional Shopping
Mall in Massachusetts. In Constructed Wetlands for Wastewater Treatment, Municipal,
Industrial, and Agricultural. Edited by D.A. Hammer. Lewis Publishers, Inc. pp. 686 -
694.
Davis, G.B. and M.M. Brinson. 1980. Responses of Submerged Vascular Plant Communities
to Environmental Change. F\VS/OBS-79/33 & 80/42.
Day, J.W., Jr. and G.P. Kemp. 1985. Long-term Impacts of Agricultural Runoff in a Louisiana
Swamp Forest. In Ecological Considerations in Wetlands Treatment of Municipal Waters.
Edited by P.J. Godfrey, E.R. Kaynor and S. Pelczarki. Van Nostrand Reinhold
Company, New York.
57
-------
Detenbeck, N.E. and C.A. Johnston. 1991. Effects of Disturbance on Water-Quality Functions
of Wetlands. Final Report to Wetlands Water Quality Program, ERL - Duluth, U.S.
Environmental Protection Agency. Natural Resources Research Institute, Duluth,
Minnesota.
Dubinski, B.J., R.L. Simpson and R.E. Good. 1986. The Retention of Heavy Metals in
Sewage Sludge Applied to a Freshwater Tidal Wetland. Estuaries 9(2): 102-111.
Esry, D.H., and D.J. Cairns. 1988. Effectiveness of the Lake Jackson Restoration Project for
Treatment of Urban Runoff. American Society of Civil Engineers (ASCE), Florida/South
Florida Section, 1988 Joint Meeting.
Folsom, Jr., B.L., and K. Preston. 1984? Heavy Metal Content of Spartina alterniflora Grown
in Sediment from Pawtucket Cove, Rhode Island. U.S. Army Engineers. Waterways
Experiment Station, CE, Vicksburg, Mississippi.
Gersberg, R.M., S.R. Lyon, B.Y. Elkins, and C.R. Goldman. 1984. The Removal of Heavy
Metals by Artificial Wetlands. EPA-600/D-84--258. Report to the U.S. Environmental
Protection Agency, Office of Research and Development, Robert S. Kerr Environmental
Researcit Laboratory, Ada, Oklahoma.
Hammer, D.A. 1989. Constructed Wetlands for Wastewater Treatment, Municipal, Industrial,
and Agricultural. Lewis Publishers, Inc. ISBN 0-87371-184-X.
Harper, H.H., M.P. Wanielista, B.M. Fries, and D.M. Baker. 1986. Stormwater Treatment by
Natural Systems. Florida Department of Environmental Regulation, Report 84-026.
Hart, J.T. 1982. Uptake of Trace Metals by Sediments and Suspended Particulates: A Review.
Hydrobiologia 91:299-313.
Henson, E.B. 1987. Biomass Production in a Nonpoint Source Wetland. Aquatic Plants for
Water Treatment and Resources Recovery. 1987.
Hey, D.L 1987. Constructing Wetlands for Stream-Water Quality Improvements. Unpublished
manuscript. Wetlands Research, Inc., Chicago, illinois.
Hey, D.L., and K.R. Barrett 1991. Hydrologic, Water Quality and Meteorologic Studies. In
The Des Plaines River Wetlands Demonstration Project, Draft Final Report to the illinois
Department of Energy and Natural Resources. Wetlands Research, Inc., Chicago,
illinois.
Hickman, S.C., and V.J. Mosca. 1991. Improving Habitat Quality for Migratory Waterfowl and
Nesting birds: Assessing the Effectiveness of the Des Plaines River Wetlands
Demonstration Project. Technical Paper No. 1. Wetlands Research, Inc., Chicago,
illinois.
58
-------
Hickok, E.A., M.C. Hannanian, and N.C. Wenck. 1977. Urban Runoff Treatment Methods.
Volume 1 - Non-Structural Wetland Treatment EPA-600/2-77-2 17. Minnehaha Creek
Watershed District, Wayzata, Minnesota.
Horak, G. 1988. An Integrated Approach to Stormwater, Wetlands and Riparian Habitat. City
of Fort Collins. In Urban Wetlands. Edited by l.A. Kusler. Proceeding of the National
Wetland Symposium held June 26-29, 1988, Oakland, California. Association of
Wetland Managers, Berne, New York.
Homer, R.R. 1988. Long-Term Effects of Urban Stormwater on Wetlands. In Design of Urban
Runoff Quality Controls, edited by L.A. Roesner, B. Urbonas, and MB. Sonnen, pp.
451-466, American Society of Civil Engineers, New York.
Homer, R.R. 1986. A Review of Wetland Water Quality Functions. In Wetland Functions,
Rehabilitation, and Creation in the Pacific Northwest: The State of Our Understanding,
edited by Richard Strickland, pp. 33-50, Washington State Department of Ecology,
Olympia, Washington.
Homer, R.R. 1991. The Puget Sound Wetlands and Stormwater Management Research
Program: Program Overview and Hydrology and Water Quality Studies. In Development
of Guidance for Managing Urban Wetlands and Stormwater. Final Report. May 1991.
Report to Washington State Department of Ecology, Coastal Zone Management Program,
by King County Resource Planning Section, Seattle, Washington.
Jolly, LW. August 1990. The Efficiency of Constructed Wetlands in the Reduction of
Phosphorus and Sediment Discharges from Agricultural Watersheds. M.S. Thesis,
University of Maine, Orono, Maine.
Kadlec, R.H., and J.A. Kadlec. 1979. Wetlands and Water Quality. In Wetland Functions and
Values: The State of Our Understanding. Edited by P.E. Greeson, J.R. Clark, and I.E.
Clark, pp. 436-456. American Water Resources Association, Minneapolis, Minnesota.
Kadlec, R.H., and D.L. Tilton. 1979. The Use of Freshwater Wetlands as a Tertiary
Wastewater Treatment Alternative. Critical Reviews in Environmental Control, pp. 185-
212.
ICing County Resource Planning. 1987. Detailed Planning of the Puget Sound Wetlands and
Stormwater Management Research Program. King County, Washington.
King County Resource Planning Section. 1991. Development of Guidelines for Managing
Urban Runoff and Stormwater. Final report to Washington State Department of Ecology
by King County Resource Planning Section, Bellevue, Washington.
ICreiger, R., Department of Veterinary Science, University of Idaho, Moscow, Idaho, personal
communication.
59
-------
Kreiger, R., D. Kreiger, K. Tomson, and W. Warner. 1986. Lead Poisoning in Swans in the
Lower Coeur d’Alene River Valley. Pres. Annual Meeting of the Society of Toxicology,
New Orleans, Louisiana, March 1986, Paper No. 651.
Lee, G.F., E. Bentley, and R. Amundson. 1975. Effects of Marshes on Water Quality.
Springer-Verlag, New York, Inc..
Lindau, C.W., R.D. DeLaune, G.L. Jones. 1988. Fate of Added Nitrate and Ammonium-
Nitrogen Entering a Louisiana Gulf Coast Swamp Forest. Journal of the Water Pollution
Control Federation, Volume 60, Number 3.
Lloyd-Evans, T.L. 1989. Use of Wetlands for Stormwater Detention - Effects on Wildlife
HabitaL Draft #1 Prepared for the New England Institute for Environmental Studies
Workshop on Stormwater Detention in Wetlands, October 13, 1989, by Ma.nomet Bird
Observatory, Manomet, Maine.
Martin, EH. and J.L. Smoot. 1986. Constituent-Load Changes in Urban Stormwater Runoff
Routed Through a Detention Pond-Wetland System in Central Florida. U.S. Geological
Survey Water Resources Investigation Report 85-4310.
Meiorin, E.C. 1986. Urban Stormwater Treatment At Coyote Hills Marsh. Association of Bay
Area Governments.
vleyer, J.L. 1985. A Detention Basin/Artificial Wetland Treatment System to Renovate
Stormwater Runoff from Urban, Highway, and Industrial Areas. Wetlands, Volume 5,
pp. 135-146.
Mitsch, WJ., and J.G. Gosselink. 1986. Wetlands. Van Nostrand Reinhold Company, New
York.
Morris, F.A., M.K.. Morris, T.S. Michaud, and L.R. Williams. 1981. Meadowland Natural
Treatment Processes in the Lake Tahoe Basin: A Field Investigation. Final Report EPA-
600/54-81-026.
Mudroch, A., and J.A. Capobianco. 1979. The Effects of Treated Effluent on a Natural Marsh.
Journal of the Water Pollution Control Federation, Volume 51, Number 9.
Newton, R.B. 1989. The Effects of Stormwater Surface Runoff on Freshwater Wetlands: A
Review of the Literature and Annotated Bibliography. Publication 90-2. The
Environmental Institute, University of Massachusetts, Amherst, Massachusetts.
Nichols, D.S. 1983. Capacity of Natural Wetlands to Remove Nutrients frOm Wastewater.
Journal of the Water Pollution Control Federation, Volume 55, Number 5.
Niering, W.A. 1989. Effects of Stormwater Runoff on Wetland Vegetation. Preliminary Draft
Prepared for the New England Institute for Environmental Studies Workshop on
Stormwater Detention in Wetlands, October 13, 1989.
60
-------
,cr -S, G.L. 1982. Impact of Wetlands on Nonpoint Source Pollution. 1982 International
Symposium on Urban Hydrology, Hydraulics and Sediment Control. Lexington,
Kentucky.
ot,cris, G.L., and R.A. Osgood. 1988. The Water Quality Effectiveness of a Detention/Wetland
Treatment System and its Effect on an Urban Lake. 8th Annual International Symposium
on Lake and Watershed Management. Nov. 15-19, 1988. North American Lake
Management Society (NALMS), St. Louis, Missouri.
G.L, PJ. Wotzka, and LA. Hartsoe. 1989. The Water Quality Performance of Urban
Runoff Treatment Systems. Part One of a Report to the Legislative Commission on
Minnesota Resources. Metropolitan Council of the Twin Cities Area, St. Paul, Minnesota
Reddy, K.R., P.D. Sacco, D.A. Graetz, K.L. Cambell, and L.R. Sinclair. 1982. Water
Treatment by Aquatic Ecosystem: Nutrient Removal by Reservoirs and Flooded Fields.
Environmental Management 6(3): 261-27 1.
Reinelt, L.E., and R.R. Homer. 1990a. Characterization of the Hydrology and Water Quality of
Palustrine Wetlands Affected by Urban Stormwater. King County Resource Planning
Section, Seattle, Washington.
inelt, L.E., and R.R. Homer. 1990b. Urban Stormwater Impacts on the Hydrology and
Water Quality of Palustrine Wetlands in the Puget Sound Region. Puget Sound Research
‘91 Proceedings, Washington State Convention Center, Seattle, Washington, January 4-
5, 1991. Edited by T.W. Ransom. Puget Sound Water Quality Authority, Olympia
Washington.
Reinelt, L.E., R.R. Homer, and H.R. Wittgren. 1990. Removal of Sediment and Nutrients in
Palustrine Wetlands Receiving Runoff from Urban and Nonurban Watershed. In Lorin E.
Reinelt, Nonpoint Source Water Pollution Management: Monitoring Assessment and
Wetland Treatment, PhD Dissertation, Linkoping University, Linkoping, Sweden, 1990.
Renfro, W.C. 1972. Radioecology of Zinc-65 in Alder Slough, an Arm of the Columbia River
Estuary and Adjacent Ocean Waters. University of Washington Press, Seattle,
Washington.
Richardson, C.J., and J.A. Davis. 1987. Natural and Artificial Wetlands ecosystem: Ecological
Opportunities and Limitations. In Aquatic Plants for Water Treatment and Resource
Recovery. Edited by K.R. Reddy and W.H. Smith. Magnolia Publishing, Inc. pp. 819-
854.
Ru ht , B.T. In press. Hydrologic and Water Quality Characteristics of a Wet Detention Pond.
In Proceedings of a Symposium on Urban Hydrology. Edited by M.E. Jennings.
American Water Resources Association, Bethseda, Maryland.
61
-------
Rushton, B.T, C.W. Dye, and C. Hall. 1989. Discharge Water Quality of Permitted Wet
Detention Systems. In Water: Laws and Management. Edited by FE. Davis.
Proceedings of a Conference, September 17 - 22, 1989. American Water Resources
Association, Bethseda, Maryland.
Rushton, B.T, and C.W. Dye. 1990. Tampa Office Wet Detention Stormwater Treatment.
Southwest Florida Water Management District. Annual Report for Stormwater Research
Program - Fiscal Year 1989 - 1990. Resource Projects Department, Environmental
Section, Engineering Section, Southwest Florida Water Management District,
Brooksville, Florida. pp. 39 - 59.
Sartor, J.D., and G.B. Boyd. 1972. Water Pollution Aspects of Street Surface Contaminants.
EPA-R2-72-08 1.
Scherger, D.A. and J.A. Davis, Jr. 1982. Control of Storn water Runoff Pollutant Loads by a
Wetland and Retention Basin. 1982 International Symposium on Urban Hydrology,
Hydraulics and Sediment Control. Lexington, Kentucky. pp. 109-123.
Seidel, K. 1966. Reinigung von Gewassern durch Hohere Pflanzen. Naturwissenschaften. 53
Jg, Heft 12.
Simpson, R.L., R.E. Good, R. Walker, and B.R. Frasco. 1983. The Role of Delaware River
Freshwater Tidal Wetlands in the Retention of Nutrients and Heavy Metals. Journal
Environmental Quality 12(1): 4 1-548.
Small, E. 1972. Ecological Significance of Four Critical Elements in Plants of Raised Sphagnum
Peat Bogs. Ecology 53:498-503.
Southwest Florida Water Management District. 1990. Annual Report for Stormwater Research
Program - Fiscal Year 1989 - 1990. Resource Projects Department, Environmental
Section, Engineering Section, Southwest Florida Water Management District,
Brooksville, Florida.
Stockdale, E.C. 1991. Freshwater Wetlands, Urban Stormwater, and Nonpoint Pollution
Control: A Literature Review and Annotated Bibliography. Second Edition. Washington
Department of Ecology, Olympia, Washington.
Teal, J.M., A. Giblin, and I. Valiela. 1982. The Fate of Pollutants in American Salt Marshes.
In Wetlands Ecology and Management. Edited by B. Gopal, R.E. Turner, R.G. Turner,
R.G. Wetzel and D.F. Whigham, pp. 357-366. National Institute of Ecology and
International Scientific Publications, Jaipur, India.
Thompson, D. and J. Minor. 1985. Wetlands/Marsh Treatment System Improves Lagoon
Effluent Quality. 52nd Annual Pacific Northwest Pollution Control Association. October
28-30, 1985, Seattle, Washington.
62
-------
rich, K.E., and T.M. Burton. 1985. The Establishment and Management. of Emergent
Vegetation in Sewage-Fed Artificial Marshes and the Effects of these Marshes on Water
Quality. Wetlands, Volume 4, 1985.
U.S. Fish and Wildlife Service (USFWS), U.S. Environmental Protection Agency (EPA),
Department of Army, and U.S. Soil Conservation Service (SCS). 1989. Federal Manual
for Identifying and Delineating Jurisdictional Wetlands. U.S. Environmental Protection
Agency, Washington, DC.
Valiela, I., J.M. Teal, and W.J. Sass. 1975. Production and Dynamics of Salt Marsh Vegetation
and the Effects of Experimental Treatment with Sewage Sludge: Biomass, Production,
and Species Composition. Journal of Applied Ecology 12: 973-98 1.
Wengrzynek, R.J., Jr., and C.R. Terrell. 1990. Using Constructed Wetlands to Control
Agricultural Nonpoint Source Pollution. Presented at the International Conference on the
Use of Constructed Wetlands in Water Pollution Control at Churchill College, Cambridge,
United Kingdom, September 24 - 28, 1990.
Wetlands Research, Inc. 1991. The Des Plaines River Wetlands Demonstration Project:
Bibliography, July 1991. Wetlands Research, Inc., Chicago, Illinois.
higham, D.F., and R.L. Simpson. 1976. The Potential Use of Freshwater Tidal Marshes in
the Purification of Water Quality in the Delaware River. In Biological Control of Water
Pollution. Edited by J. Tourbier and R.W. Pierson. University of Pennsylvania Press,
Philadelphia, Pennsylvania. pp. 173-186.
Whitlow, T.H. and R.W. Harris. 1979. Flood Tolerance in Plants: A State of the Art Review.
Technical Report E-79-2. U.S. Army Corps of Engineers, Waterways Experiment
Station, Vicksburg, Mississippi.
Windom, H.L. 1977. Ability of Salt Marshes to Remove Nutrients and Heavy Metals from
Dredged Material Disposal Area Effluents. Technical Report D-77-37. U.S. Army Corps
of Engineers, Waterways Experiment Station, Vicksburg, Mississippi.
Wisseman, R.W., and S.F. Cook, Jr. 1977. Heavy Metal Accumulation in the Sediments of a
Washington Lake. Bulletin of Environmental Contamination & Toxicology, Volume 18,
Number 1.
Woodward-Clyde Consultants. 1986. Methodology for Analysis of Detention Basins for
Control of Runoff Quality. Prepared for Office of Water, Nonpoint Source Division,
U.S. Environmental Protection Agency.
Woodward-Clyde Consultants (Demgen, F., S. Vastano and J. Strandberg). 1991. Sediment
and Storm Runoff Concentrations of Copper, Zinc and Lead in the Crandall Creek -
DUST Marsh System. Prepared for County of Alameda, Public Works Agency,
Hayward, California.
63
-------
Wotzka, P., and G. Oberts. 1988. The Water Quality Performance of a Detention Basin -
Wetland Treatment System in an Urban Area. Nonpoint Pollution: 1988 - Policy,
Economy, Management, and Appropriate Technology. American Water Resources
Association. pp. 237-247.
Wulliman, J.T., M. Maxwell, W.E. Wenk, and B. Urbonas. 1988. Multiple Treatment System
for Phosphorus Removal. In Proceedings of Engineering Foundation Conference on
Current Practice and Design Criteria for Urban Runoff Water Quality Control, Potosi,
Missouri, July 1988.
64
-------
ABAG Association of Bay Area Governments
COD chemical oxygen demand
DUST Demonstration Urban Stormwater Treatment
EMC event mean concentration
EPA U.S. Environmental Protection Agency
ER efficiency ratio
MC mean concentration
NPDES National Pollution Discharge Elimination System
PCB polychiorinated biphenyls
OL regression of loads
SCS U.S. Soil Conservation Service
SOL summation of loads
TFe total iron
TKN total Kjeldahl nitrogen
TN total nitrogen
TP total phosphorus
TPb total lead
TSS total suspended solids
TZ 1 n total zinc
USFWS U.S. Fish and Wildlife Service
USGS U.S. Geological Survey
WCC Woodward-Clyde Consultants
65
8.0
ABBREVIATIONS
-------
9.0
MEASUREMENT UNITS-ABBREVIATIONS AND CONVERSION FACTORS
MULTIPLY TO OBTAIN
acre, ac 4.05 x iO square meter, m 2
acre, ac 0.405 hectare, ha (10 m 2 )
cubic feet, ft 3 7.48 gallons
cubic feet per second, cfs 448.831 gallons per minute, gpm
cubic meter, m 3 1.31 cubic yard
cubic meter per hour, m 3 lhr 4.4 gallons per minute, gpm
feet, ft 0.305 meter, in
inches, in 2.54 centimeters, cm
meter,m 3.28 feet,ft
micron or micrometer, IL 10-6 meter, m
square feet, ft 2 9.29 x 10-2 square meter, m 2
66
------- |