691
19
Ancillary Benefits
When wetlands are used to reduce pollutant concentrations
and peak stormwater ows, ancillary benets can be achieved
through thoughtful site selection and design. This chapter
summarizes the ancillary benets of wetlands used for pol-
lution control and recommends design features to optimize
these benets. The primary objective of most wetland pol-
lution control projects is water quality enhancement through
assimilation and transformation of sediments, nutrients,
and toxic chemicals. Secondary benets that can be incor-
porated in wetland treatment designs include (1) vegetative
biodiversity; (2) protection and production of fauna; and
(3) aesthetic, recreational, commercial, and educational
human uses. Realization of these benets usually requires
extra land and extra expenditures. Some of the features that
are added are illustrated in Figure 19.1. The potentials for
including each of these benets in treatment wetlands are
described in the following text. Research and manuals on
wetland evaluation techniques (Golet, 1978; Greeson et al.,
1978; Richardson, 1981; Kusler and Riexinger, 1986) may be
used to compare the additional benets of treatment wetlands
with natural wetland functions.
19.1 VEGETATIVE BIODIVERSITY
Because of the presence of ample water, the wetland envi-
ronment is generally characterized by a high diversity and
abundance of plants. In many cases, wetland plant commu-
nities include multiple vertical strata, ranging from ground-
cover species to shrubs, and subcanopy trees to canopy tree
species. Wetland plant diversity is important in determining

wildlife diversity because of the creation of niches associated
with differing vegetative structure, reproduction strategies,
owering and seeding phenologies, gross productivity, and
rates of decomposition (Mitsch and Gosselink, 2000b).
The wetland treatment system designer should not expect
to create or maintain a system with just a few known species.
Such attempts frequently fail owing to the natural diversity
of competitive species and the resulting high management
cost associated with eliminating competition, or because
of imprecise knowledge of all the physical and chemical
requirements of even a few species. Rather, the successful
wetland designer creates the gross environmental conditions
suitable for groups or guilds of species; seeds the wetland
with diversity by planting multiple species, using soil seed
banks, and inoculating from other similar wetlands; and then
uses a minimum of external control to guide wetland devel-
opment. This form of ecological engineering results in lower
initial cost, lower operation and maintenance costs, and a
most consistent system performance.
This section presents an overview of the oristic diversity
that naturally develops in treatment wetlands as well as some
details of the growth requirements of commonly occurring
plant species in wetland treatment systems. These microbial
and plant species are typically the dominant structural and
visual components in treatment wetlands. An understand-
ing of their basic ecology will provide the wetland designer
or operator with insight into the mechanics of their “green”
wastewater treatment unit.
WETLAND PLANTS
Algae

There are multiple designations of the algal component of
wetland vegetation. The term periphyton describes the com-
munity of organisms that grows attached to emergent and
submerged plants. Although periphytons usually begin colo-
nization of new plant surfaces by attaching algal growth of
lamentous and unicellular species, this functional compo-
nent also includes mat-forming assemblages of algae, fungi,
bacteria, and protozoans. Benthic algae are attached to the
sediment–water interface. Planktonic or free-oating algae
are generally not a large component of wetland ecosystems
unless open or deep water areas are present. Plankton spend
most of their life cycle suspended in the water column and
are of potential concern as a source of total suspended solids
(TSS).
Algae are often the rst colonizers in sparsely vegetated,
newly constructed wetland basins. They are capable of rapid
colonization, and for some months may outcompete the mac-
rophytes. Eventually, emergent plants will ll in, block the
light, and reduce the algal populations.
Algae are only rarely selected as the dominant plants
in treatment wetlands. Exceptions include the technology
known as algal turf scrubbing (Adey et al., 1993; Craggs
et al., 1996; Craggs, 2000) and the use of periphyton, known
in Florida as periphyton stormwater treatment (Vymazal,
1989; Kadlec and Walker, 2004). Algae are not the target of
attempts at planting or management and would not be consid-
ered a major factor in system diversity.
Macrophytes
The term macrophyte includes vascular plants that have
easily visible tissues. A wide variety of macrophytic plants

occur naturally in wetland environments. The U.S Fish and
Wildlife Service has more than 6,700 plant species on its
list of obligate and facultative wetland plant species in the
United States. Godfrey and Wooten (1979; 1981) list more
© 2009 by Taylor & Francis Group, LLC
692 Treatment Wetlands
than 1,900 species (739 monocots and 1,162 dicots) of wet-
land macrophytes in their taxonomy of the southeastern
United States. Obligate wetland plant species are dened as
those that are found exclusively in wetland habitats, whereas
facultative species are those that may be found in upland or
in wetland areas. There are many guidebooks that illustrate
wetland plants, for example, Hotchkiss (1972) and Niering
(1985). Lists of plant species that occur in wetlands are avail-
able, for example, RMG (1992).
Wetland macrophytes are the dominant structural
component of most wetland treatment systems. The terms
emergent, oating, and submerged refer to the predomi-
nant growth form of a plant species. In the emergent plant
species, most of the aboveground part of the plant emerges
above the waterline and into the air. These emergent struc-
tures are generally self-supporting. Emergent wetland plant
species are the primary concern in this section because they
provide the dominant visual impression in wetland treat-
ment systems. Floating and submerged vascular plant spe-
cies may also occur in wetland treatment systems. Floating
species have leaves and stems buoyant enough to oat on
the water surface. Submerged species have buoyant stems
and leaves that ll the niche between the sediment surface
and the top of the water column. Floating and submerged

species are more typical of deeper, aquatic habitats within
Spreader
swales
Buried forcemain
from urban area
Interpretive
center
Picnic
areas
Boardwalk
Islands
Berm
Collector
swale
Approximate
Scale in Feet
0
250 500
Cascade
aeration
Marsh
Open
water
Overlook
Rip-Rap
Discharge to
receiving stream
FIGURE 19.1 Conceptual plan for treatment wetlands with ancillary benets. (From Kadlec and Knight (1996) Treatment Wetlands. First
Edition, CRC Press, Boca Raton, Florida.)
the wetlands, and they may dominate in wetlands when

water depth exceeds the tolerance range for rooted, emer-
gent species.
WATER REGIME
The hydropattern, or water regime, of the water is the most
important contributor to the wetland type or class (Gosselink
and Turner, 1978; Gunderson, 1989). The concept of water
regime includes two interdependent components: (1) the
duration of ooded or saturated soil conditions (the hydrope-
riod as a percentage of time with ooding), and (2) the depth
of ooding (Gunderson, 1989). Whereas hydroperiod refers
to the duration of ooding, the term water regime refers to
hydroperiod as well as to the combination of water depth and
ooding duration (depth–duration curve). The importance of
this factor as a determinant of wetland ecology cannot be
overstated, because an incorrect understanding of the hydro-
period and water regime limitations of wetland plant species
is the most frequent cause of vegetation problems and shifts
in constructed wetlands.
Plant survival in ooded environments is a balance
between the severity of oxygen limitation and the adapta-
tions available to overcome this oxygen shortage. Thus,
hydrophytic plants may be adapted to survive and ourish
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 693
in specic ooded conditions, for example, for three months
each year, or in “clean” or owing water, which might have
higher in situ dissolved oxygen concentrations (Gosselink and
Turner, 1978). However, these same plants may not be able to
grow or survive during ve months of ooding or in stagnant
or “dirty” water conditions. Likewise, plants may have adap-

tations that allow prolonged survival in 0.3 m of water, but
not in 0.6 m. It may be hypothesized that this balance is tilted
unfavorably at higher water levels because of reduced aerial
plant stem surface area to provide oxygen to the roots through
the lenticels and aerenchymous tissues. This proposed expla-
nation is supported by the nding that hydrophytes generally
respond to ooding by growing taller—a growth response
that allows a more favorable balance between emergent and
submerged plant organs (Grace, 1989).
PROPAGATION
Wetland plants increase their numbers and density through
asexual and sexual reproduction. Asexual reproduction refers
to an increase in the number of individuals of a plant spe-
cies through vegetative growth and typically occurs through
the growth of roots or rhizomes, with subsequent emergence
of new aboveground stems and leaves. Technically, a cattail
bed that developed vegetatively from a single parent plant
is a single plant. However, when these rhizomes are cut or
decay, the individual daughter plants may remain viable and
continue to spread vegetatively. A number of woody wetland
plant species can spread vegetatively through coppice growth
from viable root systems.
Most wetland plant species also may increase their num-
bers through sexual reproduction. In sexual reproduction,
two individual plants, or the male and female owers from a
single plant, contribute gametes to form seeds with new com-
binations of genetic material. Sexual reproduction is impor-
tant in providing alternative strategies for plants to survive
from year to year through seasonal extremes, to propagate
the species over large distances, to rapidly colonize new hab-

itats, and to provide genetic variants that can adapt to chang-
ing environmental and competitive conditions.
Because of the potential for year-to-year hydrological
variations in natural herbaceous wetlands with large num-
bers of annual plant species, many species produce seeds that
remain viable for years. These seeds accumulate during pro-
ductive years and remain dormant until conditions are favor-
able for germination, frequently after a period of desiccation
and rewetting. This storage of viable seeds is known as a seed
bank and has been studied in a number of marsh ecosystems
(Pederson, 1981; Leck et al., 1989). In some cases, soil from
a natural wetland with a seed bank can be used to establish
a new constructed wetland. Most seeds in seed banks are
annuals, but in some marshes, up to 50% may be perenni-
als. Numbers of seeds range from <100/m
2
to >375,000/m
2
(Leck et al., 1989). A typical freshwater marsh in Manitoba
had 4,582 seeds/m
2
and 34 species in emergent areas and
only 93 seeds/m
2
in open water areas (Pederson, 1981).
BIODIVERSITY
Constructed treatment wetlands are typically dominated by
emergent marsh, oating aquatic plant, or submerged aquatic
plant communities. Emergent marsh species are frequently
intermingled and codominant with populations of small

oating aquatic plants such as duckweed (Lemna spp.). Most
constructed treatment marshes in the United States are domi-
nated by cattails (Typha spp.) or bulrush (Scirpus spp.) and,
in Europe, by common reed (Phragmites australis) or grasses
(Phalaris arundinacea or Glyceria maxima); however, many
treatment marshes are dominated by other species or by a
complex admixture of plant species, which include cattails,
bulrush, grasses and/or reed as the dominant species. These
plants are selected because they are robust and can survive in
a wide variety of environmental conditions—in other words,
they are survivors. They are inexpensive, plentiful in supply,
and easy to propagate. However, they may also be viewed as
a nuisance species owing to their ability to crowd out other
species and assume visual dominance.
In the United States, species such as Phalaris arundina-
cea, Lythrum salicaria, Typha spp., and Phragmites austra-
lis are considered by many to be undesirable and aggressive
invaders. These may take over natural wetlands and also be
a component of treatment wetlands. Pollutant removal pro-
cesses do not require these species. They are often selected
on the basis of cost and survivability. Unfortunately, many
people now erroneously equate free water surface (FWS)
treatment wetlands with the low biodiversity and nuisance
species.
Many constructed treatment wetlands undergo plant suc-
cession during their operational life. Constructed marshes
tend to remain as marshes as long as ooding is nearly con-
tinuous and water depths exceed about 5 cm. However, if the
water is too deep, then some emergent species will give way
to submerged and oating plants. In any case, new plants nd

their way into the system. For instance, Schwartz et al. (1994)
found numerous volunteer species in a Florida constructed
wetland. The system was established by planting a diverse set
of 21 species. After one year, 185 species were found in the
system. A side-by-side study of wetland vegetation changes
was conducted at the Olentangy River wetlands site in Ohio
(Mitsch et al.
, 2004). As seen in Table 19.1, the effect of the
i
nitial planting strategy (planted or unplanted) was gone in
three years. The volunteer cattails were then eaten out by
muskrats and replaced by volunteer bulrushes in both wet-
lands. The Tres Rios, Arizona, wetlands were planted in an
elegant mosaic with Scirpus validus (soft-stem bulrush) and
Scirpus olneyi (three-square bulrush), having varying num-
bers of deep zones and islands (CH2M Hill, 1995). During
the spring and summer of the third year after startup, almost
all the vegetation died in all four wetlands for reasons that
have not been fully resolved. The three possible causes were
drowning, starvation, and sulde toxicity (Kadlec, 2006b).
Thus, it is probable that the species patterns and types set
by ecological or botanical designers will not survive in many
circumstances. Such “designer” wetlands are likely to evolve
© 2009 by Taylor & Francis Group, LLC
694 Treatment Wetlands
that hundreds of plant species occur in a variety of treatment
wetlands. Even when treatment wetlands are dominated by
cattails or bulrush, dozens of other herbaceous and woody
plant species are typically present. Thus, the issue is not so
much of diversity as it is of dominance.

It is possible to start the treatment wetland with a variety
of regionally prevalent plants so that these have an opportu-
nity to compete during the successional development of the
system. This philosophy has been implemented at a number
of treatment wetlands, including 21 species planted at the
Orange County Eastern Service Area in Florida (Schwartz
et al., 1994), 20 species at Victoria, Texas (Reitberger et al.,
2000), and 14 species at Wakodahatchee, Florida (Bays et
al., 2000). The Oregon Gardens treatment wetland used doz-
ens of native wetland plants (Figure 19.2), and the Lapeer,
Michigan, wetland used four species that did not include treat-
ment wetland species, although bulrushes were used on the
fringes (Figure 19.3).
19.2 WILDLIFE
Numerous wildlife species of all taxonomic orders depend on
wetlands as habitat. Plant structures that are lost into the aquatic
portion of the food chain are typically degraded by a complex
assemblage of small aquatic organisms that include inverte-
brates (protozoans, worms, molluscs, arthropods, and others).
These organisms, in turn, serve as the basis of the food chain
for other invertebrates, and for diverse vertebrate groups such
as sh, amphibians, reptiles, birds, and mammals. Above-water
plant materials may also be processed as food by nonaquatic
fauna that have no direct contact with the water in the wetland.
We know from qualitative observation that a diver-
sity of microscopic invertebrates such as protozoans and
into other species mixtures and patterns. Wetlands that rely on
antecedent seed banks and natural importation will develop
more slowly, but are more likely to be regionally adapted.
“Self-designed” wetlands may not be avoidable. It seems rea-

sonable that new wetlands should be established with a spe-
cies mix that provides a headstart for a number of regional
species that can tolerate the water quality and hydropatterns
deemed proper for water quality improvement.
Table 3.1 summarizes the number of plant species
reported for the treatment wetlands in the North American
Treatment Wetland Database, v.2 (NADB database, 1998).
Almost 600 species of macrophytic plants have been reported
in constructed treatment wetlands. Of these observed plant
species, 501 are emergent herbaceous macrophytes, 31 are
oating aquatic species, 12 are submerged aquatics, 17 are
shrubs, 25 are trees, and 5 are vines. While the NADB v.2
data collection effort provides some insight into the plant
diversity at constructed treatment wetland sites in North
America, the level of investigation varied widely between
sites, and these numbers should be considered as the mini-
mum number of plant species that actually occur in treatment
wetlands.
Wetland plant diversity is a poorly understood subject
both in unaffected natural wetlands and in treatment wet-
lands. Wetlands are frequently dominated by a few plant spe-
cies (for example, cattail, sedge, or sawgrass marshes) that
are best adapted to stressful environmental conditions, such
as low nutrient or soil oxygen levels and uctuating water
levels. Other unaffected natural wetlands have higher plant
diversity and greater evenness between multiple dominant
plant species. Constructed treatment wetlands cover the same
range of plant dominance and diversity as unaffected natu-
ral wetlands. Information collected for NADB v.2 indicates
TABLE 19.1

Progression of Two Side-by-Side Wetlands at the Olentangy River Wetlands Park over a Start-Up Period of Nine Years
Wetland 1 (Planted) Wetland 2 (Unplanted)
Year 1 (1994) Algal pond with few macrophytes Algal pond with few macrophytes
Year 2 (1995) Plants developed, particularly around the
perimeter, to about 13% macrophyte cover
Essentially no macrophyte cover
Year 3 (1996) Continued to develop in vegetation cover, with
about 39% cover
Developed to about 35% macrophyte cover by August, essentially catching up with
the planted wetland within 3 growing seasons
Year 4 (1997) About 54% cover About 58% cover
Year 5 (1998) Wider diversity of cover; not dominated by Typha Began to be dominated by Typha
Year 6 (1999) Dominated by Typha Dominated by 3–4 of the planted species
Year 7 (2000) Similar to 1999, except muskrats began to have a
dramatic effect
Similar to 1999, except muskrats began to have a dramatic effect
Year 8 (2001) Maximum muskrat impact, vegetation cover was
lower than 1995
Maximum muskrat impact, vegetation cover was lower than 1995
Year 9 (2002) Drawdown from April through June; developed
cover of Schoenoplectus tabernaemontani
Drawdown from April through June; developed cover of Schoenoplectus
tabernaemontani
Source: Data from Mitsch et al. (2004) In Olentangy River Wetland Research Park at The Ohio State University, Annual Report 2004. Zhang and Tuttle
(Eds.), Columbus, Ohio, pp. 59–68.
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 695
microarthropods is present, but there are no comprehen-
sive lists of even the most common species in these groups.
The macroinvertebrate groups most commonly studied are

annelids (worms), arachnids (spiders and mites), crustaceans
(amphipods, isopods, copepods, water eas, and craysh),
insects, and molluscs. Fish are an obvious component of most
aquatic ecosystems, but knowledge of their role in wetlands is
limited. Amphibian and reptile (herptile) populations in wet-
lands tend to be variable and secretive and, therefore, difcult
to quantify. Birds are extremely mobile, and many species
include wetlands in some part of their life history. Because
birds are one of the fauna groups that people appreciate the
most, study of this wildlife group in treatment wetlands has
been more extensive than for any other. Many mammal spe-
cies use wetlands as an important component of their total
habitat. Some depend greatly on wetlands, whereas others use
wetlands for specic and narrow life history requirements.
FIGURE 19.3 The Durakon treatment wetland, near Lapeer, Michigan. Most of the plants selected for the wetland are not in common use in
treatment wetlands (Carex lacustris, Sagittaria latifolia, Alisma plantago-aquatica, Pontederia cordata, and Sparganium eurycarpum). Only
narrow fringe zones were planted with species with established treatment performance (Scirpus spp.) The wastewater inow is lagoon efuent.
FIGURE 19.2 The Oregon Gardens treatment wetland, near Silverton, Oregon. A tremendous diversity of vegetation was introduced. The
wastewater is better than secondary at the wetland inow.
MACROINVERTEBRATES
Information concerning wetland macroinvertebrate popu-
lations is infrequent in the published literature. A number
of specic ndings from wetland treatment systems are
described in the following text to provide an indication of
typical macroinvertebrate population densities and diversi-
ties in these systems.
A total of 518 species of aquatic invertebrates have been
recorded from treatment wetlands in NADB v.2 (Table 19.2).
These include 14 species of aschelminthes (cavity worms),
55 species of crustaceans, 10 species of arachnids, 23 spe-

cies of molluscs, and 343 species of insects. Thirty-two treat-
ment wetland systems listed in NADB v.2 have invertebrate
data, although, in most cases, only species lists are available.
The average Shannon-Weiner diversity (H’) of benthic mac-
roinvertebrate collections was 1.36 units (with a total of 342
© 2009 by Taylor & Francis Group, LLC
696 Treatment Wetlands
species reported) for constructed treatment wetlands. Benthic
macroinvertebrate evenness averaged 0.56 for constructed
wetlands. Average benthic populations in the NADB v.2 are
6,083 individuals per square meter for constructed treatment
wetlands. Total populations of mosquito larvae and pupae
in treatment wetlands from a few projects had an average of
1,144 individuals per cubic meter for constructed treatment
wetlands (Hemet and Sacramento, California).
FISH
The number of sh families that typically occur in wetland
environments in North America is small compared to pond,
TABLE 19.2
Number of Invertebrate Species by Group in the
NADB v.2 Database for Constructed Wetlands
Benthos Subgroup Number of Species Recorded
Annelida 47
Arachnida 10
Aschelminthes 14
Bryozoa 1
Coelenterata —
Crustaceans 55
Entognatha 2
Insects 343

Mollusk 23
Platyhelminthes —
Tardigrads 1
Unknown 22
Total 518
Source: Data from NADB database (1998) North American Treatment
Wetland Database (NADB), Version 2.0. Compiled by CH2M Hill. Gaines-
ville, Florida.
TABLE 19.3
Number of Fish Species by System in the NADB v.2 Database
for Constructed Wetlands
Site Name System Name Number of Species
West Jackson County, Mississippi WJC System 2
Brookhaven, New York Meadow Marsh Pond System 1 2
Brookhaven, New York Marsh Pond System 2 2
Hayward, California Hayward 16
Hillsboro, North Dakota American Crystal Sugar Co. 3
Everglades, Florida ENRP 18
Sacramento, California Sacramento Demo Wetlands 2
Champion, Florida Champion Pilot 3
Mt. View Sanitary District, California Mt. View Marsh 33
Source: Data from NADB database (1998) North American Treatment Wetland Database (NADB), Version
2.0. Compiled by CH2M Hill. Gainesville, Florida.
lake, and stream environments. The sh species typical of
wetlands are either adapted to obtaining oxygen in low-
oxygen waters or only visit wetlands on a seasonal or shorter
basis. The most widely distributed sh species in warmer
temperate, subtropical, and tropical wetland treatment sys-
tems is the mosquitosh (Gambusia afnis) in the topmin-
now family. This species has been widely disseminated in

North America and abroad in shallow wetlands and ponds
because of its noted ability to consume the larvae and pupae
of mosquitoes. In more northern climates, the black-striped
topminnow (Notrophus fundulus) is a mosquito predator, and
the central mudminnow (Umbra limi) is an important wet-
land sh species.
Seventy-seven sh species are reported from 13 treat-
ment wetland sites in NADB v.2 (70 species from constructed
treatment wetlands and 15 species from natural treatment
we
tlands). Table 19.3 summarizes sh data by site and system
in NADB v.2.
Fish enter treatment wetlands from inlet and outlet
streams as well as from other sources, such as raptor drops.
For example, smallmouth bass (Micropterus dolomieui) and
sunsh (Lepomis macrochirus) swam up the outlet structures
at the Hillsdale, Michigan, treatment wetland, and took up
residence in all deep zones. Drawdown of the wetlands, for
purposes of maintenance and retrotting, killed the sh.
They were effectively precluded after the deep zones were
lled for muskrat control. Bass and sunsh also are found in
the Incline Village, Nevada, constructed wetlands, which are
abutted by natural wetlands with deep water habitats. Tila-
pia (Oreochromis spp.) found their way into the Tres Rios,
Arizona, wetlands, although there appeared to be no route
open to them from either downstream with entering water
(WWTP) or upsteam from receiving water (drop boxes and
cascades). Drawdown of the wetlands, for research purposes,
killed the sh. Tilapia also are prolic in the Lakeland, Flor-
ida, constructed wetlands—again with no obvious route for

them to have emigrated.
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 697
Fish may be problematic in treatment wetlands, particu-
larly those that feed or nest on the bottom. For instance, at
the Des Plaines constructed wetlands in Illinois, carp (Cypri-
nus carpio) were unintentionally introduced into the wetland
from the pumped source water (Des Plaines River), presum-
ably as very small individuals. Over the course of two years,
the sh grew in size to about 15 cm. These larger sh were
observed to stir bottom sediments and impair suspended
solids removal. The carp were removed from the wetland
through a winter drawdown and freeze. A similar phenom-
enon occurred at the Tarrant County constructed wetlands in
Texas, where oodwaters from the source river brought carp
(Cyprinus carpio), shad (Dorosoma cepedianum), and bull-
heads (Ictalurus spp.) into the system. The activity of the sh
caused severe bioturbation of the sediments during feeding
and spawning activities (APAI, 1994). The cells were dewa-
tered, and egrets and herons effectively removed the stranded
sh.
A 21-ha constructed wetland has been used for nal pol-
ishing of an industrial efuent since 1999 in Victoria, Texas.
Fish populations in the eight wetland cells have been used
as an indicator of environmental condition at or near the top
of the wetland aquatic food chain (Reitberger et al., 2000).
This constructed treatment wetland includes three different
types of cells arranged in parallel and in series. Treated
efuent from a conventional activated sludge treatment
facility is distributed into the stage 1 wetland cells, which

are colonized primarily by cattails (Typha spp.) and contain
transverse deep water zones colonized by submerged and
oating aquatic plants. Water from these cells ows into the
stage 2 cells containing a diverse marsh plant community
and transverse deep water zones. Water ows nally into a
habitat cell with diverse water depths (up to 3 m). Six sh
species have been collected; mosquitosh (Gambusia af-
nis) have the highest density, and populations of other spe-
cies are variable from year to year, presumably in response
to climatic conditions and water quality conditions. Sun-
fish (Lepomis macrochirus and Lepomis gulosus), bluefin
killifish (Lucania goodie), silversides (Menidia beryllina),
and gizzard shad (Dorosoma cepedianum) populations are
sporadic in the habitat cell.
AMPHIBIANS AND REPTILES
Amphibians and reptiles (known jointly as herptiles) are part
of the top consumer structure of wetlands, including treat-
ment wetlands. The four most important groups are frogs
(anurans), turtles, snakes, and alligators. Twenty-one anuran
species are reported from six constructed and three natural
treatment wetlands in NADB v.2 (ten from the constructed
treatment wetlands). Turtles are found primarily in wetland
treatment systems that have areas of deeper, open water, and
especially in pretreatment lagoons. Species of turtles fre-
quently found in these habitats include snapping turtles (Che-
lydra serpentina), mud turtles (Kinosternon subrubrum), and
painted turtles (Chrysemys picta). A wide variety of snakes
are associated with wetlands and wetland treatment systems
and play an important ecological role by feeding on sh,
invertebrates, birds, and small mammals. In North American

treatment wetlands, the snakes that receive the most inter-
est are the venomous species, including the cottonmouth or
water moccasin (Agkistrodon piscivorus), the copperhead
(Agkistrodon contortrix), and timber rattlesnakes (Crotalus
horridus), which occasionally inhabit wetlands. In fact, most
snakes in wetlands are nonpoisonous species of water snakes
(Genus Natrix), swamp snakes (Liodytes and Seminatrix), rat
snakes (Elaphe), and king snakes (Lampropeltis). Alligators
(Alligator mississippiensis) have been observed at relatively
high densities in wetland treatment systems having open
water areas and high sh populations (Figure 19.4). Densi-
ties in the Florida stormwater treatment constructed wetlands
have reached proportions sufcient to allow hunting.
BIRDS
About 600 different bird species (one third of the total resi-
dent bird species) are either partially or wholly dependent on
wetlands for some part of their life history in North America
(Kroodsma, 1978). The diversity and abundance of birds in
and around wetlands attract many bird watchers, who repeat-
edly have observed that their species lists are longer and their
counts higher when they include wetlands in their counting
areas. Where these water bodies are enriched by nutrients and
FIGURE 19.4 Alligators are at the top of the food chain in wetland systems across the American south. (Photo courtesy J. Bays.)
© 2009 by Taylor & Francis Group, LLC
698 Treatment Wetlands
organic matter from wastewater and stormwater discharges,
bird watchers usually nd even better success in their sport.
A number of wetland bird species are game species and
are hunted. Treatment wetlands frequently provide good
waterfowl habitat. In a few cases, these treatment wetlands

have been used for duck hunting as a recognized secondary
benet. The stormwater treatment constructed wetlands in
Florida are open to duck hunting and are deemed the best
hunting spots in the entire state. In other cases, wastewater
has been found to improve the waterfowl breeding and feed-
ing habitat as a secondary or nonessential side effect (Wil-
helm et al., 1989).
A few quantitative studies of bird populations in wet-
lands treating wastewaters are available. If the wetland is
constructed on previous upland, increases in bird species
and abundance have been observed, compared to the previ-
ous upland habitat (Hickman, 1994). During the summer of
1991, the U.S. EPA conducted a study of the environmen-
tal condition of six constructed wetland treatment systems
in the United States (McAllister, 1992; 1993a; 1993b). This
inventory included two systems in the arid west (Incline Vil-
lage, Nevada, and Show Low, Arizona), two systems along
the coastal plain of Mississippi (Ocean Springs and Collins),
and two systems in peninsular Florida (Orlando Easterly and
Lakeland). Bird surveys were conducted by local ornitholo-
gists using slightly different methods. Table 19.4 summarizes
the major ndings of this research.
The EPA bird surveys at constructed wetland treatment
systems conrmed that these systems have high species rich-
ness and population densities compared to control wetlands.
The total number of bird species observed at each site dur-
ing the 1991 study ranged from 33 to 63, with average daily
population densities of all wetland-dependent species rang-
ing from about 7 to 19 birds/ha. Densities of wading birds at
the two central Florida constructed wetland treatment sys-

tems averaged 0.29 birds/ha at Orlando and 0.38 birds/ha at
Lakeland. These densities were as high as, or higher than,
comparison data from marshes along the St. Johns River, the
central Everglades, and Central America. Highest total bird
densities were noted at the arid region sites where there are
few natural wetlands available to compete as alternative hab-
itats. The EPA studies also concluded that the constructed
treatment wetlands are important habitats for a number of
endangered or threatened wetland-dependent bird species.
TABLE 19.4
Results from Constructed Wetland Treatment System Bird Surveys by U.S. EPA during 1991
Site
Constructed Wetland Area
(ha) Total Species
Density (#/ha)
[Average (Range)]
Incline Village, Nevada 198 47 19.1 (0.8–42.2)
Show Low, Arizona 284 42 13.8 (7.8–21.7)
Collins, Mississippi 4.5 35 7.2 (5.9–8.5)
Ocean Springs, Mississippi 22 35 10.4 (6.4–14.5)
Orlando Easterly, Florida 494 141 0.29 (0.11–0.54)
Lakeland, Florida 498 63 0.38 (7.7–13.5)
Bird usage of the 16-ha DUST (Demonstration Urban
Stormwater Treatment) marsh in Coyote Hills Regional Park
near Fremont, California, was studied by Dufeld (1986).
Weekly or biweekly bird censuses were conducted in the
DUST marsh and a nearby 18.3-ha control from mid-Janu-
ary 1984 through mid-June 1985. The mean abundance of
wetland birds in the marsh areas ranged from 100 to 300
(6.25 to 18.75 birds/ha) in the DUST marsh and 90 to 420

(4.91 to 23.0 birds/ha) in the control marsh. The mean num-
ber of species per census ranged from 14 to 23 in the DUST
marsh and 10 to 18 in the control area. Highest mean species
counts were observed in both marshes during the winter and
spring seasons. During all seasons, the dabbling and diving
ducks preferred the control marsh, and shorebirds, gulls,
and terns preferred the DUST marsh. Individual sh-eating
wading birds were generally present in greater numbers in
the DUST marsh than in the control area. These populations
were enhanced by the presence of roosting black-crowned
night herons (Nycticorax nycticorax). Shorebirds preferred
the mudat areas of the overland ow cell in the DUST
marsh. The most common shorebirds were American avocets
(Recurvirostra americana), black-necked stilts (Himantopus
mexicanus), and marbled godwits (Limosa fedoa).
Bird populations at the Victoria, Texas, constructed treat-
ment wetland have been quantied by designed studies and
by birdwatcher counts (Reitberger et al., 2000). A total of 188
bird species had been observed using the wetland as of June
2000. Annual species numbers ranged from 106 to 144, with
average numbers of species per count from 34 to 53; average
annual bird densities ranged from 19 to 26 birds per hectare.
Bird population numbers were dominated by grebes, cormo-
rants, ibis, waterfowl, vultures, coots, rails, stilts, sandpip-
ers, gulls, doves, swallows, wrens, warblers, and blackbirds.
This site is open to school groups and the public, and bird
watching continues to be one of the most important ancillary
benets of the wetland.
The Wakodahatchee, Florida, wetlands are attractive
to wildlife (Bays et al., 2000). The numbers of bird species

observed within the wetlands totaled 119 and 142 in 1997
and 1998, respectively. Thirteen of these species are con-
sidered commercially exploited, threatened, or endangered
by state and federal agencies. Bird densities estimated for
1997 averaged 53 per hectare. At least 13 species of birds
were observed to have nested on the site, indicating that the
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 699
system is providing suitable avian nesting habitat for certain
species.
Bird use and water quality treatment are not always totally
compatible uses of constructed wetlands. There may be con-
icts between maintenance activities and the lifestyles of
threatened or endangered species. For example, the nesting
habits of stilts and burrowing owls lead them to prefer the levee
areas of drained or dewatered wetland cells. This has interfered
with system restarts in the stormwater treatment constructed
wetlands in Florida. Huge ocks (thousands) of yellow-headed
blackbirds (Xanthocephalus xanthocephalus) have been impli-
cated in recontamination of water by pathogens in Arizona
(Orosz-Coghlan et al., 2006) (Figure 19.5). Waterfowl have
been implicated in turbidity increases at the Columbia, Mis-
souri, constructed wetlands (Knowlton et al., 2002). Waterfowl,
particularly geese, can also be devastating on new transplants
and can effectively wipe out an entire planting.
MAMMALS
A wide variety of mammals reside in or visit treatment wet-
lands. Some of the more common for North America are listed
in Table 19.5. Shrews (Family Soricidae) are found along the
edges of wetlands and moist elds and feed on insects. But by

far the largest and most important group of mammals associ-
ated with wetlands are rodents. Small rodents are also found
in FWS systems, such as mice and voles, most of which are
herbivorous species that graze on plants and seeds and are
prey to sh, wading birds, and raptors. However, it is the
larger rodents that have proven problematic in many treat-
ment wetlands. These rodents are briey discussed herein.
Muskrats
Muskrats (Ondatra zibethica) cut large numbers of emergent
herbaceous plants, primarily cattails (Latchum, 1996), and
build feeding platforms and nests (mounds). This grazing can
change treatment wetland areas from densely vegetated to a
FIGURE 19.5 Thousands of blackbirds move in and out of the Sweetwater constructed wetlands in Tucson, Arizona.
patchwork of open and emergent areas (Kadlec et al., 2007).
Muskrats consume a portion of the annual net primary pro-
ductivity, principally rhizomes, but their mounds represent
a greater share of this production. Densities of 20 or more
animals per hectare have been found, which can destroy the
majority of the macrophyte standing crop in a given year.
At such an exacerbated scale, muskrat herbivory may be
termed as an “eatout,” and it is evidenced by the removal of
essentially all emergent plant parts (Figure 19.6). Destruc-
tion of the wetland vegetative infrastructure may create an
attendant loss of some water quality functions, but may not
harm others. No designed research studies have been con-
ducted to quantify the effects of muskrat eatout on wet-
land water quality performance, but potential impacts may
TABLE 19.5
Mammals in Treatment Wetlands in North America
Rodents

Shrews Sorex spp.
Mice Peromyscus spp.
Voles Microtus spp.
Muskrat Ondatra zibethica
Beaver Castor canadensis
Nutria Myocastor coypus
Herbivores
Rabbits Sylvilagus spp.
Deer Odocoileus virginianus
Elk Cervus canadensis
Carnivores
Raccoon Procyon lotor
Opossum Didelphus marsupialis
Skunks Mephitis spp.
Mink Mustela vison
Otter Lutra canadensis
Bobcat Lynx rufus
Coyote Canis latrans
© 2009 by Taylor & Francis Group, LLC
700 Treatment Wetlands
be speculatively identied. Two water quality measures are
very liable to change because of denudation: dissolved oxygen
(DO) and total suspended solids (TSS). Increased potential for
atmospheric reaeration is present in the plant-free wetland,
and plankton can generate oxygen within the water column
because there is sufcient light. A 2-ha wetland in Commerce
Township, Michigan, provided data to support these intuitive
concepts. In an early, fully vegetated phase, this wetland had
an annual average outow of 6 mg/L TSS and 9.8 mg/L DO.
After a 100% muskrat eatout, TSS rose to 12 mg/L and DO

rose to 10.4 mg/L. There were also possible muskrat impacts at
the Olentangy, Ohio, wetlands. An eatout occurred in the year
2000, ve years after system start-up (Mitsch et al., 2004).
Thereafter, turbidity increased upon passage, in contrast to
decreases that occurred prior to the eatout. A similar response
for total phosphorus was observed. However, such results are
not conclusive because of other factors in system operation.
The integrity of berms may be threatened by burrowing.
Impacts on wetland hydraulics are also possible. In all cases,
loss of emergent vegetation has been viewed with dismay
by owners, regulators, and the general public. The damage
caused by muskrats, primarily by burrowing in containment
and separation berms, is not a matter of conjecture. There
are several examples of compromised parallel cells in which
divider berms have been breached by burrows: Estevan,
Saskatchewan (Duncan et al., 1999); Corcoran, California
(Gao et al., 2003); and Sacramento, California (Nolte and
Associates, 1998b). At the Manitoba Interlake site 1, it was
found that muskrat burrows were extensive and were threat-
ening to breach the dikes at several locations. Roads have
been damaged by burrow collapse at Saginaw, Michigan
(unpublished), and at Sacramento.
When muskrats eat all the emergent plant parts, there is
a large visual impact, regardless of physical or water quality
damage. A common perception is that wetlands should be
lush and green to be effective in treatment. This is an intui-
tive carryover from agriculture, where lush green elds are a
FIGURE 19.6 Hillsdale, Michigan, wetland before (July 18, 1999) and after (June 29, 2001) eat-out by muskrats.
(b)
strong indicator of a good crop. Consequently, owners, regu-

lators, and the general public equate good emergent stands
of plants with the highest treatment capability. This may or
may not be the case, but the urge to “x it” is irrepressible in
many cases. From the point of view of treatment efciency,
an eatout calls for careful scrutiny of performance to ascer-
tain if the damage has, in fact, impaired pollutant removal.
If not, then muskrat control becomes an expense associated
with the ancillary benet of a green emergent appearance.
At Brighton, Ontario, where water quality was relatively un-
affected by extensive muskrat herbivory, the aesthetic aspect
became a concern. Residents of the community viewed the
partially vegetated wetland as ineffective and expressed con-
cern that there might be environmental consequences.
Nutria
Nutria (Myocastor coypus) are an introduced species from
South America, but they now range across the southern
and northwestern United States. Nutria cause all the same
problems as do muskrats, except that the animals are larger,
and each individual is capable of more damage to vegeta-
tion. Mature nutria are large and have fewer natural preda-
tors. Nutria are strictly herbivorous and feed on a broad range
of plants in treatment wetlands, including cattails, grasses,
water hyacinth, duckweed, and young tree seedlings. Nutria
cut vegetation in a manner similar to muskrats and build feed-
ing platforms and nest mounds. They are prolic and com-
monly reach damaging population densities unless they are
controlled by trapping or shooting. Nutria have occurred in
problematic numbers in many treatment wetlands, including
Halsey, Oregon; Victoria, Texas; and West Jackson County,
Mississippi.

Beavers
Beavers (Castor canadensis) are found in wetland treatment
systems from Texas to the Canadian provinces. Beavers feed
© 2009 by Taylor & Francis Group, LLC
(a)
Ancillary Benefits 701
on bark, twigs, and leaves from a variety of trees and, con-
sequently, may cause undesirable herbivory damage in and
around wetland treatment systems. For instance, the Roblin,
Manitoba, constructed wetland was integrated into the
adjoining terrestrial landscape with poplar tree (Populus
spp.) plantings (PFRA, 2002). With an abundance of their
favored aquatic habitat and favorite food (poplar), the beavers
built lodges and decimated the tree plantings.
However, beaver damage to constructed wetlands does
not stop with herbivory. When beavers nd owing water,
such as in a treatment wetland structure, they construct a
dam to contain that water, and create a deep-water habitat
that will protect their nesting lodge from predators. The ani-
mals are large, up to 40 kg body weight, and can move stones
of several kilograms, small logs, sticks, and mud to create
st
oppages in ows (Figure 19.7). Plugging of structures is
clearly inimical to the proper performance of the wetland.
Problems with beavers plugging structures have been expe-
rienced at many locations, including Fort Worth, Texas; Tres
Rios, Arizona; and Des Plaines River, Illinois. Removal of
the obstructions is not effective, because the animals can
rebuild a dam literally overnight. Removal of the animals is
the only recourse.

Beavers construct travel trails to facilitate movement with
minimum risk from predators. These may be water-lled
ditches leading to and from deeper water areas in upland mar-
gins or deepened areas in marginal shallows. In treatment
wetlands, these travel trails may connect deep zones, thus pro-
viding increased opportunity for hydraulic short-circuiting.
Beavers burrow into banks as an alternative to lodges
made of branches and small logs, or as an adjunct to such
lodges. These burrows may create signicant leakage from
the wetland or compromise the integrity of containment
berms. For instance, beaver burrows caused failure of the
containment berm of a constructed inltration bed/wetland
at Houghton Lake, Michigan, causing over 1,000 m
3
of soil
to wash out in one event.
These several effects of the presence of beavers are of
considerable concern and often warrant removal of the
animals.
Predators
The richness of the invertebrate, avian, and small mammal
populations in treatment wetlands provides good hunting for
a number of predators. Mink (Mustela vison) prefer musk-
rats as food and will harvest a sizeable fraction of the resi-
dent population in the constructed wetland. Coyotes (Canis
latrans) will prey upon waterfowl nestlings, provided the
water is not too deep. As a consequence, wetlands that are
subject to seasonal drawdown and dryout during the nest-
ing season are vulnerable to coyote predation. This situation
exists for the Incline Village, Nevada, wetlands, because water

is diverted to crop irrigation starting in the nesting season.
FIGURE 19.7 Beaver activity caused complete closure of this outlet Parshall ume structure at Des Plaines River, Illinois, wetland EW3. A
dam of soil, rocks, and sticks was constructed. Note the staff gage in the left panel is incorporated into the dam in the right panel.
(a) (b)
© 2009 by Taylor & Francis Group, LLC
702 Treatment Wetlands
Raccoons (Procyon lotor) and bobcats (Lynx rufus) are fre-
quent visitors to wetland systems, from Arizona to Michigan
in the United States. Alligators (Figure 19.4) are extremely
numerous in the South Florida stormwater treatment areas.
Incidental Visitors
There are numerous animals that may, from time to time,
visit treatment wetlands and, in many instances, create prob-
lems. The Florida stormwater wetlands are very large and
have very large inlet structures through which an occasional
manatee (Trichechus manatus) may pass into the system.
These large, gentle aquatic mammals are not a threat to the
treatment system, but the habitat is wrong, and efforts are
made to move them back to their natural habitat. The Tarrant
County wetlands in central Texas were periodically visited by
wild pigs (Sus scrofa), an abundant introduced species (more
than one million animals estimated for Texas), with perni-
cious effects on the native ecosystems (Pimentel et al., 1999).
The tubers of wetland plants were rooted up and eaten in the
pilot wetland cells. SSF wetlands are also susceptible to these
grazing problems; damage to plantings due to whitetail deer
(Odocoileus virginianus) have occurred near Lutsen, Minne-
sota, and by wild pigs (Sus scrofa) near Nassogne, Belgium.
On a worldwide basis, the list of potential troublemakers
grows to include animals such as monkeys, which ruined a

pilot treatment wetland vegetated with Cyperus immensus in
Kenya (Abira et al., 2005).
19.3 DESIGN AND WILDLIFE USE
Each new treatment wetland situation brings a need to bal-
ance the wildlife value added and the potential impairment
of the structure and function of the system for water qual-
ity improvement. To some extent, wetland design can inu-
ence this balance, but management can still be an important
feature of wildlife control and enhancement. Management
strategies are discussed under operations and maintenance
in Chapter 22.
DESIGN TO ENCOURAGE WILDLIFE
The general concept of maximizing wildlife values in treat-
ment wetlands was elucidated by Sather (1989):
In as much as plant community diversity is the key to faunal
diversity, primary attention should be devoted to basic fac-
tors responsible for plant community diversity. The magni-
tude of ancillary benets realized from wetlands constructed
for wastewater treatment will be primarily dependent upon
the degree of complexity and the amount of interspersion
of the hydrophytic plant communities. Engineering design,
therefore, should be sensitive to the hydrologic regimes and
soil characteristics required to produce these types of plant
communities.
Worrall et al. (1997) concluded that if constructed wet-
lands were to support wildlife as an integral function, they
should be structurally diverse, with variations in plant types
and cover, water depth, and substrate. Further, the wetland
should be of sufcient scale and extent to support target spe-
cies and populations: bigger is better.

The U.S. Environmental Protection Agency (U.S. EPA,
2000c) recognized that multiple values could be obtained
from treatment wetlands. The degree of wildlife habitat
provided by constructed treatment wetlands varies broadly
across a spectrum. At one end of the spectrum are those
systems that are intended only to provide treatment for an
efuent or other water source and little to no wildlife habi-
tat. At the other end are those systems that are intended to
provide water reuse, wildlife habitat, and public use, besides
providing a nal polishing function for a pretreated efu-
ent or other water source. The guidance provided by the
U.S. EPA (2000c) primarily addresses the latter end of this
spectrum. To promote wildlife values, constructed treatment
wetland designs should avoid rectangular basins and straight
channels and rather use diverse and sinuous edges and native
vegetation whenever possible. It is desirable to incorporate
the constructed wetland into the existing natural landscape
with the natural transition zones as margins, perhaps includ-
ing woody vegetated buffer areas around the site. Where
appropriate, the constructed treatment wetland should pro-
vide habitat with a diversity of native species comparable
to similar wetlands in the region. To the extent practicable,
vegetative species diversity should be maximized without
increasing the proportion of weedy, nonindigenous, or inva-
sive species at the expense of native species. The biological
diversity of the project is likely linked to, or dependent on,
physical heterogeneity. This could include providing some
areas of open water, creating nesting islands for waterfowl,
and leaving some upland and buffer areas for other nesting
species.

Waterfowl populations are enhanced if open water areas
are interspersed with deep emergent marsh and upland
islands. An approximate 1:1 ratio of wetland area devoted
to marsh and to open water will provide maximum habitat
for a variety of waterfowl species (Weller, 1978). Generally,
wetlands adjacent to upland habitats will have greater oral
and faunal diversity in proportion to the amount of the edge
between adjacent ecosystems. Extending the perimeter does,
however, carry added cost. Wading birds prefer a different habi-
tat mix. These species require shallow, sparsely vegetated, litto-
ral areas or perching substrates adjacent to open water areas. To
benet wading bird populations, constructed wetland treatment
systems can be designed to provide a broad shelf of emergent
marsh with water depth <20–30 cm. Deep, open water areas
next to a shallow marsh provide additional foraging habitat.
Open water areas and the transitional ecotones among marsh,
open water, and adjacent uplands also help promote growth
of herptile and sh populations which, in turn, provide food
for wading and diving birds.
Depending on regional occurrence and available habi-
tats, other bird varieties will colonize constructed wetlands,
too. For example, if living or dead trees are included in the
wetland, they will serve as perching and possible nesting
sites (Hair et al., 1978). Nesting boxes might attract wood
ducks and owls. Upland islands surrounded by open water
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 703
provide protection for ground-nesting bird species such as
waterfowl.
DESIGN TO DISCOURAGE INCOMPATIBLE WILDLIFE

Project plans may include mechanisms to control or elimi-
nate undesirable species. Design techniques vary, depending
on the species of animal perceived as a possible threat. For
instance, beavers may be prevented from plugging structures
if the intakes and discharges are diffuse and noiseless and
under more than a meter of water. Inlets and outlets may be
screened to block entry by larger animals, such as turtles.
Chain-link or coarse mesh fencing can block some intruders,
such as domestic livestock.
A lack of predator habitat protects muskrats from preda-
tors such as hawks and owls. Erecting dead tree stubs or util-
ity poles in an upright position that are of sufcient height
to attract raptors can serve as a perch for these aerial preda-
tors. Such structures have been very effective in the Show
Low, Arizona, wetland environments. Mink, the primary
muskrat predator, is ubiquitous across the range of muskrats.
These small animals can and will enter treatment wetlands
in search of prey and are not deterred by chain-link fencing.
Attempts to allow access to larger predators, such as foxes
and coyotes, proved ineffective at a remediation wetland in
Hillsdale, Michigan.
Exclusion fencing is an effective, but often expensive,
option for controlling access by some animals. For example,
muskrats have been effectively excluded from the Ducks
Unlimited treatment wetlands at Oak Hammock, Mani-
toba. The exclusion fencing was erected prior to vegetating
the site. Ducks Unlimited, Canada, uses 5 cm r 5 cm gal-
vanized stucco wire mesh (hardware cloth), with about 0.30
m of the mesh vertically buried, about 0.5 m upright, and
another 0.3 m set to a 45° angle sloped away from the wet-

land. It is important to ensure that the fencing is placed such
that the muskrats cannot burrow underneath or between con-
nected sections. Based on reports from the Ducks Unlimited
domestic wastewater treatment wetland at Oak Hammock,
Manitoba, similar fencing was installed at the Hillsdale,
Michigan, treatment wetland, where it was also successful
(
F
igure 19.8) (Kadlec et al., 2007).
Riprap placement along berms and bank slopes can be
very effective at preventing burrowing. This involves the
placement of up to 50−150 mm riprap in a 200−300 cm layer
that extends from the wetland bottom to the top of the berm.
Note that although this approach reduces the potential for
burrowing into the berm directly at water level, experience at
Nanticoke, Ontario, showed that the muskrats burrowed down
into the roadway at the topmost edge of the riprap, which was
3 m from the water edge along the three to one side slope. An
alternative is chain-link fencing laid into the nal deep zone of
the wetland and extended to the top of the berm, a technique
that has been used successfully at the Lambton, Ontario, wet-
lands. There is probably a trade-off between mound build-
ing and bank burrowing, because both mounds and burrows
are acceptable refugia. Thus, by denying bank access, mound
building will be encouraged, and vice versa.
ECOLOGICAL RISK ANALYSIS
The expanding use of treatment wetlands for remediation of
contaminated groundwaters and for various industrial efu-
ents poses a potential threat to resident or migratory animals.
Bluntly put, the treatment wetland should not be a health

threat to wetland creatures. For many specic chemicals
there are water quality standards set to protect the aquatic
environment, and these may serve as guidelines for FWS
treatment wetlands. There are also standards for sediments
in freshwater systems in some cases. In the event that it is
determined that a threat may exist, other treatment alterna-
tives such as SSF wetlands should be considered.
FIGURE 19.8 Human and animal exclusion may necessitate fencing. At this Hillsdale, Michigan, remediation wetland site, there is a tall
fence to exclude humans and a short fence to exclude muskrats. Whitetail deer (Odocoileus virginianus) have been observed jumping over
the two-meter chain-link and barbed wire.
© 2009 by Taylor & Francis Group, LLC
704 Treatment Wetlands
The tools for evaluating potential hazards are collectively
known as ecological risk assessment. Ecological risk assess-
ment is a process that evaluates the likelihood that adverse
ecological effects may occur or are occurring as a result of
exposure to one or more stressors. The process is used to sys-
tematically evaluate and organize data, information, assump-
tions, and uncertainties to help understand and predict the
relationships between stressors and ecological effects in a
way that is useful for decision making. This process consists
of three elements, as indicated in U.S. EPA (1998):
1. Problem formulation, the rst phase, has the pur-
pose of articulating the assessment. The problem is
dened, and a plan for analyzing and characterizing
risk is determined. Initial work in problem formula-
tion includes the integration of available information
on sources, stressors, effects, and the ecosystem and
receptor characteristics. From this information, two
products are generated: assessment endpoints and

conceptual models.
2. Analysis, the second phase, is directed by the prod-
ucts of problem formulation. During the analysis
phase, data are evaluated to determine how expo-
sure to stressors is likely to occur (characterization
of exposure) and, given this exposure, the potential
and type of ecological effects that can be expected
(characterization of ecological effects). The rst step
in analysis is to determine the strengths and limita-
tions of data on exposure, effects, and the ecosystem
and receptor characteristics. Data are then analyzed
to characterize the nature of potential or actual
exposure and the ecological responses under the
circumstances dened in the conceptual model(s).
The products from these analyses are two proles,
one for exposure, and one for stressor response.
3. During risk characterization, the third phase, the
exposure and stressor-response proles are integrated
through a risk estimation process. Risk characteriza-
tion includes a summary of assumptions, scientic
uncertainties, and the strengths and limitations of
the analyses. The nal product is a risk description
in which the results of the integration are presented,
including an interpretation of potential ecological
adversity, descriptions of uncertainty, and lines of
evidence.
As an example, suppose that a treatment wetland is proposed
for removal of trichloroethylene (TCE) from the water. One
of the endpoints of ecological risk assessment would be the
effects on muskrats that inhabit the system. The model could

be the transfer of TCE from water to soils, from soils to cat-
tails, and from cattails to muskrats that eat them. Charac-
terization of the exposure would be the computation of how
much TCE might be ingested. Characterization of the effect
would be a determination of the acute and chronic doses
that the muskrat could stand. If the estimated ingestion was
well below the harmful dose levels, say, by several orders of
magnitude, then the risk would be deemed low and ranked
accordingly. This brief example was, in fact, part of the eco-
logical risk assessment for the Hillsdale, Michigan, project
(Ecology and Environment, 1996).
Obviously, the use of partially treated water engenders
some degree of risk for humans and wildlife within the wet-
land. The most conservative course of action might be to
monitor all potential indicators of contamination that might
contribute to a problem of the health of some sector of the
ecology—invertebrates up to the carnivores. Wren et al.
(1997) outline an example of a three-tier, ultraconservative
monitoring program. Unfortunately, the cost of such a pro-
gram would usually be the fatal aw for the entire project.
19.4 HUMAN USE
Humans appreciate wetlands for their commercial values
(plant harvesting, livestock grazing, hunting, and aquacul-
ture) and nonconsumptive values (aesthetics, recreation, and
research) (Nash, 1978; Reimold and Hardisky, 1978; Sather
and Smith, 1984; Smardon, 1988).
CONSUMPTIVE ACTIVITIES
Wetland plants have been used historically for a number of
purposes. Various sedges are used for fodder (marsh hay),
and cedar (Thuja spp.) timber is highly valued. Cranberries

(Vaccinium macrocarpon) are a high-prot cash crop grown
in bogs, and blueberries (Vaccinium spp.) are grown com-
mercially in peatland soils.
In North America, at least 46 native tribes were reported
to use Phragmites historically (Kiviat and Hamilton, 2001).
There were approximately 75 different uses of Phragmites,
of which the most frequent were arrowshaft (17 tribes), ciga-
rette (13), ute (12), whistle (7), pipestem (7), and matting (6).
The number of uses per tribe was highest in the southwest-
ern quadrant of North America. In modern times, in Europe,
Phragmites is used for roof thatching, and it provides a very
high-quality structure.
Plant harvest can possibly provide an economic prot from
constructed wetlands. Solano et al. (2004) assayed two macro-
phytes, cattail (Typha) and reed (Phragmites). High levels of
biochemical oxygen demand (BOD), chemical oxygen demand
(COD), and TSS removal for all treatments were obtained, but
cattails showed the greatest production of biomass (22 tons
of dry matter per hectare). Both cattails and reeds presented
high heating values (17−20 MJ/kg). Ciri et al. (2005) specu-
lated that Typha, harvested from treatment wetlands, could be
an economic fuel. A new rice variety (Oryza sativa “Kusa-
honami”) developed for livestock feed was utilized for treating
nutrient-polluted river water (Zhou and Hosomi, 2007).
Consumptive uses of wetland animals via hunting and trap-
ping are more easily quantied than nonconsumptive functions.
For example, over 10 million ducks and $35 million worth of furs
are harvested from wetlands annually (Chabreck, 1978). Con-
structed treatment wetlands at Orlando (Florida), the stormwater
treatment wetlands of Florida, and the Incline Village (Nevada)

© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 705
system are used for waterfowl hunting. An ancillary benet of
the stormwater treatment areas (STAs) is that vegetation and
shallow waters have attracted many ducks, such as teal (Anas
discors), pintails (Anas acuta), widgeon (Anas Americana), ring-
necks (Aythya collaris), and canvasbacks (Aythya valsineria).
The South Florida Water Management District (SFWMD) and
the Florida Fish and Wildlife Conservation Commission have
opened some of the STAs to limited duck hunting by permit,
starting in 2001. The hunting has generally been excellent,
with the STAs producing an average of four to ve ducks per
hunter, which is unheard of for a public waterfowl area. As a
result, hunters have come from all over Florida, as well as other
states, to enjoy the STAs. United Waterfowlers Florida President
Newton Cook stated that “STA 5 is the #1 public waterfowling
marsh in the United States and is nationally known.”
STA-5 was opened for the rst year for alligator hunting
in 2006 (SFWMD, 2006).
PASSIVE ACTIVITIES
Nonconsumptive uses of wetlands constructed primarily for
water quality treatment include recreation, nature study, aes-
thetics, and education. Increasingly, designs of treatment wet-
lands have incorporated attractive and informative parklike
areas. For instance, stormwater treatment wetlands in urban
settings such as Greenwood Park in Orlando (Florida), the
Tollgate Wetlands in Lansing (Michigan), and Coyote Hills
east of San Francisco Bay are used frequently for eld trips
and other educational purposes (Figure 19.9; Table 19.6).
Wetlands constructed for wastewater treatment in Arcata,

California; Hillsboro, Oregon; and Orlando and Palm Beach
counties in Florida, are vital recreational areas, which offer
jogging and bird watching. In South Florida, STA 5 and
STA1W are open for guided bird-watching tours. The visitor
will typically not be disappointed because these treatment
wetlands are frequented by very large numbers of subtropical
birds, including roseate spoonbills (Ajaia ajaja).
These human uses of wetlands, including the satisfaction
of having a wetland and wildlife at the edge of town, are
perhaps the most important factors behind public support of
protection and enhancement of existing wetlands.
CONFLICTS
Although most reasonable human use is acceptable in treat-
ment wetlands, there are some potential conicts. These are
generally in two categories: destructive human activities and
liabilities of the project owner—both real and perceived.
Humans can be destructive, either maliciously or unin-
tentionally. An illustration of the unintended damage issue
FIGURE 19.9 The Tollgate urban stormwater wetland in Lansing, Michigan.
TABLE 19.6
Use Estimates for the Arcata, California, Treatment Wetlands
Use
Density
(persons/day)
Duration
(hours) Management
Picnicking, relaxing 57–175 0.3–0.8 Trafc control, parking, garbage cans, and cleanup
Birdwatching, nature study 5–50 2–4 Trafc control, parking, viewing blinds, and docent program
Walking, jogging 20–80 2–4 Designated trails, maintenance, mosquito control program
Education 2–6 2–4 Designated trails, maintenance, signage

Photography 1–2 2–4 None
© 2009 by Taylor & Francis Group, LLC
706 Treatment Wetlands
is the case of a small treatment wetland in Dixboro, Michi-
gan, which treats runoff from a botanical garden. Elementary
school children visit the facility frequently, and one of their
activities was catching frogs and dragonies in the wetland.
But about 30 children running around in the 850-m
2
wetland
caused severe damage to sediments and vegetation, and the
activity had to be cancelled.
More problematic is malicious damage. Signage has been
defaced with spray paint (Show Low, Arizona), distribution
pipes cut with an axe (Houghton Lake, Michigan), and stolen
autos driven into the wetland (Howell, Michigan).
Issues of public safety combine with issues of facility
protection to lead to limitation of access to treatment wet-
lands. Insurance against public injury has led to exclusion
fencing at many treatment wetlands. This is more important
for wetlands treating waters that may be hazardous to human
health, which include some remediation sites as well as wet-
lands treating primary efuents. Nevertheless, there are no
known instances of human injury in a treatment wetland.
19.5 DESIGN FOR ANCILLARY BENEFITS
The previous sections described the secondary features of
wetland treatment systems, which are useful for encourag-
ing wildlife. This section focuses on design considerations to
enhance ancillary benets in general, while optimizing the
primary functions of pollutant removal and ood attenuation.

Only a scattering of papers on wetland design for ancillary
benets have appeared through the past 15 years—for instance,
Knight (1992; 1997); Worrall et al. (1997); and Conner and
Luczak, (2002). As noted in an ancillary benet design review
by Conner and Luczak (2002), the U.S. EPA Design Manual
(U.S. EPA, 2000a) “pays little attention to the conservation
aspects of these treatment systems.” The same may be said for
Crites and Tchobanoglous (1998), Water Environment Foun-
dation (2001), and Crites et al. (2006). The book by Campbell
and Ogden (1999) considers sustainable development and how
wildlife considerations may be woven into design; however,
Conner and Luczak (2002) note that “the lack of detail limits
its usefulness for design purposes.’’ Outside of North Amer-
ica, there is scant attention paid as well, with the exception of
the book by Merritt (Ed.) (1994), which provides useful infor-
mation on the conservation aspects of systems in the United
Kingdom. However, virtually all treatment wetlands in the
United Kingdom are subsurface ow with Phragmites, and
contain no concessions to wildlife or humans. In summary,
there is no comprehensive manual on how to blend treatment
and ancillary benet designs. Ony a few manuals attempt to
blend the concepts of mitigation wetlands and treatment wet-
lands, such as Sutula and Stein (2003).
Wetland design considerations include siting, cell size
and conguration, water ow and depth control, planting,
and species stocking. Because this book assumes that the
treatment wetland’s primary function is pollution control for
wastewater, the water source itself is not an ancillary design
decision. However, the level of pretreatment is important to
determine the potential for secondary benets. Developing a

wide variety of habitat type within a treatment wetland will
enhance higher wildlife diversity. Controls in the design of
the treatment wetland, which affect habitat and plant diver-
sity, include seasonal hydroperiods, depth and ow changes,
vegetative succession, and accumulation of sediments (U.S.
EPA, 2000c). Habitat quality can be enhanced by providing
for areas of open water, creating nesting islands for water-
fowl, and leaving some upland and buffer areas for other
nesting species. Because the selection of plant palette is one
of the major controls on habitat quality in a treatment wet-
land, it is important that the following recommendations be
considered in the design:
Provide habitat with native species mixtures com-
parable to similar wetlands in the region (U.S.
EPA, 2000c).
Create gentle slopes to allow for plant establish-
ment and diversity. Incorporate water level uc-
tuations in the operation plan.
Maximize vegetative species diversity without
increasing use of weedy, invasive species at the
expense of native species. Project plans may
include the means to control or eliminate undesir-
able species.
F
i
gure 19.1 shows some of the features that might be incor-
porated in a constructed wetland to control water pollution
and to provide secondary wildlife and recreation benets. In
this conceptual plan, the wetlands are sized and congured
to meet specic water quality treatment goals, and operation

is based on meeting efuent criteria. However, the wetland
also may be designed to take advantage of the potential for
wildlife habitat creation by incorporating deeper, open water
areas in the marsh and providing habitat islands as a refuge
from human access and predators. Constructed wetland treat-
ment systems providing a multitude of ancillary benets now
exist at a number of project sites.
SITING
Siting wetlands to treat municipal and industrial wastewater
is relatively exible because the efuents are often pumped to
the wetland. However, wetlands to control stormwater must
be sited close to stormwater sources or further downstream in
a watershed, intercepting a tributary. The location of a storm-
water treatment wetland will determine the quantity and tim-
ing of the inuent, which will affect the ancillary functions
of the wetland. Generally, wetlands located in headwater
areas will receive more irregular and less dependable inows,
resulting in prolonged dry conditions unless soils are imper-
meable or groundwater is normally high. This relative lack
of ooding will limit ancillary benets, such as primary and
secondary production. In these cases, maintaining a healthy
stand of wetland-dependent vegetation may be difcult, and
upland or transitional species may eventually predominate.
This type of system certainly will have some upland values
•
•
•
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 707
and may support faunal assemblages seasonally. However,

overall production of wetland-dependent species likely will
be lower than in a perennially ooded wetland.
Siting the constructed stormwater treatment wetland
further downstream in the watershed may yield a different
constraint: too much water during stormwater runoff periods.
To control high stormwater volume, the downstream wetland
can be designed to be off-line (out of the main ow path) so
that it captures only some ood ow, preventing the washout
of vegetation and berms. A series of off-line constructed wet-
lands, each capturing a portion of the storm ows, can be used
to deal with high volumes. Wetlands located downstream in
a watershed are more likely than wetlands at headwaters to
have perennial water because of more constant base ows
and higher groundwater levels. In turn, downstream storm-
water treatment wetlands generally will have more wildlife
and more food-chain support.
The siting of wetland treatment systems to optimize
ancillary benets may also consider the advantages of loca-
tions with adjacent donor wetlands (sources of plants and
wildlife to colonize the area), adjacent undeveloped uplands
(habitat diversity), or proximity to humans (aesthetics).
CELL SIZE AND CONFIGURATION
The wetland cell size depends primarily on water qual-
ity treatment needs and cost considerations. Because large
cells require less berm construction per unit area and fewer
inlet and outlet structures, per-area project costs are lower.
For example, a 100-ha constructed wetland may cost about
$10,000/ha to construct, whereas a smaller constructed wet-
land may cost about $50,000/ha. Although cell size may
inuence the use of wetlands by larger wildlife, it has mini-

mal affect on plant productivity or secondary production of
most wetland animals (Sather and Smith, 1984). A higher
berm-to-cell area ratio, typical of smaller wetland cells, may
result in increased benecial edge effects. As long as berms
are infrequently mowed or visited, the larger edge area pro-
vides nesting and feeding habitat for more mammal and bird
species.
Islands surrounded by marsh or open water provide excel-
lent habitat for nesting waterfowl. In many wetlands, islands
with trees are preferred nesting habitat for wading bird rook-
eries (see Figure 19.1). Nesting islands for waterfowl should
be about 0.6 m above the normal high water, whereas higher
and lower islands may also be valuable for other species for
feeding, resting, or nesting.
Open water areas may improve the water quality treat-
ment potential of constructed wetlands, but they denitely
enhance their ancillary benets for wildlife. Mallard duck
production is maximum in wetlands with almost even areas
of marsh and open water (Ball and Nudds, 1989). Open water
areas can be created by excavating at least 1.5 m below the
normal water level. Deeper excavations can also provide
sheries habitat. To prevent hydraulic short-circuiting, open
water areas should not be connected along the ow path but
rather interspersed with densely vegetated shallow marsh
habitat (average water depth of about 0.3 m or less). In gen-
eral terms, areas will be larger when ancillary benets are
desired. Ponds, islands, and edge amenities are not required
for treatment.
Cell number and conguration in series or in parallel are
major considerations when determining treatment capabil-

ity and operational exibility. These design considerations
primarily affect ancillary wetland benets because of their
importance in depth control.
Wa
ter Depth and Flow
Water depth and the ow rate are important factors affect-
ing dissolved oxygen in wetlands. Higher ow rates resulting
from shallow water tend to provide higher dissolved oxygen
concentrations in marshes because of atmospheric reaera-
tion. These higher dissolved oxygen levels generally result
in higher secondary production of aquatic invertebrates and
vertebrates, thereby increasing ancillary wetland benets.
Although deeper water in a marsh may increase hydraulic
residence time, this longer reaction time does not always
result in enhanced water quality treatment (oxidation of
organic matter and ammonia) because of the resulting reduc-
tion of dissolved oxygen concentrations.
Water depth is one of the main factors that affect wet-
land plant growth. High water levels will stress the growth
of emergent macrophytes and encourage the dominance by
oating or submerged plants or algae. The hydrological toler-
ance range and optimum hydroperiod of any desired vegeta-
tion type should be known and closely adhered to when water
level control structures are designed. Ideal design allows
water levels to be varied from zero (drained) to the maxi-
mum depth tolerance of desired wetland plant communities.
Stop logs or weir plates should seal the leaks to help maintain
water levels during periods of limited inows. Multiple inlet
and outlet weirs between adjacent cells allow greatest hydro-
period control exibility.

VEGETATION
The plant species selected for a constructed wetland will
greatly inuence ancillary benets such as primary and sec-
ondary productivity. The selection of improper plant species
will result in low productivity, and a lengthy adaptive period
may be necessary until available plant species, either planted
or natural volunteers, rearrange themselves according to
hydrologic and water quality factors. High plant diversity can
be initially achieved by assisting existing soil seed banks to
colonize areas that are graded and shallowly ooded or by
spreading muck and propagules from a donor wetland area
(Gilbert et al., 1981). Wetland vegetation establishment is
most rapid when plants are closely spaced, less than 1 m on
centers, and planted during the growing season (Lewis and
Bunce, 1980; Broome, 1990). Wildlife use all marsh plant
species directly for food or shelter or, indirectly, through the
detritus food chain, so expensive management to exclude
“noxious species” or to select favored species may lower
© 2009 by Taylor & Francis Group, LLC
708 Treatment Wetlands
overall wildlife use in favor of optimizing specic wildlife
species. Burning marshes may be good management for
waterfowl species but poor management for sh, small mam-
mals, or other bird species.
PRETREATMENT
In addition to wetland hydroperiod, water quality is a key
determinant of a wetland’s form and function. Primary water
quality characteristics, which affect wetland plant communi-
ties, include nutrients (especially nitrogen and phosphorus),
suspended solids, salts, pH, and temperature. Except for nutri-

ent concentrations, these same water quality characteristics
also inuence faunal populations. Inows of biodegradable
solids and the ammonia form of nitrogen can affect wetland
ora and fauna indirectly by their impact on dissolved oxy-
gen concentrations.
If high loads of suspended solids are released in the
constructed wetland, the solids can smother plant growth in
inow areas (Kuenzler, 1990). Generally, solid loads can be
controlled by conventional pretreatment. In stormwater sys-
tems, the problem can be minimized with a forebay or pond
prior to the wetland (Livingston, 1989). If the pretreatment
area traps the mineral-suspended solids (clays, silt, and sand),
the wetland treatment system will need to be maintained less
frequently.
Inlet ponds also can reduce nutrient levels, although nutri-
ent reduction may not be the primary goal in constructed wet-
lands designed to enhance wildlife. As noted earlier, higher
nutrient levels generally result in higher primary production
of wetland plants which, then, support increased wildlife
populations.
As a consequence, the wetland should not be used as the
rst interceptor of urban runoff, particularly in industrial or
highly urbanized sites (Schueler, 2000). Specic recommen-
dations include the following (Sutula and Stein, 2003):
Control source of pollutants where possible (e.g.,
street sweeping, removal of sediment and debris
from storm drain inlets, improving inltration,
and other land-based best management practices
(BMPs).
Use oil and grit interceptors in industrial sites or

highways.
Install oating berms to trap trash and large
debris.
Utilize sediment collection/settling forebays.
Design and locate the forebays for ease of mainte-
nance and removal of sediment.
Segment the wetland such that the primary treat-
ment is provided in the forebay or initial pond and
the remainder of the wetland is used for polishing
and/or wildlife enhancement.
HUMAN ACCESS
Recreational facilities that provide public access to the STAs
of South Florida include asphalt parking areas, compost-
•
•
•
•
•
ing toilets, landscaping, and information kiosks. Pedestrian
gates, signage, and fencing are included as needed to dene
public access areas. Associated road improvements are
sometimes needed, such as an acceleration and deceleration
lane in the vicinity of the entrance to the parking lot. Some
facilities include a canoe launching site, access footbridges,
and boardwalks.
Although public access to a created wetland might dis-
turb wildlife populations, disturbances can be minimized
with controlled access to certain areas and design features
such as islands for roosting and nesting. Visitors typically
appreciate guidance, and signage is therefore an important

aspect of design for human use (Figures 19.10 and 19.11).
Although clearly marked walking trails sufce for wetlands
of limited size (Figure 19.10), boardwalks are the preferred
means of protecting both the visitors and the wetlands in
large systems (Figure 19.11). These are not restricted, how-
ever, to large systems (Figure 19.12). Boardwalks facilitate
nature study and outdoor recreation, and enhance the scien-
tic research benets of a constructed wetland. Observation
towers are valuable because they provide an opportunity for
visitors to get an overview of what is generally a at landscape
(Figure 19.13). Wildlife viewing blinds can greatly enhance
the ability to observe wildlife with minimal disturbance to
the animals. An interpretive center allows small groups to
assemble with shelter from sun and rain (Figure 19.14). In
some very large treatment wetlands, boat launching facilities
are provided (Figure 19.15).
It is not uncommon for scientists to visit treatment wet-
lands. This human use is typically not accompanied by any
special facilities; such visitors are deemed knowledgeable
enough to fend for themselves (Figure 19.16).
It is noteworthy that there are no reports of accidents
involving humans and treatment wetlands.
EXAMPLES
The popularity of treatment wetlands for public use is very
high. Perhaps the best way to illustrate the design features
that contribute to public use is to describe examples.
Palm Beach County, Florida
Originally conceived in 1992 as part of a cooperative efuent
reuse study between the Palm Beach County Water Utilities
Department, the City of West Palm Beach, and other munici-

palities, the Wakodahatchee Wetlands have been designed
to demonstrate the following ancillary benets of reclaimed
wat
er reuse (see Figure 19.11):
Creation of signicant wildlife habitat accessible
to the public in a parklike setting
Maintenance of biological diversity and open
space in a highly developed landscape
Recharge of local water supplies through inltra-
tion to the surcial aquifer
Provide limited public use for passive recreational
activities
•
•
•
•
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 709
FIGURE 19.10 The Whangarei, New Zealand, treatment wetlands are very visitor friendly.
(b)
(a)
FIGURE 19.11 The Wakodahatchee Wetlands in Palm Beach County, Florida, has a very user-friendly boardwalk.
(b )(a)
FIGURE 19.12 Even this short boardwalk at the University of Michigan Matthaei Botanical Gardens treatment wetland is a popular place
to stop and look.
© 2009 by Taylor & Francis Group, LLC
710 Treatment Wetlands
FIGURE 19.13 Wetland observation tower at the Hennepin–Hopper (Illinois) site.
FIGURE 19.14 Public viewing shelter at the Tucson, Arizona, Sweetwater wetlands.
FIGURE 19.15 Boardwalks and boat launching sites are features of the South Florida Stormwater Treatment Areas.

© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 711
The Wakodahatchee Wetlands are part of the Southern
Region Water Reclamation Facility, completed in 1997. The
wetland area totals 16 ha on the 23-ha site, with individual
wetland cells ranging from 0.9 ha to 4.4 ha. The marsh area is
designed to operate at an average depth of 15 cm but may be
operated normally at depths to 45 cm. High rainfall can ele-
vate wetland water levels to even greater depths of 60 cm or
more for short periods. Deep zones, consisting of ponds 1.5
m or more in depth, variable in width, and oriented transverse
to the direction of ow, are interspersed throughout each
wetland. A total of 28 deep zones and 8 habitat islands are
included. Nominal detention time varies from a maximum of
40 days at 2,000 m
3
/d of ow to ten days at 8,000 m
3
/d.
Marshes comprise about 70% of the wetland area, veg-
etated by native emergent, forested, and transitional wetland
species and designed to emulate native South Florida wetland
plant communities. Emergent marsh zones are composed of
bulrush (Scirpus validus, S. californicus), duck-potato (Sag-
ittaria lancifolia), arrowhead (Sagittaria latifolia), spikerush
(Eleocharis cellulosa), reag (Thalia geniculata), and
pickerelweed (Pontederia cordata). The wetlands are main-
tained to be free of nonnative and nuisance plant species,
and the surrounding berms and entry areas are landscaped
to maintain an aesthetic and comfortable experience for wet-

land visitors. The combined efforts to manage the vegetation,
maintain the boardwalk facility, monitor water quality, and
maintain the grounds cost on the order of $150,000/yr.
Snags and perching posts were added for wildlife use. The
numbers of bird species observed within the wetlands totaled
119 and 142 in 1997 and 1998, respectively. Thirteen of these
species were considered by state and federal agencies to be
commercially exploited, threatened, or endangered. Reptile
diversities were high and increased over the rst two years.
Mosquitosh densities were high and consistent between the
two years. Mammal and amphibian diversities were low.
The public makes extensive use of the Wakodahatchee
Wetlands. Parking facilities are available, and there is a
boardwalk through the wetlands, with associated gazebos,
which provide an attractive and comfortable setting for
walking, bird-watching, nature photography, environmental
e
d
ucation, and related passive activities (Figure 19.11). The
Audubon Society of the Everglades has identied the Wako-
dahatchee as one of the birding hot spots within Palm Beach
County. Available data on site use indicate that the numbers
of daily visitors to the wetlands have increased from 165 peo-
ple in 1997 to over 1,500 through the rst half of 2000. This
high level of interest has led the Water Utilities Department
to construct another wetland facility with an interpretive
center on 70 ha of farmland adjacent to the Wakodahatchee
Wetlands for treatment and disposal of up to 38,000 m
3
/d of

treated efuent.
Victoria, Texas
The DuPont Victoria constructed wetland was built to treat
wastewaters from a petrochemical plant that manufactures
precursors for nylon and polyethylene. The 21.5-ha wetland
receives about 12,000 m
3
/d of water that has been pretreated
in a biological plant. The wetland is congured in three
stages (Figure 19.17). Stage 1 consists of ve cells in par-
allel, which discharge into two parallel Stage 2 cells, both
stages containing deep zones. Water then passes to Stage 3
wetlands, which are primarily intended for wildlife habitat.
There are deep open-water zones for waterfowl, and islands
add protected upland features to the habitat. Twenty species
of native plants were used. Public access features include a
parking area, an outdoor education center suitable for class-
room groups, an observation platform, information kiosk,
bird watching blind, walking trails, a boardwalk across the
wetland, and a water sampling pier.
FIGURE 19.16 Scientic visits are an important human use of treatment wetlands, such as this at the leachate treatment wetlands at
Reguengo, Portugal.
© 2009 by Taylor & Francis Group, LLC
712 Treatment Wetlands
The wetland provides valuable habitat for migratory
and resident birds. In excess of 180 species were identied
in a 1999 survey, which increased to over 195 by 2002. An
additional 28 species of plants had volunteered by 1998,
and by 2002 there were 85 species of plants. Five species
of sh live in the wetland and are just as healthy as their

counterparts in the reference pond. Many larger animals
have been attracted by the abundance of the wetland, but
a problem was the invasion by nutria (Myocastor coypus),
a nonnative aquatic rodent. The wetlands suffered signi-
cant loss of vegetation because of herbivory by the nutria.
Population control methods are being employed, including
attracting alligators to the site. Hunting and trapping meth-
ods were developed that provided effective population con-
trol without using rearms. Revegetation efforts have taken
place despite continued nutria activity. Measures include
replanting, nutria exclosures, and water level adjustment,
which have successfully reestablished vegetation in most
impacted areas.
Education programs are provided to students from Grade
4 through college via the Wetland Environmental Science
Education Encounter (WE SEE). The program encompasses
treatment benets along with the inherent natural benets of
a wetland. A full-time on-site educator is the rst such person
located at an industrial wetland. The program is housed at
a state-of-the-art education center and offers a wide variety
of science education learning opportunities to area schools.
Given that many young persons will be in a natural environ-
ment, there have to be rules. There is an inherent danger in
any “natural” setting. To minimize the chance of injury, the
students are cautioned to do the following:
Deep water zone
Deep water
zone
Deep water zone
Island

Island
Island
Island
Observation
station
Parking area
Stage 2
polishing
cell
Stage 1
polishing
cells
Habitat
cell
Habitat
cell
Education
center
Stage 2
polishing
cell
200 100 0 200
6' Surface
paved
Habitat
cell
Grassed
walkway
Observation
station

Bird
Blind
N
FIGURE 19.17 Layout of the DuPont Victoria, Texas, treatment wetland. (From Reitberger et al. (2000) Achieving Multiple Benets from
a Constructed Wetland. Reddy and Kadlec (Eds.), Proceedings of the 7th International Conference on Wetland Systems for Water Pollution
Control, 11–16 November 2000; University of Florida and IWA: Lake Buena Vista, Florida, pp. 749–758. Reprinted with permission.)
Do not leave trails or enter any area of high grass
or brush.
Do not attempt to capture any animal known to
sting or bite. This includes such insects and arach-
nids as spiders, scorpions, bees, wasps, hornets,
ants, and caterpillars.
Do not attempt to capture or harass any mammal,
bird, amphibian, or reptile in the wetland. Be on
the lookout and avoid snakes.
Watch for and do not attempt to touch plants with
stems bearing leaves clustered in threes (poison
ivy or poison oak). This plant has not been found
in the wetland yet. If not sure, ask. Some plants
in the wetland have thorns, so be careful of the
brush.
Walk, and do not run anywhere in the wetland,
including on the trails, pier, boardwalk, and inside
the education center.
Follow proper eld and laboratory safety
procedures.
As of August 2006, over 27,000 students had participated.
Renton, Washington
The Waterworks Garden is a stormwater treatment system
with emphasis on not only the usual natural features of wet-

lands but also on elements conceived for their artistic value.
An environmental artist (Lorna Jordan) worked with land-
scape architects and engineers to develop the combination of
wetland treatment and art. Visitors pass through “rooms” that
•
•
•
•
•
•
© 2009 by Taylor & Francis Group, LLC
Ancillary Benefits 713
symbolize a journey from the civilized to the wild (Campbell
and Ogden, 1999). The walking path parallels the course of
the water, moving from the “Knoll” inlet structure through
more natural water forms (“Funnel,” “Grotto,” “Passage,”
and “Release”) before discharge to the river (Figure 19.18).
The planting scheme is symbolic; plants at the top of the gar-
den where the runoff is still contaminated are species Native
Americans consider bad omens (France, 2003). The entire
system is just a few hectares, but the cost of the art was $1.3
million in 1995, the year of completion.
SUMMARY
Wetlands for water quality improvement are being con-
structed throughout North America at an accelerating rate.
However, many wetland designs do not incorporate ancil-
lary benets to the extent possible. With thoughtful design,
constructed wetlands can provide benets beyond effective
water treatment, such as wildlife enhancement and recre-
ational opportunities. In fact, with the decline of the total

area of natural wetlands, constructed treatment wetlands are
a viable and cost-efcient way to compensate for the loss of
productive wetland habitat.
Pretreated wastewater and stormwater can nourish
hydrologically altered natural wetlands or create natural
wetland functions in constructed marshes. Wastewater con-
stituents can be controlled to enhance a wetland’s ancillary
benets. For example, nutrients promote plant growth at
the bottom of the food chain, which then supports other
wildlife species. By considering certain design features
such as cell conguration, hydroperiod control, vegetation
and wildlife stocking, and human access, wetlands can be
designed to meet water treatment goals as well as provide
additional benets.
FIGURE 19.18 The water enters the Waterworks Gardens urban stormwater wetlands in Renton, Washington, under a paved colonnade.
© 2009 by Taylor & Francis Group, LLC