EPA/600/3-91/058
November, 1991
EVALUATING CREATED WETLANDS THROUGH
COMPARISONS WITH NATURAL WETLANDS
by:
Mark T. Brown
Center for Wetlands
Phelps Laboratory
University of Florida
Gainesville, Florida 32611
Assistance Agreement CR-814643-01-3
Project Officer:
Eric M. Preston
Wetlands Research Program
USEPA Environmental Research Laboratory
200 SW 35th Street
Corvallis, Or 97333
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97333
-------
Regional Center for Environmental Information
US EPA Region III
1650 Arch St.
Philadelphia, PA 19103
-------
ERL-COR-
TECHNICAL REPORT DATA _,
(Pteatt ntd /mstrucriom on tht nrene before campfe'
'.REPORT NO.
EPA/600/3-91/058"
2.
PB92-111566
4. TITLE AND SUBTITLE
Evaluating Created Wetlands Through
Comparisons with Natural Wetlands
ft. REPORT DATE
10/29/91
ft. PERFORMING ORGANIZATION CODE
EPA/ORD
1. AUTMOR(S)
Mark T. Brown
ft. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM CLEMEN* NO.
». PERFORMING ORGANIZATION NAME AND ADDRESS
U of PL, Gainesville, PL
TfTCONTRACT/CRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
US Environmental Protection Agency
Environmental Research Laboratory
200 SW 35th Street
Corvallis, OR 97333
1 a. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGE NO
EPA/600/02
IB. SUPPLEMENTARY NOTES
1991. U.S. Environmental Protection Agency, Environmental Research
Laboratory, Corvallis, OR
16. ABSTRACT
, This report summarizes the evaluation and recommendations regarding
an approach to wetland sampling and characterization developed by the
Wetlands Research Program (WRP) at U.S. Environmental Protection
Agency, Environmental Research Laboratory, Corvallis, OR. Between
trials, team members discussed at length the methodology and field
protocols and modified them to reflect the conditions and difficulties
encountered in sampling herbaceous wetlands in urban areas of Florida.
Major emphasis is placed on the appropriateness of measured variables
for determining successful wetland re-creation.
Numerous physical and biological parameters were measured and
compared in the nine created and nine natural wetlands. Analysis of
these data has shown some important similarities and differences
between created and natural wetlands and lends insight into the complex
questions surrounding wetland creation and the equivalency of created
wetlands to naturally occurring wetlands. Evaluations of temporal
changes in hydrology and plant successional trends within created
wetlands seem most important in determining ultimate success.
IT.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERM!
U.S. Environmental V
Region li I Informal
Center (3PM52)
841 Chestnut Street
Philadelphia, PA 19
COSATi Field/Croup
Resource
wetlands mitigation, Section 404,
Florida
IB. DISTRIBUTION STATEMENT
Release to Public
21. NO. Of 'AGES
46 Pgs
20. SECURITY CLASS
Unclassified
22. PRICE
CPA F»rm 1220-1 (B-73)
-------
DISCLAIMER
The research described in this report has been funded wholly or in part by the
United States Environmental Protection Agency (EPA) through Assistance Agreement
CR-814643-01-3 to the Center for Wetlands of the University of Florida, Gainesville.
It has been subjected to the Agency's peer and administrative review, and it has been
approved for publication as an EPA document. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
Citation: Brown, M.T. 1991. Evaluating Created Wetlands Through Comparisons
With Natural Wetlands. EPA/600/3-91/058. U.S. Environmental Pro-
tection Agency, Environmental Research Laboratory, Corvallis, Oregon.
-------
__ TABLE OF CONTENTS _ _
DISCLAIMER ii
LIST OF TABLES v
LIST OF FIGURES vi
ACKNOWLEDGEMENTS vii
ABSTRACT viii
INTRODUCTION 1
PLAN OF STUDY 2
METHODS 6
SITE SELECTION 6
FIELD PROTOCOL 7
FIELD CREW COMPOSITION 7
FIELD MEASUREMENTS 8
TRANSECT ESTABLISHMENT . . / 8
VEGETATION MEASUREMENTS 10
Estimating Commonness 11
Estimating Percent Cover 11
VEGETATION CHARACTERISTICS 11
SOIL SAMPLES 12
WATER QUALITY SAMPLES , 12
SITE MORPHOLOGY AND WATER DEPTHS 13
DATA ANALYSIS 14
PHYSICAL CHARACTERISTICS 14
iii
-------
Surrounding Landscape Condition 14
^~Basin Topography 7 . .~7 15
Water Depth 15
Percent Open Water 15
Soil G-ganic Matter 17
FIELD DATA COLLECTION 17
RESULTS 19
PHYSICAL CHARACTERISTICS 19
WATER QUALITY 21
VEGETATION CHARACTERISTICS 21
DISCUSSION 31
COMPARISON OF NATURAL AND CREATED WETLANDS 31
SIGNIFICANCE OF MULCHING ON COMMUNITY STRUCTURE 32
UTILITY OF THE EVALUATION APPROACH 35
LITERATURE CITED 37
iv
-------
LIST OF TABLES
TABLE 1.
TABLE 2.
TABLE 3.
TABLE 4.
TABLE 5.
Physical Characteristics of Natural and Created Wetlands. .
Water Quality in Natural and Created Wetlands
Species Planted and Use of Mulching at Created Wetlands. .
Vegetation Characteristics of Natural and Created Wetlands.
Summary of the Means of the Variables Measured for Natural and Created
Wetlands
TABLE 6. Species Characteristics of Mulched and Unmulched Created Wetlands. .
20
23
25
28
29
33
-------
LIST OF FIGURES
FIGURE 1. Map of Tampa and the surrounding area in Hillsborough County showing
locations of created wetlands considered as study sites 3
FIGURE 2. General layout of 'ransects in natural and created wetlands. Both
morphology and vegetation data were collected from a total of 40 quadrats
spaced equally along the transects 9
FIGURE 3. Typical cross section of a wetland illustrating the boundaries between
upland and wetland edge, and edge slope and bottom basin 16
FIGURE 4. Major work tasks of field crew (top) and average time spent on each task
(bottom) during the field data collection in natural and created
wetlands 18
FIGURE 5. Typical cross sections of a natural (top) and created wetland (bottom)
showing physical characteristics 22
FIGURE 6. The proportion of total number of species in created wetlands that resulted
from planted species (top graph), and their importance values as
determined by relative density, relative dominance, and relative frequency
of occurrence (bottom graph) 26
FIGURE 7. The ten most important species in natural (top) and created wetlands
(bottom) as determined by relative densities, relative dominance, and
relative frequency of occurrence. For comparison, importance values are
shown for both natural and created wetlands in each graph 27
FIGURE 8. Range of weighted average (WA) scores for natural and created wetlands
using NWI (Reed 1988) classification of species 30
FIGURE 9. Number of species and proportion of upland species along transects of an
unmulched created wetland (top), and a mulched created wetland (middle),
and a natural wetland (bottom) 34
VI
-------
ACKNOWLEDGEMENTS
The author owes many thanks and much of the description of field events to R. E. Tighe,
G.R. Best, D.W. Hall, K.H. Brandt, K.E. Dollar, C.A. Raymond, S.J. Roguski and S.E.
Tennenbaum, who formed the field crew for this study. Stephanie Gwin, ManTech
Environmental Technology, Inc. (METI), coordinated activities between the Wetlands Research
Program at the USEPA Environmental Research Laboratory, Corvallis, Oregon, and the Center
for Wetlands. Ann Hairston, METI, provided technical editing and did the final revisions of
this report. Linda Crowder, Center for Wetlands, and Kristina Miller, METI, performed the
word processing that turned the manuscript into a document. Finally, thanks to Ken Bierley,
Kate Dwire, Stan Geiger, Mary E. Kentula, Steve Paulson, and Loverna Wilson who reviewed
various drafts of this report.
V1X
-------
ABSTRACT
This report describes a pilot st> dy that is part of a continuing series of wetlands studies
conducted in cooperation with tht Wetlands Research Program at the United States
Environmental Protection Agency (USriPA) Environmental Research Laboratory, Corvallis,
Oregon, (ERL-C). To date, these studies have been conducted in Connecticut, Florida, and
Oregon. In the summer of 1988, field crews from the University of Florida Center For
Wetlands in Gainesville spent ten days (two separate field trials with over one month between
the first five day period and the second five day period) sampling 18 herbaceous wetlands (nine
created and nine natural) in northern Hillsborough County, near Tampa, Florida. This report
summarizes their evaluation and recommendations regarding an approach to wetland sampling
and characterization developed by the Wetlands Research Program at ERL-C. Between trials,
team members discussed at length the methodology and field protocols and modified them to
reflect the conditions and difficulties encountered in sampling herbaceous wetlands in urban areas
of Florida. Major emphasis is placed on the appropriateness of measured variables for
determining successful wetland re-creation.
Numerous physical and biological parameters were measured and compared in the nine
created and nine natural wetlands. Analysis of these data has shown some important similarities
and differences between created and natural wetlands and lends insight into the complex
questions surrounding wetland creation and the equivalency of created wetlands to naturally
occurring wetlands. Evaluations of temporal changes in hydrology and plant successional trends
within created wetlands seem most important in determining ultimate success.
Vlll
-------
INTRODUCTION
Under Section 404 of the Clean Water Act, it may be required that wetlands be replaced if
they are degraded or destroyed. Increasingly, federal, state, and local permitting agencies have
been requiring compensatory mitigation through creation of wetlands on another portion of the
development site. The practice of creating a wetland to replace a natural wetland is appropriate
only if the created wetland will be equivalent to, or of higher quality than, the wetland that is
destroyed in terms of size, hydrology, water quality, and life support functions. Therein lies
an important concern: How best to determine successful wetland creation.
In Florida, experience has shown that controversy over the potential success of wetland
creation most often occurred, and resulted from concern over equivalency, during the permitting
process. Frequently, especially where forested wetlands were involved, created wetlands did
not resemble their naturally occurring counterparts in the initial years after creation (Clewell and
Lea 1990). Even created, herbaceous wetlands, which were quicker to revegetate, often had
different species than naturally occurring wetlands (Brown and Tighe 1990). Clearly, evaluative
procedures are needed to measure success and to eliminate some of the uncertainty concerning
the efficacy of creating wetlands as mitigation for encroachment or loss of wetlands due to
development activities.
In this study, an approach for evaluating created wetlands initially developed by the U.S.
EPA's Corvallis Environmental Research Laboratory (EPA ERL-C), Wetland Research Program
was tested and refined (see companion document, QA Project Plan for Florida Wetlands Study).
In the process of testing the methodology, numerous physical and biological parameters were
measured in both created and natural wetlands. Analysis of these data has shown some
important similarities and differences between created and natural wetlands and lends insight into
the complex questions surrounding wetland creation and the equivalency of created wetlands to.
natural wetlands. Results from this and similar studies across the country can be used to help
decision makers design new regulatory strategies for protecting the Nation's remaining wetlands.
-------
^ PLAN OF STUDY - -
As part of a series of studies begun I y the EPA ERL-C, the Florida tests were conducted
during the summer of 1988. Early on, it w?is determined that comparisons of populations of
created and natural wetlands offered the best \ otential to evaluate the success of wetland creation
on a regional basis. Much emphasis, therefore, had to be given to the process of selecting
natural wetlands as well as data collection.
Because of Florida's diverse climatic and geomorphologic character, the sites were
chosen within a relatively small, homogeneous ecoregion. The area around Tampa, Florida,
which includes Hillsborough County, was selected primarily because accurate records of species
planted, year of planting, site conditions, and follow-up site visits were kept by the county's
Environmental Protection Commission (EPC) on all created wetland projects required over the
past six years. Detailed construction plans for the created wetlands were often omitted from the
permit records, so no attempt was made to compare the current wetland condition with the
proposed wetland. The northern part of the county had more created wetlands because of the
outward expansion of the Tampa urban area, so it was chosen as the study region.
Created wetlands chosen for the study met criteria of: size (less than or equal to one
hectare); type (herbaceous vegetation); age (at least one year); and maintenance (the less
maintenance performed since creation of the site, the more desirable the site). Natural,
herbaceous wetlands that were used as reference wetlands occur most often as isolated, fully
vegetated communities throughout this ecoregion. Therefore, isolated, fully vegetated marshes
were the most desirable created sites to make relevant comparisons between natural and created
wetlands. Location of wetlands in an urbanized setting and accessibility were additional
selection criteria. Although access to use the created wetlands as study sites was granted for all
systems chosen, there was some reluctance on the part of landowners to grant access to natural
wetlands. This resulted in the inclusion of a clause in letters requesting site access guaranteeing
that landowner's names or site locations would not be used in any reports or publications.
Therefore, only general maps of the study area are presented and the wetlands are referred to
by number. Figure 1 shows the locations of wetlands from which the created study sites were
chosen.
Several physical and biological parameters were measured in nine created and nine
natural, herbaceous wetlands and comparisons were made of the characteristics of the two
populations. Comparison with natural wetlands seemed the most relevant way of evaluating
created wetland success for several reasons: (a) most qualitative evaluations by agencies, and
the public in general, involve some comparison with an "ideal" or average wetland; (b)
quantitative evaluation requires that some standard or natural characteristics be used for
comparison (characteristics of natural wetlands were used because specific goals for created
wetland projects were not given); and (c) permit conditions for the creation of wetlands usually
-------
Pasco County
R18£ '* ) HUlrtxXouoh "J* ""'»
^ r"S£,?yu8n A • ^
/ 5 . /\
( ? «4 /% \
R206
01234 5
FIGURE 1. Map of Tampa and the surrounding area in Hillsborough County showing
locations of created wetlands considered as study sites.
-------
require planting of species that occur in wetlands of the type that is being replaced, and often
require that created wetlands match specific wetland species composition.
In this study, comparisons were made between the two populations of wetlands (created
and natural) as a means of addressing the overall success or equivalency question of w .".tlands
creation rather than the success of an individual created wetland. It was not the intent of the
study to evaluate any particular wetland, but instead, to test a methodology by which indi idual
wetlands could be evaluated in the future, once an adequate data base on both created and natural
wetlands exists. This latter point is most important if evaluation is to be meaningful, since the
variation in vegetation and physical attributes between wetlands often makes one-on-one
comparisons inappropriate.
The field protocol was designed with two overriding concerns. First, it had to be quick
and economical. The field measurements had to be taken in a relatively short period of time,
by a small number of technicians and biologists. This meant that seasonal or annual monitoring
schemes were not feasible. Second, the protocol for field measurements had to minimize
damage to the wetland that might result from sampling or trampling.
Identical field measurements were made in both created and natural wetlands. Both biotic
and abiotic variables were measured within each wetland reflecting the importance of biological
success in the short run, and physical equivalence that, in the long run, would ensure continued
success of the created wetland. The variables measured were: vegetation composition, soil
organic matter, water depth, water quality (basic nutrients and common metals), and site
topography. In addition, qualitative assessments of the surrounding landscape were noted and
a photo record of the site and surroundings was made.
A word of caution here. The evaluation of a created wetland by sampling various
parameters and comparing the values obtained with those of natural wetlands is subject to two
major drawbacks. First, the sampling of a created wetland "catches" the system at one point
in time, often after a relatively short time interval since creation, and compares it to wetlands
that have existed over a much greater period of time. Second, the timing of the "snap shot",
both in relation to the season and the time since creation, has a significant impact on the utility
of the data collected as an indicator of successful creation.
To overcome seasonal variation, sampling should be conducted during the mid-to-late
growing season (in this case, sampling was conducted during late May and early July). To
overcome variation resulting from time since creation, the evaluation of created wetlands may
need to be postponed until several growing seasons have passed, reducing the likelihood of
significant changes in species composition or zonation. Our experience with created wetlands
in the phosphate district of central Florida suggests that a period of approximately three years
(depending on site treatment and existing conditions) is sufficient to eliminate early successional
variation in species composition and zonation. Balance must be achieved between the need for
early evaluation and waiting for the system to develop fioristic equilibrium.
-------
With the increased use of evaluative monitoring and greater experience in creating
wetlands, thrappropriate time interval between creation and evaluation"may be more accurately
determined. It may be necessary and productive to sample at different times for different
parameters. For instance, hydrology might be sampled in the first year after creation to
determine early on whether the created wetland has appropriate hydrologic characteristics, while
vegetation might better be sampled later.
-------
METHODS
Eighteen herbaceous wetlands (nine created, nine natural) were chosen for comparison
to test a methodology for evaluating created wetlands. Field and data analysis protocols were
developed and tested earlier at EPA ERL-C (see companion document, QA Project Plan for
Florida Wetlands Study). Suggestions for modification of the protocols were made prior to and
after field tests and data analysis in Florida. The methods given here are those used in the
Florida Pilot Study. Sample sizes were limited in this study as it was one of several pilot studies
to develop methods for evaluation of created wetlands. Following further refinement and testing
of the evaluation techniques, full-scale regional studies of created wetlands will be initiated by
the EPA ERL-C Wetlands Research Program.
SITE SELECTION
The studies were conducted in herbaceous wetlands less than or equal to one hectare in
size to coincide with previous work on the evaluation technique done in Oregon. Criteria and
a method for selection of created and natural wetlands were given in a previous paper by Brown
et al. as yet unpublished. Therefore, a general overview of the selection process follows:
1. Created wetlands were selected from a data bank of the Hillsborough County,
Florida, Environmental Protection Commission (EPC). All candidate wetlands
within the ecoregion were selected, randomly numbered, and visited in numerical
order. The first nine sites visited that met the selection criteria were chosen.
Since the sites were still in the active files of the EPC, permission to gain access
to the sites was easily obtained from the landowners.
2. Natural wetlands were selected to represent as wide a variety of environmental
settings within the same ecoregion as possible. A method of ranking the intensity
of urbanization surrounding a wetland (its environmental setting) was developed
and used as a primary selection criteria. The method, called the Landscape
Development Intensity Index (LDI Index), scaled the intensity of development
from 1 (completely natural), to 10 (completely urbanized). Permission to access
natural wetlands on private lands proved to be difficult and selection ultimately
favored access over environmental setting.
Permission to access sites was obtained for only two of the originally selected natural
sites. The selection methodology was further refined to first select natural wetlands based on
access and then to classify them according to their LDI index. The remaining seven natural
-------
wetlands were located on public lands. It was necessary to limit the candidate wetlands to those
on public lands because of the limited amount of time and the lacktrf success in obtaining access
from private landowners.
HELD PROTOCOL
Of primary concern during the field trials was the potential for trampling of vegetation
in both created and natural wetlands as a result of the number of crew members and the number
of times that they were required to enter the wetland. Every step of the evaluation protocol was
examined in light of this very important consideration, and whenever possible the number of
trips and number of individuals entering the wetland were minimized. Many suggested changes
to the field protocol resulted in decreasing the potential for trampling. For example, although
the number of field crew members was increased, trampling was decreased by reducing the
number of times each person entered the wetland to one.
FIELD CREW COMPOSITION
Based on review of ERL-C documentation and training, the original field protocol calling
for a six member work crew divided into three teams (two vegetation teams and a survey team,
each composed of two members) was modified to a seven-member work crew divided into three
teams (two vegetation teams composed of a botanist and a recorder and a survey team of three
members). In the Florida study, the addition of two alternates to the field crew was proposed
as insurance that the field work be completed within the time allotted, and the project manager
was added to observe the field team in action for later evaluations of the methodology, including
the time needed to perform the tasks.
On the first day of field trials, it was found that adding one permanent member to the
two member survey team expedited the completion of the tasks. One of the original alternates
became the third member of the team. Based on this change to the survey team protocol, one
of the team members drew the map, recorded general site information, and took photographs
while the other two members collected morphological data. Soil samples were taken following
completion of these tasks, by any two of the three survey team members, while the third person
checked and finalized data sheets.
The remaining alternate collected water samples from several sites on a given day during
the required sampling time. Since two or three sites were sampled during any given field day,
and water samples were to be collected during a relatively short period at mid-day, the additional
person was necessary to adhere to the quality assurance plan for water samples.
-------
Budget constraints and travel distance to the wetland sites required that the field work
be completed^ two, five-day periods. Since the sites were three-hours from Gainesville, the
field team stayed in Hillsborough County during each field sampling period. In addition, the
fieldwork had to be completed within a relatively short seasonal window to insure that seasonal
changes in species composition did not bias the data.
FIELD MEASUREMENTS
Field measurements were made along transects that crossed the wetlands at right angles.
Parameters measured included: plant population structure, site morphology, water depths, soil
organic matter, andjvater quaUtyj(nutrients and metals). The vegetation teams quantitatively
determined vegetation characteristics? The survey team measured site morphology and water
depth, and collected soil and water samples. In addition, they characterized surrounding site
conditions and obtained a complete photographic record of the site. The field data collection
took place in the summer of 1988 and was organized into two separate field excursions of 5 days
each separated by four weeks. After the first round of data collection, the field team evaluated
the field protocol and suggested further modifications for the secdnd round of field data
collection. Wetlands 101-110 were evaluated during the first week of field sampling and
numbers 201-208 were evaluated during the second week.
TRANSECT ESTABLISHMENT
All parameters were sampled along transects at regular intervals. Since many wetlands
exhibit zonation that results from varying hydroperiods, transects crossing the wetland and
perpendicular to each other were established to capture community level variation in measured
parameters.
Figure 2 illustrates how transects were established. In general, two transects were
established on each wetland at right angles to each other. Where the wetland was long and
narrow (length greater than twice the width) three transects were established, one lengthwise and
two crossing the long transect at right angles. Since both vegetation and basin morphology were
to be measured along the same transects, transect layout was designed to capture variation in
both features. As a consequence transect establishment in some instances favored locations that
ensured greatest topographic variation. Under most conditions this layout also resulted in
"capturing" greatest detail in vegetative zonation and variation since basin morphology
determines hydroperiod and, as a consequence, spatial manifestations of plant community
structure.
-------
FIGURE 2. General layout of transects in natural and created wetlands. Both morphology and
vegetation data were collected from a total of 40 quadrats spaced equally along
the transects.
-------
Natural, isolated wetlands in the Florida landscape are generally regular in shape, and
are usually circular, or very nearly so. Our experience indicates that created wetlands are also
generally regular in shape, although several shapes are common, ranging from circular to square
or rectangular. Both systems tend to have relatively regular morphology. From these
considerations, and the above discussion, it was concluded that most small, isolated wetland
systems in Florida—both natural and manmade--can be adequately described using only two
transects, these being measured for both morphology and vegetation. However, special or
unusual circumstances may require additional transects.
As stated earlier, trampling of vegetation was of paramount concern. Transect
establishment and subsequent measurement of variables was designed to minimize the number
of people and trips through each wetland. Once the transect locations were determined, the ends
of each transect were located 10 meters upland of the wetland edge and marked with three-meter
PVC poles and an appropriate number of flags to denote transect number. The length of each
transect was estimated and spacing of sample plots calculated so that there were at least 40 plots
per site spaced equally along the transects beginning at the wetland edge. Based on the
vegetation present and the State of Florida Wetland Plant List1, the botanists established the
wetland edge by common agreement.
To minimize the number of wetland crossings by field personnel, measuring tapes were
laid along the length of the transect as the plots were located and vegetation measurements were
made. The tapes were left in place for all subsequent measurements, and removed by the last
team through the wetland after all the data had been collected.
VEGETATION MEASUREMENTS
The composition of the vegetation team remained the same in the Florida study as
recommended by ERL-C: two teams with a botanist and a recorder on each team. Some
changes were made in the field methodology, however. Vegetation sampling basically consisted
of three parts: identification and collection of species, use of a modified Pielou Technique
(Pielou 1986), and estimation of cover. The Pielou Technique called for vegetation sampling
to be based on "k" number of species, "k" being the number of species to be sampled in a plot.
It was to be determined before field work began and was to be constant for all study sites. This
value was to be based on a pre-sampling study in wetlands similar to those to be sampled in the
actual study with the value of R large enough to sample the common species on sites. The
protocol was changed by eliminating the use of "k" number of species in favor of determining
commonness based on specified time intervals as discussed below. The reason for this being,
herbaceous wetlands in Florida vary from those dominated by only one or two species to those
with dozens of different species. We felt that commonness was not related to a pre-determined
*Fla. Admin. Code. Ch. 17-312 (July 1989).
10
-------
number of species, but rather to the number of species that could be readily observed within a
specified interval.
Nondestructive sampling was used to characterize the vegetative community structure.
Rectangular sampling frames (1-m2 quadrat frames), made of PVC pipe, were used to estimate
percent cover and "commonness" by species. A total of 40 1-m2 sample plots (quadrats) were
established at equally spaced intervals along the transects. Spacing between quadrats varied
from two meters on the smallest sites, to five meters on the larger sites.
Estimating Commonness
Relative commonness or rarity of the vegetation species at each site was determined using
the modified Pielou method. In the modified technique, the botanist identified species that were
seen within given time intervals, and the recorder assigned a numeric value to each time
interval. Specifically, species observed in the first 15 seconds were most common and given
a value of 3. Those observed in the next 15 seconds (15 - 30 seconds) were less common, and
given a value of 2. Those observed in the next 30 seconds (30 - 60 seconds) were uncommon
and given a value of 1. Finally, those species that were not observed in the first minute, were
rare, and given a value of 0.
Estimating Percent Cover
Cover was estimated using grid marks along the edges of the quadrat frame. The botanist
estimated the percent of the quadrat that was covered by all vegetative structures of each
identified species. Because of stratification of species perpendicular to the ground surface, total
percent cover estimates often were greater than 100% within a given quadrat.
We suggest that estimation of percent cover be done for differing vegetation strata,
starting with ground cover. Strata might be as follows: 0 to 5 cm above ground, 6 cm to 50 cm,
and greater than 50 cm. In this way some additional information is gained with the data set, and
one avoids the problem of having greater than 100% cover on any single plot. However, during
these field trials we did not alter the protocol to deal with percent cover in this manner.
VEGETATION CHARACTERISTICS
Using a technique suggested by Wentworth et al. (1988), a wetland weighted average
score (hereafter referred to as WA score) was calculated for each wetland. The technique uses
the NWI species categories (Reed, 1988), and assigns an ecological index from 1 to 5 as
11
-------
follows: obligate wetland = 1, facultative wetland = 2, facultative = 3, facultative upland =
4, and obligate upland = 5. Wetland weighted average scores were calculated with the
following formula:
Wetland WA Score =
Where: Q = Percent Cover of Species i
E; = Ecological index of species i
s = Number of species in the wetland
Calculated wetland WAs range from 1 (extreme wetland, 100% obligate wetland plants) to 5
(extreme upland, 100% obligate upland species).
SOIL SAMPLES
Two soil samples (5-cm and 30-cm depths) were taken from every fourth quadrat using
an 8.25-cm (3.25-inch) open-sided bucket soil auger. Munsell color was recorded and odor
subjectively determined at the vehicle and in the field, respectively. Samples were iced for later
determination of organic matter content by ignition.
The determination of soil organic matter in a variety of created wetlands over a wide
range of conditions is a valuable data set for comparative purposes. As the data on created
wetlands builds, comparisons can be made between newly sampled wetlands and the data from
all created wetlands. The value of soil organic matter data may lie more with the ability to
compare between created wetlands of various ages, rather than between created and natural
wetlands.
WATER QUALITY SAMPLES
Sample collection for measurement of water quality was generally undertaken as
prescribed by ERL-C methodology. Two minor changes resulted from previous experience.
First, sample bottles were prefixed at the laboratory selected to do the water quality analysis,
and second, samples were collected from several wetlands by an alternate field crew member
during the required time period each day.
12
-------
Where there was surface water, water samples were collected at each wetland in prefixed
250-ml samplEcontainers and iced for laboratory analysis. Two water samples, one each at the
inlet and outlet were collected in wetlands with flowing water. In isolated or "stagnant"
wetlands with no flowing water, one sample was collected. Laboratory analysis using standard
procedures was performed for the following constituents: Total Suspended Solids (TSS), Total
Phosphorous (TP), Total Kjeldahl Nitrogen (TKN), Total Organic Carbon (TOC), Lead (Pb),
Cadmium (Cd), and Aluminum (Al).
The collection of a grab sample does not allow for much inference related to long-term
water chemistry, but may give some indication of gross abnormality. Comparison of water
quality between a created wetland and other wetlands (both created and natural) that indicates
serious departure from normal ranges may be important information that would suggest a need
for closer monitoring, resampling, or reevaluation at a later date.
SITE MORPHOLOGY AND WATER DEPTHS
Site morphology was determined by measuring topographic relief across each wetland
along the transects. Topography was measured using a contractor's level and stadia rod to
determine ground elevations in each vegetation quadrat. Where relief warranted (i.e., where
slope changed abruptly), more frequent measurements of ground surface elevation were made
to adequately capture topographic relief.
At each wetland, a benchmark was established and all elevations were recorded relative
to the benchmark. Later, elevations were converted and expressed relative to the wetland edge.
When surface water was encountered, elevation was measured and surface water depth
was calculated for each quadrat by subtracting ground surface elevation from the water surface
elevation. On sites with no surface water, depth to groundwater was measured in each of the
holes augered for soil samples.
The "snap shot" of water levels obtained on sampling day, when combined with basin
morphology, begins to give some approximations of potential water depths. While it cannot be
used effectively to predict long-term hydrology, the one-time measurement does serve to
characterize the physical environment of the wetland on sampling day. It is relatively easy to
acquire water level measurements during the sampling effort and, based on the importance of
adequately characterizing the wetland, water levels should be measured.
Depths of water and periods of inundation are probably the most critical factors in
determining wetland creation success. However, for hydrologic data to be of use in
determination of hydroperiods, it must be measured over long periods of time. A minimum of
one year is necessary, and preferably, measurements over a drought-flood cycle should be
obtained. Recently, in some parts of Florida there has been a move toward requiring creation
13
-------
of mitigation sites prior to destruction of the original wetlands. Under these circumstances, it
seems quite ^Appropriate to require continuous hydrologic monitoring~to determine if proper
hydroperiods have been established.
DATA ANALYSIS
Data analysis was performed using the Statistical Analysis System (SAS) on data sets
created using dBase ffl. Data were double entered and anomalies between sets rectified until
both data sets agreed. Numerous indices were calculated for each wetland and means compared
between created and natural wetlands.
Comparisons were made between the population of created wetlands and the population
of natural wetlands rather than on a wetland-by-wetland basis, since the objectives of this study
were not to determine whether any individual wetland was successful, but rather to suggest and
test various measures and indices of wetland structure and organization using indices of
similarity and to determine the range of possibilities for successful wetland creation projects
based on natural variability. In actual practice, once a sufficient data base on both created and
natural wetlands exists, evaluations on a wetland-by-wetland basis could be conducted using
indices of similarity.
PHYSICAL CHARACTERISTICS
The physical characteristics of natural and created wetlands that were compared included:
surrounding landscape condition, wetland basin topography (depth, slope and roughness), water
depths, percent open water, and soil organic matter.
Surrounding Landscape Condition
An index of Landscape Development Intensity (LDI) was calculated for each wetland to
evaluate environmental setting. In brief, the index was calculated using the percent cover by
urban, agricultural, and natural lands in an area of 0.68 km2 surrounding each wetland. Aerial
photographs were used to estimate percent cover of each land use type. In addition, ground-
based, qualitative observations of the "condition" of the surrounding lands were noted during
the field data collection.
14
-------
Basin Topography
Figure 3 illustrates a typical wetland basin cross section divided into three segments: two
edge slopes and a bottom basin. The boundary between edge slope and bottom basin was
determined visually using cross sections of each wetland. Average slope of the edge slope area
was calculated using the difference in elevation between the wetland edge and the elevation of
the edge slope/bottom basin boundary divided by the distance between these two points and
expressed as a percent.
Average wetland depth was calculated as the average of elevation readings of the bottom
basin on each transect relative to the wetland edge.
Roughness was determined using only the elevations of the bottom basin and was
calculated as the variance about the mean elevation of the bottom basin.
Water Depth
Average water depth was calculated from the depths measured on the day of sampling.
Field measurements were conducted during the wet season and while observed water levels were
not an indication of long-term levels or seasonal fluctuation, they did provide a comparative
indication of hydrology. In addition to average water depth calculated as the average of only
those stations that had standing water, a weighted water depth was calculated by multiplying
average water depth by the percent area of the wetland that was inundated.
Percent Open Water
Percent open water was determined by subtracting vegetation percent cover from 100%
within each quadrat of the wetland, summing the total and dividing by the number of quadrats.
15
-------
Upland
Wetland
Upland
Wetland
Edge
Edge Slope Bottom Basin i Edge Slope
FIGURE 3. Typical cross section of a wetland illustrating the boundaries between upland and
wetland edge, and edge slope and bottom basin.
16
-------
Soil Organic Matter
Soil samples collected at 5-cm and 30-cm depths at every fourth vegetation plot were
oven dried to 0% moisture content, weighed and then fired at 550° F for 30 minutes. Organic
matter content as percent of total was calculated as the difference between weights before and
after ignition divided by the pre-ignition weight and multiplied by 100.
FIELD DATA COLLECTION
Field data were collected by a seven-member work crew divided into three teams as
described in the section on field crew composition. The vegetation team established transects,
quadrat spacing and collected both percent cover and plant species commonness data. The
survey team measured topography, assessed surrounding environmental conditions, took
photographs, collected soil and water samples, and mapped the wetland.
The length of time spent at each site was recorded by task and is summarized in Figure
4. In the top portion of Figure 4, a flow diagram for each of the field teams is given. The
length of time given for each task in the bottom diagram of Figure 4 is somewhat generalized
and is based on the average time spent at each task. Field data collection in created wetlands
was more time consuming than in the natural sites. The average total length of time at created
sites was 3.8 hours, while the time spent at natural sites was about 2.6 hours. Total person
hours averaged 26.6 and 18.2 at each site for created and natural wetlands, respectively.
17
-------
Project Crew Arrives At Site
Si Assembles
Survey Equip.
S2 Collects
Water Sample
S3 Distibutes
Form Packets
S1.S2 Conduct
Perimeter Survey
S1.S2 Measure
Transect Elevations
S1 Completes
Forms H and R
B1.B2 Conduct Site
Reconnaissance.
Establish Transects
S3 Takes Site Photos
R1.R2 Assemble
Equipment
S3 Prepares Site Map
and FiHs Out Site
Description Form
Plant LD.:B1.B2
Determine Species Present,
R1.R2 Fill Out Forms D1 and 02
S2.S3 Collect
Soil Samples
Vegetation Teams Conduct
Pietou and % Cover Procedures
All Teams Disassemble Equipment and Check Forms
B1
B2
R1
R2
S1
S2
S3
0:0
Site Rec.
Establish
Transect;
Assem.
Equip.
Assem
Equip.
Water
Sample
Plant I.D.
Record
Species
Survey Basin
Perimeter
KM ""«
Conduct Vegetation Transects
Survey Transect
Elevations
Draw Map
Check Forms
Soil Samples
Check
Forms
I 1 1 I l 1
0 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:1
Check
Forms
Disassemble
Equipment
5 2:3
Start
End
FIGURE 4. Major work tasks of field crew (top) and average time spent on each task
(bottom) during the field data collection in natural and created wetlands.
18
-------
RESULTS
PHYSICAL CHARACTERISTICS
Created wetlands ranged in age from 1 to 8 years with the majority 3 years of age (Table
1). Age was taken from permit records and corresponds to the number of growing seasons since
site construction was completed.
Several characteristics that are related to basin morphology as well as percent open water
and organic matter in soils of natural and created wetlands are given in Table 1. Maximum
depth (the deepest point measured from the wetland/upland edge) in natural wetlands ranged
from 0.46 to 1.58 m and mean depths ranged from 0.13 to 1.10 m. Created wetlands, on the
whole, had a similar range of maximum depths (0.33 to 1.66 m) but a larger portion of the
created wetlands exceeded 1 m in depth. Created wetlands also had greater mean depths,
ranging from 0.36 to 1.33 m.
Roughness and edge slope are measures of basin morphology that may have a large
influence on long-term vegetative community structure. Roughness, a measure of micro-
topographic relief was defined as the variance of the measurements of ground surface elevation
around the mean ground surface elevation of the bottom basin. Roughness varied in natural
wetlands from a low of 0.069 to a high of 0.723, and from 0.152 to 0.897 in created wetlands.
Edge slopes were greater in created wetlands (varying from 3.31% to 17.70%) than in natural
sites (from 1.41% to 5.92%).
Percent open water was evaluated to determine differences in overall percent cover for
natural and created wetlands. The variation in percent open water for natural and created
wetlands was 0.00% to 2.5% and 0.51% to 3.03%, respectively, suggesting little difference
between the two populations.
The last two columns in Table 1 give the mean percent organic matter in soil samples
obtained at 5-cm and 30-cm depths for each wetland. The 5-cm sample from natural wetlands
had percent organic matter content that was significantly higher than the deeper samples at the
same sites, and from the samples taken at created sites. With the exception of three sites (103,
105 and 106, all created sites), results indicated normal wetland conditions, as all had lower
percent organic matter in the deeper samples (30 cm) than in the surface samples.
19
-------
TABLE 1. Physical Characteristics of Natural and Created Wetlands.
ho
O
Wetland
Number
Natural
107
108
109
110
201
202
203
206
207
Created
101
102
103
104
105
106
204
205
208
Age
3
1
4
8
3
3
3
2
3
Maximum
Depth
1.19
0.67
0.46
0.76
1.58
0.51
0.56
1.04
0.83
1.17
0.33
1.05
1.66
1.43
1.48
1.62
0.53
1.18
Mean
Depth
0.70
0.13
0.22
0.33
1.10
0.22
0.27
0.58
0.37
0.67
0.40
0.57
0.78
1.08
0.93
1.33
0.36
0.67
Roughness1
.553
.069
.088
.190
.723
.066
.154
.310
.173
.383
.152
.191
.450
.588
.323
.897
.188
.488
Edge
Slope2
2.02
1.9
1.8
3.35
5.92
. 1.41
2.13
2.56
2.28
21.69
16.82
3.71
3.31
17.70
11.03
9.69
8.78
9.69
% Open
Water
1.23
0.00
1.05
2.50
2.22
2.50
2.50
2.50
2.38
3.03
2.38
1.89
2.50
2.50
2.50
2.50
0.51
2.50
%O.M.
(5cm)
42.0
12.7
11.7
37.3
3.6
10.3
3.1
16.7
2.22
8.6
6.3
5.3
7.9
3.1
3.6
3.9
1.8
13.8
% O.M.
(30 cm)
40.5
4.5
3.8
12.9
2.1
4.5
8.7
6.3
10.9
3.4
3.7
5.4
3.5
4.2
5.5
3/»
0.7
4.9
1 Roughness is calcuated as the variance about the mean depth of the bottom basin.
1 Edge slope is calculated as the difference in elevation between the wetland/upland edge and the edge/bottom basin boundary divided by the
horizontal distance.
-------
Figure 5 summarizes the physical differences between natural wetlands and created
wetlands usifif typical cross sections of each. Natural wetlands were shallower (0.47 m mean
depth) compared to created wetlands (0.79 m mean depth) and were smoother (mean roughness
equal to 0.258 and 0.407 for natural and created wetlands respectively). Created wetlands had
greater maximum depths (1.16 m) than did natural wetlands (0.84 m) and steeper edge slopes
(11.4% in created versus 2.6% in natural wetlands).
WATER QUALITY
Table 2 summarizes water quality data for the natural and created wetlands. Water was
present in seven of the nine natural wetlands and in all nine of the created sites. Water levels
in all wetlands sampled presumably were lower than normal since the region had been
experiencing a relatively serious drought prior to and during field data collection.
In several wetlands more than one sample was taken either as part of the quality
assurance (QA) program, or because there was flowing water. Wetlands 103, 108, and 207 had
two samples taken as part of the QA program as detailed in the ERL-C work plan. The results
from the two samples taken from sites 103 and 207 reflect similar trends in concentrations of
nutrients and metals, but site 108 had significant differences between the two samples, probably
the result of inadvertently collecting suspended organic matter during sampling.
Wetland 204 was part of a flowing water system where two water samples were taken,
one at the inflow directly from a lake (sample 204) and the second sample at the outflow from
the created wetland (sample 204a). The higher total suspended solids and total phosphorus in
the inflow than in the outflow may be indicative of the filtering action of the created wetland.
Water quality differed somewhat between the natural sites and the created sites.
Generally the natural sites exhibited higher concentrations of nutrients and metals than the
created sites. Since most of the natural wetlands were experiencing unseasonable dry-down, the
result of low rainfall, the higher concentrations may be a function of oxidation of organic
substrate and the concentrating influence of the dry-down.
Since only one sample was taken from most wetlands and the three duplicates show
considerable variation, these results should be viewed cautiously.
VEGETATION CHARACTERISTICS
Table 3 summarizes the species planted for the nine created wetlands. In addition, six
of the nine sites were mulched with organic matter from a "donor wetland." By far the most
prevalent species planted were Pontederia cordata and Sagittaria lancifolia, two rather showy
21
-------
- r- Natural Sites
E,
c
"5
_o
UJ
o
o
cc
Mean Depth: 0.47 m
Average Roughnesss:0.258
Average Max, Depth: 0.84 m
Average Edge Slope: 2.6%
Distance (m)
_o
"5
UJ
o
2
o
cr
Created Sites
Mean Depth: 0.79 m
Average Roughness: 0.407
Average Max. Depth: 1.16m
Average Edge Slope: 11.4%
Distance (m)
FIGURE 5. Typical cross sections of a natural (top) and created wetland (bottom) showing
physical characteristics.
22
-------
TABLE 2. ^~ Water Quality in Natural and Created Wetlands.
Wetlands #
TSS
TotP
TKN
TOC
Pb
Cd
AJ
Natural Wetlands
107
108
108a
110
201
206
207
207a
11.0 .
800.0
4.0
66.0
5.0
3.0
270.0
280.0
0.05
1.50
0.05
0.11
0.05
2.10
6.10
8.70
2.20
15.00
2.20
2.50
1.90
6.40
10.00
13.00
26
37
28
54
18
41
43
37
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.003
0.003
0.003
0.003
0.003
0.003
0.003
" •-. 0.003
0.50
• 1.30
0.50
0.50
0.50
0.50
1.40
1.30
Created Wetlands
101
102
103a
103
104
105
106
204
204a
205
208
21.0
13.0
13.0
10.0
1.0
24.0
20.0
43.0
4.0
50.0
59.0
0.42
0.13
0.05
0.06
0.05
0.19
0.18
0.44
0.05
0.05
0.05
1.40
1.30
0.65
1.30
1.20
2.40
1.10
3.90
7.50
1.20
6.60
18
30
17
9
12
21
14
26
19
27
20
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.003
0.003
0.003
0.003
0.003
0.003
0.003
0.003
0.003
0.003
0.003
3.70
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.72
23
-------
species that tend to flower much of the growing season and are relatively easy to propagate and
transplant. An average of five species were planted at each created site. Site 104 had no
species planted, while site 204 had 11 species planted. Comparisons between the species planted
and species occurring on each created site are given in Figure 6. With the exception of site 204,
the number of species planted at each site was less than 10% of the total number of occurring
species.
Figure 7 graphs the importance values of the 10 most dominant species in natural sites
(top graph) and created sites (bottom graph) as determined by relative density, relative
dominance and relative frequency of occurrence. For comparative purposes, the importance
values for the same species in the other population of wetlands is given for each. The two most
dominant species in natural wetlands were Panicum hemitomon and Pontederia cordata.
Pontederia cordata was also important in created sites. Four species in the top ten were shared
by both populations: Panicum hemitomon, Pontederia cordata, Juncus effusus, and
Hydrocotyle umbellata. Of these species only Panicum hemitomon and Pontederia cordata
were planted in created wetlands.
Table 4 summarizes the characteristics of plant communities of natural and created
wetlands. In the first three columns, details of species composition are given. The total number
of species found in natural and created wetlands was comparable, although, created wetlands
tended to have greater numbers of species and greater variation in number (13 to 93) than did
natural wetlands (23 to 48). When species were categorized using the NWI ecological groups
(Reed 1988) where obligate, facultative wetland, and facultative were considered wetland
species, while facultative upland and upland were considered upland species, the numbers of
wetland and upland species in the second and third columns resulted. While the general trend
was for greater numbers of upland species in created wetlands representing a greater proportion
of the total number of species, the trend was not significant.
The fifth column in Table 5 lists the WA score for each of the wetlands. Figure 8
graphically depicts the range of WA scores for natural wetlands (above the scale line) and
created wetlands (below the scale line). The WA scores ranged from 1.06 to 1.84 for created
wetlands; from .08 to 2.5 for natural wetlands. There was no significant difference between
WA scores for the populations of natural and created wetlands. With the exception of natural
wetland 107, all wetlands scored below 2.0, with the majority of sites scoring below 1.5. Six
of the nine natural wetlands scored below 1.5, while five of the created wetlands were below
1.5. The lowest score was that of site 105, an 8 year old mulched, created site. The highest
score, site 107 was a wetland in a residential setting that was described by its owners as
becoming increasingly dryer over the past several years, presumably the result of changes in
ground water conditions.
The last two columns in Table 4 give indices of plant species diversity. Comparisons
between the two populations of wetlands revealed no significant differences.
24
-------
TABLE 3. Species Planted and Use of Mulching at Created Wetlands.
Sites
Treatment 101 102 103 104 105 106 204 205 208
Mulch XXX XXX
Species Planted:
Bacopa caroliniana X X
Blechnum serrulatum X X
Cannaflaccida X XXX
Hypericum fasciculatum X
Iris spp. X
Iris virgirdca X XX
Juncus effusus X XXX
Ludwigia spp. X
Nymphaea odorata XX X
Osmunda regalis X
Peltandra virgirdca X
Pontedaria cordate XXX XXXXX
Sagittaria lancifolia XXX X X X X
Saururus cernuus X
Spartina bakeri XX XXX X
Thalia geruculata X X
25
-------
120
3 10°
"o ~~
0)
°- 80
Iff Planted Species
l£3 Naturally Occurring Species
100
o
^
CO
O
O
c
CO
*•<
O
Q.
E
CO
O
101 102 103 104 105 106 204 205 208
Site Number
FIGURE 6. The proportion of total number of species in created wetlands that resulted from
planted species (top graph), and their importance values as determined by relative
density, relative dominance, and relative frequency of occurrence (bottom graph).
26
-------
Panicum hemitomon
Pontoderia cordata
Eleocharis baldwinii
Proscrplnaca pectlnata
Juncus effusus
Nympholdes aquatlca
Nuphar luteum
Eupatorium leptophyllum
Hydrocotyle umbellata
Bacopa carollnlana
Natural Sites
Natural Sites
Created Sites
'////////A
7ZX
Pontederla cordata
Juncus effusus
Ludwigla peruvfana
Hydrocotyle umbellata
Saglttarta lancifolfa
Centella aslatica
Typha domlngonsls
PtiKmnium capIHacoum
Panicum hamJtonxjn
Polygonum nydroplperolde
'/////////////////////////////////A
'///////////A
'////////A
///7/7/A
//////A
Created Sites
Natural Sites
Created Sites
'/////A
/////A
V/7/A
0.0 2.0 4.0 6.0 8.0 10.0
Importance Value
12.0
FIGURE 7. The ten most important species in natural (top) and created wetlands (bottom) as
determined by relative densities, relative dominance, and relative frequency of
occurrence. For comparison, importance values are shown for both natural and
created wetlands in each graph.
27
-------
TABLE 4. Vegetation Characteristics of Natural and Created Wetlands.
Wetland
Number
Total
Species
Wetland
Species
Upland
Species
WA
Score
Shannon
Diversity
Simpsons
Diversity
KJ
00
Natural
107
108
109
110
201
202
203
206
207
Created
101
102
103
104
105
106
204
205
208
27
39
48
23
41
32
28
34
47
50
78
49
38
13
35
93
33
59
25
36
42
22
38
31
26
31
45
42
69
46
37
13
32
83
28
56
2
3
6
1
3
1
2
3
2
8
9
3
1
0
3
10
5
3
2.57
1.11
1.38
1.08
1.43
1.49
1.55
1.55
1.25
1.84
1.78
1.29
1.54
1.06
1.24
1.82
1.31
1.35
4.14
4.63
4.91
3.50
4.58
4.30
3.99
4.51
4.79
4.87
5.55
4.95
4.82
3.00
4.44
5.92
4.05
4.92
0.923
0.944
0.951
0.874
0.939
0.929
0.894
0.937
0.941
0.952
0.971
0.942
0.957
0.834
0.932
0.977
0.897
0.952
-------
TABLE 5.-;- Summary of the Means of the Variables Measured^for Natural and Created
Wetlands.
Parameter
Natural
Created
Mean Depth
Maximum Depth
Roughness
Edge Slopes*
% Open Water
% O. M.-
LDI*
Shannon Diversity
Simpsons Diversity
WA Scores
Species Richness
0.47
0.844
0.258
2.58
1.88
13.8
2.79
4.37
0.926
1.49
35.4
0.79
1.16
0.407
11.4
2.26
4.86
5.32
4.73
0.935
1.47
49.8
*indicates significant difference
by an analysis of variance.
.05 confidence level determined
29
-------
Range of Weighted Average Scores
Wetland -^ 1 ^ Upland
Natural Wetlands
3.0
Extreme wetland
(100 % obligate
hydrophytes)
3.5
Created Wetlands
4.0
5.0
4.5
Extreme upland
(100 % obligate
upland species)
FIGURE 8. Range of weighted average (WA) scores for natural and created wetlands using
NWI (Reed 1988) classification of species.
30
-------
DISCUSSION
COMPARISON OF NATURAL AND CREATED WETLANDS
Table 5 summarizes the results of analysis of the data for the populations of natural and
created wetlands. Several parameters were significantly different at the .05 confidence level
(determined by an analysis of variance) between the two populations, they included: edge slope,
percent organic matter in both the 5-cm and 30-cm soil samples, and the Landscape
Development Intensity (LDI) index. Other parameters, while showing differences in the means
for each sample were not significantly different at the .05 confidence interval, owing in part to
the small sample size and variability within each population.
Differences in edge slope between the created and natural wetlands were observed during
field data collection. The unnaturally steep slopes in the created wetlands most probably resulted
from a combination of inadequate design and economic realities associated with the value of real
estate. All the created wetlands were built in residential or commercial developments, often
"tucked" into comers, beside roadways or were associated with stormwater systems. The lack
of space and possible unwillingness of the landowner to commit larger land areas to wetland
creation (resulting from the high development value of land) may have contributed to the
designer and contractor minimizing the "footprint" of the wetland by increasing the slope of the
wetland edge. Gentle slopes require greater area to achieve adequate depth and result in larger
transitional area surrounding each wetland. While no attempt was made in this study to evaluate
the widths of transitional areas between uplands and wetlands, observations suggested that
transitional areas were smaller in the created sites, often completely absent. Most small
herbaceous natural wetlands in Florida exhibit zonation of plant species as a result of differing
hydroperiods along the moisture gradient extending from upland edge to the topographic low
point of the wetland. Analysis of the edge slopes showed that created wetlands had steeper edge
slopes and thus less potential for development of transitional areas and zonation.
The low soil organic matter in created wetlands was expected since most of these
wetlands were less than 4 years old. Created wetlands that were mulched had higher organic
matter than those that were not mulched, although because of the small sample size (six of the
nine sites were mulched) the difference was not statistically significant.
Mulch increases organic matter content of the surface soil horizon, but can easily oxidize
under dry conditions, or wash away under flowing water conditions. Previous studies of the use
of mulch in wetland reclamation on phosphate mined lands (Brown et al. 1985) estimated the
costs of digging, transporting and spreading of the mulch to be between $237 and $408 per
acre/inch per mile of round trip travel distance. In light of these costs, inundation soon after
31
-------
application of mulch material and control of water velocity as well as tillage in flowing water
wetlands may^help to lower loss of the added organic matter.
The LDI index was significantly higher for created wetlands than for natural wetlands
since most created sites were within residential and commercial developments, within regions
of Hillsborough County, Florida that were under considerable development pressure. Every
effort was made during natural site selection to choose natural wetlands that had the full range
of LDI indices, yet difficulties in obtaining permission to access sites in urbanizing locations
ultimately shifted site selection to more remote locations. The long-term consequences of the
differences in landscape setting as indicated by the LDI index on the structural and functional
aspects of created wetlands are yet to be understood, but observations (Brown 1986) of wetlands
within urban settings would suggest decreasing wetland function with increasing level of
urbanization in the surrounding landscape.
SIGNIFICANCE OF MULCHING ON COMMUNITY STRUCTURE
The lack of significant differences between created and natural wetlands related to species
diversity, and the nearly identical WA scores suggested that created sites were well stocked with
wetland species. Since six of the nine created sites were mulched with organic matter from
donor wetlands and only 16 species were planted in all the sites, the relatively high number of
species (mean = 49.8) probably resulted from introduction of species with mulch material.
Mean number of species for the three created wetlands that were not mulched was 32.6, while
the mean for the six mulched wetlands was 58.3 (Table 6). Planted species in mulched created
sites varied from 11 (site 204) to none (site 104). While planting helped to increase the number
of species, mulching alone had a greater effect (site 104, where no species were planted, had
32.6 species sampled).
The three graphs in Figure 9 show the number of wetland and upland species in plots
along one transect of an unmulched site (top), a mulched site (middle), and for comparison, a
natural site (bottom). The unmulched site exhibited lower overall numbers of species per plot
and no upland species occurred. The mulched site (middle graph) had higher overall numbers
of species per plot, exhibited zonation with decreasing numbers of species from the upland edge
toward the center of the wetland, and had several upland species along the wetland edge. This
graph illustrates that although the potential for introducing undesireable upland or exotic species
is increased with mulching, the zonation and species richness of the mulched site tends to mimic
the conditions found in natural sites (bottom graph).
The potential for introduction of noxious or exotic species is increased with mulching.
In Florida, Typha spp. (cattails) are considered an undesirable species in created wetlands. Since
they are rather ubiquitous within wetland systems, especiallys those that have some level of
disturbance, the possibility of transferring cattails and the long term consequences related to
maintenance should be considered when selecting "donor" wetlands. None of the mulched
32
-------
TABLE 6. Species Characteristics of Mulched and Unmulched Created Wetlands.
Parameter Mulched Unmulched
Richness 58.3 32.6
Shannon Diversity 5.04 4.1
Simpson Diversity 0.949 0.906
WA Score 1.52 1.38
Upland Species 5.2 3.67
33
-------
Upland Species
Wetland Soecies
Unmulched
Created Sites
YA Upland Speaes
Wetland Spades
Mulched
Created Site
Upland Speaes
Wetland Species
Natural Sites 109
6 8 10 12 14 16 18
Plot Number
20 22 24
FIGURE 9. Number of species and proportion of upland species along transects of an
unmulched created wetland (top), and a mulched created wetland (middle), and
a natural wetland (bottom).
34
-------
wetlands in this study exhibited problems with undesirable or exotic species invasions, nor did
the permit records indicate that problems had occurred in the past
UTILITY OF THE EVALUATION APPROACH
The evaluation approach tested in this study was designed to minimize impacts to
wetlands and to be relatively quick, yet yield sufficient data to evaluate the created wetland. The
average time spent at each site was about 16 person hours. The field crew left relatively few
visible traces of their presence. Therefore, using these criteria the approach was successful.
Evaluation of created wetlands should take into account not only short-term structural
similarity to naturally occurring wetlands but long-term functional equivalency as well. Long-
term functions are more difficult to evaluate since they require long-term monitoring, or at the
least, several site visits and data gathered over a period of years. The evaluation technique used
in this study did not address the long term, but could form the basis for data collection and
comparative analysis over a period of years to measure temporal trends and long-term success.
Seven of the nine created wetlands evaluated in this study were less than four years old
(one site was eight years old), and while experience has shown that four years is sufficient time
for the establishment of herbaceous wetlands, successional trends after a period of years could
result in a community with few wetland characteristics. Long-term stability, and therefore the
long-term success of the creation of wetlands, is impossible to determine with only one synoptic
field event. Yet, the measured parameters, including vegetation and soil organic matter, if
monitored on a yearly basis for several years could provide sufficient trends data to extrapolate
long-term success. For instance, trends in vegetative community structure and spatial
distributions of species could indicate whether the created wetland was shifting from domination
by wetland vegetation to domination by more upland vegetation. Changes in organic matter
content of soils could provide inference of hydrologic condition. If soil organic matter decreased
over time, it might be inferred that the wetland had low water levels and dry periods sufficient
for oxidation, and thus, no net accumulation of soil organics. If corroborated with vegetation
data, the inference could be strengthened without the necessity to monitor long-term water
levels.
Probably the most important class of functions that was not evaluated in this study was
related to wetland hydrology. Synoptic water levels were taken but proved little, other than
water was or was not present on the day of measurement. A better approach, of course, would
involve measuring water levels over a complete hydrologic cycle. Hydroperiod (period and
frequency of inundation) and water depth are key factors that control vegetation and most
chemical and biologic processes. Without some evaluation of hydrologic characteristics over a
sufficient period of time, it is difficult to ascertain equivalency in this most important category
of wetland functions. The use of iron rods implanted in created wetlands and left for a sufficient
period of time may help to evaluate minimum and maximum water levels (Harenda, et al. 1991).
35
-------
In closing, evaluations of temporal changes in hydrology and plant successional trends
within created wetlands, seem most important to determine ultimate success. The costs
associated with several site visits may be justified until sufficient data and experience have been
gained that one site visit after an appropriate number of growing seasons will yield enough
information to accurately estimate the success of the created wetlands.
36
-------
LITERATURE CITED
Brown, M.T. 1986. Cumulative impacts in landscapes dominated by humanity. In Estevez,
E.D., Miller, J., and Hamman, R. (eds.), Managing Cumulative Effects in Florida
Wetlands: Conference Proceedings, pp. 33-50. Sarasota, Florida: New College
Environmental Studies Program, Publication No. 37.
Brown, M.T., H.T. Odum, F. Gross, J. Higman, M. Miller, and C. Diamond. 1985. Studies
of a Method of Wetland Reconstruction Following Phosphate Mining. Final Report
to the Florida Institute of Phosphate Research, Bartow, FL. Gainesville, FL: Center for
Wetlands, University of Florida. 76 pp. (CFW-85-05)
Brown, M.T. and R.E. Tighe (eds.). 1990. Techniques and Guidelines for Reclamation of
Phosphate Mined Lands. Final Research Report to the Florida Institute of Phosphate
Research. Gainesville, Florida: Center for Wetlands, University of Florida.
Clewell, A.F. and R. Lea. 1990. Creation and restoration of forested wetland vegetation in the
southeastern United States. Wetland Creation and Restoration: The Status of
Science, pp. 195-232. Kusler, J.A. and M.E. Kentula (eds.) Island Press, Washington,
D.C.
Harenda, M.G., M.E. Kentula, J.C. Sifneos, and J. Honea. Evaluating the "Rusty Rod"
Technique for Use in Monitoring Subsurface Water levels in Wetlands. In Press
Wetlands Ecology and Management.
Pielou, E.G. 1986. Assessing the diversity and composition of restored vegetation. Can. J.
Bot. 64: 1344-1348.
Reed, P. B., Jr. 1988. National List of Plant Species That Occur hi Wetlands: 1988
Florida. Biological Report NERC-88/18.09. National Wetlands Inventory, U.S. Fish
and Wildlife Service. St. Petersburg, Florida.
Sherman, A.D., S.E. Gwin, and M.E. Kentula, in conjunction with M. Brown. 1991. Quality
Assurance Project Plan: Florida Wetlands Study. Companion document, Environmental
Research Laboratory-Corvallis, Oregon.
Wentworth, T.R., G.P. Johnson, and R.L. Kologiski. 1988. Designation of wetlands by
weighted averages of vegetation data: A preliminary evaluation. Water Resources
Bulletin, 24(2): 389-396.
37
-------
•a
^ Q
o 8 "53 •
^ ^ u &
•^ U
£ % %
S-a
liH CX
O V
3'
o o
------- |