Content uploaded by Mirela G Tulbure
Author content
All content in this area was uploaded by Mirela G Tulbure
Content may be subject to copyright.
ORIGINAL PAPER
Environmental Conditions Promoting Non-native
Phragmites australis Expansion in Great Lakes
Coastal Wetlands
Mirela G. Tulbure &Carol A. Johnston
Received: 24 August 2009 / Accepted: 15 February 2010 / Published online: 18 May 2010
#Society of Wetland Scientists 2010
Abstract The invasion and expansion of the non-native
Phragmites australis in Great Lakes coastal wetlands is of
increasing concern, but quantitative studies of the extent,
rate, and causes of invasion have been lacking. Here we
revisited 307 plots in 14 wetlands along the Great Lakes
coast in 2005 that had previously been sampled for
vegetation in 2001–2003. During the 2–4 years between
sample events, Phragmites occurred in 101 plots. Genetic
analysis revealed that none of the Phragmites samples
collected at the 14 wetlands belonged to the native
genotype. Decreases in water depth and bare soil area were
associated with the greatest increases in Phragmites cover.
Phragmites invasion was greater on Lakes Michigan,
Huron, and Erie than it was on Lake Ontario, and occurred
predominantly on sandy substrates. Soil water concentra-
tions of NO
3
-N, NH
3
-N, and soluble reactive P did not
differ significantly between plots with and without Phrag-
mites. Monitoring coastal wetlands where water level has
dropped and controlling Phragmites at early stages of
invasion are essential for maintaining healthy Great Lakes
coastal wetlands of high species diversity and wildlife
habitat. This becomes important as water levels in the Great
Lakes have reached extreme lows and are expected to
decline with future climate change.
Keywords Exotic species .Invasion .Invasive species .
Nutrients .Soil .Water level
Introduction
Phragmites australis (Cav.) Trin. ex Steud. (hereafter
Phragmites) has expanded tremendously in North America
with the introduction of the M Eurasian genotype, and has
become one of the most invasive plants in wetlands along
the Atlantic coast (Ailstock et al. 2001; Minchinton 2002;
Bart et al. 2006). The species has become a problem more
recently along the Great Lakes coast, where rapid Phrag-
mites expansion has been noted on Lake Erie (Wilcox et al.
2003; Ghioca-Robrecht et al. 2008) and Green Bay (Pengra
et al. 2007; Tulbure et al. 2007). Lacking the salinity
gradients that limit the seaward expansion of Phragmites on
ocean coasts (Meyerson et al. 2000; Bart and Hartman
2003), wetlands of the Great Lakes coast may be
particularly susceptible to future invasion by Phragmites.
Research is needed to understand the mechanisms and
factors that make coastal wetlands in this region susceptible
to Phragmites invasion.
Great Lakes coastal wetlands are dynamic, with average
annual water levels changing by a meter or more over
decadal time scales (Wilcox et al. 2007). These water level
fluctuations are a major factor controlling the composition
of native vegetation in Great Lakes coastal wetlands,
producing and maintaining plant species diversity by cycles
of drawdown and flooding (Keddy and Reznicek 1986;
Hudon 1997; Keddy 2000; Gathman et al. 2005). When
non-native aggressive taxa such as Phragmites germinate,
proliferate and form monotypic stands, the germination of
native species dependent on fluctuating water levels is
obstructed (Haslam 1972; Thiet 2002; Herrick and Wolf
M. G. Tulbure (*):C. A. Johnston
Department of Biology and Microbiology,
South Dakota State University,
Brookings, SD 57007, USA
e-mail: Mirela.Tulbure@sdstate.edu
Present Address:
M. G. Tulbure
Geographic Information Science Center of Excellence,
South Dakota State University,
Brookings, SD 57007, USA
Wetlands (2010) 30:577–587
DOI 10.1007/s13157-010-0054-6
2005). This reduces plant species diversity and wetland
functionality (Chambers et al. 1999; Galatowitsch et al.
1999).
Average annual lake levels on Lakes Michigan and
Huron were >0.5 m below the long-term average between
2000 and 2008 (http://www.glerl.noaa.gov/data/now/wlevels/
levels.html), which may have contributed to the recent
expansion of Phragmites on their coasts. Prolonged low
Great Lakes levels exposed unvegetated lake bottom sedi-
ments, providing a substrate for colonization by new plants,
such as invasive Phragmites (Pengra et al. 2007; Tulbure et
al. 2007). Repeat visits to a Green Bay coastal wetland in
2001 and 2004 showed a 100-fold increase in Phragmites
cover (Tulbure et al. 2007).UnlikethelakelevelofLake
Ontario, which has been partially regulated since 1958,
lakes Erie, Michigan, and Huron are not regulated and
their lake levels therefore respond more to climatic and
other large scale changes (Sellinger et al. 2008). Wetland
values such as diversity, primary productivity, and habitat
for wildlife and waterfowl will be affected if an increase in
frequency and duration of low water levels results from
future climate change (Mortsch 1998).
A number of factors have been shown to facilitate
Phragmites invasion, including land use changes in a
wetland’s watershed (King et al. 2007), increases in nitrate
and the concentration of other nutrients (Marks et al. 1994;
Bertness et al. 2002; Minchinton and Bertness 2003), the
presence of exposed mineral soil (Tulbure et al. 2007), and
alterations of the hydrologic regime by roads, dikes, and
ditches ( McNabb and Batterson 1991; Bart and Hartman
2000,2003;Herrick and Wolf 2005; Maheu-Giroux and de
Blois 2007; Johnston et al. 2008). The use of dikes is a
prevalent feature in coastal wetlands on the southwestern
shore of Lake Erie (Kroll and Gottgens 1997; Gottgens et
al. 1998).
Once it invades, Phragmites expands rapidly via
rhizomes and seeds (Haslam 1973; Bart and Hartman
2003) and becomes monodominant in a wetland, displacing
the native flora (Marks et al. 1994; Chambers et al. 1999;
Meyerson et al. 2000). Phragmites expands vegetatively by
horizontal rhizome growth at a rate of 1–2 m/yr (Haslam
1972), grows taller and has higher biomass than other
marsh species (Meyerson et al. 2000), outcompeting co-
occurring plant species by shading and litter mat formation
(Haslam 1973). The native subspecies Phragmites australis
subsp. americanus Saltonstall, P.M. Peterson & Soreng,
which is endemic to the Great Lakes, is considered to be
less aggressive than the non-native M Eurasian genotype
(Saltonstall et al. 2004).
Despite anecdotal accounts of recent Phragmites expan-
sion on the Great Lakes, quantitative studies of the rate and
extent of Phragmites invasion are lacking. The EPA-funded
Great Lakes Environmental Indicators (GLEI) project
collected data on plant species of Great Lakes coastal
wetlands during 2001–2003 (Johnston et al. 2007a). This
provided baseline vegetation data and a unique opportunity
to assess Phragmites shifts in time. In the present study, we
returned in 2005 to 14 Great Lakes coastal wetlands in the
Eastern Broadleaf Forest ecoprovince (Keys et al. 1995)
that had some Phragmites present when first sampled, and
quantified short-term changes in the extent of Phragmites
cover.
The overarching aim of this study was to determine what
characteristics caused coastal wetlands in the southern
Great Lakes region to be susceptible to Phragmites
invasion. Specific objectives were to: (1) determine the
nativity of Phragmites, (2) determine whether Phragmites
expanded since first sampled at the 14 revisited wetlands;
(3) identify natural and anthropogenic factors that influ-
enced invasive Phragmites success at these wetlands; (4)
assess whether short-term change in Phragmites cover
differed on different soil types, and whether nutrient
concentrations were different in Phragmites versus non-
Phragmites stands, and (5) examine whether the increase in
Phragmites cover was greater on Great Lakes wetlands in
which there was a drop in water levels between 1999 and
2001 (i.e., Lakes Michigan, Huron, Erie) than on the lake
where there was an increase in water levels between 1999
and 2001 (i.e., Lake Ontario).
Methods
Site Selection and Vegetation Sampling
The GLEI project sampled 35 Great Lakes coastal wetlands
within the Eastern Broadleaf Forest (EBF) ecoprovince,
which encompasses Lake Ontario, Lake Erie, and southern
Lakes Huron and Michigan (GLEI sites listed in Johnston
et al. 2007a). Of these wetlands, we revisited 14 of the 16
that contained Phragmites in 2001–2003 when they were
first sampled as part of the GLEI project; time constraints
prevented us from revisiting the other two (Fig. 1, Tables 1
and 2). We focused on the EBF region because the invasive
genotype was documented to occur in numerous coastal
wetlands of the southern Great Lakes but not the northern
Great Lakes (Saltonstall 2002). A total of 307 1-m × 1-m
plots were sampled at the 14 wetland sites and the number
of plots sampled per wetland ranged from 8 to 96
proportional to wetland size (Table 1).
Geographic coordinates of each 1-m × 1-m plot within
the wetland study sites had been recorded using a hand held
Garmin GPSMAP 76 unit (Garmin International Inc.,
Olathe, KS) during initial sampling in 2001–2003, and
were used to return to the same locations during the
summer of 2005. Plot locations were the average of
578 Wetlands (2010) 30:577–587
hundreds of readings recorded by a tripod-mounted GPS set
to continuously record as field researchers were collecting
vegetation data at each plot (Johnston et al. 2009a). All the
sites were herbaceous wetlands in level terrain, therefore
errors introduced by blocked or multipath signals due to
topography or canopy cover were minimal.
Plots were distributed along randomly placed transects
within areas mapped as emergent vegetation (Cowardin et
al. 1979) by national and state wetland inventories along
the Great Lakes. Transects were established with a
geographic information system (GIS) prior to field
campaigns, using a program called Sample (http://www.
quantdec.com/sample) to randomize transect placement.
Each transect intersected a randomly selected point
generated by the Sample program, and was oriented
along the perceived water depth gradient, extending from
open water to the upland boundary. Transect length and
target number of plots were determined in proportion to
the size of the wetland to be sampled (20 plots/60 ha,
minimum transect length= 40 m, minimum plots/site = 8).
Plot locations were established in the field by dividing
each transect into 20 m segments and randomly locating
a plot in each segment using a random number table
(Bourdaghs et al. 2006). Within each plot percent cover of
Phragmites was estimated visually according to modified
Braun-Blanquet cover class ranges (ASTM 1997): < 1%, 1
to<5%, 5 to < 25%, 25 to<50%, 50 to< 75%, 75 to 100%.
Prior to data analyses, cover classes were converted to the
midpoint percent cover of each class using the algebraic
mid-points of the six cover class ranges (0.5, 3.0, 37.5,
62.5, 87.5).
Leaf samples were collected from Phragmites stands at
each wetland revisited and identified as native or non-native
genotype using the genetic analysis described in (Tulbure
et al. 2007) and based on Saltonstall’s(2002)protocol.
Table 1 Comparison of Phragmites cover at 14 Great Lakes wetland study sites between first and second sampling. Invaded = plots where
Phragmites was not present in 2001, but expanded in 2005; Increased/decreased = plots where Phragmites was present in 2001 but increased/
decreased in cover by 2005; Remained unchanged = plots where Phragmites cover did not change from 2001 to 2005
Site name Lake Mean Phragmites
cover per plot (%)
at each site
Number of
samples (plots), n
Number of plots
with Phragmites
Number of plots where Phragmites
2001–03 2005 2001–03 2005 Invaded Increased Decreased Remained
unchanged
Kalamazoo River Michigan 0.03 0.65 96 1 1 1
White Feather Creek Huron 0.09 7.63 27 5 8 3 4 1
Neuman Road Huron 0.04 0.08 13 1 2 1 1
Blind Pass Huron 4.34 24.48 22 10 12 5 5 3 2
Wildfowl Bay Huron 8.07 24.00 22 10 12 5 4 3 3
Caseville Huron 35.63 78.44 8 8 8 6 2
Otter Creek Erie 12.73 21.88 24 7 6 1 3 2 2
Toledo Beach Erie 26.25 26.25 10 3 3 3
Bay Creek Erie 17.50 32.15 13 4 8 5 1 2 1
Little Lake Creek Erie 64.05 73.80 10 10 10 3 7
Kelly Doty Drain Erie 52.50 39.14 11 10 7 2 4 4
Presque Isle Erie 2.70 19.30 15 2 10 8 2
Braddock Bay Ontario 0.60 2.50 25 1 1 1
Fox Creek Ontario 0.27 0.05 11 1 1 1
Total number of plots (n) 307 73 89 28 32 15 26
Table 1 Comparison of Phragmites cover at 14 Great Lakes wetland
study sites between first and second sampling. Invaded = plots where
Phragmites was not present in 2001, but expanded in 2005; Increased/
decreased = plots where Phragmites was present in 2001 but
increased/ decreased in cover by 2005; Remained unchanged = plots
where Phragmites cover did not change from 2001 to 2005
Fig. 1 Location of the 14 wetland sites (black triangles) re-sampled in
2005
Wetlands (2010) 30:577–587 579
Abiotic Factors
Thirteen abiotic environmental factors were measured in
the field or derived from mapped data for each of the 14
study sites (Table 3). At each plot, vegetation data were
collected, water depth was measured, and bare soil area was
estimated using the same six cover class ranges used for
plants. Water depth was measured using a meter stick,
which was inserted until we felt resistance from bottom
sediments without applying pressure. The values obtained
in 2005 were subtracted from comparable values for water
depth and bare soil area measured initially by the GLEI
project to compute change in water depth (WaterDepth-
Diff) and bare soil (BareSoilDiff). The substrate at each
plot was examined to a depth of 30 cm below the litter
layer using a soil probe, and assigned to one of the
following broad categories: organic, sand, silt, clay.
“Organic”soils were those composed of organic soil
material (peat or muck) in a histic epipedon (Soil Survey
Staff 1999); undecomposed plant litter overlying the soil
surface was excluded when making this determination.
The texture of mineral soils (i.e., sand, silt, clay) was
determined by feel using standard field methods (Soil
Survey Staff 1951).
In 2005 we measured nitrate (NO
3
-N), ammonia (NH
3
-
N), and soluble reactive phosphorus (SRP) in soil water in
Phragmites versus non-Phragmites stands at each wetland
site. Phragmites stands were chosen that were 100%
covered with Phragmites and along or close to the initial
GLEI sample transect, but not necessarily coinciding with
it. Non-Phragmites stands were chosen that had no
Phragmites present and were at the same elevation and as
close as possible to the Phragmites stands so as to minimize
confounding differences due to edaphic conditions. We dug
holes using an auger in the plant rooting zone, and water
that accumulated in the holes was used for nutrient analysis.
All nutrients were measured in the field using a Digital
Hach PortableDR/890 Colorimeter (Hach Company, Love-
land, CO). Nutrient measurements were taken at three
locations in each stand type (Phragmites vs. non-Phrag-
mites). Nitrate was measured using the Chromotropic Acid
Method (estimated detection limit, EDL, of 0.3 mg/L NO
3
-
N); ammonia was measured using the Salicylate Method
(EDL of 1 mg/L NH
3
-N), and SRP using the Ascorbic Acid
Method (EDL of 0.05 mg/L PO
4
; Hach Company 2005).
We examined digital orthophotos taken 1–8 years prior
to the initial field work (USGS 2007) to determine if sites
were previously diked. The agriculture and urban indices
were developed by the GLEI project using methods
described by Danz et al. (2007), but were calculated for
watersheds draining to the specific wetlands studied
(Hollenhorst et al. 2007). The agricultural index was
derived from 21 variables characteristic of the major types
of stresses in a wetland’s watershed associated with
agricultural activities (e.g., nutrient runoff, fertilizers,
pesticide application, and erosion), and the urban index
was derived from 14 variables associated with human
population, road density, and developed land in the
Table 2 Description of water level at 14 wetland sites. Lake level changes based on July average monthly levels (see Table 4) except where
indicated. Negative numbers indicate declines in water levels
Site Lake Initial sample
year
Lake level
change, m
Diked? Number of plots where water level
increased decreased same
Kalamazoo River Michigan 2002 −0.14 N 4 18 74
White Feather Cr. Huron 2003 0.15 N 21 1 5
Neuman Road Huron 2003 0.15 N 9 0 4
Blind Pass Huron 2003 0.15 N 8 6 8
Wildfowl Bay Huron 2003 0.15 N 7 0 15
Caseville Huron 2003 0.15 N 1 2 5
Otter Creek Erie 2002 −0.02 N 4 9 11
Toledo Beach Erie 2002 −0.02 Y 7 1 2
Bay Creek Erie 2002 −0.02 N 0 4 9
Little Lake Creek Erie 2003 0.04 Y 0 3 7
Kelly Doty Drain Erie 2003 0.04 Y 3 2 6
Presque Isle
a
Erie 2003 −0.04 N 0 8 7
Braddock Bay
ab
Ontario 2002 −0.53 N 0 18 7
Fox Creek
a
Ontario 2001 −0.17 N 0 5 6
a
2005 lake level from August used to correspond with final sampling date
b
lake level from June used to correspond with initial sampling date
580 Wetlands (2010) 30:577–587
watershed (Danz et al. 2007). Principal component analysis
was carried out for each category of variables, and the
resulting principal components were standardized so that
the mean equaled 0 and the standard deviation equaled 1.
Index values for all GLEI EBF wetland study sites ranged
from 0.11 (least stress) to 1.14 (most stress) for the
agricultural index and from −2.09 to 1.82 for the urban
index. The values of the agricultural index at our revisit
sites ranged from 0.25 to 1.14, while the urban index
ranged from −1.11 to 1.82. Areal proportions of row crop
and development within watersheds around the National
Wetlands Inventory boundary were computed to account
for anthropogenic disturbance at different spatial scales
(Brazner et al. 2007).
Data Analysis
To gain insight into the drivers of change in Phragmites
cover between the two sampling events, we used mixed
models incorporating both fixed and random effect varia-
bles with PROC MIXED, SAS Version 9.1 (SAS Institute
Inc. 2001), with plots nested within wetland sites. We used
13 explanatory variables as fixed effects, which were
hypothesized to influence Phragmites change (Table 3).
We assessed the degree of multicollinarity among the
continuous explanatory variables by computing variance
inflation factors (VIF, Neter et al. 1990). Five variables
(agriculture PC, watershed to wetland area, watershed area,
row crops, and development in the watershed) with the
highest VIFs were eliminated from the model to reduce
collinearity, which resulted in VIFs <= 10 in the new mixed
model.
A one-way ANOVA followed by Tukey’s HSD compar-
ison test was used to analyze: a) initial Phragmites cover on
four different soil types (i.e., sand, silt, clay, and organic); b)
change in Phragmites cover at wetland sites located on
Great Lakes in which there was a drop in water levels
between 1999 and 2001 (i.e., Lakes Michigan/Huron, and
Erie) versus lakes where there was an increase in water
levels between 1999 and 2001 (i.e., Lake Ontario). The
Table 3 Coefficients and p-values from the mixed model of difference in Phragmites cover as a function of explanatory variables
Variables Description Unit of
measure
Type
a
Scale Reference Mixed model
with 13 variables
Mixed model
with 8 variables
Estimate p-value VIF Estimate p-value VIF
Intercept −55.92 30.54
Lake Great Lakes (Michigan-
Huron or Erie-Ontario)
unitless C see text 0.98 0.02
Substrate clay,organic, sand, silt unitless C plot see text 0.67 0.57
WaterDepthDiff difference in water depth
between first and second
sampling
cm N plot see text −0.96 0.19 1.33 1.31 0.05 1.07
BareSoilDiff difference in bare soil area
between second and first
sampling
% N plot see text −0.26 0.12 1.26 −0.3 0.05 1.04
UrbanIndex urban PC unitless N watershed Danz et al.
2007
82.58 0.12 7.69 4.65 0.36 1.46
AgIndex agriculture PC unitless N watershed Danz et al.
2007
11.04 0.53 10.25
Nitrogen field-measured average
inorganic nitrogen
mg/L N wetland see text 33.96 0.18 9.34 12.06 0.13 1.76
Phosphorus field-measured soluble
reactive phosphorus
mg/L N wetland see text −3.67 0.87 2.61 1.39 0.89 1.27
Ratio watershed to wetland
area ratio
unitless N wetland Brazner et al.
2007
0.21 0.13 13.34
watershed area ha N watershed Brazner et al.
2007
0.01 0.15 13.78
WetlandArea wetland area ha N wetland Brazner et al.
2007
0.01 0.53 1.62 0.01 0.19 1.26
RowWatershed row crops areal
fraction
N watershed Brazner et al.
2007
76.45 0.56 38.29
UrbanWatershed development areal
fraction
N watershed Brazner et al.
2007
−82.93 0.37 28.61
Variables with bolded VIF values were eliminated from the initial mixed model to reduce multicollinearity
a
The type of variable is denoted by N = numerical, C = categorical
Wetlands (2010) 30:577–587 581
years 1999–2001 were chosen as a surrogate for exposure
of bare soil during the first year of sampling. We used
Wilcoxon non-parametric two-sample tests to examine
differences in nutrients between Phragmites versus non-
Phragmites stands. All tests were conducted in SAS ® 9.1
(SAS Institute Inc. 2001). All data were arcsine transformed
prior to data analysis (none of the data sets was normally
distributed).
Results
Phragmites Nativity and Expansion
Phragmites occurred in about one-third of the 307 plots
sampled at the 14 wetland sites. PCR/RFLP genetic
analysis of leaf samples revealed that none of the
Phragmites sampled at the 14 wetland sites belonged
to the native genotype. Out of the 307 plots revisited at
14 wetlands along the Great Lakes, Phragmites occurred
in 101 plots. At the other 206 plots, Phragmites was
absent during the first and second sampling events. Out of
the 101 plots that had Phragmites, Phragmites invaded 28
plots, increased cover in 32 plots, disappeared or
decreased cover in 15 plots, and remained unchanged in
26 plots (Table 1).
Eight sites exhibited increases in Phragmites cover of
7.5 percentage points or more. This ranged from 7.54
percentage points at White Feather Creek wetland, which
had 0.09% mean Phragmites cover (as mean plot value per
site) in 2003 and increased to 7.63% by 2005. A high
increase of 42.8 percentage points occurred at Caseville
wetland, which started with a mean Phragmites cover of
35.63% in 2003 and reached 78.44% by 2005. The change
is even more dramatic given that it took place only 2 or
3 years after the initial sample date (Table 1). One site
(Kelly Doty Drain) first sampled in 2003 exhibited a 13.4%
decrease in Phragmites cover. The remaining five sites
(Kalamazoo River, Neuman Road, Toledo Beach, Braddock
Bay, and Fox Creek) remained essentially unchanged
between first and second sampling times (Table 1). Four
out of the five Lake Huron sites experienced large increases
in Phragmites, as did four sites on Lake Erie, but the Lake
Michigan and Lake Ontario sites did not change substan-
tially. The increase in Phragmites was due either to
invasion of new plots or increase in cover at plots where
it already occurred (Table 1). At two sites, Bay Creek and
Presque Isle, the number of plots newly invaded by
Phragmites was greater than the number of plots where it
merely increased in cover (Table 1). Phragmites disap-
peared completely from most of the plots that exhibited a
decrease in Phragmites cover between the initial and 2005
sampling (12 out of 15 plots).
Water depth changes in the plots sampled were generally
consistent with the overall lake level trends from initial to
2005 sampling: increases occurred primarily at sites where
the lake level was at least 10 cm higher in 2005 than during
the initial sampling (e.g., Lake Huron), and decreases
occurred primarily at sites where the lake level was at least
10 cm lower in 2005 (e.g., Lake Ontario, Table 2; USCOE,
http://www.lre.usace.army.mil/greatlakes/). All plots that
lacked water standing above the soil surface were assigned
a water depth of 0, regardless of depth to the water table
beneath the soil surface, which explains the predominance
of “same”values recorded. Plot-scale water level decreases
in Little Lake Creek despite lake level increases are
probably due to artificial water level control by dikes. The
relatively few undiked plots in which the plot-scale change
in water depth is inconsistent with the lake level change
may be due to changes in bottom configuration (e.g.,
dredging, siltation).
Natural and Anthropogenic Factors that Influenced
Phragmites Success
The VIF-adjusted, 8-variable mixed model related change
in Phragmites cover with difference in water depth,
difference in bare soil area, urban PC, nitrogen, phospho-
rus, wetland area, lake, and substrate (Table 3). Three of
these variables were individually statistically significant:
difference in water depth, difference in bare soil area, and
lake. An increase in Phragmites cover was associated with
decreasing water depth and less bare soil, and Phragmites
cover increased more on Lake Michigan-Huron than on
Lake Erie-Ontario (t-value=2.35, df = 74, p= 0.02). Of these
variables, the lake and decrease in water depth are
environmental factors promoting Phragmites invasion,
whereas the decrease in bare soil area is a consequence of
that invasion. The greatest decrease in bare soil occurred at
plots where Phragmites increased in cover (Fig. 2a),
suggesting that the newly exposed substrate was colonized
by Phragmites. Plots invaded by Phragmites experienced a
decrease in measured water depth from first to second
sampling, plots where Phragmites increased in cover had
relatively stable water levels, plots exhibiting a decrease in
Phragmites cover had water depth increases averaging
7 cm, whereas other plots experienced little change in water
depth over the same time period (Fig. 2b). Even though
Urban Index was not a significant predictor of change in
Phragmites cover, Phragmites first appeared where the
Urban Index was low, but it increased where it was
dramatically higher (Fig. 2d). This might suggest that
Phragmites expansion is facilitated by disturbance and
urban land uses along wetland-terrestrial borders, as
previous studies have found (Bertness et al. 2002;Minchinton
and Bertness 2003;Kingetal.2007).
582 Wetlands (2010) 30:577–587
Substrate and Nutrients
Phragmites cover was significantly greater on sand than on
clay and organic soils (F=5.19, df= 4, p< 0.01), with the
highest average cover of 13.5% on sand, 6.2% on silt, 2.2%
on clay, and 1.7% on organic soil. Although initial substrate
type was not a significant predictor of Phragmites cover
change, 59% of the plots where Phragmites occurred had
sandy substrate, whereas 52% of the plots where Phrag-
mites did not occur had organic soils (Fig. 3).
Average NO
3
-N concentration was 0.28 mg/L in
Phragmites stands compared to 0.32 mg/L in non-Phrag-
mites stands; both values were close to the detection
limit, and their difference was not statistically significant
(χ
2
=0.65, df=1, p=0.42). Average NH
3
-N concentrations
were also close to the detection limit, and no statistically
significant difference was noted between the two groups
(χ
2
=0.22, df=1, p=0.64) in Phragmites (1.05 mg/L)
versus non-Phragmites (1.02 mg/L) stands. Average
concentrations of SRP were well above detection limits,
but no statistically significant difference was observed in
SRP concentrations (χ
2
=0.33, df=1, p=0.56) between
Phragmites (0.3 mg/L) versus non-Phragmites stands
(0.43 mg/L) when data were analyzed together for all
fourteen sites.
Antecedent Lake Levels
Between 1999 and 2001, prior to the initial sampling event,
average July lake levels in Lakes Erie and Michigan/Huron
dropped 19 and 35 cm, respectively, to levels that were 28
to 55 cm below average (Table 4). The rapid water level
decline exposed large areas of lake bottom along their
shorelines. Lake Ontario experienced no such decrease; its
lake level increased slightly between 1999 and 2001 and
was close to the long-term average in 2001. Phragmites
cover increased an average of only 1.25% in the Lake
Ontario sites, as opposed to 7.4% in the sites on Lakes
Michigan, Huron, and Erie (F=2.74, df = 1, p= 0.09).
Fig. 3 Substrate type of Phragmites plots on the first sampling event
Fig. 2 Plot environmental
conditions in relation to Phrag-
mites cover change categories
(A = invaded, B = increased,
C = decreased, D = same, and
E = was zero). Mean values and
standard errors for (a) Change in
bare soil area (%). Negative
values indicate a decrease in
bare soil area between first
sampling date and 2005, (b)
Agricultural index, (c) Change
in water depth (cm). Negative
values indicate a decrease in
water levels between the first
sampling date and 2005. (d)
Urban index
Wetlands (2010) 30:577–587 583
Discussion
Our research demonstrated that Phragmites invasion can
occur very rapidly: Phragmites cover greatly expanded at
four of our five Lake Huron sites in only 2 years (all first
sampled in 2003). All of the Lake Huron sites are in
Saginaw Bay, which is characterized by lakebed wetlands
with gently sloping bathymetry (Burton et al. 2002; Mink
and Albert 2002). All of our Lake Huron sites were open-
coast wetlands, with emergent vegetation relatively
exposed to wave action (Johnston et al. 2007a). Wetlands
with direct exposure to wave action usually have sandy
soils, which is conducive to Phragmites expansion. The
gentle slope of Saginaw Bay plus the presence of mineral
substrates creates very fertile sand flats when lake levels
drop. In healthy coastal wetlands these newly exposed flats
would be colonized by native, early successional species that
are adapted to these cycles, such as Bidens cernua L. (Wilcox
et al. 2007). However, the newly exposed flats created by
rapidly receding water levels, as occurred in Lakes Michigan
and Huron, provided excellent substrate for introduced
Phragmites colonization. In other regions, Phragmites seed-
lings can germinate on exposed bottoms without standing
water, but cannot colonize submerged or densely vegetated
substrates (Weisner and Ekstam 1993).
Neuman Road is the only Lake Huron site we revisited
that did not experience increase in Phragmites cover. In
contrast, the nearby White Feather Creek site had increases
in Phragmites cover despite comparable agricultural index
in the watershed, water level change, and amount of
change in bare soil. One of the major differences between
the two sites is the small watershed area of Neuman Road,
which might result in lower total nutrient inputs from the
watershed.
Previous studies have suggested that once it has
invaded, Phragmites alters soil properties and nutrient
pools (Windham and Lathrop 1999;Ehrenfeld2003;
Windham and Meyerson 2003). In the present study we
did not detect significant differences in nutrients from soil
water of vegetation (total nitrogen and SRP) between
Phragmites and non-Phragmites stands. This could be due
to the fact that the invasion process is in its early stages
(some of the sites were invaded in periods as short as
2 years) and it is too early to detect any changes.
On Lake Erie, Phragmites increased in cover at Presque
Isle, Otter Creek, Bay Creek, and Little Lake Creek,
remained unchanged at Toledo Beach, and decreased in
cover at Kelly Doty Drain. All of our Lake Erie sites with
the exception of Presque Isle were located at the western
end of Lake Erie, which has been altered by construction of
dikes for wetland water level control (Johnston et al.
2007b). These structures disrupt the natural flow and water
level fluctuations and may have influenced changes in
Phragmites cover at these sites.
Phragmites did not change in cover at our sites located
on Lake Ontario. This is one of the Great Lakes where
water level did not drop between 1999 and 2001, and our
findings show that Phragmites expanded more at Great
Lakes coastal wetlands where there was a drop in water
levels early in the decade. Lake Ontario sites had primarily
organic and clay soils (Johnston et al. 2009b). The GLEI
project showed that the invasive Typha Xglauca is an
indicator species of organic soil and a common species on
Lake Ontario wetlands (Johnston et al. 2009b; Vaccaro et
al. 2009). Typha spp. could also generate their own organic
matter substrate, as they are known to form floating mats of
organic matter (Hogg and Wein 1988). The competition
between the two species might prevent Phragmites from
invading those wetlands.
Although Phragmites tolerates most soil conditions
(Global Invasive Species Database), we found that Phrag-
mites occurred predominantly on sandy soils (60% of the
plots where it occurred), similar to our previous study
(Tulbure et al. 2007). This could also be due to the fact that
Phragmites invades newly exposed substrates where sandy
soils are prevalent.
Our focus on wetlands that already contained Phrag-
mites in 2001–2003 was intentional, because those wet-
lands would be expected to experience more rapid invasion
by Phragmites than wetlands lacking an internal source of
Phragmites propagules. However, the exclusion of Phrag-
mites-free wetlands from our sample design means that
these findings cannot be extrapolated to all Great Lakes
coastal wetlands without further research.
The fact that none of the samples we analyzed in this
present study came from native Phragmites populations
suggests that non-native genotypes are common in wetlands
Table 4 July Great Lakes water level from 1999–2006 (USCOE,
http://www.lre.usace.army.mil/greatlakes/)
July water level (meters above IGLD 1983)
Year Michigan/Huron Erie Ontario
1999 176.40 174.23 74.81
2000 176.13 174.27 75.24
2001 176.05 174.04 74.97
2002 176.33 174.25 75.19
2003 176.04 174.19 75.05
2004 176.37 174.35 75.09
2005 176.19 174.23 74.95
2006 176.14 174.29 74.98
Record High 177.39 175.03 75.66
Record Low 175.78 173.45 74.14
Long term average 176.60 174.32 74.99
584 Wetlands (2010) 30:577–587
of the southern Great Lakes coast. The non-native genotype
displays an aggressive behavior and has a greater ability to
ventilate the root system with atmospheric oxygen (Tulbure
2008). To our knowledge, there is no evidence to suggest
that the native Phragmites inhabited the wetlands that were
sampled in this study. However, there still are native
Phragmites populations in the region, at Bark Bay (Lynch
and Saltonstall 2002) and other Lake Superior wetlands
(Natalie Wright, Univ. of Minnesota, personal communica-
tion),andinOhio(JohnMack,OhioEPA,personal
communication). This underlies the need to identify the
origins of populations especially in areas where they are
sympatric (Saltonstall 2003).
Mixed model results showed that a combination of plot
and site level variables were the most useful predictors of
Phragmites change. Decreases in bare soil at plots with
high increase in Phragmites suggest that the bare substrate
was colonized by Phragmites rather than other species.
Change in water depth was another predictor of Phragmites
increase in cover, with a greater increase in Phragmites
cover at plots where there was a decrease in water level.
Phragmites is generally found in shallower water or areas
not permanently inundated, but it is also favored by wide
water-level fluctuations and is known to survive in water as
deep as 2 m (Squires and van der Valk 1992; Herrick and
Wolf 2005).
Phragmites invades and spreads when water levels drop
and temperatures rise in Lake Erie (Wilcox et al. 2003) and
Lake Michigan-Huron (Pengra et al. 2007; Tulbure et al.
2007). Recently, Brisson et al. (2008) found evidence of
sexual reproduction in the non-native Phragmites at sites
with newly exposed substrate of eastern Canada. The
authors attributed the phenomenon to the recent climate
change towards warmer years (Brisson et al. 2008). The
substrate exposed by declining water levels in the Great
Lakes provides new germination opportunities for rapid
colonization by invasive wetland plant species. Under
most climate models Great Lakes water levels are
projectedtodecline(Chaoetal.1999). Stream runoff will
also drop (International Joint Commission 2003). Water
level in Lake Michigan-Huron Basin is anticipated to drop
by as much 1.38 m due to decreased precipitation and
increased air temperature and evapotranspiration (Lofgren
et al. 2002). The frequency and duration of low water
levels could increase, dropping water levels below historic
lows (International Joint Commission 2003). Given these
projections it is very likely that Phragmites is going to
expand and thrive at most Great Lakes coastal wetlands
where water levels drop, adding to the multiple stresses
that these ecosystems are facing. Once established,
Phragmites populations cause biodiversity loss and are
extremely difficult to eradicate (Havens et al. 1997).
Monitoring coastal wetlands where water level has
dropped and controlling Phragmites at early stages of
invasion are essential for maintaining healthy wetlands
along the Great Lakes coast. Our work underscores the
need to collect baseline vegetation data and revisit sites
often to monitor Phragmites invasion.
In contrast to previous studies of Phragmites invasion in
individual wetlands, this study is, to our knowledge, the
first to investigate rates and causes of Phragmites invasion
in multiple Great Lakes wetlands spanning a large range of
geographic and abiotic conditions. Understanding how
natural and anthropogenic abiotic factors drive changes in
coastal wetlands is important and can help managers focus
their efforts in areas where they are needed the most.
Documenting vegetation shifts in time is especially impor-
tant in dynamic systems such as Great Lakes coastal
wetlands that experience water level fluctuations and
changes in emergent vegetation.
Acknowledgments We thank the associate editor and two anony-
mous reviewers for helpful comments on an earlier draft. This research
has been supported by a grant from the United States Environmental
Protection Agency’s Science to Achieve Results Estuarine and Great
Lakes (EaGLe) program through funding to the Great Lakes
Environmental Indicators Project, US EPA Agreement EPA/R-
828675. Although the research described in this article has been
funded wholly or in part by the U.S. Environmental Protection
Agency, it has not been subjected to the Agency’s required peer and
policy review and therefore does not necessarily reflect the views of
the Agency and no official endorsement should be inferred. We thank
Dr. Donald Auger for help with the genetic analysis. We thank Heidi
Walking, Lynn Vaccaro, Christin Frieswyk DeJong, Michael Aho,
Kathy Bailey Boomer, Michael Bourdaghs, K. Cappillino, Randy
Clark, Spencer Cronk, D. James, Angie Marsh, and Cindy Williams
for field assistance.
References
Ailstock MS, Norman CM, Bushmann PJ (2001) Common reed
Phragmites australis: control and effects upon biodiversity in
freshwater nontidal wetlands. Restoration Ecology 9:49–9
ASTM (1997) Standard guide for sampling terrestrial and wetlands
vegetation. ASTM International, West Conshohocken
Bart D, Hartman JM (2000) Environmental determinants of Phrag-
mites australis expansion in a New Jersey salt marsh: an
experimental approach. Oikos 89:59–69
Bart D, Hartman JM (2003) The role of large rhizome dispersal and
low salinity windows in the establishment of common reed,
Phragmites australis, in salt marshes: New links to human
activities. Estuaries 26:436–443
Bart D, Burdick D, Chambers R, Hartman J (2006) Human Facilitation
of Phragmites australis Invasions in Tidal Marshes: A Review
and Synthesis. Wetlands Ecology and Management 14:53–65
Bertness MD, Ewanchuk PJ, Silliman BR (2002) Anthropogenic
Modification of New England Salt Marsh Landscapes. Proceed-
ings of the National Academy of Sciences of the United States of
America 99:1395–1398
Bourdaghs M, Johnston CA, Regal RR (2006) Properties and
performance of the Floristic Quality Index in Great Lakes coastal
wetlands. Wetlands 26:718–735
Wetlands (2010) 30:577–587 585
Brazner JC, Danz NP, Niemi GJ, Regal RR, Trebitz AS, Howe RW,
Hanowski JM, Johnson LB, Ciborowski JJH, Johnston CA,
Reavie ED, Brady VJ, Sgro GV (2007) Evaluation of geographic,
geomorphic and human influences on Great Lakes wetland
indicators: a multi-assemblage approach. Ecological Indicators
7:610–635
Brisson J, Paradis E, Bellavance ME (2008) Evidence of sexual
reproduction in the invasive common reed (Phragmites Australis
subsp Australis; Poaceae) in eastern Canada: A possible
consequence of global warming? Rhodora 110:225–230
Burton TM, Stricker CA, Uzarski DG (2002) Effects of plant
community composition and exposure to wave action on
invertebrate habitat use of Lake Huron coastal wetlands. Lakes
& Reservoirs: Research & Management 7:255–269
Chambers RM, Meyerson LA, Saltonstall K (1999) Expansion of
Phragmites australis into tidal wetlands of North America.
Aquatic Botany 64:261–273
Chao PT, Hobbs BF, Venkatesh BN (1999) How climate uncertainty
should be included in Great Lakes management: modeling
workshop results. Journal of the American Water Resources
Association 35:1485–1497
Company H (2005) Hach DR/890 Colorimeter Procedures Manual.
Cowardin LM, Carter V, Golet C, LaRoe ET (1979) Classification of
wetlands and deepwater habitats of the United States. U.S.
Department of the Interior, Fish and Wildlife Service, Wash-
ington, D.C
Danz NP, Niemi GJ, Regal RR, Hollenhorst T, Johnson LB, Hanowski
JM, Axler RP, Ciborowski JJH, Hrabik T, Brady VJ, Kelly JR,
Morrice JA, Brazner JC, Howe RW, Johnston CA, Host GE
(2007) Integrated measures of anthropogenic stress in the US
Great Lakes basin. Environmental Management 39:631–647
Ehrenfeld JG (2003) Effects of exotic plant invasions on soil nutrient
cycling processes. Ecosystems 6:503–523
Galatowitsch SM, Anderson NO, Ascher PD (1999) Invasiveness in
wetland plants in temperate North America. Wetlands 19:733–755
Gathman JP, Albert DA, Burton TM (2005) Rapid plant community
response to a water level peak in northern Lake Huron coastal
wetlands. Journal of Great Lakes Research 31:160–170
Ghioca-Robrecht DM, Johnston CA, Tulbure MG (2008) Assessing
The Use Of Multiseason Quickbird Imagery For Mapping
Invasive Species In A Lake Erie Coastal Marsh. Wetlands
28:1028–1039
Gottgens JF, Swartz BP, Kroll RW, Eboch M (1998) Long-term GIS-
based records of habitat changes in a Lake Erie coastal marsh.
Wetlands Ecology and Management 6:5–17
Haslam SM (1972) Phragmites Communis Trin. (Arundo Phragmites
L.,? Phragmites Australis (Cav.) Trin. ex Steudel). Journal of
Ecology 60:585–610
Haslam SM (1973) Some aspects of the life history and autecology of
Phragmites communis Trin. A review. Polskie Archiwum Hydro-
biologii 20:79–100
Havens KJ, Priest WI, Berquist H (1997) Investigation and long-term
monitoring of Phragmites australis within Virginia’s constructed
wetland sites. Environmental Management 21:599–605
Herrick BM, Wolf AT (2005) Invasive plant species in diked vs.
undiked Great Lakes wetlands. Journal of Great Lakes Research
31:277–287
Hogg EH, Wein RW (1988) The Contribution of Typha Components
to Floating Mat Buoyancy. Ecology 69:1025–1031
Hollenhorst TP, Brown TN, Johnson LB, Ciborowski JJH, Host GE
(2007) Methods for generating multi-scale watershed delineations
for indicator development in great lake coastal ecosystems.
Journal of Great Lakes Research 33(SI3):13–26
Hudon C (1997) Impact of water level fluctuations on St. Lawrence
River aquatic vegetation. Canadian Journal of Fisheries and
Aquatic Sciences 54:2853–2865
International Joint Commission 2003. Climate change and water
quality in the Great Lakes Region. Risks, opportunities, and
responses. Great Lakes Quality Board of the International Joint
Commission
Johnston CA, Bedford BL, Bourdaghs M, Brown T, Frieswyk C,
Tulbure MG, Vaccaro L, Zedler JB (2007a) Plant species
indicators of physical environment in Great Lakes coastal
wetlands. Journal of Great Lakes Research 33(SI3):106–
124
Johnston CA, Watson T, Wolter PT (2007b) Sixty-three years of land
alteration in Erie Township. Journal of Great Lakes Research
33(SI3):253–268
Johnston CA, Ghioca DM, Tulbure MG, Bedford BL, Bourdaghs M,
Frieswyk CB, Vaccaro L, Zedler JB (2008) Partitioning vegeta-
tion response to anthropogenic stress to develop multi-taxa
wetland indicators. Ecological Applications 18:983–1001
Johnston CA, Brown T, Hollenhorst T, Wolter P, Danz N, Niemi N
(2009a) GIS in Support of Ecological Indicator Development.
Manual of Geographic Information Systems, pp. 1095–1113.
American Society for Photogrammetry and Remote Sensing,
Bethesda, MD
Johnston CA, Zedler JB, Tulbure MG, Frieswyk CB, Bedford BL,
Vaccaro L (2009b) A unifying approach for evaluating the
condition of wetland plant communities and identifying related
stressors. Ecological Applications 19:1739–1757
Keddy PA (2000) Wetland ecology: Principles and conservation.
Cambridge University Press, Cambridge
Keddy PA, Reznicek AA (1986) Great Lakes vegetation dynamics: the
role of fluctuating water levels and buried seeds. Journal of Great
Lakes Research 12:25–36
Keys JE, Carpenter CA, Hooks SL, Koeneg FG, McNab WH, Russell
W, Smith ML (1995) Ecological units of the eastern United
States: first approximation.Map (scale 1:3,500,000). Technical
Publication R8-TP 21 U.S. Department of Agriculture, Forest
Service, Atlanta, Georgia, USA
King RS, Deluca WV, Whigham DF, Marra PP (2007) Threshold
effects of coastal urbanization on Phragmites australis (common
reed) abundance and foliar nitrogen in Chesapeake Bay. Estuaries
and Coasts 30:469–481
Kroll RW, Gottgens JF (1997) Wild rice to rip rap: 120 years of
habitat changes and management of a Lake Erie coastal wetland.
In Trans. 62nd N. Am. Wild. and Nat. Resour.Conf, pp. 490–500
Lofgren BM, Quinn FH, Clites AH, Assel RA, Eberhardt AJ,
Luukkonen CL (2002) Evaluation of potential impacts on Great
Lakes water resources based on climate scenarios of two GCMs.
Journal of Great Lakes Research 28:537–554
Lynch EA, Saltonstall K (2002) Paleoecological and genetic analyses
provide evidence for recent colonization of native Phragmites
australis populations in a lake superior wetland. Wetlands
22:637–646
Maheu-Giroux M, de Blois S (2007) Landscape ecology of Phragmites
australis invasion in networks of linear wetlands. Landscape
Ecology 22:285–301
Marks M, Lapin B, Randall J (1994) Phragmites australis (Phrag-
mites communis): threats, management and monitoring. Natural
Areas Journal 14:285–294
McNabb CD, Batterson TR (1991) Occurrence of the common reed,
Phragmites australis, along roadsides in Lower Michigan.
Michigan Academician 23:211–220
Meyerson LA, Saltonstall K, Windham L, Kiviat E, Findlay S (2000)
A comparison of Phragmites australis in freshwater and brackish
marsh environments in North America. Wetlands Ecology and
Management 8:89–103
Minchinton TE (2002) Precipitation during El Nino correlates with
increasing spread of Phragmites australis in New England, USA,
coastal marshes. Marine Ecology-Progress Series 242:305–309
586 Wetlands (2010) 30:577–587
Minchinton TE, Bertness MD (2003) Disturbance-mediated competi-
tion and the spread of Phragmites australis in a coastal marsh.
Ecological Applications 13:1400–1416
Mink LD, Albert DA (2002) Great lakes coastal wetlands: Abiotic and
floristic characterization. http://www.epa.gov/glnpo/ecopage/
wetlands/glc/ Accessed 24 Aug 2008
Mortsch LD (1998) Assessing the impact of climate change on the
Great Lakes shoreline wetlands. Climatic Change 40:391–416
Neter J, Wasserman W, Kutner MH (1990) Applied linear statistical
models: regression, analysis of variance, and experimental
designs. Irwin, Homewood
Pengra BW, Johnston CA, Loveland TR (2007) Mapping an invasive
plant, Phragmites australis, in coastal wetlands using the EO-1
Hyperion hyperspectral sensor. Remote Sensing of Environment
108:74–81
Saltonstall K (2002) Cryptic invasion by a non-native genotype of the
common reed, Phragmites australis, into North America.
Proceedings of the National Academy of Sciences of the United
States of America 99:2445–2449
Saltonstall K (2003) Genetic variation among north American
Populations of Phragmites australis: implications for manage-
ment. Estuaries 26:444–451
Saltonstall K, Peterson PM, Soreng R (2004) Recognition of
Phragmites australis subsp. americanus (Poaceae: Arundinoi-
deae) in North America: evidence from morphological and
genetic analyses. Sida 21:683–692
SAS Institute Inc (2001) SAS. Cary, North Carolina
Sellinger CE, Stow CA, Lamon EC, Qian SS (2008) Recent water
level declines in the Lake Michigan-Huron system. Environmen-
tal Science & Technology 42:367–373
Soil Survey Staff (1951) Soil Survey Manual. Agricultural Handbook
18. United States Department of Agriculture.Soil Conservation
Service, Washington D.C., USA
Soil Survey Staff (1999) Soil taxonomy. United States Department of
Agriculture, Washington D.C
Squires L, van der Valk AG (1992) Depth tolerance of emergents.
Canadian Journal of Botany
Thiet RK (2002) Diversity comparisons between diked and undiked
coastal freshwater marshes on Lake Erie during a high-water
year. Journal of Great Lakes Research 28:285–298
Tulbure MG (2008) Invasion, environmental controls, and ecosystem
feedbacks of Phragmites australis in coastal wetlands. Disserta-
tion, South Dakota State University
Tulbure MG, Johnston CA, Auger DL (2007) Rapid invasion of
a Great Lakes coastal wetland by non-native Phragmites
australis and Typ h a . Journal of Great Lakes Research 33
(SI3):269–279
USGS (2007) Photo Finder. URL: http://edcsns17.cr.usgs.gov/Earth
Explorer/?dataset_name=NAPP [last checked May 2010]
Vaccaro LE, Bedford BL, Johnston CA (2009) Litter accumulation
promotes the dominance of invasive species of cattails (Typha
spp.) in Great Lakes wetlands. Wetlands 29:1036–1048
Weisner SEB, Ekstam B (1993) Influence of germination time on
performance of Phragmites australis on temporarily exposed
bottoms—implications for the colonization of lake beds. Aquatic
Botany 45:107–118
Wilcox KL, Petrie SA, Maynard LA, Meyer SW (2003) Historical
distribution and abundance of Phragmites australis at Long
Point, Lake Erie, Ontario. Journal of Great Lakes Research
29:664–680
Wilcox DA, Thompson TA, Booth RK, Nicholas JR (2007) Lake-level
variability and water availability in the Great Lakes: U.S.
Geological Survey Circular 1311, p 25
Windham L, Lathrop RG (1999) Effects of Phragmites australis
(common reed) invasion on aboveground biomass and soil
properties in brackish tidal marsh of the Mullica River, New
Jersey. Estuaries 22:927–935
Windham L, Meyerson LA (2003) Effects of common reed (Phrag-
mites australis) expansions on nitrogen dynamics of tidal
marshes of the northeastern US. Estuaries 26:452–464
Wetlands (2010) 30:577–587 587
