Torrey Botanical Society

advertisement
Torrey Botanical Society
Plant Community Patterns of Low-Gradient Forested Floodplains in a New Jersey Urban
Landscape
Author(s): Myla F. J. Aronson, Colleen A. Hatfield, Jean Marie Hartman
Source: Journal of the Torrey Botanical Society, Vol. 131, No. 3 (Jul. - Sep., 2004), pp. 232-242
Published by: Torrey Botanical Society
Stable URL: http://www.jstor.org/stable/4126953
Accessed: 01/03/2010 15:41
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=tbs.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Torrey Botanical Society is collaborating with JSTOR to digitize, preserve and extend access to Journal of the
Torrey Botanical Society.
http://www.jstor.org
Journal of the Torrey Botanical Society 131(3), 2004, pp. 232-242
Plant community patterns of low-gradient forested
floodplains in a New Jersey urban landscape1
Myla F. J. Aronson2,3, Colleen A. Hatfield, and Jean Marie Hartman
GraduateProgramin Ecology and Evolution,RutgersUniversity,New Brunswick,NJ 08901-1582
AND J. M. HARTMAN. (Graduate Program in Ecology and Evolution,
ARONSON,M. E J., C. A. HATFIELD,
RutgersUniversity,New Brunswick,NJ 08901-1582). Plantcommunitypatternsof low-gradientforestedfloodplains in a New Jerseyurbanlandscape.J. TorreyBot. Soc. 131: 232-242. 2004.-This study characterizedthe
vegetationof floodplainforests along a 66 km stretchof the upperPassaic River in northernNew Jersey,USA.
Althoughthe study wetlandslie in a highly disturbedregion of New Jersey,they are intact and well-buffered
floodplains.A characterizationof wetlandsin this region is imperativeto properlyassess and restorenatural
lands in this ever increasinglydeveloped landscape.As in similar floodplainsystems, there was a change in
canopy compositionalong the 66 km stretch.Specifically,Quercuspalustris dominatedforests shifted to Acer
saccharinumdominatedforests near the midpointof the sampledriver section. Sub-canopy,shruband ground
vegetation were sampledbut clear patternswere not detectedwith respect to position along the sampledriver
section. Species richness was lower than any other publisheddescriptionsfor this type of floodplainsystem.
The low species richness in all stratamay be attributedto urbaninfluences althoughthis requiresadditional
study. Exotic flora representedup to 20% of the total flora,but did not appearto correlatewith river position
or canopy composition.Ourdataprovidequantitativevegetationdescriptionsof referencewetlandstandardsfor
a hydrogeomorphicmodel for this river system.
Key words: floodplain forests, Passaic River, New Jersey, river gradient,urban landscapes,exotic species,
hydrogeomorphicmodels.
Studies of riparian vegetation patterns provide
a valuable way of quantifying the increasingly
changing landscapes of urbanizing regions. As
natural lands and farmlands are converted to urban/suburban development, riparian forests are
often the only areas left undeveloped. Even if
riparian systems remain relatively intact, urbanization in the surrounding landscape can change
the long-existing conditions of the aquatic and
terrestrial margins by increasing surface runoff,
nutrient and sediment loads, and decreasing biotic diversity (Karr and Schlosser 1978, Nagasaka and Nakamura 1999). New Jersey provides
an especially strong opportunity to study riparian communities in a changing environment
since much of the state is experiencing some degree of development pressure (Hasse and Lathrop 2003).
Riparian zones contain a unique array of
plants adapted to the environmental characteristics and dynamics related to the flood regime
I
Fundingfor this projectwas providedby the New
Jersey Departmentof EnvironmentalProtectionand
the United States EnvironmentalProtectionAgency.
2 The authorswould like to thankPatrickRyan,Jennifer Mokos, JenniferMomsen and various graduate
andundergraduate
studentsin the Hartmanlab for their
assistance in field work; ZacharyLong and Dr. Peter
Morin for statisticaladvice; and Dr. Steven Handel.
3 Author for correspondence.E-mail: mfjohnso@
rci.rutgers.edu
Received for publicationJanuary13, 2003, and in
revised form January24, 2004.
(Naiman et al. 1993). Compositional shifts in
species, life forms and functional groups along
a river gradient from upstream to downstream
reflect changes in the extent and magnitude of
interactions among the river, the riparian zone
and the adjacent upland (Poole 2002, Ward et al.
2002). Longitudinal trends of plant species richness and composition along rivers tend to show
the greatest species richness occurring at the
midreaches of a river system (Nilsson et al.
1989, Nilsson et al. 1994, Planty-Tabacchi et al.
1996) and this pattern has been linked to the
intermediate disturbance hypothesis (Connell
1978, Sousa 1979). Other studies have found
compositional shifts in plant species with greater
proportions of ruderal and invasive species in
the downriver direction, possibly reflecting
greater disturbance rates downriver (Nilsson et
al. 1994, Planty-Tabacchi et al. 1996, Tabacchi
et al. 1996).
In New Jersey, most characterizations of riparian plant communities have focused on floodplains of the Raritan River (Buell and Wistendahl 1955, Wistendahl 1958, Frye and Quinn
1979) or the Millstone River (Van Vechten and
Buell 1959) in the central portion of the state.
As with many floodplain studies, these have focused on plant and environment relationships
within relatively small sections of the river (Buell and Wistendahl 1955, Wistendahl 1958, Van
Vechten and Buell 1959) or an individual floodplain site (Frye and Quinn 1979). These studies
232
2004]
ARONSONET AL: NEW JERSEYFLOODPLAINFORESTS
also pre-dated the recent intense and rapid urban
and suburban development throughout the state.
As urbanization has become a prevalent factor
in ecological systems in New Jersey, more research has examined the urban influences on riparian plant communities (Ehrenfeld and
Schneider 1991, 1993) and wetland soil and water quality (Ehrenfeld and Schneider 1991, 1993;
Vedagiri and Ehrenfeld 1992). None of this
work has explicitly described vegetation characteristics along a river gradient from upstream
to downstream.
The objective of this study was to characterize
the vegetation of forested floodplain wetlands
along a 66 km stretch of river within the upper
Passaic River watershed. Concomitant with the
river gradient is an urbanization gradient; the
upper portion of the watershed is relatively less
urbanized than the lower portions of the watershed (Johnson 2002). The upper Passaic River
system is unique in that most of the floodplains
in this system are buffered by large annually
flooded forest and marsh complexes while much
of the surrounding uplands have been converted
to urban and suburban land cover. This study
provides a baseline of information that will allow future ecologists to evaluate whether there
are clear interactions between plant community
structure, urbanization patterns in the surrounding landscape and river gradient position.
The sites described in this study represent the
reference standard, or the best ecological functioning of sites, that provides the basis by which
to compare other wetlands of the same wetland
type (Smith 1997). This study was part of a larger hydrogeomorphic (HGM) model study designed to evaluate the wetlands as reference wetlands for low-gradient floodplain forested wetlands (Hatfield et al. 2002). The principle idea
behind establishing reference standards is to
identify and characterize the most undisturbed
sites in the region (Smith 1997). The proximity
of the Passaic River watershed to major urban
centers has led to rapid urban and suburban expansion in this region that creates increased potential for these wetlands to be further degraded.
Therefore, it becomes even more critical to identify and characterize what reference conditions
currently exist. Development of reference sites
such as these are useful for describing restoration targets as well as invasive and exotic species management and reserve management.
Materials and Methods. STUDY AREA.
Study sites were located in floodplains of the
233
Bay
sPark
NiSol
East
4
0
4
8
12 km
FIG.1. Study site locations within the upper Passaic River basin, also the boundaryof the glacial Lake
Passaic. Inset:Locationof the upperPassaicRiverbasin within New Jersey,USA.
upper Passaic River basin (Fig. 1) in the Piedmont physiographic region of New Jersey
(Wolfe 1977). Most of the upper basin was the
site of the glacial Lake Passaic, which was
formed during the Wisconsin glaciation (Wolfe
1977). This area has very little relief, allowing
large floodplains and marsh and swamp complexes to develop. The Passaic and its tributaries
in this region are low-gradient rivers. Most of
the floodplains studied are within large forested
wetland and marsh complexes surrounded by
upland suburban or urban development. Mean
annual peak flow for the upper basin has been
estimated at the confluence with the Pompton
River as 32.5 m3/s for water years 1989-93
(Buxton et al. 1998).
The Passaic River (Fig. 1) extends approximately 161 km from Morristown to Newark
Bay, New Jersey (Bartlett 1984). The drainage
basin encompasses approximately 2,398 km2 in
north-central New Jersey. As of 1993, land use
in the lower half of the basin, extending from
the confluence with the Pompton River to Newark Bay, is almost entirely developed land, including high-density residential, industrial and
commercial areas. The upper Passaic River ba-
JOURNALOF THE TORREYBOTANICALSOCIETY
234
[VOL.131
Table 1. Site and vegetationcharacterization
of forestedfloodplainwetlandsin the upperPassaicRiverBasin,
New Jersey,USA. The inner floodplainwidth was measuredin the field. Floodplainarea includesthe entire2year floodplainand was measuredusing NJ State 1995/97 Land use/Landcover GIS coverage. The structural
characteristicsof the canopy and sub-canopywere sampledin five 100 m2plots at each site. Sub-canopydata
includes shrub species. Sites are arrangedfrom most upstreamto most downstream.Structuralcharacteristics
are reportedas averages standarderror.
_?
Density
Ratio
Canopy
Ave.
Sub-canopy
Sub-canopy
Diameter
Density
Canopy
Stem
Site
Great Swamp
Dead River
Passaic River
South Main
Roosevelt
East Orange
Sommers Park
Hatfield Swamp
Horseneck
Two Bridges
Width(m) Area(ha)
32.8
146
153.2
67.1
49.1
89.4
70.3
106.4
117.8
65.9
9.1
38.8
64.8
19.4
59.3
79.7
19.4
275.1
622.4
348.3
Density
(#stems/ha)
500
420
620
340
400
100
625
400
340
380
? 141
? 124
+ 188
+ 93
89
+_ 55
390
_ 148
+_ 117
+ 128
sin, from the source of the river near Morristown
to the confluence with the Pompton River, is approximately 50 percent undeveloped, including
agriculture and natural lands, and 50 percent urban, including residential and commercial lands
(Buxton et al. 1998). As the Passaic River basin
lies within the New York metropolitan region,
there is high development pressure for residential and corporate complexes on the remaining
undeveloped open space.
VEGETATIONSAMPLING.Ten sites were sam-
pled for vegetation in the upper Passaic River
basin (Fig. 1). Site selection criteria included:
annual flooding, forest cover, and relatively undisturbed and intact floodplains. Eight sites were
located on floodplains adjacent to the Passaic
River. Two sites were located adjacent to tributaries of the Passaic, one at Great Brook near
the confluence with the Passaic River, and the
other at Dead River.
At eight sites, a 125 m baseline was established parallel to the river at the transition from
the inner to the outer floodplain. At Sommers
Park, the baseline was only 100 m long because
the length of the floodplain was less than 125
m. At Two Bridges, the baseline was established
perpendicular to the river in an old ox-bow connected to the river by over bank flow. At all
sites, the baseline was established parallel with
the direction of flow of floodwaters. The location of the baseline was selected using visual
observations of changes in vegetation, such as
Basal area
(m2/ha)
41.3
15.7
26.2
33.8
44.6
6.8
16.6
28.4
6.0
26.1
8
?_ 6
+ 6
+ 15
+ 4
+ 4
+ 12
10
+_ 1
+ 10
(cm)
33
22
23
36
38
29
13
15
15
30
(#stems/ha)
360
480
360
220
360
20
50
3700
320
580
+
+
+
+_
+
+
+
+
+
187
156
287
128
186
20
50
1409
168
22
stems/#
stems
0.7
1.1
0.6
0.6
0.9
0.2
0.1
9.3
0.9
1.5
increased shrub cover, and quite discernible evidence of changes of flood frequency and depth,
and elevation. The distance from the river to the
baseline varied among sites due to variations in
the width of the inner floodplain at each site (Table 1). Five transects were established, 25 m
apart, from the baseline to the river except at
Sommers Park, where only four transects were
established.
Vegetation sampling was conducted in 10 x
10 m plots established at a random point on each
transect. In each 100 m2 plot, canopy, sub-canopy and shrub species were measured. Canopy
trees, defined as greater than 5 cm dbh (diameter
at breast height), were identified to species and
measured for dbh only. Sub-canopy and shrub
species were defined as less than 5 cm dbh and
greater than 50 cm in height and were identified
and measured for height, dbh or number of
stems (for shrubs with more than one stem), and
length and width of cover. Canopy cover was
estimated by placing a spherical densiometer
above the soil surface at 1 m height at the midpoint and at each corner of the 100 m2 plot.
Ground vegetation was sampled in two 1 X 1 m
quadrats randomly established within each 100
m2 plot. Percent cover was measured for ground
vegetation using modified Braun-Blanquet cover
classes (Mueller-Dombois and Ellenberg 1974,
Barbour et al. 1987). When specimens were impossible to identify, due to physical damage
such as deer browse, they were identified to genus or family or categorized as unknown.
2004]
ARONSONET AL: NEW JERSEYFLOODPLAINFORESTS
Canopy, sub-canopy and shrub species were
measured once during 1999 or 2000. Ground
vegetation was recorded in both the spring and
summer in order to examine seasonal variation.
The spring sampling for all sites occurred in
May or June of 2000. The summer sampling occurred in August of 1999 or 2000. Authority for
all species identification is from Gleason and
Cronquist (1991).
DATAANALYSIS.For ground, sub-canopy and
shrub vegetation, importance values were calculated using relative cover, relative frequency
and relative density. Sub-canopy and shrub data
were combined for calculating importance values as well as multivariate analyses. Relative
basal area, relative frequency and relative density were used to calculate tree canopy importance values (Barbour et al. 1987). To evaluate
plant community patterns, several approaches
were utilized including cluster analysis, which
was performed on importance values for canopy
and sub-canopy vegetation and on presence/absence data for summer and spring ground vegetation. We used a modified Morisita's similarity
coefficient with the unweighted pair-group
method using arithmetic averages (UPGMA) to
cluster importance value data (Krebs 1999). To
cluster presence/absence data, we used the Baroni-Urbani and Burser similarity coefficient with
UPGMA. These analyses were performed with
the Multi-Variate Statistical Package (Kovach
2000).
The non-parametric ordination technique,
nonmetric multidimensional scaling (NMDS),
was also used to examine seasonal patterns in
the ground vegetation as well as variation in
canopy composition. In both cases, NMDS was
preformed on arcsin transformed importance
values using PC-ORD (McCune and Medford
1997). The multi-response permutation procedure (MRPP) in PC-ORD (Ver 4.10) was used
to further examine the strength of the groupings
indicated by the canopy cluster analysis and the
NMDS (McCune and Medford 1997).
To evaluate replacement potential of the canopy, Pearson correlation coefficients between
canopy and sub-canopy species importance values were calculated (SAS Institute 2001). Pearson correlation coefficients were also used to determine relationships between canopy species
(SAS Institute 2001).
Exotic species representation was measured
as the percentage of exotic species within the
total floral richness. Total floral richness was
235
measured as a combined total of canopy, subcanopy, and ground species richness. Due to human disturbances to permanent plots and spring
flooding events, Sommers Park and Hatfield
Swamp were not included in this analysis. Total
and exotic species richness were calculated using only those specimens identified to species.
Results.
CANOPY,
SUB-CANOPY
AND
SHRUB
PATTERNS.
Forest structural characteristics varied greatly across the river gradient with no consistent patterns (Table 1). Canopy cover ranged
from 83% at Two Bridges to 92% at Sommers
Park. Basal area ranged from 6.0 m/ha at Horseneck to 44.6 m/ha at Roosevelt. Tree density was
greatest at Sommers Park (650 stems/ha) and
this site had one of the lowest shrub/sub-canopy
densities (25 stems/ha). In contrast, Hatfield
Swamp had the highest shrub/sub-canopy density (3700 stems/ha), due to dense shrubby
clumps of Toxicodendron radicans, with intermediate canopy tree density (400 stems/ha).
Density for both the canopy (100 stems/ha) and
shrub/sub-canopy (20 stems/ha) was lowest at
East Orange.
Plant species richness was low in the canopy,
sub-canopy and shrub layers across all sites.
Species richness for the forest canopy in the
study area was a total of eight species, with a
maximum richness of five species at any individual site (Table 2). Sub-canopy species richness reached a total of five species across all
sites. The maximum sub-canopy richness at any
individual site was four species. Four canopy
species were not represented in the sub-canopy,
Acer rubrum, Betula nigra, Quercus bicolor and
Crataegus sp. One species in the sub-canopy,
Liquidambar styraciflua, was not represented in
the canopy. Fraxinus pennsylvanica was by far
the most dominant sub-canopy species and occurred in eight sites. There were five shrub species observed across all sites, but six of the ten
sites had no shrubs.
With correlation between canopy and subcanopy vegetation as a measure of replacement
potential of the canopy species, the dominant
canopy trees, Acer saccharinum and Quercus
palustris, appeared to have a moderate replacement potential. Acer saccharinum in the canopy
was significantly correlated with A. saccharinum
in the shrub/sub-canopy (r = 0.66, P = 0.039),
and canopy Q. palustris was also significantly
correlated with sub-canopy Q. palustris (r =
0.63, P = 0.05). In addition, Ulmus americana,
a common sub-dominant canopy species, was
JOURNALOF THE TORREYBOTANICALSOCIETY
236
[VOL.131
Table 2. Species importancevalues of tree canopy, sub-canopyand shrubstrataof forestedfloodplainwetlands in the upperPassaicRiver basin, New Jersey.Study sites are abbreviatedas follows: GS = GreatSwamp,
DR = Dead River,PR = PassaicRiver,SM = SouthMain,RO = Roosevelt,EO = East Orange,SP = Sommers
Park, HA = Hatfield Swamp, HO = Horseneck, TB = Two Bridges.
Site
Stratum
Species
Acer rubrum
canopy
Acer saccharinum
Betula nigra
Crataegussp
Fraxinuspennsylvanica
canopy
canopy
canopy
canopy
GS
DR
PR
SM
0
22.8
16.1
0
RO
0
0
9.4 12.7
0
0
0
0
0
0
9.5 0
0
15.1 11.7 15.6 19.6
44.7
0
0
9.6
EO
SP
0
0
34.8 100
26.7
0
0
0
38.8
0
HA
0
HO
TB
0
36.2 79.2
0
5.4
0
0
0
43.6
0
44.5
0
0
20.5
Quercus bicolor
Quercus palustris
Ulmus americana
canopy
canopy
canopy
10.6 14.0
44.6 43.5
20.2
8
0
32.4
26.5
0
61.1
6.6
0
37.7
8.1
0
0
0
0
0
0
0
10.2
4.7
0
20.8
0
0
35.0
0
Acer saccharinum
Fraxinus pennsylvanica
Liquidambar styraciflua
Quercus palustris
Ulmus americana
sub-canopy
sub-canopy
sub-canopy
sub-canopy
sub-canopy
shrub
shrub
shrub
shrub
shrub
0
0
82.8 41.6
0
0
0
14.7
17.2 22.0
0
67.7
0
0
32.3
19.9
43.9
0
36.2
0
0
100.0
0
0
0
0
0
0
0
0
100
0
0
0
0
14.7
36.4
0
4.4
0
0
100.0
0
0
0
19.7
45.6
23.6
4.6
0
Cornus racemosa
Lindera benzoin
Rosa multifloraa
Rubus sp.
Toxicodendron radicans
Species Richness
a
0
0
0
0
0
6.5
0
4.6
10.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100.0
0
0
0
0
0
0
0
0
0
0
0
44.6
0
0
0
0
0
0
6.5
0
0
0
canopy
sub-canopy
shrub
5
2
0
5
4
2
5
2
0
4
3
0
4
1
0
3
0
1
1
1
0
5
3
1
2
1
0
3
4
1
Total
5
8
5
4
4
4
1
6
3
5
Non-native species (Gleason and Cronquist 1991, Anderson 1979, D. Snyder, unpublished manuscript).
significantly correlated with U. americana in the
sub-canopy (r = 0.85, P = 0.002). Fraxinus
americana, the most common sub-canopy species, was not significantly correlated with F.
americana in the canopy.
HA
EO
HO HO
SP
saccharinum
TB
RO
PR
DR
SM Q.palustris
GS
0.28
0.4
052
0.64
0.76
0.88
1
ModifiedMorisita'sSimilarity
FIG. 2. UPGMA cluster analysis of canopy vegetation of forested floodplain wetlands in the upper Passaic River basin, New Jersey, USA. Sample sites are
divided by a change along the sampled river gradient
from Q. palustris dominated forests to A. saccharinum
dominated forests, as indicated by the division between
the two groups. Sites are abbreviated as follows: GS
= Great Swamp, DR = Dead River, PR = Passaic
River, SM = South Main, RO = Roosevelt, EO = East
Orange, SP = Sommers Park, HA = Hatfield Swamp,
HO = Horseneck, TB = Two Bridges.
Canopy species importance values revealed a
subtle pattern in canopy dominance across the
sampled river gradient. Quercus palustris dominated the tree canopy from Great Swamp
through South Main; Roosevelt represented the
transition from Q. palustris to A. saccharinum,
which dominated or co-dominated the lower
portion of the sampling gradient (Table 2). Roosevelt and Two Bridges were co-dominated by
Q. palustris and A. saccharinum. Acer saccharinum co-dominated with F. pennsylvanica at
East Orange and Hatfield Swamp while it
strongly dominated Sommers Park and Horseneck. Ulmus americana tended to co-occur with
Q. palustris while F. pennsylvanica was a member of the canopy in a majority of the sites (Table 2). The importance value of A. saccharinum
was negatively correlated with Q. palustris (r =
-0.65, P = 0.041) and U. americana (r =
-0.66, P = 0.038).
Cluster analysis, based on canopy importance
values, divided sites into two groups that corresponded to a compositional shift of the dominant canopy species from upstream to downstream (Fig. 2). The first cluster, referred to as
the Acer saccharinum community, was charac-
ARONSONET AL: NEW JERSEYFLOODPLAINFORESTS
2004]
237
A. rubrnum
Q palustris
Q palustris
SDR
SKA.saccharinum
+ PR
0.5
+ GS
Q bicolor
Q palustrs0
U.
americana -15
xHO
x ROI
-1,,I
-0.
SM
A. saccharinum
00.5
1
1.5
X EO
a
1.5
B ngra
NMDS A
NMDS Axis 2
FIG.3. NMDS ordinationplot of canopy vegetationin forestedfloodplainsof the upperPassaicRiver Basin,
New Jersey.Symbols denote canopy classes classifiedby clusteranalysisof tree canopy importancevalues (see
Fig. 2). Sites are abbreviatedas follows: GS = GreatSwamp, DR = Dead River, PR = Passaic River, SM =
South Main, RO = Roosevelt, EO = East Orange, SP = Sommers Park, HA = Hatfield Swamp, HO =
Horseneck,TB - Two Bridges.
terized by relatively high importance values
(34.48 to 100%) of A. saccharinum. This cluster
included the sites located at Roosevelt, East
Orange, Sommers Park, Hatfield Swamp, Horseneck and Two Bridges (Fig. 1). The second
cluster, referred to as the Quercus palustris community, was characterized by low importance
values (zero to 12.71 %) of A. saccharinum and
by relatively high importance values (32.4% to
61.1%) of Q. palustris. This cluster included the
sites located at Great Swamp, Dead River, Passaic River, and South Main. MRPP analysis also
divided the sites based on the same groupings
as the cluster analysis (A = 0.247, P = 0.002).
An NMDS ordination of the data agreed with
the above community divisions and more clearly
defined the species responsible for this division
(Fig. 3). Acer saccharinum was positively correlated with NMDS axis two (r = 0.9866, P <
0.0001), while Q. palustris (r = -0.7348, P =
0.0155), Q. bicolor (r = -0.6982, P = 0.0247),
and U. americana (r = -0.73484, P = 0.0155)
were negatively correlated with axis two. Quercus palustris (r = 0.68403, P = 0.0292) and A.
rubrum (r = 0.67871, P = 0.0309) were both
positively correlated with axis one. Betula nigra
(r = -0.79376, P = 0.0061) was negatively correlated with axis one. Axes one and two accounted for 62.3% of the variation in the data,
with axis one accounting for 42.7% and axis two
accounting for 19.6%.
GROUND VEGETATION.Ground
species
rich-
ness ranged from 9 species at Sommers Park to
34 species at East Orange. Lack of seasonal data
at Sommers Park and Hatfield Swamp clearly
contributed to low species richness at these sites.
These sites were not used in statistical analysis.
There was no difference in compositional trends
in the ground vegetation between sites according
to the NMDS and MRPP analyses. Subtle compositional differences appeared between spring
and summer vegetation, but there was no statistically defined pattern. According ground species percent frequency, Leersia oryzoides and
Toxicodendron radicans were present at all sites
and were most common at the majority of sites
(Table 3). Other species that were present at the
majority of sites and common in many included:
Boehmeria cylindrica, Cinna arundinacea,
Geum canadense, Lysimachia nummularia, Pilea pumila, and various Carex, Polygonum, and
Viola species. Seedlings of Acer saccharinum,
Fraxinus pennsylvanica, and Quercus palustris
were also common in the ground vegetation of
most sampled floodplains.
EXOTIC SPECIES. Exotic
species
represented
between 5% and 20% of the total floral richness
of each site (Fig. 4). The greatest representation
of exotic species occurred in the most upriver
sites, Great Swamp (16.7%) and Dead River
(15.4%), and at the mid-portion of the river gra-
JOURNAL OF THE TORREY BOTANICAL SOCIETY
238
[VOL. 131
Table 3. Percent frequency of ground species in each study site in the upper Passaic River basin, New Jersey,
USA. Percent frequency was calculated as the number of quadrats a species occurred in divided by the total
number of quadrats sampled at a site times 100. GS = Great Swamp, DR = Dead River, PR = Passaic River,
SM = South Main, RO = Roosevelt, EO = East Orange, SP = Sommers Park, HA = Hatfield Swamp, HO =
Horseneck, TB = Two Bridges.
Site
Species
Acer rubrum
Acer saccharinum
Allium vinealea
Artemisia vulgarisa
Aster spp.
Aster racemosus
Bidens frondosa
Blephilia hirsuta
Boehmeria cylindrica
Cardamine pratensis
Carex crinita
Carex grayi
Carex lupulina
Carex lurida
Carex scoparia
Carex spp.
Carex vulpinoidea
Cinna arundinacea
Claytonia virginica
Cuscuta gronovii
Elymus sp.
Elymus virginicus
Fraxinus pennsylvanica
Galium palustre
Geum canadense
Gratiola neglecta
Impatiens capensis
Leersia oryzoides
Lysimachia ciliata
Lysimachia nummulariaa
Lycopus uniflorus
Lycopus virginicus
Microstegium vimineuma
Oxalis stricta
Oxalis spp.
Panicum clandestinum
Parthenocissus quinquefolia
Peltandra virginia
Phalaris arundinacea
Pilea pumila
Poa palustris
Polygonum arifolium
Polygonum caespitosuma
Polygonum erectum
Polygonum hydropipera
Polygonum hydropiperoides
Polygonum pensylvanicum
Polygonum persicariaa
Polygonum spp.
Potentilla sp.
Quercus palustris
Ranunculus abortivus
Rorippa nasturtium-aquaticuma
Rosa multifloraa
Rubus spp.
Saururus cernuus
Solidago caesia
GS
DR
PR
30.0
15.0
SM
RO
EO
SPb
5.0
15.0
25.0
12.5
HAb
HO
TB
62.5
50.0
5.0
22.2
15.0
5.0
5.0
5.0
35.0
50.0
10.0
11.1
11.1
11.1
22.2
22.2
5.6
5.0
10.0
5.0
25.0
5.0
20.0
12.5
38.9
5.0
5.0
40.0
5.0
75.0
15.0
5.0
12.5
55.6
11.1
10.0
5.0
40.0
15.0
5.0
40.0
5.0
20.0
20.0
20.0
30.0
15.0
65.0
25.0
5.0
20.0
50.0
6.3
20.0
10.0
18.8
25.0
10.0
5.6
5.6
16.7
20.0
33.3
31.3
40.0
33.3
10.0
66.7
12.5
75.0
11.1
81.3
10.0
11.1
27.8
16.7
16.7
16.7
5.6
100.0
5.6
83.3
25.0
20.0
10.0
30.0
50.0
15.0
5.0
30.0
20.0
25.0
5.0
5.0
25.0
30.0
5.0
15.0
5.0
70.0
5.0
25.0
20.0
60.0
10.0
70.0
5.0
10.0
5.0
5.0
80.0
5.0
35.0
25.0
45.0
10.0
10.0
30.0
10.0
75.0
11.1
27.8
16.7
22.2
10.0
5.0
10.0
10.0
11.1
10.0
15.0
55.6
11.1
11.1
20.0
5.0
15.0
15.0
10.0
30.0
20.0
30.0
5.0
25.0
30.0
35.0
25.0
20.0
5.0
20.0
35.0
5.0
5.0
10.0
25.0
5.0
16.7
25.0
50.0
5.0
5.6
38.9
27.8
10.0
50.0
40.0
35.0
10.0
12.5
6.3
12.5
12.5
31.3
5.6
18.8
11.1
5.6
11.1
6.3
10.0
25.0
10.0
20.0
5.0
5.0
5.0
6.3
20.0
16.7
12.5
5.6
5.6
16.7
5.0
5.0
5.0
12.5
35.0
ARONSONET AL: NEW JERSEYFLOODPLAINFORESTS
2004]
239
Table3. Continued.
Species
Solidago spp.
Toxicodendronradicans
Ulmusamericana
Viola spp.
Total Richness
GS
DR
77.8
45.0
PR
SM
Site
RO EO
5.0
50.0
20.0
70.0
10.0
45.0
10.0
45.0
90.0
5.0
55.0
33
30
34
38.9
15.0
5.0
20.0
10.0
20.0
28
23
22
SPb
HAb
HO
TB
62.5
62.5
100.0
77.8
37.5
12.5
20.0
16.7
9
15
24
19
aNon-nativespecies (Gleasonand Cronquist1991, Anderson1979, D. Snyder,unpublishedmanuscript).
b Only sampled during one season. Sommers Park sampled only in spring due to human disturbancesto
permanentplots. HatfieldSwamp sampledonly in summerdue to springflooding events.
dient, Roosevelt (20%) and East Orange
(16.7%). Interestingly, the two sites with the
lowest exotic species richness, Passaic River
(5.6%) and South Main (9.7%), were bounded
upriver and downriver by sites that tended to
have the highest percentage of exotic species.
Lysimachia nummularia was the most common
exotic species across the study area and was present at all sites analyzed for exotic species richness. Other exotic species included: Allium vineale, Artemisia vulgaris, Microstegium vimineum, Polygonum caespitosum, Polygonum hydropiper, Polygonum persicaria,
Rorippa
nasturtium-aquaticum, and Rosa multiflora.
Many of these species are considered invasive
in the state of New Jersey (D. Snyder, unpublished manuscript).
there are some important differences. Canopy
species richness (eight species) in the Passaic
River floodplain sites is very low compared to
other northern floodplains on low-gradient rivers. For example, along the Raritan River, within
the same physiographic region of New Jersey as
the Passaic River, tree species richness was reported as 15 species at one site (Frye and Quinn
1979) and as high as 20 species along a stretch
of the river (Buell and Wistendahl 1955). In
low-gradient floodplains in southern Illinois, tree
species richness ranged from 19 to 24 species
(Bell and del Moral 1977). Overall woody species richness, including tree and shrub species,
in low-gradient floodplains ranged from 48 in
Texas (Nixon et al. 1977) to 29 in Illinois (Bell
1974) to 32 (Buell and Wistendahl 1955) and 24
(Frye and Quinn 1979) in New Jersey. This conDiscussion. The floodplains of the Passaic trasts with the Passaic
River, where the total
River are somewhat similar in vegetation comrichness
is
woody species
only 14.
position and structure to floodplains in New JerThe species composition of the tree canopy in
sey as well as other northern floodplains, but this study is similar to those in other northeastern floodplains. In their study of a floodplain
250
along the Raritan River, Frye and Quinn (1979)
found Acer negundo, Fraxinus spp. and Quercus
20
0
palustris as dominant species. Acer saccharinum
and Ulmus americana were also common. Quercus palustris and A. saccharinum dominated
15a
floodplains along the Millstone River (Van
100
Vechten and Buell 1959). In northern Illinois,
floodplains are dominated by A. saccharinum,
50
Populus deltoides, Salix spp. or Q. palustris
(Hosner and Minckler 1963). Yin (1998) docu00
mented the dominance of A. saccharinum and
DR
PR
GS
SM
RO
EO
HO
TB
Site
Q. palustris along the upper Mississippi River.
The compositional change of dominant canFIG.4. Percentof exotic species to total floralrichness across study sites in the upperPassaic River ba- opy species at the Roosevelt site suggests that
sin, New Jersey,USA. Sites are orderedfromupstream there may be a change in environmental condito downstreamandabbreviatedas follows: GS = Great
tions along the river gradient. Those sites in the
Swamp, DR = Dead River, PR = Passaic River, SM
= South Main, RO = Roosevelt, EO
East
Orange, upriver portion of the gradient, which were dom=
HO = Horseneck, TB = Two Bridges.
inated by Q. palustris, tended to have less mac-
240
JOURNALOF THE TORREYBOTANICALSOCIETY
ro-topographic complexity and more poorly
drained soils (M. Aronson, pers. observation).
Sites in the downriver portion of the gradient,
which were dominated by A. saccharinum, tended to have more macro-topographic complexity
(M. Aronson, pers. observation) and sandier
soils than those upriver (Johnson 2002).
While species composition of the understory,
sub-canopy and shrub strata, in this study is similar to those in other northeastern floodplains, we
found that the floodplains of the upper Passaic
River are characterized by a depapurate understory with only an average of 6.6 + 2.0 individuals per plot. Floodplains have a tendency for
severe flood scouring and damage to the understory, which likely has contributed to the lack
of development of the shrub/sub-canopy in these
sites (Frye and Quinn 1979, Ehrenfeld 1986).
The shrub/sub-canopy species richness of Passaic River floodplains, 10 species, is comparable
to other New Jersey floodplains where between
9 and 12 species were found in the understory
(Frye and Quinn 1979, Buell and Wistendahl
1955).
A further limitation to understory development of floodplains along the Passaic River is
herbivory by deer. At all sites, there was obvious
damage on most saplings, seedlings, shrub and
herbaceous vegetation from deer browse. Deer
sightings were also common during field sampling; the average minimum density of deer in
the upper Passaic River basin ranged from 11 to
15 deer/square km in 1999 (Anon 1999). This
population estimate is based on hunt check stations and is most likely an underestimation of
the actual deer population in this region (D.
Drake, pers. comm.). New Jersey has an overabundant deer population (Anon. 1999) and it is
well known that overabundant deer populations
have negative effects on forest regeneration (e.g.
Tilghman 1989, Waller and Alverson 1997).
The species composition of the herbaceous
layer does not appear to vary widely across
northern floodplains. The Passaic River sites
contain similar herbaceous species as floodplain
forests along the Raritan (Wistendahl 1958) and
Millstone Rivers (Van Vechten and Buell 1959)
in New Jersey. Herbaceous species composition
in these sites is also similar to northern floodplains outside of New Jersey. In oak-hickory
forests of northern Mississippi River floodplains,
Shelford (1954) found herbs to be few in both
species richness and density, but two species of
Polygonum dominated. Herbaceous species
composition of Wisconsin floodplains included
[VOL.131
several similar species to those found along the
Passaic River, including Carex lupulina, C. grayi, Cinna arundinacea, Leersia oryzoides, Phalaris arundinacea, Polygonum pensylvanicum
and Impatiens spp. (Menges and Waller 1983).
These similarities across all northern floodplains
may explain the failure of cluster analysis,
MRPP and NMDS to successfully divide Passaic
floodplain sites by composition. In addition, the
distribution of herbaceous species may be better
explained by smaller scale processes such as
moisture gradients, microtopographic variation,
occasional anoxic conditions, and scouring by
floodwaters and ice flow.
Exotic plant species were present across the
entire sampled river gradient. In contrast to
Planty-Tabacchi et al. (1996), which found an
increase in exotic vegetation with downriver position, there was no clear pattern of exotic richness across our sites. The fact that we did not
find this pattern may be a function of our relatively short river gradient; Planty-Tabacchi et al.
(1996) examined river reaches from 150 to 350
km long while our river reach was only 66 km
long. It is more likely that the factors influencing
exotic vegetation at the sites studied here were
local factors such as interior disturbances, for
example, recreational use and ditching, or adjacent disturbances, such as roads and bridges. Although the floodplains studied here are considered relatively well buffered, interior and adjacent disturbances were common. It is important
to note that most of the exotic species are considered invasive in this region and that further
study of the importance of exotic invasive species may show that they dominate cover of the
herbaceous and shrub layers.
The floodplains of the upper Passaic River basin are unusual in that they are well buffered by
large swamp and marsh complexes surrounded
by suburban and urban developed uplands. The
majority of these floodplains are unlikely to be
developed because of the magnitude of the annual overbank flood events and lengthy periods
of inundation. These floodplains offer important
ecological, hydrological and recreational functions and values (Ehrenfeld 2000). Of the many
functions they provide, the most important economic and social functions may be storing floodwaters and therefore ameliorating the effects of
flooding on downriver communities and improving water quality. In addition, the wetlands function as biodiversity sources for urban landscapes
(Naiman and Decamps 1997) and frequently
provide the only natural habitat for both biodi-
2004]
ARONSONET AL: NEW JERSEYFLOODPLAINFORESTS
versity conservation and recreational uses in urban areas (Ehrenfeld 2000).
These results provide interesting possibilities
for future research. The location of large wetlands in the urban/suburban region may result in
a shift in the plant community diversity either
by lowering the number of species or by replacing native species with exotic species. Although
both of these patterns are indicated in our current study, considerably more sampling directly
testing this relationship is needed. Although we
did see a change in canopy dominance from
downstream to upstream, it is impossible to
tease apart potential urban effects from natural
changes along the river without further research.
The contribution of forest wetland systems to the
regional biodiversity of densely populated regions is believed to be considerable. Broad surveys of urban wetlands such as this one are
needed to understand their significance in terms
of diversity and importance to society.
Literature Cited
K. 1979. A check list of the plants of New
ANDERSON,
Jersey. Rancocas Nature Center, Mount Holly, NJ.
1999. Governor's report on deer manageANONYMOUS.
ment. New Jersey Department of Environmental
Protection, Division of Fish and Wildlife, and Department of Agriculture. Trenton, NJ.
M. G., J. H. BURKANDW. D. PITTS.1987.
BARBOUR,
Terrestrial plant ecology, second edition. The Benjamin/Cummings Publishing Company, Inc. Menlo
Park, CA.
R. A. 1984. Rolling rivers: an encyclopedia
BARTLETT,
of American rivers. McGraw-Hill Book Company,
New York, NY.
BELL,D. T. 1974. Tree stratum composition and distribution in the streamside forest. Am. Midl. Nat.
92: 35-46.
BELL, D. T., AND R. DEL MORAL. 1977. Vegetation gradients in the streamside forest of Hickory Creek,
Will County, Illinois. Bull. Torrey Bot. Club 104:
127-135.
1955. FloodBUELL,M. E, ANDW. A. WISTENDAHL.
plain forests of the Raritan River. Bull. Torrey Bot.
Club 82: 463-472.
ANDR. E. HIVKBUXTON, D. E., K. HUNCHAK-KARIOUK
MAN. 1998. Relations of surface-water quality to
streamflow in the Hackensack, Passaic, Elizabeth,
and Rahway River basins, New Jersey, water years
1976-93. U. S. Geological Survey, West Trenton,
NJ.
J. H. 1978. Diversity in tropical rain forests
CONNELL,
and coral reefs. Science 199: 1302-1310.
J. G. 1986. Wetlands of the New Jersey
EHRENFELD,
pine barrens: the role of species composition in
community function. Am. Midl. Nat. 115: 301313.
J. G. 2000. Evaluating wetlands within an
EHRENFELD,
urban context. Ecol. Eng. 15: 253-365.
J. G., AND J. P. SCHNEIDER.
1991. ChaEHRENFELD,
241
maecyparis thyoides wetlands and suburbanization:
effects on hydrology, water quality and plant community composition. J. Appl. Ecol. 28: 467-490.
EHRENFELD,J. G., AND J. P. SCHNEIDER.1993. Responses of wetland vegetation to perturbations of water
chemistry and hydrology. Wetlands 13: 122-129.
FRYE, R. J., AND J. A. QUINN. 1979. Forest development in relation to topography and soils on a floodplain of the Raritan River, New Jersey. Bull. Torrey
Bot. Club 106: 334-345.
GLEASON, H. A., AND A. CRONQUIST.1991. Manual of
vascular plants of northeastern United States and
adjacent Canada, second edition. The New York
Botanical Garden, Bronx, NY.
2003. A Housing-Unit
HASSE, J., ANDR. G. LATHROP.
Level Approach to Characterizing Residential
Sprawl. Photogramm. Eng. Rem. S. 69: 10211030.
HATFIELD, C. A., P W. RYAN, M. E JOHNSON,AND J.
M. HARTMAN.
2002. Assessing the feasibility of using hydrogeomorphic wetlands assessment in New
Jersey-Riverine forest example. Draft for Review
submitted to the Department of Environmental Protection, Trenton, NJ.
HOSNER, J. E, AND L. S. MINCKLER. 1963. Bottomland
hardwood forests of southern Illinois-regeneration
and succession. Ecology 44: 29-41.
JOHNSON,M. E 2002. Floodplain forest vegetation,
soils, and land use of the upper Passaic River basin,
New Jersey. M.S. Thesis. Rutgers University, New
Brunswick, NJ.
KARR, J. R., AND I. J. SCHLOSSER.1978. Water resources and the land-water interface. Science 201: 229234.
KOVACH,W. 2000. Multi-Variate Statistical Package,
Version 3.12c. Kovach Computing Services, Anglesey, Wales.
KREBS, C. J. 1999. Ecological methodology, second
edition. Addison-Welsey Educational Publishers,
Inc., New York, NY.
MCCUNE,B., AND M. J. MEDFORD. 1997. Multivariate
Analysis of Ecological Data, Version 3.20. MjM
Software, Gleneden Beach, OR.
MENGES,E. S., ANDD. M. WALLER.1983. Plant strategies in relation to elevation and light in floodplain
herbs. Am. Nat. 122: 454-473.
MUELLER-DOMBOIS,
D., AND H. ELLENBERG.1974.
Aims and methods of vegetation ecology. Wiley,
New York, NY.
NAGASAKA, A., AND E NAKAMURA. 1999. The influence
of land-usechanges on hydrologyand riparianen-
vironment in a northern Japanese landscape. Landscape Ecol. 14: 543-556.
NAIMAN, R. J., AND H. DICAMPS. 1997. The ecology
of interfaces:riparianzones. Ann. Rev. Ecol. Syst.
28: 621-658.
NAIMAN, R. J., H. DECAMPS AND M. POLLOCK. 1993.
The role of corridorsin maintainingregional biodiversity. Ecol. Appl. 3: 209-212.
NILSSON, C., A. ELKBLAD, M. DYNESIUS, S. BACKE, M.
B. CARLBERG,
AND R.
S. HELLQVIST
GARDFJELL,
JANSSON.1994. A comparison of species richness
and traits of riparian plants between a main river
channel and its tributaries. J. Ecol. 82: 281-295.
AND U.
NILSSON,C., G. GRELSSON,M. JOHANSSON
SPERENS.1989. Patterns of plant species richness
along riverbanks. Ecology 70: 77-84.
JOURNAL OF THE TORREY BOTANICAL SOCIETY
242
NIXON, E. S., R. L. WILLETT, AND P. W. Cox.
1977.
Woody vegetation of a virgin forest in an eastern
Texas river bottom. Castanea 42: 227-236.
PLANTY-TABACCHI,A. M., E. TABACCHI,R. J. NAIMAN,
C. DEFERRARI, AND H. DECAMPS. 1996. Invasibility
of species-rich communities in riparian zones. Conserv. Biol. 10: 598-607.
POOLE,G. C. 2002. Fluvial landscape ecology: addressing uniqueness within the river discontinuum.
Freshwater Biol. 47: 641-660.
SAS INSTITUTE.
2001. The SAS system for windows,
version 8e. SAS Institute Inc., Cary, NC.
SHELFORD, V. E. 1954. Some lower valley Mississippi
valley flood plain biotic communities; their age and
elevation. Ecology 35: 126-142.
SMITH, R. D. 1997. Developing the reference wetland
system. In Hydrogeomorphic approach to assessing
wetland functions: guidelines for developing regional guidebooks. Wetlands Research Program,
Technical Report WRP-DE.
SOUSA,W. P. 1979. Disturbance in marine and intertidal boulder fields: the nonequilibrium maintenance
of species diversity. Ecology 60: 1225-1239.
TABACCHI,E., A. M. PLANTY-TABACCHI,M. J. SALINAS
AND H. DECAMPS.1996. Landscape structure and
[VOL. 131
diversity in riparian plant communities: a longitudinal comparative study. 1996. Regul. River 12:
367-390.
TILGHMAN, N. G. 1989. Impacts of white-tailed deer
on forest regeneration in northwestern Pennsylvania. J. Wildl. Manage. 53: 524-532.
VAN VECHTEN
III, G. W., ANDM. E BUELL.1959. The
floodplain vegetation of the Millstone River, New
Jersey. Bull. Torrey Bot. Club 86: 219-227.
VEDAGIRI, U. AND J. G. EHRENFELD. 1992. Partitioning
of lead in pristine and runoff-impacted waters from
acidic wetlands in the New Jersey pinelands. Water
Air Soil Poll. 64: 511-524.
WALLER, D. M., AND W. S. ALVERSON. 1997. The
white-tailed deer: a keystone herbivore. Wildlife
Soc. Bull. 25: 217-226.
D. B. ARSCOTT,AND C.
WARD, J. V., K. TOCKNER,
CLARET.2002. Riverine landscape diversity. Freshwater Biol. 47: 517-539.
WISTENDAHL,W. A. 1958. The floodplain of the Raritan River, New Jersey. Ecol. Monogr. 28: 129-153.
WOLFE,P E. 1977. Geology and Landscapes of New
Jersey. Crane, Russak, New York, NY.
YIN, Y. 1998. Flooding and forest succession in a modified stretch along the upper Mississippi River. Regul. River 14: 217-225.
Download