 

ha.“ h: ..

It}

|||

 

,
5.... .
. , \. v...

A In...
an .ywzga

, .
.z an?

v2

‘ﬂnﬁ...
ﬂ w.

b . 3
2..

“no
4::

it
. 1‘.

Rut:

.meﬁzwmm
. Whammy“. . 1.5.
rwé

l
a

,,.. -i.
p...

s. u

:1 It: - A .!

A. ‘ .

é... . wiymlxwuss
‘ I l

(4 1L .
, VD;- - . I
iﬁQHIHmws ﬁmvpmﬁ . 1.
. 5. :11“.
‘ .‘ Waugh» 5

E3 k’g‘§.$
g'm. I.)

A

5M ”1“) “Va

 

n—r :-- This is to certify that the
I RARY dissertation entitled
Mlch 3r: State
Ur versity

 

 

 

Assessing wetlands and their restoration using algae

presented by
Lei Zheng

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Zoology/ EEBBirogram

7%,.Eim ,...._

 

Major Professor’s Signature

4314M /, 4005

\J

 

Date

MSU is an Afﬁnnativo Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

EP242007
r3240?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDue.p65—p.15

ASSESSING WETLANDS AND THEIR
RESTORATION USING ALGAE

By

Lei Zheng

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
For the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology
and
the graduate program in Ecology, Evolutionary biology, and Behavior
(EEBB)

2003

ABSTRACT
ASSESSING WETLANDS AND THEIR RESTORATION USING ALGAE
By

Lei Zheng

This dissertation develops a better understanding of relationship among a variety of
ecological factors within wetlands, evaluates a number of valuable ecological attributes to
develop indicators for wetland bioassessment and wetland restoration, and establishes
biological and environmental criteria to protect wetland ecosystems. Ecological
assessment in wetlands has to develop habitat or class speciﬁc attributes to reduce the
spatial and temporal variability. I ﬁrst evaluated spatial and temporal variation of
environmental and algal attributes among habitats within wetlands and between restored
and extant wetlands in Michigan. The results indicated non-diatom algal assemblages
differed among habitats, but sediment and epiphytic diatom assemblages were not
signiﬁcantly different. Epiphytic assemblages in the restored and extant wetlands were
relatively distinct, while both sediment and phytoplankton assemblages were not as
different between restored and extant wetlands. The differences in algal assemblages
among habitats within wetlands and between restored and extant wetlands were related to
a number of environmental characteristics, but particularly nutrients and sediment
organic matter. In my study of salt marsh restoration, organic matter in sediments and the
similarity of diatom species composition between paired restored and reference salt
marshes increased with age of restored marshes during spring and summer. Primary
production by epiphytic and sediment algae in summer showed Site-speciﬁc changes and

did not change consistently with marsh age. In this study, pairwise comparison of a

restored marsh and a nearby reference marsh reduced regional variation among wetlands
and helped identify the successional pattern of ecological development after restoration.
Because nutrients were one of the most important factors regulating periphyton growth
and species composition in wetlands, experimental and ﬁeld surveys were conducted to
determine the limiting factors in wetlands. Experiments indicated that algae were
nitrogen limited and limitation was directly related to nutrient concentrations and nutrient
ratios. Although individual diatom taxa responded to nutrient additions differently, the
responses of indicator taxa in the nutrient bioassays were positively correlated with
natural nutrient gradients among natural wetlands indicating variation in species
composition among restored wetlands was related to N-limitation. Based on the
establishment of a cause-effect and stressor-response relationships between nutrient
stressors and algal attributes, I evaluated the response of a number of valued ecological
attributes to nutrient stressors. The similarity in species composition to reference
conditions and the number of native species decreased and trophic index calculated with
species composition increased with increasing nutrient concentrations. Diatoms as
indicators of wetland nutrient pollution and human disturbance were developed using
paleoecological diatom assemblages to characterize reference conditions. Biological
criterion was then established based on the historic trophic status indicated by diatom
species composition, and a nutrient criterion to protect or restore the biological integrity
of wetlands was then established from the relationship between trophic status index and
nutrients and using the predicted nutrient concentration. My studies indicated that algal
assemblages were valuable indicators for assessing biological condition of wetlands,

monitoring restoration, and diagnosing stressors.

This thesis is dedicated to
My wife
Jingfang Chen
And
My parents

For their inﬁnite love and constant inspiration

iv

ACHNOWLEDGEMENTS

First, I would like to thank my major professor, Dr. R. J an Stevenson, for his constant
support and encouragement throughout my graduate studies. His academic guidance was
essential to the completion of this dissertation and has provided important insights to my
scientiﬁc research. I would also like to express the gratitude for my committee members,
Dr. Thomas Burton, Dr. Katherine L. Gross, and Dr. Donald Hall, for reading early drafts
of this dissertation and providing valuable comments that helped to improve the
presentation and contents of this dissertation. I am also grateful to all the people in the
algae lab for their great help to accomplish this research.

I would like to thank my wife Jenny Chen for her support and love during the
dissertation writing. Her encouragement made this dissertation possible. My parents

receive my deepest gratitude and love for their dedication and many years of support.

TABLE OF CONTENTS

List of Tables .......................................................................................... iv
List of Figures ......................................................................................... xi

Chapter 1. Algae as indicators of Biological Integrity in restored and natural wetland
ecosystems
1.1 Wetland bioassessment...
1.2 Algal indicators...
1. 3 Wetland biogeochemistry and primary producers
1.4 Objectives of the dissertation...
1.5 References...

bdbbb

Chapter 2. Algal assemblages in multiple habitats of restored and extant wetlands

2.1 Abstract .......................................................................................................... 11

2. 2 Introduction... 13

2. 3 Methods... ....16
2.3.1 Site description 16
2. 3. 2 Sampling... 17
2.3.3 Samples analySis 18
2. 3. 4 Statistics ................................................................................................. 19

2. 4 Results... 21
2.4.1 Environmental vanables 1n reStored and extant wetlands... .......21
2. 4. 2 Non- diatom algal assemblage and diatom assemblages 1n restored and
extant wetlands ................................................................................................ 26
2. 4. 3 Difference of taxonomic composition among different habitats and between
restored and extant wetlands... 3O
2. 4. 4 Algal assemblages and env1ronmental vanables . . . . . . . ... ...33

2. 5 Discussion... ...37
2.5.1 Wetland restoration and env1ronmental change In reStored wetlands ........ 37
2. 5.2 Difference of algal species composition between restored and extant" ...38
2.5.3 Difference of algal species composition among different habitats ............ 41

2.6 References ...................................................................................................... 44

Chapter 3. Changes in benthic algal attributes during salt marshes restoration and use in
bioassessment
3.1 AbstractSO
3.21ntroductlon52
3.3 Method554
3.3.1 Site location ........................................................................................... 54
3.3.2 Field samplmg55
3.3.3 Laboratory analysesS7
3. 3. 4 Statistical analyse359
3 .4 Results... 61
3.4.1 Biomass development 61

vi

3. 4. 2 Algal species diversity and composition.................................................64
3.4.3Productivity.... 69
3. 4. 4 Algal attribute development 1n Virginia salt marshes . . . . . . ... .69
3.5 Discussion” 72
3. 6 Acknowledgments ........................................................................................... 79
3. 7 References... 80

Chapter 4. Relations between nutrient conditions, algal growth, and species composition
in the restored wetlands of southern Michigan

4.1 Abstract .......................................................................................................... 86

4. 2 Introduction... 87

4. 3 Methods... ..89
4.3.1 Sample collection 8.9

4. 3. 2 Nutrient diffusion experiment .................................................................. 91

4. 3. 3 Laboratory analyses ................................................................................ 92

4.3.4 Data analyses .......................................................................................... 94

4.4 Results ............................................................................................................ 95
4.4.1 Wetland habitats ...................................................................................... 95

4.4.2 Effects of nutrient enrichment and nutrient concentrations.......... . . . . . . . ......97

4.4.3 Algal species composition in natural and artiﬁcial substrata....... . . . . . . . .....101

4.4.4 Algal assemblage response to nutrient additions.....................................106

4.5 Discussion ..................................................................................................... 109
4.5.1 Algal nutrient limitation ......................................................................... 109

4.5.2 Algal species composition and nutrient concentrations.......................115

4.6 Conclusions ................................................................................................... 116
4.7 Acknowledgments118
4.8 References ..................................................................................................... 119

Chapter 5. Assessing wetland response to land use using diatoms and paleoecological
characterization of reference condition
5.1 Abstracts ....................................................................................................... 127
5. 2 Introduction ................................................................................................... 128
5.3 Methods... .... 131
5.3.1 Material arid methods 131
5.3.2 Land use coverage ................................................................................. 133
5.3.3 Statistical analyses ................................................................................ 133
5.4 Results ...................................................................................... 135
5.4.1 Environmental variables ........................................................................ 135
5.4.2 Current diatom assemblages and environmental conditions............. . . . . ...139
5.4.3 Comparison between reference and current conditions.......... . . . . . ......143
5.5 Discussion ..................................................................................................... 147
5.6 ReferenceslSZ

Chapter 6. Conclusion ............................................................................................. 158

vii

Appendices
Appendix 1 ....................................................................................... 164
Appendix 2 ....................................................................................... 165

viii

LIST OF TABLES

Table 2.1 Correlation matrix of environmental variables measured and principal
component analysis (PCA) loadings of environmental variables in the 45
wetlands. When a correlation coefﬁcient r >0.30, Dunn-Sidak P<0.05; when
r>0.39, P<0.01. Correlation between age and environmental variables had a
degree of freedom 24 and none of the correlation was signiﬁcant ................ 23

Table 2.2 Mean alpha diversity (Number of species), Gamma (landscape) diversity, and
beta diversity between habitats with a wetland and between restored and extant
wetlands .................................................................................. 29

Table 2.3 MRPP comparisons of species composition in non-diatom algae and diatom
between different habitats ..................................... - ........................ 31

Table 2.4 MRPP comparisons of species composition of non-diatom algae and diatoms
between restored and extant wetlands ............................................... 32

Table 2.5 Pearson correlation matrix of dissimilarities in species composition of non-
diatom and diatom between restored and extant wetlands in different habitats.
A * indicates a P value <0.05. ...................................................... 33

Table 2.6 Regression by forward stepwise selection of independent variables. Dependent
variable is dissimilarity among epiphytic, phytoplankton, and sediment non-
diatom algae, and dissimilarity between epiphytic and sediment diatom
assemblages, independent variables are uncorrelated environmental
variables .................................................................................. 35

Table 2.7 Linear regressions by forward stepwise selection of independent variables.
Dependent variable is average Bray-Curtis dissimilarity of algal species
compositions in a restored wetland with extant wetlands. Independent variables
are average dissimilarities of a restored wetland with all extant wetlands ...... 36

Table 3.1 Sampling sites and their properties during the summer sampling season ...... 56

Table 3.2 Results of regression analyses between algal attributes and environmental
variables ................................................................................. 64

Table 3.3 Number of species found in all sites, habitats and seasons ....................... 64

Table 3.4 The 10 most dominant diatom species in all sites (sorted by relative abundance
rank, habitat type, marsh type, and season) ........................................ 70

Table 4.1 Correlation matrix of ambient nutrients in 22 wetlands. A * indicates a
signiﬁcant correlation (Dunn-Sidak P<0.05) ....................................... 97

ix

Table 4.2 Algal Nutrient Limitation and Corresponded Water Nutrient and sediment
nutrients. Value in the parentheses are correlations between log-transformed
value and algal nutrient limitation. A * indicates signiﬁcant correlation (Dunn-
Sidak
P<0.05) ................................................................................. 100

Table 4.3 Weighted correlation matrix from Canonical Correspondent Analysis
(CCA) .................................................................................. 103

Table 4.4 Twenty-six most dominant taxa and their responses to nutrient additions. The
4.digit number represents comparisons of the relative abundance of a diatom
taxa in a treatment with control, N-enriched, P—enriched, and NP-enriched
treatments. A signiﬁcant change (P<0. l) of relative abundance between
treatments was presented by 1, and insigniﬁcant effect was 0. * is empty when
compared with itself .................................................................. 107

Table 4.5 Correlations between relative abundance of nutrient indicator species and
nutrient concentration in different habitats ........................................ 108

Table 5.1 Pearson correlation matrix of environmental variables of 35 wetlands across
Lower Peninsula of Michigan in 1994.995. A signiﬁcant correlation (Dunn-
Sidak Probabilities P<0.05) is indicated by * ...................................... 138

Table 5.2 The most dominant diatom taxa used in CCA and the correspondent taxa
code ..................................................................................... 140

Table 5.3 Regression of observed environmental variables against predicted values based
on diatom species composition. Simple regression and jackkniﬁng cross-
validation indicators based on weighted average (WA) and tolerance down-
weighted WA used to compare the precision. MSE represents mean square
error and RMSE represent residual of MSE ......................................... 142

Table 5.4 Correlation among community characteristics and environmental
variables ................................................................................. 145

LIST OF FIGURES

Figure 1.1 Stressors-response relationship between human disturbance and response of
ecological valuable attributes .............................................. _ ........... 8

Figure 2.1 Wetland sampling locations in southern Michigan ................................ 16

Figure 2.2 Wetland physical and chemical characteristics and comparison between
restored and reference wetlands. .................................................... 24

Figure 2.3 Wetland nutrient concentration and comparison between restored and extant
wetlands ................................................................................. 26

Figure 2.4 Non-metric multidimentional scaling (NMDS) of the restored and extant
wetlands based on non-diatom and diatom assemblages in different habitats.
The circles are the sample ellipse which is centered on the sample means of
the x and y variables. The size the ellipse is speciﬁed a probability value of
0.6827 .................................................................................... 30

Figure 2.5 Non-metric multidimentional scaling (NMDS) of the restored and extant
wetlands based on non-diatom and diatom assemblages in different habitats.
The circles are the sample ellipses which are centered on the sample means of

the x and y variables. The size the ellipse is speciﬁed a probability value of
0.6827 .................................................................................... 31

Figure 3.1 Map of sampling sites between Morehead City and Wilmington along the
North Carolina coast. Age (yr) of restored marshes is reported in
parentheses .............................................................................. 54

Figure 3.2 Box-plot of dry mass (DM), ash free dry mass (AFDM), the organic matter
proportion of sediments (0MP), algal biovolume, and Ch] a in reference

marshes in surface sediments and on macrophytes during spring and summer
1998. .................................................................................... 61

Figure 3.3 Relationship between the attribute restoration ratio and age of restored marsh
in surface sediment and on macrophytes during spring and summer 1998. The
OMRR is the comparison of 0MP of sediments in a restored site with that of
its reference marsh. The ACRR is the comparison of algal chlorophyll a
concentration in a restored site with that of its reference marsh. The AVRR is
the comparison of algal biovolume in a restored site with that of its reference
marsh. The line represents a signiﬁcant relationship (linear regression
P<0.05). The correlation coefﬁcients (R2) for each relationship are: 3a: 0.727;
3b, 0.931; 3c, 0.969; 3d, 0.235; 3e, 0.304; 3f, 0.647; 3g, 0.577; 3h, 0.749; 3i,
0.442; 3j, 0.741 ......................................................................... 63

xi

Figure 3.4 Relationship between the percentage similarity index (PSI) of algal functional
groups (4a, 4b) and diatom assemblages (4c, 4d) and the age of restored
marshes in surface sediments and on macrophytes during spring and summer
1998. PSI was calculated for algal functional groups or diatom assemblages
between restored and reference marshes. The line represents a signiﬁcant

relationship (linear regression P<0.05). The triangles represent sites in Virginia,
and dots represent sites in North Carolina. ............................................ 67

Figure 3.5 Relationship between algal productivity ratio and age of restored marshes in
summer 1998. Productivity ratio compared between algal GPP between
restored and reference sites in surface sediments (A) and on macrophytes

(B) ....................................................................................... 71

Figure 3.6 Conceptual model of algal structural and functional development during marsh
restoration ............................................................................... 78

Figure 4.1 Map of Lower Peninsula of Michigan, showing the Upper Grand River and
Maple River watersheds where sampling sites were located. .................. 90

Figure 4.2 Box-plot of environmental variables in the 22 wetlands ......................... 96

Figure 4.3 Algal biomass/production in artiﬁcial substrata of control, N-enriched, P-
enriched, and NP-enriched treatments. They are represented by different
attributes ................................................................................ 98

Figure 4.4 Algal cellular nutrient contents and molar N:P ratios in control, N-enriched, P-
enriched, and NP-enriched treatments on artiﬁcial substrata .................... 99

Figure 4.5 Chl a: AFDM and Ch] azDM ratios in control, N-enriched, P-enriched, and NP-
enriched treatments on artiﬁcial substrata ........................................ 100

Figure 4.6 Comparisons of algal N and P limitation in all 22 wetlands. Primary N or P
limitation is deﬁned as algal biomass of that nutrient treatment minus control
treatment then weighted by control treatment. ................................. 101

Figure 4.6 The relationships between algal primary nutrient limitation and nutrient
concentrations in the water column and the sediments. Primary nutrient
limitation is deﬁned as algal biomass of that nutrient treatment minus control
treatment then weighted by control treatment ................................... 102

Figure 4.8 CCA biplots of algal assemblages and environmental variables. Diatom taxa
on macrophytes, in sediments, and on artiﬁcial substrata in control treatments,
and algal genera of phytoplankton were used for the CCA .................... 104

Figure 4.9 Box-plots of similarities of diatom assemblages between different habitats and
treatments. E-epiphytes, S-sediment, C-control, N-N treatment, P-P treatment,

xii

NP- NP treatment. X-axis shows the comparison of two paired habitats. Y-
axis shows the Percentage similarity between the two habitats .............. 105

Figure 5.1. Distribution of the 35 wetlands sampled in fall 1994 and summer 1995 in the
lower peninsula of Michigan along a north-south and an east-west
transect ................................................................................. l3 1

Figure 5.2. Box plot of environmental variables in 1994.1995 in the Michigan
wetlands ............................................................................... 136

Figure 5.3 Relationship between average TP and % agricultural land use in sampled
wetlands. Regression tree analysis was used to explore the relationship
between TP and land use. A threshold effect occurred when agriculture
occupied 57.9% land use around the wetland watershed ...................... 137

Figure 5.4 CCA biplots of sites and fall environmental variables from 35 palustrine
wetlands in Michigan. (a) Dots show site scores in ordination space, (b) dots
shows species scores in ordination space, and arrows indicate the strength and
correlation between environmental factors and the ordination axes, The dotted
circle shows sites with above 59% agricultural land cover. The dotted circle
shows sites with above 57.9% agricultural land cover. Diatom species are
indicated by eight-digits code (see Table 5.2 for complete taxon names)..140

Figure 5.5 Regression models of diatom inferred environmental variables against
observed environmental variables. 5b shows the relationship between diatom
inferred TP and agricultural land use ............................................. 143

Figure 5.6 Non-metric multidimensional scaling ordination diagrams of paleoecological
diatom assemblages and current diatom assemblages. Reference
(paleoecological diatom assemblages) and current (surface diatom
assemblages) were indicated by different symbols. The centroids with 68.5%
conﬁdent ellipse were plot around each region .................................. 144

Figure 5.7 Relationships of diatom assemblage similarity between current assemblages
with nutrient concentration (TP) (a), and relationship of native taxa richness
with nutrient concentration (TP) (b) in wetlands. S1 is the Bray-Curtis
similarity of relative abundance between surface sediment and the centroid of
presettlement samples ............................................................... 144

Figure 5.8. Relationship between van Dam trOphic index and palustrine wetland TP (a),
and land use (b). Box—plot Shows paleoecological diatom trophic index from
14 palustrine wetlands to represent reference wetland condition. (a)Surface
diatom trophic index (dots) respond linearly to TP concentration in wetlands
(R2=0.407, P=0.001). The intercept of biological criteria and regression line
represented nutrient criteria (stressor criteria=46 ug/L). (b). Dots represent
change of trophic index with agricultural land use; a threshold effect was

xiii

observed when agriculture exceeds 56% total land use, with van Dam trophic
index increasing from 2.46 to 4.85 ................................................ 146

xiv

Chapter 1

Algae as indicators of biological integrity in restored and extant wetland ecosystems

1.1 Wetland bioassessment

Wetlands have suffered serious impairment and loss since European settlement in
North America. Approximately 53% of the wetlands in the United States have been lost
since the late 17003 (Mitsch and Gosselink 1999), and a large proportion of wetlands
have been identiﬁed as impaired by various pollutants such as nutrients and other
environmental stressors. The US. Congress passed the Clean Water Act (CWA) in 1972
to preserve and improve the quality of the nation’s water resources, including wetlands.
Over the past 30 years, wetland mitigation has been increasingly used to counteract loss
of wetland habitat and the associated ecosystem functions. As a result, thousands of
wetlands have been restored and created to fulﬁll the goal of the “no-net-loss” policy
(Section 307 of the Water Resources Development Act). However, even for the
elementary purpose of maintaining water quality, the success rates of wetland mitigation
have been very low (USEPA 2000).

To attain the stated goal of the CWA to “ restore and maintain physical, chemical, and
biological integrity” of wetlands, a set of structural and functional criteria for extant and
restored wetlands needs to be developed and evaluated. Traditional wetland assessment
techniques evaluate physical and chemical conditions, especially chemical inﬂuences
from point-source discharges. Increased human activities have contributed to nonpoint
source runoff, introduced species, and other anthropogenic impacts that have imposed

increasing damage to wetland ecosystems. New approaches based on structural and

functional assessment of biological communities have been developed during the last two
decades. Two of the most common approaches for assessing wetlands are the
Hydrogeomorphic (HGM) approach, which focuses on functional assessment (Hauer &
Smith 1998), and the Index of Biotic Integrity (IBI) approach of bioassessment
(Danielson 2001). The HGM functional assessment approach is based on the
hydorgeomorphic concepts for classifying wetlands and was speciﬁcally deve10ped for
rapid wetland assessments in response to the requirements of “no overall net-loss” of
wetland area and functions (Conservation Foundation, 1988). In contrast, the [BI
approach of bioassessment was largely developed to meet water quality and biocriteria
requirements and is based on the conventional Cowardin et a]. classiﬁcation systems
(Cowardin et a1. 1978). These two approaches both contrast and complement each other
and both contribute to assessment of wetland structure and function (Stevenson and
Hauer 2002).

In the past 20 years, biological assessment has been developed to evaluate the integrity
of aquatic systems. Biological integrity is a comprehensive reﬂection of physical and
chemical factors and is deﬁned as “the ability of an aquatic ecosystem to support and
maintain a balanced, adaptive community of organisms having a species composition,
diversity, and functional organization comparable to that of natural habitats within a
region” (Karr and Dudley 1981). By evaluating the condition of one taxonomic
community relative to physical and chemical conditions across a range of sites, biological
assessment can identify stressors that impair ecological health and also diagnose the
cause of environmental impairment. Bioassessment can also be used to evaluate wetland

restoration and changes in ecosystem functions over time.

Evaluation of wetland restoration and wetland impairment involves assessment of
nutrient enrichment at multiple spatial and temporal scales and how they affect biological
communities. Numerous studies have documented wetland structural and functional
changes after restoration (Middleton 1999; Streever 1999). Most of these studies have
focused on changes in biogeochemistry, macrophyte community structure, and
invertebrate species composition. It is the general expectation that restored wetlands
should increase in species diversity, nutrient retention, systems sustainability, and
productivity over time. The direct impact of nutrient accrual and recycling during
wetland development is a major regulator of wetland structure and function.

1.2 Algal indicators

As pointed out by Landis and McLaughlin (2000), indicators have three basic
purposes: to measure current status, to predict future changes in the system, and to predict
progress toward the restoration of a system. The most common use of indicators is to
establish the current status of a system.

Algae are one of the most sensitive indicators for biological assessment of aquatic
systems. The importance of algal assemblages and function has been summarized in a
numbers of reviews (Dixit et al. 1992, Stevenson and Pan 1999, Stevenson 2001). Algal
characteristics that contribute to their effectiveness in assessment, include great diversity,
presence in almost all aquatic ecosystems, sensitivity to chemical and physical changes,
and taxon-speciﬁc responses to environmental changes (Shubert 1984, Steneck & Dethier
1994, Stoermer 1990). Environmental stressors could lead to degradation of valued
ecological attributes and thus the sensitive response of algal assemblages could be useful

indicators of environmental degradation (Figure 1). Over the past several decades,

evaluation of streams and lakes using algal attributes has proved successful (Lowe & Pan
1996). Although the biological communities and the underlying biogeochemical
processes of wetlands are more complex than streams and lakes, the general approaches
and methods for streams and lakes assessment can also be applied to wetland systems
(McCormick and Stevenson 1998, Danielson 2001). Attempts to evaluate environmental
gradients using algal attributes have been successful in Kentucky wetlands and the
Everglades (Pan et al. 1996, McCormick 1998, Pan et a1. 2000) and have proved to be
successful.
1.3 Wetland biogeochemistry and primary producers

Wetland systems have complicated biogeochemical characteristics that vary both
within and among different types of wetlands (Odum 1988, Mitsch and Gosselink 2000).
Salt marshes differ from freshwater marshes because they receive chemicals from both
oceanic and freshwater input. Anaerobic decomposition is predominated by sulfur
reduction and fermentation in salt marshes. Large amounts of sulfur gases and C02, low
amounts of methane, and unknown amounts of nitrous oxides are the main biogenic
gases. Nitrogen is the major limiting factor for primary production in salt marsh
ecosystems, while phosphorus is accumulated. Salt marshes have been shown to be both
sources and sinks of nutrients, particularly nitrogen. The general model of salt marsh
nutrient ﬂux suggests that salt marshes act primarily as transformers of nutrients by
importing dissolved oxidized inorganic forms and exporting dissolved and particulate
reduced forms (ammonium, organic nitrogen, and phosphorous compounds). Salt
marshes tend to be a net importer of nutrients during spring and summer and a net

exporter in the autumn and winter (Odum 1988).

The water and soil chemistry of freshwater marshes is dominated by a combination of
factors associated with the underlying mineral soils, overlain with autochthonous inputs
of organic matter from vegetation production (Mitsch and Gosselink 2000). The
inﬂowing water has high amounts of dissolved materials and nutrients, derived from
stream, groundwater, and surface water inﬂow. Organic matter in freshwater marshes can
vary widely, but is generally higher compared to salt marshes (Mitsch and Gosselink
2000). Dissolved inorganic nitrogen and phosphorus vary seasonally from very low
concentration in the summer to high concentration in the winter. Thus, nutrient limitation
varies seasonally. Sulfur reduction is an important anaerobic decomposition process, but
is of less signiﬁcance in freshwater marshes than in salt marshes. Anaerobic
decomposition is predominated by methanogenesis and fermentation in freshwater
marshes. Large amounts of methane and C02, small amounts of sulfur gases, and
unknown magnitude of nitrous oxides are the main biogenic gases. Most studies Show
that freshwater wetlands are nutrient sinks and transformers. Nitrogen and phosphorus are
generally limiting factors and are probably related to each other (Mitsch and Gosselink
2000).

Both salt marshes and inland freshwater wetlands are among the most productive
ecosystems in the world. Marsh grasses, periphyton, and phytoplankton compose the
three major autotrophic units in salt marshes (Odum 1988). Periphyton production can be
as high as or higher than that of the macrophyte community’s production. Vascular plant
communities are always dominated by a few species, whereas algal species attain high
species diversity and production. Freshwater marshes, on the contrary, have a much

higher diversity of vascular plants but a relatively lower diversity of algae than salt

marshes (Odum 1988). The plant productivity of inland marshes can be as high as salt
marshes (Mitsch and Gosselink 2000).
1.4 Objectives of the dissertation

The central goal of this dissertation was to apply algal biological assessment
approaches that are broadly used in lakes and streams to evaluate ecological health of
wetlands and to assess wetland restoration. Based on this objective, my research focused
on how algal assemblages in both coastal salt marshes and inland freshwater wetlands
respond to environmental changes and human disturbance. I conducted both ﬁeld
observations and experiments to delineate the pattern of algal responses along human
disturbance gradients, especially nutrient gradients. I evaluated algal indicators for
wetland nutrient enrichment and wetland restoration in both coastal salt marshes and
freshwater wetlands. Both simple statistics and multivariate approaches were used to
identify, develop, evaluate, and integrate a suite of algal indicators of ecological
condition. Assessment of ecological conditions in least disturbed conditions and stressor-
response relationships were used for development of biological criteria and
corresponding stressor criteria for environmental monitoring and management (Fig. 1).

This dissertation includes the following major components as separate chapters:

Part 1 (chapter 2): Freshwater wetland restoration was evaluated by: examining the
difference of environmental variables and algal assemblages between restored and extant
wetlands; identifying what environmental variables causing the difference in algal
assemblages between restored and extant wetlands; and comparing algal assemblages
among different habitats and how similarities among habitats change with change of

environmental conditions.

Part H (chapter 3): The success of salt marsh restoration was assessed by: using the
reference approach and pairing a restored wetland with a nearby extant wetland to
minimize regional difference among wetlands; evaluating a suite of algal attributes and
ﬁnding the best indicators for salt marsh restoration; and evaluating the structural and
functional development of wetland ecosystem development over time.

Part III (chapter 4): Nutrient enrichment experiments were conducted in a set of
restored freshwater wetlands to determine: nutrient limitation, how algae respond to
nutrient enrichment in artiﬁcial habitats, and how these responses are different or similar
in experimental examination compared to natural environmental gradients, and thus
establish indicators based on cause-effect relationship for wetland bioassessment.

Part IV (chapter 5): A set of wetlands across Michigan were used to: determine
stresses and stressor gradients through diatom ecological response in wetland ecosystems;
investigate how these stressors are related to human disturbance gradients based on
diatom ecological responses; and set biological and stressor criteria to protect restored
wetlands.

The ultimate goal of this research was to provide new tools for diagnosing and
quantifying risks from pollution that threaten wetland ecosystems. The results of my
research should support ecologically sensitive decision-making in land planning and
wetland restoration. Speciﬁcally, I was able to deﬁne ecological thresholds based on
stressor-response relationships and reference approaches. These results can be applied for
watershed management by providing targets for decreasing anthropogenic pollution and
thereby increasing the rate of successful restoration projects. I also developed easy and

rapid biotic and abiotic indicators that serve as qualitative and quantitative criteria to

assess water quality and ecosystem health. These studies help to identify ecosystems at

risk due to excess nutrients in the environment.

High
0, 5 Goal: 9.9. Number of Native spp.
C o
E s:
8 0.0.0.?0§0:000 ......OOOOOOOOOOOOOOO00......
: \ Response Criterion:
13’ : \ e.g. assemblage
‘>" E similarity to reference
_ ' \
8 3 x
'55 3 ‘— Indicator
2 : Stressor Criterion \
O o \
f”) yset to protect goal
Low I ° \
Low (Fief) High

Stressor or Human Disturbance Gradient

Figure 1.1 Stressors-response relationship between human disturbance and response of
ecological valuable attributes

1.5 Reference

Cowardin, L.M., V. Carter, F.C. Golet, E.T. LaRoe. 1979. Classiﬁcation of wetlands and
deep water habitats of the United States. US. Fish and Wildlife Service, Washington,
DC.

Conservation Foundation. 1988. Protecting America’s wetlands: An action agenda. Final
Report of the National Wetlands Policy Forum. The Conservation Foundation,
Washington, DC.

Danielson, T. J. 2001. Methods for evaluating wetland condition: Introduction to wetland
biological assessment. EPA 822-R-01-007a. US. Environmental Protection Agency,
Ofﬁce of Water, Washington, DC.

Dixit, 8.8., J. P. Smol, J. C. Kingston, and D. F. Charles. 1992. Diatoms: powerful
indicators of environmental change. Environmental Science and Technology. 25: 22-
33.

Hauer, F. R., and RD. Smith. 1998. The hydrogeomorphic approach to functional
assessment of riparian wetlands: evaluating impacts and mitigation on river
ﬂoodplains in the USA. Freshwater Biology 40: 517-530.

Karr, J. R., and D. R. Dudley. 1981. Ecological perspective on water quality goals.
Environmental Management 4:55-68.

Landis, W. G., and J. F. McLaughlin. 2000. Design criteria and derivation of indicators
for ecological position, direction, and risk. Environ. Toxicol. Chem. 19: 1059-65.

Lowe, R. L. and Y. Pan. 1996. Benthic algal communities as biological monitors. In R. J.
Stevenson, M. L. Bothwell and R. L. Lowe [eds] Algal ecology: Freshwater benthic
ecosystems. California Academic Press, San Diego, pp 705-739.

McCormick, P. V., and R. J. Stevenson. 1998. Periphyton as a tool for ecological
assessment and management in the Florida Everglades. J. Phycol. 34: 726-33.

Middleton, B. 1999. Wetland restoration: Flood pulsing and disturbance dynamics. John
Wiley & Sons, Inc. New York. pp6-7.

Mitsch, W. J ., and J. G. Gosselink. 2000. Wetlands. John Wiley & Sons, Inc. New York.

Odum, W. E.1988. Comparative ecology of tidal freshwater and salt marshes. Annual
review of ecological system, 19: 147-176.

Shubert, L. E. 1984. Algae as Ecological indicators. Academic Press, INC.

Stevenson, R. J ., and Y. Pan. 1999. Assessing environmental conditions in rivers and
streams with diatoms. In The diatoms: applications for the environmental and earth
sciences. Edited by .E. F. Stoermer and J. P. Smol. Cambridge University Press,
Cambridge, UK. pp 11-40.

Stevenson, R. J ., R. P. Sweets, Pan Y., and R. E. Schultz. 1999. Diatom community
patterns in wetlands of Michigan (USA) and their use as indicators of ecological
conditions. Pages 517-527 in A. J. McComb, and J. A. Davis (editors). Proceedings of
INTECOL’S Vth International Wetland Conference. Gleneagles Press, Adelaaide,
Australia.

Stevenson, R. J ., McCormick, P. V., Frydenborg, R. 2001. Methods for evaluating
wetland condition: using algae to assess environmental conditions in wetlands. EPA
843-B-00—002k. US. Environmental Protection Agency; Ofﬁce of Water,
Washington, DC.

Stevenson, R. J. and F. R. Hauer. 2002. Integrating Hydrogeomorphic and Index of Biotic
Integrity approaches for environmental assessment of wetlands. J. N. Am. Benthol.
Soc.

Stoermer, E. F. 1990. Quantitative characteristics of phytoplankton assemblages. In
Shubert L. E. [ed] Algae as ecological indicators. Academic Press, Inc. New York.

Streever, W. 1999. An international perspective on wetland rehabilitation. Kluwer
Academic Publishers. Boston.

USEPA. 1992. Framework for ecological risk assessment. EPA 630/R-92/001. US.
Environmental Protection Agency. Washington, DC.

US. EPA. 2002. Methods for evaluating wetland condition, #11 Using algae to assess

environmental conditions in wetlands. United States Environmental Protection Agency,
Washington, DC, USA. EPA-822-R-02-021.

10

CHAPTER 2
ALGAL ASSEMBLAGES IN MULTIPLE HABITA TS 0F RES TORED AND EXTANT

WETLANDS

2.1 Abstract The functional and structural attributes of algal assemblages were studied
in 25 restored and 20 extant depressional, palustrine wetlands in southern central
Michigan. Environmental conditions and algal assemblages were compared between
restored and extant wetlands and among habitats within wetlands. Restored marshes
generally had lower shading by macrophytes, nutrient concentrations, and organic matter
in sediments. Relative biovolume of non-diatom algae was signiﬁcantly different among
habitats in restored wetlands, but did not differ between macrophytes and sediments in
extant wetlands. Species composition of diatom assemblages was not signiﬁcantly
different between epiphytes and sediment in both restored and extant wetlands. The
discovered differences in non-diatom algae could not be attributed to any environmental
variables measured; however, diatom assemblage differences between habitats increased
with light irradiance. Differences in sediment diatom assemblages were observed
between restored and extant wetlands and were related to differences in nutrients, pH,
and canopy cover. Differences were also observed between epiphytic diatom assemblages
in restored and extant wetlands and they were related to light and dissolved oxygen.

Key index words: diatoms; non-diatom algae; epiphyton; epipelon; phytoplankton;
habitats; restored wetlands; extant wetlands; periphyton

Abbreviations: DIN, dissolved inorganic nitrogen; DO, dissolved oxygen; LEC, light

extinction coefﬁcients; MANOVA, multivariate analysis of variances; MRPP, multi-

ll

response permutation procedure; NMDS, nonmetric multi-dimensional scaling; 0MP,
organic matter proportion; PCA, principle component analyses; SRP, soluble reactive

phosphorus; TN, total nitrogen; TP, total phosphorus

12

2.2 INTRODUCTION

Wetland restoration has been increasingly used to mitigate loss and anthropogenic
impact of wetland systems (McDonnell and Pickett 1990, Holland et a1. 1995, Detenbeck
et al. 1999, Zedler 2000). Beginning in the early 1990's, a large effort was initiated in
Michigan to compensate wetland losses by establishing wetland restoration programs
(MDEQ 2001a). Hundreds of wetland restoration projects have been carried out since
then. Most of the wetland restoration projects involve simple techniques, such as
plugging agricultural ditches or breaking ﬁeld tiles to restore the hydrology of an area,
and then restore the biological community. Restored wetlands support aquatic and
terrestrial habitats and sustain many plant, invertebrate, and wildlife species (NRC 1995,
USEPA 2000).

Ecological development after wetland restoration has been studied extensively
(Whigham et a1. 1990, Patten 1994, Middle 1999, Keddy 1999), and many environmental
variables change during transition from one successional state to another (Gosz 1992).
Nutrient accumulation is one of the most noticeable changes in biogeochemical process
affecting the biological community in wetlands. In human impacted landscapes, nutrient
enrichment from human activities affects the wetland restoration process, and its
detrimental effects on wetland systems have been frequently reported (Koerselman and
Beltman 1988, Blair 1996, Magee et a1. 1999). Nutrient accumulation and storage from
agricultural and urban runoff affect wetland productivity, species diversity, water quality,
as well as the restoration rate of wetlands (Rader and Richardson 1992, Craft and Casey

2000).

13

Algae are one of the most important groups of primary producers in wetlands, and
they play crucial roles in wetland biogeochemical cycles (Vymazal 1995). Algae,
especially diatoms, are effective indicators of ecological conditions (Zedler 1993,
Adamus 1996, Cottingham and Carpenter 1998, Detenbeck 2001, Stevenson et al. 2001)
and have been used as indicators of ecological conditions in wetlands (Pan and Stevenson
1996, Stevenson et al. 1999, Pan et al. 2000, Stevenson 2001). Algae are also sensitive to
nutrient concentration change (Tilman 1982). Soil organic C, N and P accumulation
during wetland restoration can affect the biomass and species composition of the algal
community (McCormick et al. 1996, Pan et al. 2000). Therefore, algal species
assemblages could be indicators of restoration success because algal species composition
and diversity would differ in low- and high-nutrient wetlands (Mayer and Galatowitsch
1999).

Habitat differences, along with a number of other environmental factors, contribute to
spatial and temporal variability of algal assemblages in aquatic ecosystems (Leland 1995,
Stevenson 1996, Stevenson 1997). The extent and conditions in four major algal habitats,
phytoplankton, epiphyton, epipelon, and metaphyton, vary temporally at different stages
in wetland cycles (Goldsborough and Robinson 1996). A number of studies have reported
that different habitats support different algal assemblages in aquatic systems (Stevenson
and Hashim 1989, Pringle 1990, Burkholder 1996, McCormick et al. 1998, I-Iillebrand
and Kahlert 2001, Lim et al. 2001). Thus the similarity of algal assemblages between
different habitats should increase with the increase of wetland age and increasing

similarity of habitat characteristics.

14

Environmental factors such as, substratatum, substratratum stability, and nutrient
availability have been suspected as major determinants of species composition
differences among habitats (Stevenson 1996, Burkholder 1996). Changes in
environmental conditions of algal habitats during wetland development ﬂowing creation
or restoration may regulate heterogeneity in species composition among habitatis. The
lack of substratum should distinguish the plankton from the two benthic habitats of all
stages of wetland development, but heterogeneity between phytoplankton and benthic
habitats could decrease if more benthic algae were suspended in the water during either
early or late stages of development. Algae on plants and sediments get some nutrients
direcetly from the substratum. Nutrient supply from sediments directly to epipelon may
be available earlier during wetland development than the indirect supply from sediments,
through plants, to epiphyton. Differences between epiphyton and epipelon probably
decrease with wetland development as nutrient availabitliyt increases and differences
between habitats decrease.

To evaluate the success of wetland restoration and investigate the impact of human
disturbance to both natural and restored wetland systems, I selected 45 (25 restored, 20
extant) freshwater depressional wetlands in the Maple and Upper Grand watersheds in
southern Michigan. The environmental characteristics and algal assemblages of
phytoplankton, epiphyton, and epipelon on different habitats in these wetlands were
examined. I hypothesized that: environmental variables would differ between restored
and extant wetlands; algal assemblages would differ among habitats and between restored
and extant wetlands; differences in algal assemblages among habitats and between

restored and extant wetlands would decrease with increasing age of restored wetlands;

15

and that decrease would be due to increasing similarity of environmental characteristics. I
also expected to identify environmental factors that determine the distribution of algal

assemblages and biodiversity among habitats and between restored and extant wetlands.

2.3 METHODS

Site description. Forty-ﬁve (25 restored, 20 extant wetlands) depressional, palustrine
wetlands (Cowardin et al. 1979) were selected in the Maple and Upper Grand River
watersheds of Southern Michigan for this study (Figure 2.1). All restored wetlands were
reestablished within the past 15 years as part of wetland mitigation program. All wetland
sites were located in landscapes dominated by agricultural and urban land. The sampled
wetlands were mostly Typha or
Nuphar dominated marshes,
but also included several
swamps. Watershed land use
data were obtained from the
Michigan DNR spatial data
library to determine land use
around these wetlands.
Wetland boundaries for each

site were determined based on

 

100 100 200 Miles

 

 

 

shifts from wetland to upland
Figure 2.1. Wetland sampling locations in southern vegetation and changes in slope
Michigan. 45 sites were selected in Maple and Upper
Grand River watersheds. between the wetlands and the

16

adjacent upland. Duckweed covered at least some proportion of all wetlands during the
sampling period.

2.3.] Sampling. The selected wetlands were sampled and assessed during mornings of
July 2000. Water temperature and dissolved oxygen (DO) were measured with a YSI®
DO meter (YSI Inc., Yellow Springs, Ohio, USA) at the beginning and at the end of the
sampling period, allowing at least one hour between measurements. Wetland net
production was calculated based on DO change over one hour period times the estimated
average depth of the wetland. Conductivity and pH were measured using a YSI® model
33 SCT meter with a conductivity and pH electrode (Denver Instrument Company).
Canopy cover by emergent vegetation was estimated with a spherical canopy densiometer
held at water level and by averaging canopy cover from 4 directions. Duckweed coverage
was estimated based on the proportional coverage of the wetland surface. Irradiance at
the water surface (10) and 30 cm below the water surface ([2) were recorded with a
LICOR light meter. Light extinction coefﬁcients (LEC) were calculated as ln(Io)-
ln(Iz)/0.3 m (Wetzel 2001). Water samples from open water area in each wetland were
collected in 2 125-ml acid-washed polyethylene bottles, stored in ice, transported to the
laboratory, and frozen until chemical analysis. Phytoplankton samples from 10 cm below
the water surface were collected in 2 l-L polyethylene bottles for chlorophyll a and algal
cell counts. Sediment samples were collected randomly from eight different locations in
each wetland using hard plastic tubes (internal diameter = 3.8 cm, length = 20 cm). Only
the top 1 cm of sediment was collected. All 8 sediment samples from a wetland were
composited in a Whirl-pak® bag. Epiphytic algae were collected by cutting 4-5

macrophyte stems, brushing epiphytes from each macrophyte surface, and putting them

17

in a Whirl-pak® bag. Metaphyton were collected for qualitative identiﬁcation. Algae were
preserved with M3 (APHA 1998) after subsampling for chl a analysis.

2.3.2 Sample analysis. One 250-m] water sample from each wetland was analyzed for
total nitrogen (TN) and total phosphorus (TP). Another bottle was ﬁltered through
Coleman® glass ﬁber ﬁlters (0.45um diameter) and the ﬁltrate was analyzed for nitrate-
nitrite (NO3'-N + NOz'-N) and soluble reactive phosphorus (SRP) using a Spectronic®
GenesysTM 2 Spectrophotometer. TP and SRP were measured using the ascorbic acid
method, while TN and nitrate and nitrite (NOx-N) were measured using cadium reduction
methods (APHA 1998). Ammonia (NH4-N) was analyzed using a Wuick-Chem 8000
autoanalyzer at University of Michigan Biological Station (UMBS). Dissolved inorganic
nitrogen (DIN) was calculated as the sum of ammonia and nitrate-nitrite concentrations.
Sediment samples were rinsed with 200 ml of deionized water and homogenized using a
Biospec® homogenizer. Appropriate dilution was made for sediment samples to measure
sediment TN and TP concentration following the same methods as the water samples.

Chl a in water sample was assayed by extracting in 90% buffered acetone overnight
at 4°C, reading absorbance on a Spectronic® GenesysTM 2 Spectrophotometer, and
calculating pheophytin-corrected chl a concentration (APHA 1998). Dry mass (DM) and
ash free dry mass (AFDM) of periphyton were determined according to standard methods
(APHA 1998). Organic matter proportion in sediment (OMP) was the AFDM/DM ratio
of surface sediments.

Phytoplankton, epiphytic algae, and sediment algae were diluted or condensed as
necessary before counting. Algal densities and non-diatom species composition were

determined using a Leica® microscope and a Palmer-Maloney (0.1 ml) counting chamber

18

under 400X and by counting 300 natural units. Algal taxa were identiﬁed to the lowest
possible taxonomic level. Biovolume was estimated for at least 15 cells of each taxon by
assigning them a geometric shape and measuring appropriate cell dimensions (Hillebrand
et al. 1999, Charles et al. 2002). Diatom taxa and relative abundance were determined
from permanent Naphax mounts of acid-cleaned diatoms (Stosch and Reimann 1970). At
least 500 diatom valves were counted. Alpha diversity (species richness per locality),
gamma diversity (total species diversity within landscape, all wetland sampled), and beta
diversity (variation in species composition between 2 communities or habitats) were
calculated for non-diatom algae and diatoms (Beilman 2001).

2.3.3 Statistics. To compare environmental variables between restored and extant
wetlands, both individual variables and scores of principle component analyses (PCA)
were used. Environmental variables were log-transformed to evenly distribute the
variance before analysis, if necessary. The Student t-test was used to compare individual
variables between extant and restored wetlands. PCA summarized major environmental
patterns in this region. Multiple analysis of variance (MANOVA) was also used the test
overall differences between restored and extant wetlands. The scores of the ﬁrst three
PCA axes were used as composite variables for MANOVA. All statistics were performed
using SYSTAT® v. 10.

Attributes of non-diatom algae and diatoms were analyzed separately. The relative
biovolume of non-diatom algae and relative abundance of diatom species were calculated

and used for comparisons. Only diatom species with a relative abundance 2 1% in a

minimum of three sites or 2 5% in at least one site were included in the analysis.

19

Nonparametric multidimensional scaling (NMDS), a multivariate ordination
technique (Clarke 1993), was used to summarize and illustrate patterns of non-diatom
and diatom distribution among different habitats and between restored and extant
wetlands. Multi-response permutation procedure (MRPP), a non-parametric procedure for
testing the hypothesis of no difference between two or more groups (Biondini et al.
1985), was used to assess differences in taxonomic composition of non-diatom and
diatom assemblages among different habitats and between restored and extant wetlands.
A p-value was provided by the MRPP to characterize the signiﬁcance of between group
differences. An A value (0-1) characterizes within-group homogeneity compared to the
random expectation. When all items are identical within groups, then A = 1; if
heterogeneity within groups equals the expectation by chance, then A = 0. Both MRPP
and NMDS were performed using PC-ORD ® v.4.0 (McCune and Mefford 1997).

To relate differences in taxa composition to environmental variables, Bray-Curtis
similarity matrics (srrrilarity in relative abundances or biovolumes) were calculated for
species composition of non-diatom and diatom assemblages among different habitats and
between restored and extant wetlands. Stepwise automatic linear regressions were used to
explore the relationship between non-diatom or diatom assemblages and environmental
variables. This was done by using similarity of non-diatom or diatom assemblages
between habitats as separate dependent variables and a set of 19 uncorrelated variables as
independent variables. Using these same procedures, similarities of species composition
between restored and extant wetlands were calculated and expressed as a function of
differences in environmental variables between wetlands. The dependent variables were

the average Bray-Curtis similarities in species composition of a restored wetland with all

20

the extant wetlands. The Euclidean distance matrix for each environmental variable
among different wetlands was calculated to indicate environmental differences between
wetlands and the average Euclidean distances of individual environmental variables
between a restored wetland with all extant wetlands were used as independent variables.
Those variables that had the highest correlations with other variables were screened out,
so only the most uncorrelated environmental variables were included for the regression
analyses. For example, the distance of conductivity between restored and extant wetlands
was positively correlated with the distance of sediment TP (1:0. 424, P=0.035),
negatively correlated with pH distance (r=0.430, p=0.032), and related to several other
variables. Therefore, it was excluded from the model. All these analyses were performed

using Systat® v10.

2.4 RESULTS
2.4.1 Environmental variables in restored and extant wetlands.

Land cover in the region was dominated by agriculture (78%) with smaller portions
of urban (9%), forest (6%), and other land types. Most of the restored wetlands were
Typha-dominated marshes. In contrast, extant marshes in this area were seldom
dominated by one macrophyte Species. Phragmites, Typha, Nuphar, Ceratophyllum
demersum, Myriophyllum, and Scirpus were co-dominant in many of the extant marshes.

Three PCA scores accounted for 42% of environmental variance in the data set (Table
2.1). The ﬁrst axis was positively correlated with major nutrients, such as TP (r=0.621),
ammonium (r=0.784), and DIN (r=0.675) in water column, as well as light supplies

(duckweed cover and light extinction coefﬁcient). The second axis was negatively

21

correlated with N, P content in sediments and shading by vegetation, and positively
correlated with conductivity. The third axis was positively correlated with SRP and DO.
Overall, the scores of the ﬁrst 3 axes were signiﬁcantly different between restored and
extant wetlands (MANOVA, Wilks’s lamba F3, 42: 5.938, P=0.002). Univariate F test
indicated that PCA axis 1 score was signiﬁcantly different between restored and extant
wetlands (P=0.012), while the scores of PCA axis 2 and axis 3 were not different (P>0.1).
Many variables did not differ between the two types of wetlands (P>0.1). Water
temperature varied from 18.1 to 322°C among wetlands during the sampling period.
Conductivity was highly variable (from 149 to 1003, 499i34 pmhos/cm). pH varied
greatly among sites (from 6.8 to 9.6, 7.92i0.9 (Average :tS.E.)), but was not signiﬁcantly
different between restored and extant wetlands (P=0.110). D0 in the morning could be as
high as 11 mg/L and as low as 0.7 mg/L in some of the wetlands (4.567i0.381mg/L).
Wetland net production based on DO change ranged from -1.0 to 3.17 gom'z-h'l
representing heterotrophic to autotrophic wetland types.

Most restored wetlands had relatively low shading by vegetation (26.2i3.8%, mean i
standard error). Extant wetlands had higher shading than restored sites (t-test, P=0.031)
(Figure 2.2). Duckweed and other ﬂoating plants appeared in at least 2/3 of the sampled
sites with an average coverage of 38.2% (i31.5%) of water surface during sampling
period. Duckweed coverage was much higher in extant wetlands than restored ones
(P=0.003). LEC at 0.3m depth in open water (if available) of restored wetlands (1.207
i012] umol m'3s‘1) was much lower than in extant wetlands (3.173i0.777 umol m'3s")

(P=0.01) indicating that extant wetlands had much lower light avalaibility in the water

column and at the bottom.

22

 

5m.0 NN.0- w0.0- $0- 20- m0.0 3.0- m~.0- v0.0 mad- 2.0- mmd 0N.0 52.0 Nv.0- 02.0 05.0 0m.0- 2N0- v0.0 mom

vm.0- N00 00.0- 05.0- 0N0 2.0 2.0 2.0 2.0 2.0 2.0 20.0 $0 2&0- wvd w0.0- 2.0 02.0- 52.0 20.0- NOAH
NN.0- No.0 w0.0 3.0 00.0 00.0 mmd- mud- 3.0 05.0 00.0 N00 and 00.0- 52.0- M500 mm.0- .500 00.0 3.0 2052
00.2 50.0 v0.0 00.0 N00 N00 N00 N00 N00 30- 30- no.0 02.0- 2.0 ~0.0- 02.0 50.0- 50.0 2.0 NN.0 ow<
00.2 NN.0- 20.0 00.0 No.0 3.0 00.0 00.0- 00.0- 2.0- 2.0- m0.0 50.0 w0.0 vmd 2N0. 3H0 02.0 00.0- $2 .Uom

. 00.2 05.0 2.0 v0.0- 000- 00.0 2.0 2.0 02.0 V00 v0.0- 020- 52.0- 00.0- 50.0- 02.0 N0.0- 0N0 0.00m

. 00.2 0N0 n0.0- 00.0 ~00- 00.0 m0.0- 20.0 50.0- 2.0- 2.0 0N0- 00.0- w0.0- 02.0 m0.0- wad 2.00m

. 00.2 0V0 00.0- 2.0- 3.0 and N00 mud #00 20.0- 52.0- 00.0- 2.0- 3.0 and 52.0 #20

. 00.2 2.0- 2.0- 0N0 0~.0 2.0 mmd 2.0 00.0 :0. mad N00 wmd wm.0 80 «EU

. 00.2 0nd 3.0- 00.0- 200 3.0- 2.0 2.0- 20.0 00.0 20.0- 00.0 3.0- 2.0- AFHZH

. 00.2 mud- 0N.0 30 05:0- 2.0 N00- 020 00.0 00.0 02.0- 02.0- 0N.0- Amy—minn—

. 00.2 05-0 mmd 0V0 5m.0 NN.0- 2.0- 20.0- 02.0 2.0 3.0 02.0 ﬂaw

00.2 00.0 mmd 3.0 #00 50.0- 02.0- 200- 5m.0 20.0 3.0 .22

. 00.2 mN.0 vmd m0.0 5.0- M00- 0N0- wmd mvd 0V0 7:9

. 00.2 00.0 52.0 m0.0- 50.0 3.0- 5N.0 0Y0 2.0 at.

. 00.— 52.0- 20.0- 00.0- 20.0- 50.0 00.0 00.0 2.5

. 00.2 3.0- 000 2.0 20.0 00.0- N00 20

. 00.— 2.0 02.0- 3.0- 2.0- 3.0- .950

. 00.2 2.0 50.0 52.0- 2.0 .095

00.2 who. 5m.0- 3.0- CD

. 00. 2 00.0 02 .0 083x059

00.2 no.0 Um:
. 8; macaw
ow< muz dam 0 .com 2 dew E0 a EU AFHZP AEWHZHQ mam M22 ED “2. 2.5 In .950 .005 On— coospo—osn— Um: weaver—m
.Hgocmcmmm we? coca—oboe 05 me one: 05 em 888.0 00 8&8 a 0m: 833:2, 28:08:83.8 0.3 own 5053 5022.50
.20.0vm .300 can? ”m0.0vm xaEm-ccsa .0m.0A e 220508 scum—oboe a :25, £28003 we 05 E 833:; 3:82.835
me $532 20% $3028 “5.59:8 3905.5 0.8 35.0.38 mpg—oat? 25:08:83.5 no 5.52: "802280 .2.N 033.

 

23

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A100. , .- lm I 40 I I
5:3: : .-
° " r - 30- J
§°40* ”£75
0 20" ' BSOF .1520" "
>~ 1‘-
8- :3 L ,
g 025 T - 10- g -
U 0
0 r l 0 1 1 O 1
1100 T I 10 I I 40 I r
E - - t
§9OOI I 9~ 3 ~53
t v -
3700' _‘ m 530 I T
.. ' ‘ q.- 'a #2
3:500" ‘
E - « 820' -
8300- - 7' “E3
'2 ~ -
8100 ‘ ' 6 . 4 10 - .
12 l I 6 F i m: I i ‘
10~ - is- - 60E
g 8r- - 34L -§30- I 1
DD v
5 6* l 53- -§
8 4+ 1 82- ~01” 1
. <1
2. J. .. . 1
0 1 O I l I
R E R E R E

Figure 2.2. Comparisons of physical and chemical characteristics between restored and
reference wetlands.

Nutrient concentrations in these two types of wetlands were different (Figure 2.3).
TP (1.858i0.314 mg/L) and SRP (0.406i0.102 mg/L) were high in most of the wetlands,
and had signiﬁcantly higher concentrations in extant wetlands than in restored wetlands
(P=0.007 and 0.027 respectively). TN (0.779zt0.069 mg/L overall) in extant wetlands was
not different from restored wetlands (P>0.1); the major form of inorganic N was NHI,
and N03 ‘was almost depleted in most of the wetlands (<0.010mg/L). DIN (005410.007)

was signiﬁcantly higher in extant wetlands (P=0.023) than restored wetlands. N:P and

24

 

DIN: SRP ratios were generally lower than 17. N: P ratio was higher in restored wetlands
than in extant wetlands (P=0.002) while DIN:SRP ratio was not signiﬁcantly different
between these two types of wetlands (P=0.210). Most of wetlands had mineral soils with
an average 0MP 17.4 (i2.3)%, which was higher in extant wetlands than in restored sites
(P=0.035). TN (3.911i1.106 mg/L) and TP content (0.757i0. 186 mg/g) in soil surface
was higher in extant wetlands than in restored wetlands (P=0.025 and 0.058 respectively),
but N: P ratio (3.78:1.86) in sediment were not different (P>0.128).

Many of the environmental variables related to trophic condition correlated with each
other (Table 2.1). LEC at 30 cm depth positively correlated with duckweed cover, chl a,
DIN, ammonia, TP concentration in the water colurrm, and sediment organic matter, and
negatively correlated with D0. D0 in the water column negatively correlated with
duckweed cover, LEC, and ammonia concentration. Most nutrient factors were also
correlated with each other. TN in water column positively correlated with SRP, DIN,
NH], and sediment TN. TP in water column positively correlated with SRP, ammonia,
chl a. Sediment TN and TP content were positively correlated with each other. LEC,
Duckweeds, SRP, and TP positively correlated with age of wetlands.

Age of restored marshes was not a signiﬁcant factor in differentiating the ﬁrst 3 PCA
axes. Environmental conditions between marshes less than 5-year-old and 5-10 year old
were not signiﬁcantly different (MANOVA, Wilks’s lamba F3,25=O.833, P=O.48) based
on the three PCA axes. None of the environmental variables correlated with age of

wetlands.

25

p—s
C9.

N:DM(mg/g)
‘2

h
C?

 

———'-1

 

 

 

 

 

 

 

 

 

 

(mg/L)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g

L l
'6

E

{

s_L

_. T
RE

SRP
y—s
O

 

V‘Vv

V—T‘IVVVI'" *'_' m".

 

 

 

 

 

 

 

y—a
9..

N:P__in sediment
‘2.

t—b
C?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:’ l
.r i a
.F i.
a :1:
=' 1
E a
i i
F .
E 1.
E 3
F 1.
R E

Figure 2.3. Wetland nutrient concentration and comparison between restored and extant

wetlands.

2.4.2 Non-diatom algal assemblage and diatom assemblages in restored and extant

wetlands

Of the 311 epiphytic algal taxa identiﬁed, diatoms had the highest taxa richness

(157), followed by Chlorophyta (99 taxa) and cyanobacteria (34). Other groups

(Euglenophyta, Dinophyta, and Chrysophyta) appeared but with less abundance.

Chlorophyta were the most abundant, with an average of 48.2% (i6.9% SE) of the total

cell density and 46.1% i5.0% of total biovolume. Stigeoclonium, Mougeotia, and

26

Oedogonium were the most abundant genera. The next abundant group in epiphytes was
diatoms with 30.3% (i4.5%) of total density and 26.8% (i 4.3%) of total biovolume.
Navicula minimum, Cocconeis placentula, Achnanthidium minutissimum, and Navicula
cryptocephala were the most dominant diatom taxa. Cyanobacteria, Euglenophyta,
Dinophyta, and Chrysophyta appeared but were low in density in epiphytic samples.
Total number of algal taxa (gamma diversity) and average taxa richness (alpha diversity)
were higher in restored wetlands (250 and 30.0 respectively) than in extant wetlands (189
and 25.8, respectively), mostly because of higher diatom gamma diversity in restored
wetlands (Table 2.2). Beta diversity of diatoms was also higher in restored wetlands than
in extant wetlands.

Of the 235 epipelon taxa identiﬁed, diatoms had both the highest taxa richness (199) and
abundance with an average of 54.4 % (i6.6% SE) of total cell density and 48.8% (:1:
6.5%) of total biovolume. Navicula minima, A. minutissimum, N. cryptocephala, and
Nitzschia palea dominated diatom communities. Chlorophyta was the next dominant
group with 19 taxa and 22.3 (i4.3%) of total density and 12.9% (i4.1%) of total
biovolume because of dominance of Scenedesmus in some of the wetlands. Only 8
cyanobacteria taxa were found, but they remained in low density (13i2.3%) and
biovolume (16.4i4.7%). Other groups (Euglenophyta, Dinophyta and Chrysophyta)
appeared but with relatively low abundance. Although the alpha diversity of non-diatom
algae and diatoms were pretty similar in restored (3.2 and 28.2 respectively) and extant
wetlands (2.9 and 29.7 respectively), the gamma diversity of diatom taxa in restored
wetlands (165) was much higher than in extant wetlands (129) and caused higher beta

diversity in restored wetlands (Table 2.2).

27

Phytoplankton biomass was low in most sampled wetlands (0.030i0.009 mg chl a/L)
and was not signiﬁcantly different between restored and extant wetlands (P=0.113). The
direct correlations of phytoplankton biomass with major nutrient gradients were weak
(r=0.368 with TP, r=0. 126 with TN), but were signiﬁcantly related to OMP in sediments
(r=0.499, P=0.025).

Phytoplankton taxa richness was relatively low compared to other habitats. A total of
153 non-diatom taxa were found in phytoplankton samples, but diatom density was very
low in most wetlands. Most of the phytoplankton were Chlorophyta (42.3%) and
cyanobacteria (33.1%) on the basis of cell density, only 15.2(:|:3.3%) of total algal
abundance was diatom. The average relative biovolume was dominated by dinoﬂagellates
(37.9i 4.7%), Englenophyta (26.5i3.7%), Chrysophyta (10.5i4.4%) and green algae
(9.0i2.9%). Cyanobacteria (3.5i1.6%) and diatom biovolume (2.3:l:O.7%) were very low.
Restored wetlands had higher alpha, beta, and gamma diversity than extant wetlands
(Table 2.2). Planktonic diatoms were not used for quantitative analysis because of their
extremely low density in half of the wetland sites.

Metaphyton appeared in the majority of the wetlands and completely covered the
basin of at least M: of wetlands. The most commonly found macroalgae were
Stigeoclonium, Oedogonium, and Mougeotia. Because of the difﬁculty of quantifying
their biomass, metaphyton were not used for further analysis.

Non-diatom and diatom diversity among different habitats were quite different.

Alpha, beta, and gamma diversity of non-diatom epiphyton and phytoplankton were

28

 

 

 

m6 M2: mmm :m WEN—m mgﬂﬁwm 3w?
0.2 5Q 0.2 m2 mm #2 mmdﬁﬁw Ndﬁod odﬂd 8036-52 86 =<
ﬁe >6 9.: 52 @303 :Hmdm 89me
wé ms 52 am: vdﬁcdm ﬁmﬁwwm 8w?
v; d ed Nd co mm aw modﬁos QNHQN v. 356 8036-52
mg» Nam 92 com Wmﬁhdm @392 SEED Eﬁxm
3 2 § own 3%.: 33.8 8% 5..
ed“ mm 0.2 _ _ _ on w: ondﬁmd mdﬂmﬁ. m. 35w 8036-52
m6 ﬁe mg mg wAHNwm 030.3 9:8me 62.38%
3:96:53 E0863 23:33 2:03:33 EoEEom oath—35m oEoE—EE EoEEom oﬂnmmdm
Sum 2555 69m + owﬁozwv 23¢.

 

aﬁcm—HOB uguxu ﬁg GDHOHMOH GDQBHB Cam “g..—6:03 N

551$ 38an 5953 .0936 Son 98 56.33% Acmaowucws «8:30 A86 com 38on mo “38:75 ﬂaccid «an? :32 .N.N 0331

29

higher than epipelon in all wetlands. Sediment diatoms had higher alpha diversity, higher

gamma diversity, but lower beta diversity than epiphytic diatoms (Table 2.2).

2.4.3 Difference of taxonomic composition among habitats and between restored and
extant wetlands

Differences in algal assemblages were commonly observed among habitats within
wetlands. Most comparisons of non-diatom algae among habitats were different in both
restored and extant wetlands, except that the difference between epiphyton and epipelon
assemblages was not signiﬁcantly different in extant wetlands (P=0.113) (Table 2.3,
Figure 2.4). Similarly, epiphytic diatom assemblages and sediment diatom assemblages
were not signiﬁcantly different in restored or extant wetlands (Table 2.3). The sample
ellipse of epiphytic diatom assemblages (Figure 2.4) was larger than that of sediment
diatoms, indicating a larger beta diversity for epiphytic diatom assemblages than
sediment diatoms.

Differences in species composition of algal assemblages between restored and extant
wetlands revealed habitat speciﬁc patterns (Table 2.4, Fig. 2.5). Both epiphytic non-
diatom algae and diatom assemblages in restored wetlands were signiﬁcantly different
from those in extant wetlands (Table 2.3, Fig. 2.5). However, sediment algae and

phytoplankton were not different between these two types of wetlands.

30

Restored wetlands Extant wetlands

 

 

1.0 it
Q 0.5 05
U)
D L
2 0.0 0.0 -
2

-0.5 '

   
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_10 -10 l .10 r 1
-1.0 -0.5 0.0 0.5 1.0 -l.0 -0.5 0.0 0.5 1.0 1.5
u 2.0 r u
'0 1.0 r a
0.0 " ‘ ‘
’1-0 h o . Epiphytes . °
A Phytoplankton '10 " -
if Sediments
‘21) i 1 1 1 ‘2.0 J 4
-l.0 -0.5 0.0 0.5 1.0 1.5 -1 0 1 2
NMDSI NMDSI

Figure 2.4. Non—metric multidimentional scaling (NMDS) of the restored and extant
wetlands based on non-diatom and diatom assemblages in different habitats. The circles
are the sample ellipse which is centered on the sample means of the x and y variables.
The size of the ellipse is speciﬁed at a probability value of 0.6827.

Table 2.3. MRPP comparisons of species composition in non-diatom algae and diatom
between different habitats in restored and extant wetlands

 

 

Comparisons Wetland T statistic A P
tI/Pe

Based on non-diatom algal biovolume

Epiphytes vs. Sediments Restored -3.940 0.0145 0.002
Extant - l .241 0.0051 0.1 13

Phytoplankton vs. Epiphytes Restored -5.333 0.0174 0.000
Extant -4.876 0.0276 0.001

Phytoplankton vs. Sediments Restored -8.660 0.0371 0.000
Extant -6.944 0.0442 0.000

Based on diatom relative abundance

Epiphytes vs. Sediments Restored -0.138 0.0006 0.389
Extant -1.1 18 0.0057 0.133

 

31

epiphytes

 

 

 

 

 

 

 

1.5 1 I r
g P=0.023 -
or) 1.0"
'3 N
a Eff .
8
.3 .0- ,
1':
2 -O.5 '
-10 1 1 1
-2 -1 0 l
2 I l
P=0.026
(D
E
O
a
.53
Q 3

 

 

 

NMDS]
Figure 2.5. Non-metric multidimentional scaling (NMDS) of the restored and extant
wetlands based on non-diatom and diatom assemblages in different habitats. The circles
are the sample ellipse which is centered on the sample means of the x and y variables.
The size of the ellipse is speciﬁed at a probability value of 0.6827.

 

 

    

 

 

 

 

 

 

 

    

 

 

 

sediment phytoplankton
1.5 I I LUV I 3 ﬁt
1. 0r P=0.762- - . 1 0'5 .
0.5” . " e
0.0’ . tr
0.0r -
_0.5_ 1'. _ -0.5- -
, P=O.594 ,
-10 ‘ -l.0 ‘ L ‘
“-2 -1 1 -1.0 -0.5 0.0 0.5 1.0
1'0 ' 110352
, " - - ' O Restored
0.5 i Q - X Extant
0.0 ' ' ) '
'05-05 0.0 0.5 1.0
NMDSl

Table 2.4. MRPP comparisons of species composition of non-diatom algae and diatoms
between restored and extant wetlands

 

 

habitats T statistic A P
Non-diatom Epiphytes -2.219 0.0070 0.023
algae Sediments 0.760 -0.0042 0.762
Phytoplankton 0.466 -0.0033 0.594
Diatom Epiphytes -2.322 0.0100 0.026
Sediment -0.279 0.0010 0.352

 

32

Table 2.5. Pearson correlation matrix of dissimilarities in species composition of non-
diatom algae and diatoms between restored and extant wetlands in different habitats. A *
indicates a P value <0.05.
Epiphytic Sediment Epiphytic non- Planktonic Sediment non-
diatom Diatom diatom algae non-diatom diatom algae

 

Epiphytic 1.000
diatom
Sediment 0613* 1.000
Diatom
EP‘Ph-V‘” 0560* 0440* 1.000
Non-dratom
Planktonic 0691* 0.377 0656* 1.000
non-dratom
Sedfmem 0.394 0.163 0418* 0477* 1.000
non-diatom

 

Differences in species composition of algal assemblages between restored and extant
wetlands consistently showed the same pattern in different habitats (Table 2.5). The
dissimilarities of non-diatom algae between restored and extant wetlands and the
dissimilarity of diatoms between these two types of wetlands were correlated (P<0. 1)
(Table 2.5). The only correlation not statistically signiﬁcant was between the dis-

similarities of sediment diatom and non-diatom algae.

2.4.4 Algal assemblages and environmental variables.

Multiple linear regression of non-diatom algal similarities among habitats indicated
that little variance could be explained by environmental variables (Table 2.6). P and
sediment TP had very weak effect on algal assemblages among habitats. However, none
of the regression models was signiﬁcant.

Bray-Curtis similarity between epiphytic and sediment diatoms, however, revealed

signiﬁcant regression model with shading by vegetation, LEC, and DO as predictors

33

(R2=0.346, P=0.001). With the increase of plant shading, LEC, and DO, similarity in
diatom species composition between epiphytes and epipelon increased.

Similar to inter-habitat comparison, similarity between restored vs. extant wetlands
using non-diatom algae could not be as easily explained by environmental differences as
diatom assemblages (Table 2.8). Only similarity of planktonic non-diatom algae between
restored and extant wetlands slightly responded to canopy coverage (P=0.132). Non-
diatom algae in other habitats did not respond to environmental variables.

Similarity in diatom species composition between restored and extant wetlands in
different habitats were related to different environmental factors (Table 2.8). Similarity of
epiphytic diatom species composition between restored and extant wetlands could be
negatively affected by LEC difference and positively to differences in plant shading.
Similarity in sediment diatom species composition related to a number of
factors, such as differences in pH, plant shading, and nutrient differences between
restored and extant wetlands (R2=0.540, P=0.007). With the decrease of nutrient
difference between restored and extant wetlands, similarity in diatom species

composition increased.

34

 

 

 

 

Sod End 3 nommmocwom
:2. $2- 485 um:
83 Ram. Rob 09228 386 35868 .3 maraam
03 o 95 N- Re C OD 2835
83 RE 23. “2%qu
Sod woman 33 052286
god Sod- 0%
Red $56 com; m Ed- m6 53626.80 8.886% .w> song—goinm
ovﬂo Sod- In ....me 8036-52
3 3 Sad 23- RS 2 asses 5 a 35863. as 3.2%....—
EE m8: 8wa 43¢. a. $5. 8.45392“. as 83%.QO
:3. NE 2:? H E30580 z :oEmanoo
mm assa> 935.": 868m

 

.moBatg 3588560 622283: 03 333.3.»

E65365 acme—988$ 8066 82:68 6.3 069360 :33qu b56636 6cm Jaw—a Eoﬁ6éo= 82:68 65 .=o§§_ao§ﬁ
.osxnmmao wcoEm 5528? mm 03mg; “counomon— $2.662, 6065365 .6 868.8 83%.; 6.822 .3 :ommmocwom dd 033.

35

 

 

 

36

 

Sod 02 .o 8 8088mm
3 3 ~88 owed 8828 825
Rod 83- 83. um.—
Sod 88.2 mm; 2588 8:83
Bed 9.2 8 80888 8585
wood :88 83- A; 82:88
God 23- 85 88260 888
Sod 83 80.0- 28
8. .o 83 :3- E
Eco v8.7 23 :8
Sod 38.8 83 08080 35868
o 8 05568
83 2 S 82- 83 8 8828 >856 8883 ...8.“ 898.82
o 8 88880
m 0.838 m 020?“ 806F600 Z 0030.8; 3860: 80688800 8623

 

.6820? 88x0 :0 :65 68603 606808 s 800309 0608;
a 6 0008306 0w80>0 80 00308; 80680806.: .6560? 80:0 63» 68603 606808 a 8 80688800 860% 6%? .6 58:88
6.8043}! 09808 8 038.8.» 8068082“ @0588; 806808068 6 8060060 083806 68308 .3 82000800 .8085 gum 030E

2.5 DISCUSSION

Distinct ecological differences were evident between the restored and extant wetlands
in southern-central Michigan. In addition to difference in shading by vegetation and
duckweeds on water surface, nutrient concentrations and soil organic matter proportion
also differed between these two types of wetlands. Differences in algal species
composition between two types of wetlands depended on habitat. Epiphytic assemblages
were more different between restored and extant wetlands than sediment and
phytoplankton assemblages. Algal assemblages also differed among habitats. Abiotic
environmental variables were related to many of the changes in algal assemblages among
habitats in wetlands and between restored and extant wetlands. Age of a restored site was

not a determinant factor for attributes in these wetlands.

2.5.1 Wetland restoration and environmental change in restored wetlands
Althought the initial goal of restoration is usually to restore the original physical and
chemical characteristics and biological communities of ecosystems, the actual
consequences of wetland restoration and construction are variable (Mitsch and Wilson
1996, Kaiser 2001). Wetland restoration can recover some aspects of the lost function
and structure of pristine wetlands (USEPA 2001), but most evidence (van der Valk 1981,
Galatowitsch and van der Valk 1996, Magee et a1 1999) indicates that restored wetlands
differ from extant wetlands in a number of ways. Restored wetlands generally have
different hydrogeomorphology, more ﬂood area with expanse of open water (Gwin et al.
1999), narrow borders of vegetation (Kentula et al. 1992, Bedford 1996), low ﬂoristic

diversity, and different ﬂoristic composition (Galatowitsch and van der Valk 1996,

37

Magee et al. 1999) than natural wetlands. In this study, the extant wetlands in southern
Michigan generally had complete emergent, submergent, and ﬂoating zones. Vegetation
in these wetlands was more diverse, and vegetation shading was higher than restored
wetlands. Restored wetlands were mainly composed of submergent zone with fewer
sedge meadows. Restored vegetation was composed of one or fewer species, and canopy
cover was lower. The higher canopy cover and duckweeds cover in extant wetlands
indicated light limitation for algal growth.

A number of studies have shown that extant wetlands in human impacted landscape
had much higher nutrient concentration and sediment organic matter than restored
wetlands, which has been related to long—term accumulation and wetland’s function in
nutrient retention (Craft et al 1991, Holland et al. 1995, Craft and Richardson 1997).
Studies (Bishel-Machung et al. 1996, Craft and Casey 2000) also found reference
wetlands had higher C and N content in sediments than in constructed wetlands.
Continuous saturation near the soil surface should limit decomposition of organic matter
and retain N in natural wetlands (Stolt et al. 2000). My results also showed that nutrient
concentration in the extant wetlands were well above average level (MDEQ 2001b) in
southern Michigan, while restored marshes were at the beginning of succession and had
relatively low nutrient concentration and organic accumulation. I also found signiﬁcant
differences in organic matter and nutrient concentrations between restored and extant
wetlands, which may cause different biological communities.

2.5.2 Difference of algal species composition between restored and extant

Presumably, algal assemblages in restored wetlands would increase its similarity with

extant wetlands and eventually would have the same assemblages as the extant wetlands

38

due to an increasing similarity of environmental characteristics. However, the starting
point of a restored wetland may be different from its original state. The physical,
chemical, and biological conditions during succession could also differ from natural
wetland succession. Human perturbation during succession could also signiﬁcantly
change the path of succession (Detenbeck et al. 1999). Thus, a restored wetland may not
necessarily develop to a state similar to mature natural wetlands (Mitsch and Wilson
1996, Lockwood and Pimm 1999, Kaiser 2001).

Although high variability in community composition among extant wetlands
suggested a high degree of environmental heterogeneity in this study, both environmental
conditions and epiphyton in restored wetlands still were signiﬁcantly different from those
in extant wetlands. None of the environmental variables or algal assemblage changes
correlated with age of restored wetlands in this study. No linear relationship between age
of wetland and organic carbon content in constructed wetlands was found. Previous
studies (Bishel-Machung et al. 1996) also indicated that many environmental variables
did not reach restoration goals within 10 years of initiating restoration. However, some
authors have suggested that wetland nutrient content might recover after more than 15
years of restoration (Craft and Richardson 1998). According to MDEQ (2001a), only
29% of the mitigation projects in Michigan were considered biologically successful, and
26% of the mitigation sites in southern Michigan were observed to have poor water
clarity. Mitsch and Wilson (1996) suggested that mitigation projects should be given as
much as 15-20 years rather than 5 years before judging their success. The apparent
difference of biogeochemical conditions between restored and extant wetlands make

diatom assemblages great indicators of wetland conditions.

39

Generally, restored wetlands had higher diatom gamma diversity, beta diversity, and
alpha diversity on epiphytes than observed in extant wetlands. Other studies have also
observed that restored wetlands did not necessarily have less biodiversity of plant and
invertebrates than reference sites (Scatolini and Zedler 1996, Bedford et al. 1999, Magee
et al. 1999, Knutson 1999, Williams and Zedler 1999), and no evidence indicated a lower
diatom species diversity in restored wetlands (Mayor et al. 1999). Two reasons might
have caused low diversity in extant wetlands. First, eutrophication decreases algal species
diversity and evenness (Detenbeck et al. 1999, Hillebrand and Sommer 2000). The
signiﬁcantly higher nutrient levels in extant wetlands may be the cause of lower alpha
diversity. The other possible explanation was that higher spatial heterogeneity of restored
wetlands at different stages of succession provided greater habitat variety for different
algal species. The lower beta diversity of diatoms in extant wetlands suggested that
diatom species dispersed evenly in extant wetlands, while the higher beta diversity
suggested heterogeneous distribution of diatom assemblages among habitats in restored
wetlands.

The patterns in similarity of diatom species composition in restored wetlands with the
extant wetlands were related to environmental factors that may regulate biological
differences during the ecological succession. Differences in algal species composition on
macrophytes between these two types of wetlands were complexly related to light
variables, but the model only explained 20% of the variance of diatom similarity among
wetlands. On the other hand, the similarity in sediment diatom species composition
between extant and restored wetlands could be explained by difference in pH, nutrients,

and light effects. This model was much stronger (explaining 54% of the variance in

40

 

diatom similarity) and indicated the importance of nutrient accumulation and light

regime, particularly for regulating algal assemblages during wetland succession.

2.5.3 Algal species composition among different habitats within wetlands

Phytoplankton communities were very different from benthic algal assemblages. As
expected, phytoplankton and benthic algal communities had different growth forms
which were adapted to planktonic and benthic habitats (Stevenson 1996). Phytoplankton
communities contained few diatoms, despite their abundance in the benthos. The majority
of southern Michigan wetlands in this study were in open and shelter states (Goldsbough
and Robinson 1996), which should be dominated either by epiphyton or metaphyton.
Phytoplankton density was generally very low. Duckweeds and macrophytes shading had
not only caused light limitation for phytoplankton, but may also have competed with
phytoplankton for nutrients as suggested in other studies (Herbst and Hartman 1981).
Phytoplankton had high beta diversity, and the structural features of algal assemblages
lacked a common pattern among sites. Grazing, allelopathy, light limitation, and
alternating self-organization processes in these wetlands make phytoplankton dynamics
more complex than in lakes due to the large number of interacting factors (Rojo 2000a,
b).

The composition of non-diatom algae differed much more between benthic habitats
than diatom species composition. The difference of non-diatom algae between
macrophytes and sediments in restored wetlands was probably due to the very
complicated interactions of algae and their substrata, involving both physical and

chemical processes (Burkholder 1996). The physical and chemical habitats for benthic

41

algae on organic substrata is very different than on ﬁrm substrata of macrophytes. Lower
light due to denser macrophytes and duckweed cover may have constrained the
development of differences between non-diatom epipelon and epiphyton growth in extant
wetlands, the only comparison of non-diatom algae that was not signiﬁcantly different
among habitats.

Diatom assemblages on macrophytes and in sediments, however, were not different in
restored and extant wetlands. Lim et al. (2001) claimed that sediment diatom
communities reﬂect changes in the diatom communities on epiphytic and rock substrates
in response to limnological conditions. This study also demonstrated that sediment
diatom assemblages contained the majority of epiphytic diatom taxa within a wetland, but
included a wider range of species than epiphyton. Although epiphytic and sediment
diatom did not differ in restored and extant wetlands, the regression model indicated that
differences between these two habitats were a function of several different factors
including light, DO, and nutrient ratio.

Light and D0 are inter-correlated and may present different perspectives to diatom
assemblages. The effect of light on diatom assemblages and productivity has been
reviewed by Hill (1996). In addition to the demonstrated effects of conductance and pH
on diatom assemblages, % open canopy was another important variable that determined
patterns of algal taxa among sites in streams (Carpenter and Waite 2000). The direct
effect of increased macrophyte shading and duckweed cover in Michigan wetlands could
be providing light limited environment for epiphytes, thereby increasing the similarity of
epiphytic algae to those in the limited-light environment of sediments. The indirect effect

of decreasing light was to decrease DO concentrations. D0 is a comprehensive variable

42

and could be determined by algal productivity and plant and bacteria respiration (Kemp
and Dodds 2002). DO also related to many other environmental variables. The negative
relation between DO and NI-I4+ concentration may be related to high decomposition,
animal excretion, nitriﬁcation of ammonia, and decomposition of high N organic matter.
Duckweed cover might cause light limitation for algal productivity and low D0 in the
water column, but algal biomass was not eliminated, which was indicated by the positive
relation between duckweed cover and chl a concentration in the water column. Relatively
higher DO environment for epiphytic diatoms and lower DO for sediment diatoms may
be another regulating factor for autotrophic and heterotrophic diatom taxa and thus affect
diatom species composition.

In summary, the physical, chemical, and biological features of restored wetlands
differed from extant wetlands in a number of ways. In particular, light reduction and
nutrient concentrations, were distinctly greater in extant wetlands, and probably
contributed to differences in algal communities among habitats within wetlands and
between restored and extant wetlands. I found no direct evidence for increasing similarity
in algal assemblages between restored and extant wetlands with increasing age of
restored wetlands. However, response of algal assemblages to nutrient concentrations and
light, two factors distinguishing restored and extant wetlands, indicated indirectly that
successional processes in restored wetlands were regulating algal assemblages. Restored

wetlands < 10 years old may need more time to develop mature qualities.

43

2.6 REFERENCES

Adamus, PR. 1996. Bioindicators for assessing ecological integrity of prairie wetlands.
US Environmental Protection Agency, Western ecological division, Corvallis, OR.
USA. EPA/600/R-96/082.

American Public Health Association (APHA). 1998. Standard Methods for the Examination
of water and wastewater. 20th edition by LS. Clececeri, A.E. Greenberg, and AD. Eaton.
Washington D. C., USA.

Bedford, B. 1996. The need to deﬁne hydrologic equivalence at the landscape scale for
freshwater wetland mitigation. Ecol. Appl. 6: 57-68.

Beilman, D.W. 2001. Plant community and diversity change due to localized permafrost
dynamics of bogs in western Canada. Can. J. Bot. -Rev. Can. Bot. 79: 983-93.

Bishe-Machung, L., Brooks, R. A., Yates, S. S., & K. L. Hoover. 1996. Soil Properties of
reference wetlands and wetland creation projects in Pennsyvania. Wetlands 16: 532-
41.

Biondini, M. E., Bonham, C. D., & Redente, E. F. 1985. Secondary successional patterns
in a sagebrush (Artemisia tridentata) community as they relate to soil disturbance and
soil biological activity. Vegetatio 60: 25-36.

Burkholder, J .M. 1996. Interactions of benthic algae with their substrata. In Stevenson,
R.J., Bothwell, M.L., and Lowe, R.L. [eds] Algal Ecology: Freshwater benthic
ecosystems. Academic Press, San Diego. pp3-30.

Charles, D.F., Knowles, C., & Davis, RS. 2002. Protocols for the analysis of algal
samples collected as part of the US. Geological Survey National water-Quality
Assessment Program. Report No. 02-06: 215/299-1000. The Academy of Natural

Sciences, Patrick Center for Environmental Research, Phycology Section,
Philadephia, PA.

Clarke, K. R. 1993. Non-parametric multivariate analyses of changes in community
structure. Aust. J. Ecol. 18: 1 17-43.

Cowardin, L., Carter, V., Golet, F. C., & LaRoe, E. T. 1978. Classiﬁcation of wetlands
and deepwater habitats of the United States. U. S. Fish and wildlife Serveice,
Washington, DC, USA. FWS/OBS-79/31.

Cottingham, K. L. & Carpenter, S. R. 1998. Population, community, and ecosystem
variates as ecological indicators: Phytoplankton responses to whole lake enrichment.
Ecol. Appl. 8: 508-30.

Craft, C. B. & Richardson, C. J. 1997. Relationships between soil nutrients and plant
species composition in everglades peatlands. J. Environ. Qual. 26: 224-32.

Craft, C.B, Seneca, E. D., & Broome, SW. 1991. Poewater chemistry of natural and
created marsh soils. J. Exp. Mar. Biol. Ecol. 152: 187-200.

Craft, C.B., & Casey, WP. 2000. Sediment and nutrient accumulation in ﬂoodplain and
depressional freshwater wetlands of Georgia, USA. Wetlands 20: 323-32.

Detenbeck, N. E. 2001. Methods for evaluating wetland condition: Wetlands
classiﬁcation. EPA 822-R-01-007g. US. Environmental Protection Agency; Ofﬁce of
Water; Washington, DC.

Detenbeck, N. E., Galatowitsch, S. M., Atkinson, J. & Ball, H. 1999. Evaluating
perturbations and developing restoration strategies for inland wetlands In the Great
Lakes Basin. Wetlands 19: 789-820.

Galatowitsch, S. M. & van der Valk, A. G. 1996. The vegetation of restored and natural
prairie wetlands. Ecol. Appl. 6: 102-112.

Goldsborough, L. G., & Robinson, G. G. C. 1996. Pattern in wetlands. In R.J.Stevenson,
M.L.Bothwell, and R.L.Lowe eds “Algal Ecology: freshwater benthic ecosystems.
Academic Press, San Diego. Ppl21-48.

Gosz, J .R. 1992. Gradient analysis of ecological changes in time and space: implications
for forest management. Ecol. Appl. 2: 248-61.

Gwin, S.E., Kentula, M.E., & Shaffer, P.W. I999. Evaluating the effects of wetland
regulation through hydrogeomorphic classiﬁcation and landscape proﬁles. Wetlands
19: 477-89.

Herbst, R.P., & Hartman, RT. 1981. Phytoplankton distribution of a duckweed covered
pond. J. F reshw. Ecol. 1:97-111.

Hill, W. 1996. Effects of light. In R.J.Stevenson, Bothwell, ML, and Lowe R.L [eds.]
Algal Ecology: freshwater benthic ecosystems. Academic Press, San Diego. Pp121-
48.

Hillebrand, H., Durselen, C., Kirschtel, D, Pollingher, U., & Zohary, T. 1999. Biovolume
calculation for pelagic and benthic microalgae. J. Phycol. 35:403-24.

Hillebrand, H. & Sommer, U. 2000. Diversity of benthic microalgae in response to
colonization time and eutrophication. Aquat. Bot. 67: 221-36.

45

Hillebrand, H. & Kahlert, M. 2001. Effect of grazing and nutrient supply on periphyton
biomass and nutrient stoichiometry in habitats of different productivity. Limnol.
Oceanogr. 46: 1881-98.

Holland, C. C., Honea, J. E., Gwin, S. E., & Kentula, ME. 1995. Wetland degradation
and loss in the rapidly urbanizing area of Portland, Oregon. Wetlands 15:336-345.

John, J. 1993. The use of diatoms in monitoring the development of created wetlands at a
sandmining site in western Australian. Hydrobiologia 269/270: 427-436.

Kaiser, J. 2001. Recreated wetlands no match for original. Science 293: 25.

Keddy, P. 1999. Wetland restoration: The potential for assembly rules in the service of
conservation. Wetlands 19: 716-32.

Kemp, M.J. & Dodds, W.K. 2002. The inﬂuence of ammonium, nitrate, and dissolved
oxygen concentrations on uptake, nitriﬁcation, and denitriﬁcation rates associated
with prairie stream substrata. Limnol. Oceanogr. 47: 1380-93.

Kentula, M E., Brooks, R. P., Gwin, S. E., Holland, C. C., Sherman, A. D., & Sifness, J.
C. 1992. Wetlands: An approach to improving decision making in wetland restoration
and creation. Island Press, Washington, DC. 151.pp.

Knutson, M. G., Sauer, J. R., Olsen, D. A., Mossman, M.J., Hemesath, L.M., & Lannoo,
M.J. 1999. Effects of landscape composition and wetland fragmentation on frog and

toad abundance and species richness in iowa and Wisconsin, USA. Conserv. Biol.
13: 1437-46.

Koerselman, W., & Beltman, B. 1988. Evapotranspiration from fens in relation to
penmans potential free-water evaporation (e0) and pan evaporation. Aquat. Bot. 31:
307-20.

Leland, H. V. 1995. Distribution of phytobenthos in the Yahima River Basin,
Washington, in relation to geology, land use, and other environmental factors. Can. J.
Fish. Aqua. Sci. 52: 1108-29.

Lockwood, J .L., & Pimm, S.L.1999. When does restoration succeed? In Weiher, E. &
Keddy, P. [eds] Ecological Assembly Rules: Perspectives, advances, retreats. Pp363-
92.

Magee, T. K., Ernst, T. L., Kentula, M.E., & Dwire, K. A. 1999. Floristic comparison of
freshwater wetlands in an urbanizing environment. Wetlands 19: 517-34.

Mayer, P. M. & Galatowitsch, S. M. 1999. Diatom communities as ecological indicators
of recovery in restored prairie wetlands. Wetlands 19: 765-74.

46

McCormick, P. V., Rawlik, P.S., Lurding, K., Smith, E.P., & Sklar, F. H. 1996.
Periphyton-water quality relationships along a nutrient gradient in the northern
Florida Everglades. J. N. Am. Benthol. Soc. 15:433—49.

McCormick, P. V., Shuford, R. B. E., Backus, J. G. & Kennedy, W. C. 1998. Spatial and
seasonal patterns of periphyton biomass and productivity in the northern Everglades,
Florida, USA. Hydrobiologia 362: 185-208.

McCune, B. & Mefford, J. 1997. PC-ORD. Multivariate Analysis Of Ecological Data.
Version 3.0 For Windows. Mjm Software Design, Gleneden Beach, OR, USA.

McDonnell, M. J. & Pickett, S. T. A. 1990. Ecosystem Structure and Function Along
Urban Rural Gradients - an Unexploited Opportunity for Ecology. Ecology 71: 1232-
37.

Michigan Department of Environmental Quality (MDEQ). 2001a. Michigan wetland
mitigation and permit compliance study, ﬁnal report. Land and water management
division, Michigan Department of environmental Quality, Lansing, MI.

Michigan Department of Environmental Quality (MDEQ). 2001b. MDEQ wetland
identiﬁcation manual: a technical manual for identifying wetlands in Michigan. Land
And Water Management Division, Michigan Department Of Environmental Quality.

Middleton, B. 1999. Wetland restoration: ﬂood pulsing and disturbance Dynamics. John
Wiley & Sons, Inc. New York. Pp6-7.

Mitsch, W.J., & Wilson., B. F. 1996. Improving the success of wetland creation and
restoration with know-how, time, and self-design. Ecological applications 6:77-83.

NRC. 1995. Wetlands: Characteristics and Boundaries. Nation Research Council.
National Academy Press, Washington, DC, USA.

Pan, Y., Stevenson, R. J ., Vaithiyanathan, P., Slate, J ., & Richardson, C. J. 2000.
Changes In Algal Assemblages Along Observed And Experimental Phosphorus
Gradients In A Subtropical Wetland, USA. F reshw. Biol. 44: 339-53.

Pringle, C. M. 1990. Nutrient Spatial Heterogeneity - Effects on Community Structure,
Physiognomy, and Diversity of Stream Algae. Ecology 71: 905-20.

Rader, R. B., & Richardson, C.J. 1992. The effects of nutrient enrichment on algae and
macroinvertebrates in the everglades: a review. Wetlands 12: 121-35.

Rojo, C, Ortega-Mayagoitia, E., Alvarez-Cobelas, M. 2000a. Lack of pattern among

phytoplankton assemblages. Or, what does the exception to the rule mean?
Hydrobiologia 424: 133-39.

47

Rojo, C. E., Ortega-Mayagoitia, E., Rodrigo, M.A., & Alvarez-Cobelas, M. 2000b.
Phytoplankton structure and dynamics in a semiarid wetland, the National Park "Las
Tablas de Daimiel" (Spain). Archiv Fur Hydrobiologie 148: 397-419.

Sayer, C., Roberts, N., Sadler, J ., David, C. & Wade, P. M. 1999. Biodiversity changes in
a shallow lake ecosystem: a multi-proxy palaeolimnological analysis. J. Biogeogr. 26:
97-1 14.

Scatolini, S.R. & Zedler, J .B. 1996. Epibenthic invertebrates of natural and constructed
marshes of San Diego Bay. Wetlands 16: 24-37.

Stevenson, R. J. 1996. An introduction to algal ecology in freshwater benthic habitats. In
R.J.Stevenson, Bothwell, M.L, & Lowe, R.L [eds.] Algal Ecology: freshwater benthic
ecosystems. Academic Press, San Diego. Pp3-30.

Stevenson, R. J. & Hashim, S. 1989. Variation in Diatom Community Structure among
Habitats in Sandy Streams. J. Phycol. 25: 678-86.

Stevenson, R. J. 1997. Scale-dependent determinants and consequences of benthic algal
heterogeneity. J. N. Am. Benthol. Soc. 16: 248-62.

Stevenson, R. J ., Sweets, R. P., Pan, Y., & Schultz, R. E. 1999. Diatom community
patterns in wetlands of Michigan (USA) and their use as indicators of ecological
conditions. In McComb, A. J. & Davis, J. A.[eds.] Proceedings of INTECOL’s Vth
International Wetland Conference. Gleneagles Press, Adelaaide, Australia. Pp. 517-
27.

Stevenson, R. J ., McCormick, P V., & Frydenborg, R. 2001. Methods for evaluating
Wetland Condition: Using Algae to assess environmental conditions in wetlands.
EPA 843-B-00-002k. US. Environmental Protection Agency; Ofﬁce of Water,
Washington, DC.

Stolt, M.H., Genthner, M.H., Daniels, W.L., Groover, V.A., Nagle, S., & Haering, KC.
2000. Comparison of soil and other environmental conditions in constructed and
adjacent palustrine reference wetlands. Wetlands 20: 671-83.

Stosch, H. A. von & Reimann, B.E.F. 1970. Subsilicea fragilarioide gen. Et spec. nov., eine
Diatomee (Fragilariaceae) mit vorwiegend organischer membran. Nova Hedwigia Beih.
3121-36.

Tilman, D. 1981. Tests of resource competition theory using four species of lake
Michigan algae. Ecology 62: 802-15.

USEPA, 2000. Principles for the Ecological Restoration of Aquatic Resources. EPA841-

F-00-003. Ofﬁce of Water (4501F), United States Environmental Protection Agency,
Washington, DC. 4 pp.

48

van Dam, H. Mertens, A. & Sinkeldam, J. 1994. A coded checklist and ecological
indicator values of freshwater diatoms from the Netherlands. Neth. J. Aquat. Ecol.
28:1 17-33.

van Der Valk, AG. 1981. Succession in wetlands: a gleasonian approach. Ecology 62:
688-96.

Vymazal, J. 1995. Algae and element cycling in wetlands. Lewis Publishers, Boca,
Raton, FL. 689 pp.

Whigham, D. F., Good, R. E., & Kvet, J. 1990. Wetland ecology and management: case
studies. Kluwer Academic Publishers. Netherlands.

Williams, G. D. & Zedler, J. B. 1999. Fish assemblage composition in constructed and
natural tidal marshes of San Diego Bay: Relative inﬂuence of channel morphology
and restoration history. Estuaries 22: 702-16.

Wu, X. Y. & Mitsch, W. J. 1998. Spatial and temporal patterns of algae in newly
constructed freshwater wetlands. Wetlands 18: 9-20.

Zedler, J. B. 1993. Canopy architecture of natural and planted Cordgrass marshes:
selecting habitat evaluation criteria. Ecol. Appl. 3: 123-38.

Zedler, J. B. 2000. Progress in wetland restoration ecology. Trends Ecol. Evol. 15: 402-
07.

49

Chapter 3
CHANGES IN BENTHIC ALGAL ATTRIBUTES DURING SALT

MARSH RESTORATION AND USE IN BIOASSESSMENT

3.1 Abstract: Although salt marsh restoration has been used to mitigate loss of habitat
and ecosystem function and hundreds of restored marshes have been built along the US
coast, few effective indicators have been developed to determine success of ecological
restoration in salt marshes. To assess attributes of algal assemblages as indicators of salt
marsh restoration, I chose 8 pairs of salt marshes in North Carolina, each pair with one
restored marsh (from 1 to 28 year old) and a nearby reference salt marsh. Algae on
Spartina altemiﬁora and sediments (sediment algae) were collected in each marsh during
spring and summer 1998 for assaying biomass (dry mass (DM), ash free dry mass
(AFDM), chl a content, algal biovolume), species composition and diversity, and gross
primary production. An attribute restoration ratio was calculated by dividing attribute
values from each restored marsh by values from the paired reference marsh. Controlling
for regional variation in reference marshes substantially increased precision in relations
between attributes and age of restored marshes. Organic matter restoration ratios of
sediments increased with age of restored marshes in spring and summer. Algal biomass
restoration ratios of epiphytes, calculated with algal biovolume and chl a, increased with
restored marsh age in summer, but not during spring. In contrast, sediment algal biomass
was not related to marsh age. The species diversity of sediment algae in summer showed
an asymptotic relationship with sediment nutrient concentration. The similarity of diatom

species composition between paired restored and reference sites increased with age of

50

restored marshes during both spring and summer. Primary production by epiphytic and
sediment algae in summer showed site-speciﬁc changes and did not vary consistently
with marsh age. Algal biomass, algal diversity, and diatom species assemblages during
summer were positively correlated with sediment nitrogen and phosphorus concentration.
Overall structural and functional development of restored wetlands, especially nutrient
storage in sediments, regulates algal community structure and function, which can be

used to evaluate marsh restoration success.

Key Words: salt marsh, restoration, algal attributes, succession, algal diversity, diatoms,

biomass, nutrients, environmental variation

51

3.2 INTRODUCTION

Ecosystem restoration is one of the most pressing problems for applied ecologists
(NRC 1992). With expanding urbanization and other human activities along coasts, many
salt marshes have been degraded. A large effort has been mounted to stabilize these
coastal ecosystems (Adam 1990). Along the US coast, hundreds of degraded or lost salt
marshes have been restored in the past decades. Currently, few criteria exist to guide
assessment of wetland restoration, so their evaluation is often superﬁcial and subjective.
Furthermore, very few reference data sets exist that can be used to compare restored
habitats with natural habitats (Lockwood and Pimm 1999, Keer and Zedler 2002).

The most commonly used criterion for judging the success of ecosystem restoration is
whether or not the restored community resembles the original (Adamus 1988).
Developing criteria to evaluate progress in restoration can be done by characterizing
development of a salt marsh ecosystem following initial restoration and identifying
attributes in that marsh that increase in similarity relative to mature, existing marshes.
Alternatively, multiple restored marshes with different ages could be compared to
reference marshes to relate successional processes in restored ecosystems to a reference
system (Lockwood and Pimm 1999). Living and nonliving organic matter, species
diversity, and nutrient retention increase during ecosystem development (Odum 1969).
Previous studies of restored salt marshes suggest that some of these attributes develop
very slowly (e.g., 15 years for soil nutrient pools, Craft et al. 1988), while others recover
within 1-5 years (Craft et al. 1991). Slow changes in biogeocherrristry, water chemistry

and plant and invertebrate species composition demonstrate the importance of long-terrn

52

research to characterize ecosystem succession (Hook 1988, Craft et al. 1991, Sacco et al.
1994, Scatolini and Zedler 1996).

Algae are important primary producers in salt marsh habitats and affect secondary
production, nutrient retention, and biogeochemical cycling (Pomeroy 1959, Leach 1970,
Gallagher and Daiber 1974, Van Raalte 1976, Zedler 1980, Sullivan and Moncreiff 1988,
1990, Pinckney and Zingmark 1993a, b). Algal photosynthesis can account for a large
proportion of primary production in salt marshes, ranging from 25% to as much as 70%
in some coastal marshes (Adamus 1988). Teal (1962) estimated that 47 percent of the
total net primary productivity in salt marshes was lost through respiration by microbes,
however, algal debris still contributed to more than one third of soil organic matter. Algae
are also a major energy resource and provide the primary diet for metazoans and other
invertebrate grazers (Kreeger and Newell 2000). Finally, nitrogen ﬁxed by cyanobacteria
is a very important nutrient resource in nitrogen limited salt marsh ecosystems (Jones
1974, Raalte et al. 1976, Sage and Sullivan 1978).

Algal species diversity, species composition, biomass, chemical composition, and
productivity are most important attributes for assessing ecosystem structure and function
in aquatic systems (Stevenson and Pan 1999). Algae, especially diatoms, have been
widely used as indicators of ecological conditions in freshwater, marine, and terrestrial
ecosystems (Shubert 1984, Steneck and Dethier 1994, Stevenson and Pan 1999, Sullivan
and Currin 2000, Stevenson et a]. 2001), but they have never been used to assess salt
marsh condition. Because of their species-speciﬁc response to biogeochemistry (Shubert
1984), changes in diatom community structure and function could be used to assess

restoration of wetland structure and function.

53

In this project, algal attributes in restored salt marshes of different ages were
compared with reference marshes to answer the following key questions: 1) what
timescales are required to achieve full restoration of algal-related attributes of salt marsh
structure and function; 2) which indicators can be used to determine whether full
restoration of salt marsh structure and function has been achieved; and 3) can ecological
succession theory apply to salt marshes restoration? I hypothesized that l) algae would
rapidly colonize restored salt marshes and establish high primary production; 2) algal
species diversity and organic matter on the surface of sediments Would increase slowly
over time and became more similar to reference sites with greater age of restored
marshes; and 3) diatom species assemblages and organic matter in sediments would be

important indicators of successful restoration.

 

 

3.3 METHODS
3.3.1 Site Locations
J; 30 80 90 120 150 Kilometers ; M
an“ 7 $-
, ‘ . . w E
Morehead Clty 2 5
A OI
‘V / \\
‘> .. *‘ Marine Lab Marsh (26)
~ Consultant Marsh (3)
" W"'“"‘9‘°" . Port Marsh (3)
. 4.2-5% , _ ' Dill’s Creek (13)
. ‘ DOT Marsh (1)
‘ V \ (I Fine Knoll shores (24)
1
“ Swansboro (1 1)
4
Snow’s Cut (28)

 

 

 

Figure 3.1. Map of sampling sites between Morehead City and Wilmington along the
North Carolina coast. Age (yr) of restored marshes is reported in parentheses.

54

Salt marshes were sampled in the urbanizing area between Morehead City and
Wilmington along the North Carolina coast (Fig. 3.1). The restored marshes examined
were established for shoreline stabilization, mitigation of wetland loss, or research. The
reference marshes were natural marshes near the restored marshes so that differences in
water chemistry factors that affect algal assemblages, such as salinity and nutrient
concentrations, would not vary due to regional factors. Each pair of wetlands received
water from the same source. Eight pairs of salt marshes were selected and sampled in
March 1998 (Figure 3.1): Long Beach (LB), Consultant Marsh (CS), Port Marsh (PT),
Swansboro Marsh (SB), Dill Creek (DC), Pine Knoll (PN), Marine Lab (ML), and
Snow’s Cut (SC). All restored marshes were planted with Spartina altemiﬂora Loisel at
the time of restoration. Long Beach and Pine Knoll marshes were dropped from the study
in July 1998 because of the poor macrophyte growth and an additional pair of salt
marshes (Department of Transportation (DOT)) was added.

To determine whether changes in developing salt marshes of North Carolina were
similar to those in other regions, samples were collected from 3 pairs of restored and
reference salt marshes in Chesapeake Bay at Virginia Beach, VA during summer 1999.
The restored VA marshes were 2, 10, and 20 years old. These marshes represent very
different habitats than those in North Carolina because they were far from the ocean, had

salinity below 10 ppt, and were brackish to oligohaline.

55

3.3.2 Field Sampling

An area referred to as "the front marsh" was sampled. This area was approximately 1-
2 meters from the levee away from the water and relatively low and ﬂat. Sampling
transects 30 to 50 m long were established parallel to the beach in this zone. Because
some of these marshes did not have well-developed levee zones and several of them
didn’t have Spartina in back marshes, only the front marsh area was included in this
study. Spartina stem density was measured in 0.25 m2 plots (10 per marsh) in October
1998. Plant stem density in the front marsh zone of reference sites ranged from 201 to
316 stems/m2 except in Pine Knoll marsh, which had much higher stem density (855/m2)
than any other reference sites. Restored sites had stem density ranging from 240 to 498
stems/m2 (Table 3.1). Salinity was measured in the ﬁeld using a refractometer. Surface
soils (0-10 cm, n=10 per marsh) were collected in June 1998 and analyzed for total N and
P as described in Craft et a1. (1999). N and P concentrations in younger restored marshes
were generally lower than its nearby reference marshes (Table 3.1).

Table 3.1. Sampling sites and their properties during the summer sampling season
(from Robert Freese and Christopher Craft, unpublished data).

 

 

Salinity(ppt) Plant Stem density (lmz) Nitrogen (mg/g) in Phosphorus (mg/g) in
surface sediment surface sediment
Age Site Reference Restored Reference Restored Reference Restored
1 DOT 20-30 266 498 2. 1 0.2 0.264 0.21 l
3 CS 17-32 208 304 2 0.2 0.368 0.143
8 PT 18-30 255 342 1.2 0.3 0.955 0.263
1 1 SB 20-30 316 398 4.8 0.4 0.545 0.409
13 DC 14-33 280 240 3 1 0.358 0.236
24 PN 20-30 855 512 0.3 0.6 0.453 0.370
26 ML 20-30 201 423 2. 1 3 0.369 0.408
28 SC 5-20 271 297 4.4 3 0.492 0.371

 

56

Algal productivity on sediments and plants was measured during summer 1998. Eight
sediment samples were collected from each transect by using plastic tubes (internal
diameter = 3.8 cm, length = 20 cm) along each restored and reference site. Samples were
taken early in the morning and brought back to the shore immediately. The sediments in
the corers were covered with 30-ml of estuary water and then sealed air-tight with a thin,
transparent plastic sheet and rubber bands. The cores from each transect were incubated
in a light or dark chamber for 4 hours (4 cores in each treatment). Subsamples of 15 ml
seawater were taken from each core to measure dissolved oxygen concentration (DO)
using Winkler methods (APHA, 1998). The top l-cm of sediment from each core was
sampled and stored on dry ice for later analysis of algal biomass.

Epiphytic samples were collected from front marshes by cutting 1-3 randomly
selected stems of Spartina from just above the ground and placing them into a zip-lock
bag. Eight bags of stems were randomly collected in each zone of each marsh. All stems
in a bag were washed using 200 ml of estuary water and removed from the bag. The
epiphytes were left in the bag. All bags were then sealed underwater to avoid trapping air
bubbles. The surface area of macrophytes from where the samples were taken was
measured using a ruler. Four samples from each site were put in a light chamber, four
were put in a dark chamber, and all were incubated for 4 hours. The experiments were
done under light irradiance ranging from 800 to 2,200 umole quanta'm‘zs’l, 32 ppt
salinity, and a water temperature of 30il°C. DO was measured using 60 ml subsamples
before and after incubations. The remaining portions of samples were frozen on dry ice

for later analysis of algal biomass.

57

3.3.3 Laboratory Analysis

In the laboratory, the frozen algae sample was homogenized and split into different
volumes for assay of organic mass (30 ml), chlorophyll (15 ml), and separate microscopic
assays for all algae (10 ml) and diatoms (10 ml). 200 ml of ﬁltered seawater were added
to sediment samples to enable subsampling. Dry mass (DM) and ash free dry mass
(AFDM) were analyzed using standard methods (APHA 1998) to estimate sediment and
epiphytic mass and organic matter. The organic matter proportion of sediments (OMP)
was then calculated based on the organic proportion of total sediment (AFDM/DM).
Chlorophyll a (chl a) contents of sediment algae and epiphytes in summer were measured
using a Spectronic GenesysZ® spectrophotometer (Spectronic Instruments) following
standard methods (APHA 1998). Ch] a and phaeophytin were related to area of ground
and plant surface sampled.

Algal biomass and species composition of all algae and diatoms were determined by
microscopically identifying and counting algae using two preparation techniques. After
suitable dilution and settling, 300 natural units of algae in the chamber were identiﬁed
and counted within a known portion of the samples. Non-diatom algae were recorded and
identiﬁed to genus; diatoms were divided into 5 groups according to their length and
width (less than 10, 11-20, 21-40, 41-60, and >60 um) and counted. Soft algal volume
was determined by measuring at least 10 individuals or colonies. Average volumes of
diatom size groups were estimated by measuring 30 cells for each group. Total algal
biovolume was calculated to provide a second measure of algal biomass with chl a.

Diatom samples were mounted in N aphrax® after digestion using HNO3 and KzCr207

58

(Stosch and Reimann 1970). Diatoms were identiﬁed to species and at least 500 valves
were counted for each sample.

Among all pairs of restored and nearby salt marshes, I determined that Snow’s Cut
(28 year old) was a poor reference site for the nearby restored marsh, which was located
in the middle of the Cape Fear River channel. Differences in algal assemblages between
these marshes were probably related to large differences in salinity, nutrient input, and
sedimentation. Therefore I decided to exclude this site from analyses that involved

diatom species composition.

3.3.4 Statistical analyses

I used SYSTAT® 10 to calculate summary statistics, t-test for means, F statistics for
variance, and Analysis of variance (AN OVA) to compare attribute characteristics
between restored and reference marshes. Analysis of Covariance (ANCOVA) was used to
compare slopes of two different regression lines.

Attribute restoration ratios were calculated to indicate relative condition in a restored
site compared to the nearby mature reference marsh. A ratio equal to one meant that the
attribute of the restored marsh was the same as in reference marsh. Thus, the organic
matter restoration ratio (OMRR) was deﬁned as (OMP in restored marsh)/(OMP in
reference marsh). An algal chlorophyll restoration ratio (ACRR) was similarly calculated
to relate chlorophyll concentrations in restored and reference marshes. The algal
biovolume restoration ratio (AVRR) was deﬁned as the ratio of total algal biovolume per
unit surface area in a restored marsh vs. biovolume in the respective reference marsh and

was calculated for both epiphytes and sediment algae. A primary production restoration

59

ratio (PRR) was deﬁned as the ratio of gross primary production per unit surface area in a
restored marsh vs. that in its reference marsh for both epiphytes and sediment algae.

A percentage similarity index (PSI) was used to indicate the similarity of diatom
species composition in restored and nearby reference marshes (Gauch 1982). The index
was calculated as

PSI=E minimum (pn, p2,),
where PSI: percentage similarity between samples 1 and 2, p1, = percentage of the ith
species in the restored marsh (i = 1,. . ..., k species), and pzi = percentage of species i
corresponding reference marsh.

PSI of algal functional groups was also calculated by comparing the biovolumes of
functional groups in restored and reference marshes. Functional groups were deﬁned by
algal division and growth form (Steneck and Dethier 1994). All algae were divided into
9 functional groups: 5 diatom size groups, cyanobacteria, green algae, brown algae and
red algae.

To evaluate successional patterns of change of algal attributes, linear regression
analyses (SYSTAT ®10) were used to relate algal attributes, restoration ratios, and
diatom community similarities to age of restored marshes in both spring and summer.
Nonlinear regression analyses were also performed to relate changes in attributes to
environmental variables (SYSTAT® 10). Canonical correspondence analysis (CCA) was
performed using the computer program CAN OCO (ter Braak, 1987) v4.02 to relate
diatom species data to environmental variables. Sediment and epiphytic diatom taxa with
relative abundances (taxa proportions of assemblages) greater than 1% in one or more

sites during summer were included in our CCA with respect to the available

environmental variables: sediment total nitrogen, total phosphorus, salinity, stem density,
OMP, and chl a. A forward-selected CCA was then used to determine which
environmental variables accounted for the greatest amount of variance in the distribution
of the diatom taxa. The signiﬁcance of each variable added in this fashion was tested
using a Monte Carlo permutation test with 99 unrestricted permutations (P _<. 0.05) (ter
Braak 1990).

To determine the importance of regional factors, three pairs of restored and reference
salt marshes from the Virginia coast were compared to the North Carolina summer

sample set to determine how similar restoration patterns were between different regions.

3.4 RESULTS

3.4.1 Biomass Development

08 '- u u v i d 0 r u u u
06 I- ¢ .. 03 $ 1

 

 

l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a? 0.4 “L? p B d
E g .. .
. . 0.02 .
in 0.2 [F §
2 z .. .
Q 0 E 0.01 .
O
r: $
I T I I 50 I t v 30 1 y .
A1000 [ B . 1 b , a; 20 . 0 ¢ .
NE : i o E 10 "
E 10.0 f H g 1:; 30 - i E o
g r i g 20 P ' "g: 5 p a
g 1.0 I 1.33.
a a 10' .43. 5
r ' 2
0.1 n l l O Q" ‘ .
as" 59‘“ 59" 9“" ‘39 5““ $3 9““ 9““ 9°“
Epiphytes Sediment Epiphytes Sediment Epiphytes Sediment

Figure 3.2. Box-plot of dry mass (DM), ash free dry mass (AFDM), the organic matter
proportion of sediments (OMP), algal biovolume, and chl 3 samples from epiphytes and
sediments in reference marshes during spring (8 samples) and summer 1998 (7 samples).

61

Characteristics of sediments (DM, AFDM, and OMP) and algal biomass (chl a and
biovolume) in reference marshes differed between macrophytes and surface sediments
(Figure 3.2). During spring, DM and algal biovolume on macrophytes were much lower than
in surface sediments (t-test, P<0.05), AFDM on macrophytes was not different from sediment
(P>0.05), but the OMP on macrophytes (27.3128 %) was higher than in sediment (14.2:l:3.7
%)(P<0.05). During summer, DM, AFDM, algal biovolume, and chl a on macrophytes were
all signiﬁcantly lower than in surface sediments (P<0.05), but OMP on macrophytes did not
differ from sediment (P>0.05).

Sediment mass (DM and AFDM), algal biovolume, OMP on macrophytes, and algal
biomass in surface sediment were higher in spring than summer (P<0.05). Measures of
sediment mass (DM and AFDM) were less in spring than summer on surface sediment
(P<0.05) (Figure 3.2).

DM, AFDM, and OMP in restored marshes were not related well to restored marsh
age (P>0.05). However, the organic matter restoration ratio (OMRR) on macrophytes and
in sediment during spring was linearly related to age of restored marshes (Figure 3.3).
During summer, the OMRR on macrophytes was not related to age of restored marshes
and the OMRR in sediments again signiﬁcantly increased with age of restored marshes in
a manner not statistically different from that observed in spring (ANCOVA, P=0.450).

Algal biomass in young restored marshes was about half that observed in mature
reference marshes. In general, algal biomass restoration ratios increased with age of
restored marshes. AVRR was not related to restored marsh age during spring either on
macrophytes and in sediments (P>0.05) (Figure 3.3e-f). During summer, sediment AVRR
and sediment ACRR were not related to restored marsh age, but epiphytic AVRR and

ACRR were signiﬁcantly related to age of restored marshes (Figure 3.3g-j). The AVRR

62

patterns were statistically similar to the ACRR pattern for epiphytic and sediment algae

in summer (ANCOVA, P=0.375 and P=0.453 respectively).

 

 

 

 

 

 

OMRR AVRR ACRR
1.5 . . , , .
3e

1.0 , ..................................................... _
0.5 ' ° .
0.0 :7 :' % If :
1.5 3f .
1.0 , ................. .
0.5 ° . .

00 ‘ l l l l l r r r 1 l

 

 

 

 

 

 

0 51015 20 25 30 0.00 51015 20 25 30

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

2.0 I I r T I 1.5 I T I I I 1.5 F I I I I
O .
3C . 3g 31
I- II .
E 1 5 / 1.0 ...; .................................................... ....“ 1.0 ¢ ' 0
g o
.5 1 O " o .
ran) ' . 0.5 - ° - 0.5 .o -
E- 0.5 " ‘ . . '
0.0 l l l 4 0-0 l A r 4 1
(’5) 2.0 r 1 I I T 0'0 v t v v 1.5 .V F I l I
. 3d 1.5

m 1.5 " 7
v2, . 1.0 - 1-0‘
£1 1 0 - .
8‘ ° . ' 0-5 0.5

0.5 l‘ ‘

O O l 1 1 l I 00 1 l l l 1 0.0 O V = 5 : =

' 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30
Age of restored marsh(yr) Age of restored marsh(yr) Age of restored marsh(yr)

Figure 3.3. Relationship between the attribute restoration ratio and age of restored marsh
in surface sediment and on macrophytes during spring and summer 1998. The OMRR is the
comparison of OMP of sediments in a restored site with that of its reference marsh. The
ACRR is the comparison of algal chlorophyll a concentration in a restored site with that of its
reference marsh. The AVRR is the comparison of algal biovolume in a restored site with that
of its reference marsh. The line represents a signiﬁcant relationship (linear regression
P<0.05). The correlation coefﬁcients (R2) for each relationship are: 3a: 0.727; 3b, 0.931; 3c,
0.969; 3d, 0.235; 3e, 0.304; 3f, 0.647; 3g, 0.577; 3h, 0.749; 3i, 0.442; 3j, 0.741.

63

In summer, sample OMP increased signiﬁcantly with increasing nitrogen and
phosphorus concentration of surface sediments (Table 3.2). Sediment chl a concentration
was asymptotically related to sediment nitrogen concentration and was barely correlated
with surface phosphorus concentration (Table 3.2). Sediment algal biovolumes showed
the same correlation with N and P as chl a. All three attributes were not signiﬁcantly
related with macrophyte stem density (Table 3.2).

Table 3.2. Summary of regression analyses between algal attributes and environmental
variables. Each marsh (both restored and reference marshes) was treated as one site.

 

 

Attributes Environmental variables Adj R2 n Equation P(2 tail) Model
0.372 14 5.409+0.192X 0.012 linear
OMP TP 0.531 13 -0.870+30.306X 0.003 linear
Stem density 0.464 14 0.465 none
TN 0.459 14 23.278 «1» (X /(0.311+ )0) 0.039 asymptotic
Chlorophyll TP 0.232 13 5.501+33.551X 0.055 linear
Stem density 0.119 14 8.558+0.029 X 0.227 linear
TN 0.505 14 4.841+4.502X 0.003 linear
Biovolume TP 0.564 13 -8.015+15.990X 0.002 linear
Stem density 0.039 14 2.317+0.037X 0.241 linear
TN 0.540 14 3.546 * (X / (X+0.024)) asymptotic
Biodiversity TP 0.419 13 3.714 * (X / (X + 0.026)) asymptotic
Stem density 14 none

 

3.4.2 Algal Species Diversity and Composition

Table 3.3. Number of species found in all sites, habitats and seasons.
Habitat No. of algal species No. of diatom taxa

 

Season type Res. Ref. Total Common Res. Ref. Total Common

 

Spring Sediment 204 218 292 l37(46.9%) 212 197 277 132(47.7%)
Epiphyte 167 170 225 104(46.2%) 155 149 205 98(47.8%)
Summer Sediment 164 159 205 1 17(57. 1 %) 154 156 198 1 14(57.6%)

Epiphyte 131 118 161 85(52.8%) 112 123 150 81(54%)

 

A total of 420 algal taxa, including 400 diatom taxa, were found on all substrates in
the spring and summer samples. Of them, 362 algal taxa were found during spring and
the majority (342 and 94.5%) were diatom taxa. Generally, more algal species were
observed in sediment samples (292 taxa) than on plants (225 taxa) (Table 3.3). Fewer
algal species were found in summer than in spring. Two hundred ﬁfty-two algal taxa with
241 diatom taxa were found in all sampling sites in summer with more algal species (205
algal taxa) in sediments than on plants (161 species) (Table 3.3).

Restored and reference marshes shared fewer algal taxa (about 47%) and fewer
diatom taxa (about 48%) in spring than in summer (>52.8%)(Table 3.3). Shannon
diversity indices of epiphytes varied from 2.02 to 3.10 in restored marshes and 2.59 to
3.00 in reference marshes in summer, Shannon indices of sediment algae ranged from
2.93 to 3.55 in restored marshes and from 3.36 to 3.58 in reference marshes. Generally,
the index values for restored sites in summer were more variable than for reference
marshes (F test, P=0.002 for epiphytes and P=0.016 for sediment algae). However,
neither the epiphytic nor sediment diversity indices differed between restored sites and
reference sites (P>0.05). Asymptotic relationships were observed between sediment algal
diversity (H’ ) and sediment nitrogen and phosphorus concentrations (Table 3.2) in
summer.

Algal species were mostly composed of diatoms, ﬁlamentous green algae, and
cyanobacteria. The most frequently observed soft algal genera were Enteromorpha spp.,
Rhizoclonium spp., Oscillatoria spp., and Schizothrix spp. Enteromorpha was distributed
patchily in most marshes; however, because of the difﬁculty of quantifying the

occurrence of macroalgal species such as Enteromorpha spp. and relative few numbers of

65

macroalgae taxa, they were not used to compare species composition of algal
assemblages. The most common epiphytic algal group was diatoms with an average of
75% (il3% SE) of total algal biovolume in spring and 89% (i4.8%) in summer. Diatoms
also were dominant in sediment assemblages, with an average of 73% (21:12%) of total

biovolume in spring and 93% (i4%) in summer. Epiphytic cyanobacteria algae were
0.3% to 35.2% of total algal biovolume in restored marshes and from 0.04 to 22.3% in
reference marshes. Sediment cyanobacteria were less than 10% of algal biovolume in
both restored and reference marshes in spring.

PSI of functional group biovolume between restored and its respective reference
marshes did not change with restored marsh age in spring or summer (Figure 3.4a, b).
Most of the paired sites had high similarity of functional group composition (larger than
50%) on both macrophytes and sediments.

Diatom composition on macrophytes and in sediment shifted seasonally. The most
dominant epiphytic species on macrophytes in spring were Melosira nummuloides Agardh,
Denticula subtilis Grunow, Synedra fasiculata Grunow, and Navicula penninuta Grunow.
During summer, M. nummuloides became much less abundant, while Nitzschia nana
Grunow, Amphora coffeaeformis (Agardh) Kutzing, and Nitzschia frustulum Grunow became
the most dominant epiphytic species. Dominance of sediment diatom assemblages changed
little from spring to summer. With M. nummloides, A. cojfeaeformis, and Rhaphoneis

surirella (Ehrenberg) Grunow being most common (Table 3.4).

66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Spring 4a Summer 4b
0.8 I I I I I 1.0 I I t j
0 o
a, jg 0.6 - - 0-3 l . ' '
8 g 0.6 ' ° .-
9 g 0.4 " ‘
3’ a) ' 0.4 " '
g 02 II . .u
,9 ' 0.2 ' "
E 0.0 . . . . . 0,0 : ; + . :
§ ‘0 0.8 I I I I I 1.0 . . . .
g g 0.6 I- . II 0.8
.9- 0 0.6 ' ' "
°' 13 0.4 - J
J 0.4 'L' '
0.2 r 0.2 .
0.0 L l n l n 0.0 n n n ' n n
0 5 10 15 20 25 30 0 5 10 15 20 25 30
4d
0.8 . . - --4L1 0,3 . . . . .
E 0-5 " 0.6
“e’
-- 0.4 - - ,4
($3 0
8 0.2 0.2
‘9 00
32 -
g 0.8 833
‘9 06 06
a a '
E‘ 0.4 0.4
.9-
13 02 ' 02 '1
' R2=0.65 ' R2=o.850
0.0 J L ‘ J 0.0 ' ‘ ‘ ‘ ‘
0 5 10 15 20 25 30 0 5 1O 15 20 25 30
Age of restored sites (yrs) Age of restored sites (yrs)

Figure 3.4. Relationship between the percentage similarity index (PSI) of algal functional
groups (4a, 4b) and diatom assemblages (4c, 4d) and the age of restored marshes in surface
sediments and on macrophytes during spring and summer 1998. PS1 was calculated for algal
functional groups or diatom assemblages between restored and reference marshes. The line
represents a signiﬁcant relationship (linear regression P<0.05). The triangles represent sites
in Virginia, and dots represent sites in North Carolina.

PSI of diatom species composition between restored and reference marshes increased

with age of restored salt marshes both on macrophytes and in sediments in the spring

67

(Figure 3.4 c). The rate of change in PSI between restored and reference marshes with
age of restored marshes did not differ between epiphytic and sediment algae (ANCOVA,
P=0.602). The PSI changed from 37% to 70% on epiphytes and from 30% to 50% in
sediment.

Similarity in epiphytic and sediment diatom communities between restored and
reference marshes during summer also increased with increasing age of restored salt
marshes (Figure 3.4d). The rate of change in PS1 between restored and reference marshes
with age of restored marshes did not differ between epiphytic and sediment algae
(ANCOVA, P=0. 160). The similarity changed from 30% to 73% on macrophytes and
from 36% to 65% in sediment.

Canonical correspondence analysis of epiphytic diatom assemblages showed that 26.8
of the total variance in the species data was explained by the ﬁrst two axes. The ﬁrst axis
was correlated with OMP (r=0.769), nitrogen (r=0.495), chl a (r=0.524), and salinity (r =
-0.467); the second axis was correlated with stem density (r=0.493), salinity (r = -0.474)
and OMP(r=0.394). Similarly, CCA of sediment diatom assemblages and environmental
variables showed that 26.2% of the species variance was explained by the ﬁrst two axes.
The ﬁrst axis was negatively correlated with OMP (r=0.617), nitrogen (r=0.503), and chl
a (r=0.490); the second axis was negatively correlated with salinity (r = -0.737).

In epiphytic samples, Achnanthes exigua var. elliptica Hustedt, Actynoptychus splendens
Ralfs, and Skeletonema costatum (Greville) Cleve dominated in high N concentration sites,
while Navicula pygmaea Kutz., Nitzschia clausii Hantzsch, and Amphora proteus Gregory
dominated in low N environments. In sediment samples, Plagiogramma pygmaeum Grev.,
Actinoptychus splendens (Shadbolt) Ralfs in Pritchard, Gyrosigma balticum (Ehrenberg)

Rabenhorst, Diploneis papula (A. Schmidt) Cleve, Amphora gigantean Grunow ex A.

68

Schmidt were negatively correlated with nitrogen concentration. With an increase of
nitrogen concentration, diatom species composition shifted to Nitzschia debilis (Amott)
Grunow.
3.4.3 Productivity

Gross primary production (GPP) of algae had a wide range during summer. The GPP
of epiphytes ranged from 3.832 to 75.338 mg C rn'2 h’1 in reference marshes, and from
6.503 to 137.068 mg C m’2 h'1 in restored marshes. The sediment algal GPP showed less
variation, ranging from 4.56 to 16.56 mg C m'2 h'1 in reference marshes and from 4.08 to
46.981 mg C m'2 h'1 in restored marshes. PRR did not change signiﬁcantly with age of
restored marshes either for epiphytes or sediment algae (Figure 3.5). PRR was generally
higher in younger marshes. Great variation among different pairs of marshes was
observed, but age of restored marshes was not a determinant of algal productivity in this

study.

3.4.4 Algal Attribute Development In Virginia Salt Marshes

DM, AFDM, OMP, and algal biovolume in Virginia salt marshes varied among
different marshes and did not change with age of restored marshes. Sediment DM (0.182
1: 0.043 g/cmz) of Virginia reference marshes was lower than BM in North Carolina
reference marshes (P<0.05); AFDM (0.033 i 0.010 g/cmz) and OMP in sediment (18.2 i

3.5%) were similar to North Carolina marshes. DM on plants (0.43 :l: 0.12mg/cm2),

69

 

30:80 000.000.0000.: 0.000802

 

30:80 000.000.0000.: 0.00.4002 30:80 30:80 0000:: 0.00302
2090: 0000.000 0.0000002 30:80 3:00:00an 0:00.200 000000002 30:80 000.000 0000000000.
8053— 0000000000000 0000.302 @853: 0000.20.00 0.0000002 30:80 000000.000. 0.0000002 30:80 00000000 00000000040
30:80 00.00000 0000000030 £80m< 0000000000000 00.20002 80: 30:80: 00%000 00000000 30:80 00.00000 00.00.000.000
30:80 00000000 0000000000. 885. 000000000000 0000.302 30:80 000.000.0000.: 0.0000002 30:80
30:80 000.000.0000.: 0.0000002 30:80 000000§030802 0>0_0 3000:0008 000.000 0.000% 3000:0008 0:00.000“. 0000000002

m8~8¥ 0000 0.000.302 835— 30:80 00.0300 00000000000 006.0 0088.6
.8. 0000.302 2000?: 00000000080 0000050 30:80 000.00 0000000030 8:00:85: 00000». 0.00002
8502 .w000 000000000v 0000008 30:80 0.0.0000 0000.000Q 30:80 0.0.0000 0000.000Q
A5809: 00000000080 0000008 .8. 000.302 8820— 850M
30:80 000000000\ 0.0000002 30:80 0.0.0000 0000.000Q £0009: 0.800000%.00 00000000. 3000?: 0.800000%.00 00000050
30:80 0.0.0000 0000.000Q 30:80 0000 0.0000002 583.4. 000000000000 00.0.0002 £00090. 00000000000 00.0.0002
30:80 0.0.0000 0000.000Q .2 00 .m A5803:
MENSM 000003.— 00000000 0.0000002 000000000 0000000000000
A€8w<v 000000000500 00000030 30:80 00.000000 0000.302 mid—2v— 0302000 0000.302 30:80 00050000 0000.302
30:80 30:80 000000000x00000~02 883— 30:80 00.00000 00000000000
8:53: 0000.20.00 0.000.302 30:80 00002000 000000000< £0009: 0000000000000 00000000. 200800-823. 00.000000. 0.900000%
0885-823 00.0000x 00000000m 30:80 38803008005: @53—
000.00: 0.3000 0.500002 $000588 0:00.000“. 000000002 0200.000 0000000002 £00003 000000.000§00 0000050
.8. 0000.302 808:: 0.3000 0.30802 8000:: 0000000000000 0.0000002 30:80

30:80 0000000000 0000.302 30:80 0000000000 0000.302 30:80 3853: 0000.20.00 0.000302 wEEm
833— 8. 0000.302 3833: 000000.00 0.0000002 3889908088
38093 0000000000R00000§m 850M 000 0:00.000 0000000002
585 .3 00083000 0.0200002 £000w<00000000000x 00000.00 30:800 00.000000 0000.302 30:80 000.0300 0000000000:

30:80 0.0.0000. 00000000
£00030 0000000000000 00.0.0002

30:80 0.0.0000 0000.000Q
spam/.00 0000000000000 00.0.0002

30:80 00002000 0000000000.
0000900 00000000000000 00.20002

850M 00002030 0000.302
£80w< 0000000000000 00.0.0002

 

8:808 000083—

moF—mhwe OUn—vuouom

8:808 00003.30

mun—Pas 00.09—0qu

 

080000 08308.0

080000 02:03

 

800.00.“. 0:0 .090 £808 .090 0030: .088 00:00:30 030200 .3 00803 000m :0 8 860% 80.80 008800 508 o— 05. .0.m 030.0.

70

 

2.5
2.0

1.5
1.0
0.5
0.0

 

PRR

 

3.0

2.0

1.0

0.0

 

-1.0

l

 

 

l

 

 

0

5

1O

15

20

25 30
Age of restored marsh(yrs)

Figure 3.5. Relationship between
algal productivity ratio and age of
restored marshes in summer 1998.
Productivity ratio compared between
algal GPP between restored and
reference sites in surface sediments

(A) and on macrophytes (A).

AFDM (0.03 :t 0.01mg/cm2), and OMP (<20%) were lower than in North Carolina

marshes. OMRR in Virginia did not ﬁt well in the OMRR pattern found in North

Carolina.

Shannon’s species diversity indices in reference marshes varied from 1.72 to 2.64 on

macrophytes and from 2.86 to 3.43 in sediments, which were similar to that in North

Carolina marshes. The diversity index in a restored marsh was most similar to its

reference marsh but did not vary with age of restored marshes. PSI of diatoms in Virginia

marshes ﬁt well into the PSI pattern with restored marsh age for North Carolina (Figure

3.4d).

71

3.5 DISCUSSION

The species composition of algae in salt marshes is determined by multiple
environmental variables and clearly responds to age of restored marshes. Previous studies
(Pinckney and Zingmark 1993b, Mayer and Galatowitsch 1999) questioned the use of
diatom assemblages as indicators of ecological restoration in wetlands because of the
high degree of environmental heterogeneity in reference wetlands. Controlling for
regional and habitat variation by comparing restored marshes to nearby reference
marshes enabled us to increase the precision in relations between algal attributes and the
age of restored marshes. Algal biomass, algal species composition, and organic matter
responded, like invertebrate, soil, and plant attributes (Craft 1988, 1991, Sacco et al.
1994, Scatolini and Zedler 1996), to age of restored wetlands. Observed relations
between algal attributes and other environmental factors provided the base for hypotheses
as to how changing wetland processes that regulated algal attribute restoration.

A number of spatial and temporal factors are account for variation of organic matter
in soil sediment and sediment on plants. Organic matter in Mississippi salt marshes
ranged from 14.3 to 23.3% by weight of the soil surface sediments (Sullivan and
Moncreiff 1988). OMP in our reference marshes in North Carolina was more variable
among different marshes than in Mississippi but remained constant with depth in soil
cores. OMP decreased with depth in restored marshes indicating gradual accumulation of
organic matter over time (Craft et al. 1999). Sediments on macrophytes had higher OMP
than soil surface sediments during spring, probably because of a larger algal proportion of

total mass on plants and less inorganic matter (silt) than in sediment. So I presumed that

72

the OMP on plants could be a different indicator of algal biomass change over time than
sediment OMP. Seasonal variation of OMP in NC salt marshes showed higher OMP
during spring than summer, which was probably caused by high organic matter input
from runoff and input from rivers and high tides during spring (Fong and Zedler 1993).

Algal biomass in North Carolina salt marshes, represented by algal biovolume and Ch]
a concentration, was higher in spring than summer on both macrophytes and in surface
sediments. Seasonal succession of the algal community from large green macroalgae to
mainly diatom-dominated microalgal communities may contribute to that variation
(Mitsch and Gosslink 2000). In North Carolina salt marshes, the diatom proportion of
biovolume was much higher during summer than in spring. In addition to changes in
algal growth form, low nutrient concentrations in summer in the water column of North
Carolina salt marshes and the importance of sediment nutrients in regulating algal growth
(Larned 1998, Fong et al. 2001) also indicated that nutrients were important in regulating
algal biomass and caused the seasonal variation. Relatively high nutrient availability and
low grazing pressure by meiofauna, junvenile macrofauna, and ﬁddler crabs during
spring are commonly related to seasonal variability in algal biomass (Sullivan and Daiber
1975, Darley et al. 1981, Montagna et al. 1983, Wear et al. 1999).

Algal biomass also varied among habitats in salt marshes (Pinckney and Zingmark
1991, 1993b). Sediment algal biomass was much higher than epiphytic algal biomass in
the North Carolina salt marshes. Vertical migration of algae in sediments enables
occupation of a 3-dimensional habitat that can extend to at least the upper 5 mm. Thus a
much thicker algal accumulation in surface sediments (Leach 1974) can support a much

higher areal algal biomass than on macrophytes. The increase of Spartina canopy density,

73

water column turbidity, and different species of marsh plant (Pip and Robinson 1984)
may also affect the light availability for algal growth on macrophytes (Colijn 1982,
Zedler 1993) and lead to varied distribution of algal biomass in different salt marshes.

Algal diversity and species composition were highly related to N concentrations and
sediment organic matter in North Carolina salt marshes as in other marshes (McCormick
and O'Dell 1996, Stevenson et al. 1999, Pan et al. 2000). Shannon diversity of sediment
algae was positively correlated to sediment nutrient concentration, especially TN,
indicating a release of constraints on species membership in assemblages by N limitation.
Algal species diversity was higher during spring than summer, which may have been due
to higher N runoff to upstream watersheds. Shifts in diatom species composition on
plants and sediments were also related to OMP and TN concentration in sediments.

The relationships between algal biomass and species composition with sediment N
and OMP in NC salt marshes are likely due to organic matter sources of N and release of
algae from N limitation. Algae as well as macrophytes in many estuarine habitats can be
N limited (Cargill and Jefferies 1984, Pedersen and Borum. 1996, Kiehl et al. 1997, van
Wijnen 1999). Increases in sediment organic matter increase N supply to macrophytes
such as Spartina (Kiehl et al 1997, Craft 2001). Benthic algae can acquire nutrients that
leak from plants and sediments as well as from the water column (Moeller et a1 1988,
Burkholder and Wetzel 1990, Fong and Desmond 1997, Fong et al. 2001). Sediment
algae probably get higher proportions from their substratum because of proximity to the
two nutrient supplies. If nutrient concentrations in the water column are lower during
summer than spring due to runoff patterns and if nutrient ﬂuxes from sediments stay

relatively constant, then seasonal effects of nutrient supply should be greater for

74

epiphytic algae than sediment algae, as I observed. In addition, the relationships between
algal biomass and nutrients were stronger for epiphytic than sediment algae and during
summer than spring. Thus N limitation was probably related to organic matter in
sediments and was greater during summer than spring and for epiphytic than sediment
algae in NC marshes.

Despite variability in salt marsh algal assemblages related to season, habitat, OMP
and N, I was still be able to detect relations between OMRR in sediment, AVRR and
ACRR on macrophytes during summer, and PSI of diatoms on plants and sediments
increased with age of restored marshes. Despite regional differences, the similarities of
diatom assemblages between 3 pairs of restored and reference marshes in Virginia ﬁt
well into the similarity-age relationship found in North Carolina. Most processes of
benthic algal community development probably occur on much shorter time scales than
the 10-15 years required for the algal community development observed in restored NC
salt marshes. Algal immigration, growth, and community development occur during
times scales ranging from weeks to months, not years (Fisher et al. 1982, Stevenson
1990, McCormick and Stevenson 1991). Therefore, exogenous processes at scales
greater than the periphyton community, perhaps at the wetland scale, probably regulate
succession of algal assemblages after the ﬁrst year of wetland restoration (Figure 3.6).
Relations between organic matter, algal biomass on macrophytes during summer, and
algal species composition indicated a linkage between algal community development in
microhabitats of marshes and the larger-scale successional process of organic matter
accrual in wetlands. Organic matter accrual in restored NC salt marshes requires at least

15 years before it reaches the level of reference marshes (Craft et al. 1999, Craft 2000).

75

The increase in organic matter in salt marshes probably regulated benthic algal
succession indirectly during marsh restoration through control of N supply. CCA of
species-environment relationships showed that the difference in nutrient concentration
between restored and reference marshes was one of the main causes of species
assemblage differentiation. The increased nutrient concentration in restored marshes with
age of marsh correlated with similarity of diatom assemblage to reference marshes.
Epiphytic algal biomass was constrained by nutrient limitation during summers, probably
due to N limitation. Thus, development of algal species composition on macrophytes and
sediments and algal biomass on plants during summer was probably regulated by
biogeochemical processes in wetlands and may be valuable indicators of those processes
(Figure 3.6)(Sullivan 1975).

Due to the high spatial and temporal variation of algal biomass and productivity, no
relations were found between algal primary productivity or respiration and age of
restored marshes. A number of studies have been done to accurately measure the biomass
and production in estuarine habitats (Sullivan and Moncreiff 1990, Pinckey and
Zingmark 1993a, b). Sediment grain size, sediment moisture, ambient irradiance, salinity,
H2S concentration, temperature, pH, and nutrients all affected algal production and
biomass (Frey and Basan 1985, Sullivan 2000). Estimates of annual sediment primary
production varied from 4 to 267 g C 111’2 yr'1 depending on different marsh types and
nutrient conditions (Leach 1976, Zedler et al. 1978, Zedler 1980, Pinckney and Zingmark
1993b). Although algal productivity is a very important attribute of algal community
function, it is also a complicated attribute and can easily be affected by microbial activity,

macrophyte shading, irradiance, and algal biomass (Sullivan 2000). Algal primary

76

production on macrophytes and in sediments in my studies were also subject to a wide
range of variation, but all were within the range of previous results. Spatial and temporal
variability and difﬁculties in measuring primary production in ﬁeld settings probably
contributed to variation in this study and the lack of strong relations between algal
productivity, biomass, and wetland development.

I hypothesize that marsh-wide algal productivity should increase relatively rapidly
during restoration. Algal productivity on macrophytes during summer should have
increased with algal biomass on plants and restored marsh age due to release from N
limitation. However, algal biomass in sediments was much higher than on plants and was
not as sensitive to age of restored marshes as algae on macrophytes. Sediment algae
should be exposed to higher nutrient concentrations than algae on macrophytes. In
addition, irradiance and photosynthetic rates should be higher during early stages of
restoration because plant density is lower. Thus, limitation of algal biomass and
productivity on a marsh-wide scale is probably not as sensitively related to organic matter
build-up and development of biogeochemical cycling of nutrients as algal species
composition and epiphytic algae biomass (Figure 3.6).

In summary, diatom species composition, epiphytic algal biomass in summer, and
organic matter proportions in sediments are good indicators for success of salt marsh
restoration when regional variables are controlled by comparing a restored marsh to a
nearby mature, reference marsh. Historic problems with use of algal assemblages as
indicators of wetland restoration were related to variability among wetlands (Pinckney

and Zingmark 1993b, Mayer and Galatowitsch 1999), which is a common problem

77

Stages of wetland restoration for algae

 

 

 

 

 

 

 

Invasive Habitat Development Restored
I j
i i
i i
i i
c 1 0 i ; W 1 0
0: O : : . I ”a .
E I K,"
5 z 'r’
a ! l
3., . .
g ! !
g ! ! — Algal production
I I
i i — - Algal community
' i composition
: — - Biogeochemical
! c cling
i ........ rganrcMatter
i Accrual
0.0 0.0

Log(Time .............. ’)

Figure 3.6. Conceptual model of algal structural and functional development during
marsh restoration.

in ecological assessment. Precise characterization of expected reference conditions calls
for classifying wetlands based on the natural regulatory factors that affect biota, such as
climate, hydrology, geochemistry, and salinity in coastal environments (Hawkins et al.
2000, USEPA 2002a, Stevenson et al. in review). Thus, precision of algal assessments of
wetland restoration, as any wetland assessments, should be greatly enhanced by
identifying the class of restored wetland and comparing observed conditions to conditions
found in relatively natural, reference wetlands of the same class (Stevenson 2001,
USEPA 2002b). With increasing understanding of algal ecology in wetlands, algal

sensitivity to environmental conditions, and new developments in assessment protocols,

78

great potential exists for diagnosing factors that retard wetland restoration, establishing

criteria for wetland restoration, and improving assessments.

3.6 ACKNOWLEDGMENTS

Although the research described in the article has been funded wholly or in part by
the US. Environmental Protection Agency’s Science To Achieve Results (STAR)
program through grant #826111-01-0, it has not been subjected to any EPA review and
therefore does not necessarily reﬂect the views of the Agency, and no ofﬁcial
endorsement should be inferred.

Brian Wade, Carrie DeJaco, and Richard Schultz participated in some of the ﬁeld
work and sample collection. Kalina Manoylov provided assistance in diatom
identiﬁcations. Bruce Wang, Scott Rollins, and Vanessa Loughead provided valuable

comments on an early draft of the manuscript.

79

3.7 REFERENCES
Adam, P. 1990. Salt marsh Ecology. Cambridge University Press, Cambridge, NY, USA.
Adamus, RR. 1988. The FHWA/Adamus (WET) Method For Wetland Functional

Assessment. p. 128-133. In D.D. Hook, W.H. McKee, Jr., H.K. Smith, J. Gregory, V.G.
Burell, Jr., M.R. DeVoe, R.E. Sojka, S. Gilbert, R. Banks, L.H. Stolzy, C. Brooks,
T.D. Matthews, and TH. Shear (eds.) The Ecology And Management Of Wetlands.
Volume 2. Management, Use And Value Of Wetlands. Crom Helm, London,
England.

American Public Health Association (APHA). 1998. Standard Methods for the
Examination of water and wastewater. 20th edition by LS. Clececeri, A.E. Greenberg,
and AD. Eaton. Washington D. C., USA.

Burkholder, J .M., R.G. Wetzel, and KL. Klomparens. 1990. Direct comparison of
phosphate uptake by adnate and loosely attached microalgae within an intact bioﬁlm
matrix. Applied Environmental Microbiology 56: 2882-2890.

Cargill, SM. and R.L. Jefferies. 1984. Nutrient limitation of primary production in a sub-
arctic salt-marsh. Journal of Applied Ecology 21: 657-668.

Colijn, F. 1982. Light absorption in the waters of the Ems-Dollard estuary and its
consequence for the growth of phytoplankton and microphytobenthos. Netherlands
Journal of Sea Research 15: 196-216.

Craft, C.B., S.W. Broome, and ED. Seneca. 1988. Soil nitrogen, phosphorus and organic
carbon in transplanted esturine marshes. p.351-358. In D.D. Hook (ed.) The ecology
and management of wetlands. Volume 1: Ecology of wetlands. Croom Helm,
London and Sydney, Timber Press Portland, Oregon, USA.

Craft, C.B., ED. Seneca, and S.W. Broome. 1991. Poewater chemistry of natural and
created marsh soils. Journal of Experimental Marine Biology and Ecology 152: 187-
200.

Craft, C.B., J .M. Reader, J .N . Sacco and S.W. Broome. 1999. Twenty ﬁve years of
ecosystem development of constructed Spartina altemiﬂora (Loisel) marshes.
Ecological Applications 9: 1405-1419.

Craft, CB. 2000. Co-development of Wetland Soils and Benthic Invertebrate

Communities Following Salt Marsh Creation. Wetlands Ecology and Management 8:
197-207.

80

Craft, CB. 2001 . Soil organic carbon, nitrogen and phosphorus as indicators of recovery
in restored Spartina marshes. Ecological Restoration 19: 87-91.

Darley, W., C. Montague, F. Flumley, W. Sage, and A. Psalidas. 1981. Factors limiting

edaphic algal biomass and productivity in a Georgia salt marsh. Journal of Phycology
17: 122- 128.

Fisher, S.G., L. Gray, N. B. Grimm, and DE. Busch. 1982. Temporal succession in a
desert stream ecosystem following ﬂash ﬂooding. Ecological Monographs 52: 93-
1 10.

Fong, P. and J. B. Zedler. 1993. Temperature and light effects on the seasonal succession
of algal conununities in shallow coastal lagoons. Journal of Experimental Marine
Biology and Ecology 171: 259-272.

Fong, P. and J .S. Desmond. 1997. The effect of a horn snail on Ulva expansa
(Chlorophyta): Consumer or facilitator of growth? Journal of Phycology 33 : 353-
359.

Fong, P., K. Kamer, K.E. Boyer, and K.A. Boyle. 2001. Nutrient content of macroalgae
with differing morphologies may indicate sources of nutrients for tropical marine
systems. Marine Ecological Progress series 220: 137-152.

Frey, R.W. and PB. Basan. 1985. Coastal Salt Marshes. p. 225-302. In R.A. Jr. Davis
(ed) Coastal Sedimentary Environments. Springer Verlag, NY, USA.

Gallagher, J. L. and F. C. Daiber. 1974. Primary production of edaphic algal communities
in a Delaware salt marsh. Limnology and Oceanography 19: 390-395.

Gauch, H.G. 1982. Multivariate analysis in community ecology. p.298. Cambridge
University Press, Cambridge, NY, USA.

Hawkins, C.P., RH. Norris, J. Genitsen, R.M. Hughes, S.K. Jackson, R.K. Johnson, and
RJ. Stevenson. 2000. Evaluation of landscape classiﬁcations for biological
assessment of freshwater ecosystems: synthesis and recommendations. Journal of the
North American Benthological Society 19:541-556.

Hook, D.D. (ed.) 1988. The ecology and management of wetlands. Volume 1: Ecology of
wetlands. Croom Helm, London and Sydney, Timber Press Portland, Oregon.

Jones, K. 1974. Nitrogen ﬁxation in a salt marsh. Journal of Ecology 62:553-565.

Lamed, ST. 1998. Nitrogen- versus phosphorus-lirnited growth and sources of nutrients
for coral reef macroalgae. Marine Biology 132: 409-421.

81

Keer, OH. and J .B. Zedler. 2002. Salt marsh canOpy architecture differs with the number
and composition of species. Ecological Applications 12: 456-473.

Kiehl, K., P. Esselink, and J .P. Bakker. 1997. Nutrient limitation and plant species
composition in temperate salt marshes. Oecologia 111: 325-330.

Kreeger, DA. and R.I.E. Newell. 2000. Trophic complexity between producers and
invertebrate consumers in salt marshes. p. 187-220. In M.P. Weinstein and D. A.
Kreeger (eds.) International Symposium: Concepts and controversies in Tidal Marsh
Ecology. Kluwer Academic Publishers, Dordrecht, The Netherlands.

Leach, J .H. 1970. Epibenthic algal production in an intertidal mudﬂat. Limnology and
Oceanography 15:5 14-521.

Lockwood, J .L. and S.L. Pimm. 1999. When does restoration succeed? p. 363-392. In E.

Weiher and P. Keddy (eds) Ecological Assembly Rules: Perspectives, advances, retreats.
Cambridge University Press, Cambridge, NY, USA.

Mayer, RM. and SM. Galatowitsch. 1999. Diatom communities as ecological indicators
of recovery in restored prairie wetlands. Wetlands 19: 765-774.

McCormick, RV. and MB. O’Dell. 1996. Quantifying periphyton responses to
phosphorus in the Florida Everglades: A synoptic-experimental approach. Journal of
the North American Benthological Society 15: 450-468.

McCormick, RV. and RJ. Stevenson. 1991. Grazer control of nutrient availability in the
periphyton. Oecologia 86:287-291.

Moeller, R.E., J .M. Burkholder, and R.G. Wetzel. 1988. Signiﬁcance of sedimentary
phosphorus to a rooted submersed macrophyte (Najasﬂexilis) and its algal epiphytes.
Aquatic Botany 32:261-281.

Montagna, P., B. Coull, T. Herring, and B. Dudlwy. 1983. The relationship between
abundances of meiofauna and their suspected microbial food (diatoms and bacteria).
Estuarine, Coastal and Shelf Science 17: 381-394.

National Research Council (NRC). 1992. Restoration of aquatic ecosystems. National
Academic Press, Washington, D. C., U.S.A.

Odum, E. P. 1969. The strategy of ecosystem development. Science 164:262-270.
Pan, Y., R]. Stevenson, P. Vaithiyanathan, J. Slate, and C]. Richardson. 2000. Changes

in algal assemblages along observed and experimental phosphorus gradients in a
subtropical wetland, U.S.A. Freshwater Biology 43: 1-15.

82

Pedersen, M. F. and J. Borum. 1996. Nutrient control of algal growth in estuarine waters:
Nutrient limitation and the importance of nitrogen requirements and nitrogen storage
among phytoplankton and species of macroalgae. Marine Ecological Progresse Series
142: 261-272.

Pinckney, J. and R.G. Zingmark. 1991. Effects of tidal stage and sun angles on intertidal
benthic microalgal productivity. Marine Ecological Progress Series 76: 81-89.

Pinckney, J. and R.G. Zingmark. 1993a. Modeling the annul production of intertidal
benthic microalgae in estuarine ecosystems. Journal of Phycology 29: 396-407.

Pinckney, J. and R.G. Zingmark. 1993b.Biomass and Production of benthic microalgal
communities in estuarine habitats. Estuaries 16: 887-897.

Pip, E. and G.G.C. Robinson. 1984. A comparison of algal periphyton composition on
eleven species of submerged macrophytes. Hydrobiological Bulletin 18: 109-118.

Pomeroy, LR. and R.G. Wiegert. 1981. The ecology of a salt marsh. Springer-Verlag,
NY, USA.

Sacco, J. N., E. D. Seneca, and T. Wentworth. 1994. Infaunal community development of
artiﬁcially established salt marshes in North Carolina. Estuaries 17: 489-500.

Sage, W.W. and M.J. Sullivan. 1978. Distribution of bluegreen algae in a Mississippi
gulf coast salt marsh. Journal of Phycology 14: 333-337.

Scatolini, SR. and J. B. Zedler. 1996. Epibenthic invertebrates of natural and constructed
marshes of San Diego Bay. Wetlands 16: 24-37.

Shubert, LE. 1984. Algae as Ecological indicators. Academic Press, INC. London, UK.

Steneck, RS. and MN. Dethier. 1994. A functional group approach to the structure of
algal-dominated communities. Oikos 69: 476-498.

Stevenson, R.J. 1990. Benthic algal community dynamics in a stream during and after a
spate. The Journal of North American Benthological Society 9: 277-288.

Stevenson, R.J. 2001. Using algae to assess wetlands with multivariate statistics,
multimetric indices, and an ecological risk assessment framework. p 113-140. In: R.
R. Rader, D. P. Batzger, and S. A. Wissinger (eds.) Biomonitoring and Management
of North American Freshwater Wetlands. John Wiley and Sons, Inc. New York,
USA.

Stevenson, R.J., J. Alba-Tercedor, B. Bailey, M. Barbour, C. Couch, S. Dyer, F. Fulk, J.

Harrington, M. Harass, C.J. Hawkins, C. Hunsaker, R. Johnson, and K. Thornton.
accepted. Design, implementation, and interpretation of ecological assessments. In M.

83

Barbour, K. Thornton, and R. Preston (eds.) Ecological Assessment. Society of
Environmental Toxicology and Contamination Publication, USA.

Stevenson, R.J. and Y. Pan. 1999. Assessing ecological conditions in rivers and streams
with diatoms. p. 11-40. In E.F. Stoerrner and J. P. Smol (eds.) The Diatoms:
Applications To The Environmental And Earth Sciences. Cambridge University
Press, Cambridge, UK.

Stevenson, R.J. and RV. McCormick, R. Frydenborg. 2001. Use of algae to assess
wetland condition: Using Algae to assess environmental conditions in wetlands. EPA
843-B-00-002k. US. Environmental Protection Agency; Ofﬁce of water;
Washington, DC, USA.

Stevenson, R.J., P.R. Sweets, Y. Pan, and RE. Schultz. 1999. Algal community patterns
in wetlands and their use as indicators of ecological conditions. p. 517-527. In A. J.
McComb and J. A. Davis (eds.) Proceedings of INTECOL's Vth International
Wetland Conference. Gleneagles Press, Adelaide, Australia.

Stosch, H. A. von and B.E.F. Reimann. 1970. Subsilicea fragilarioide gen. Et spec. nov.,
eine Diatomee (Fragilariaceae) mit vorwiegend organischer membran. Nova
Hedwigia Beih. 31:1-36.

Sullivan, M. J. 1975. Diatom communities from a Delaware salt marsh. Journal of
Phycology 14:333-337.

Sullivan, M.J. and F. Daiber. 1975. Light, nitrogen and phosphorus limitation of edaphic
algae in a Delaware slat marsh. Journal Of Experimental Marine Biology And
Ecology: 18:79-88.

Sullivan, M.J. and CA. Moncreiff. 1988. Primary production of edaphic algal
communities in a Mississippi salt marsh. Journal of Phycology 24:49-58.

Sullivan, M.J. and CA. Moncreiff. 1990. Edaphic algae are an important component of
salt-marsh food-webs - evidence from multiple stable isotope analyses. Marine
Ecological Progress Series 62 : 149-159.

Sullivan, M.J. and CA. Currin. 2000. Community structure and functional dynamics of
benthic microalgae in salt marshes. p. 81-106. In M.P. Weinstein and DA. Kreeger
(eds)Concepts and controversies in Tidal marsh ecology. Kluwer Academi
Publishers, Dordrecht, The Netherlands.

Teal, J .M. 1962. Energy ﬂow in the salt marsh ecosystem of Georgia. Ecology 43: 614-
624.

ter Braak, C]. F. 1987. CANOCO-a FORTRAN program for canonical community
ordination by (partial)(detrended)(canonical) correspondence analysis, principle

84

components analysis and redundancy analysis (versin 2.0). TNO Institute of applied
computer science, Wageningen, The Netherlands.

ter Braak, C.J.F. 1990. Update Note: CANOCO v 3.1. Agricultural Mathmatics Group,
Wageningen, The Netherlands.

US. EPA. 2002a. Methods for evaluating wetland condition, #7 Wetlands classiﬁcation.
United States Environmental Protection Agency, Washington, DC, USA. EPA-822-
R-02-017.

US. EPA. 2002b. Methods for evaluating wetland condition, #11 Using algae to assess
environmental conditions in wetlands. United States Environmental Protection
Agency, Washington, DC, USA. EPA-822-R-02-021.

Wear, D.J., M.J. Sullivan, and AD. Moore. 1999. Effects of water-column enrichment on
the production dynamics of three seagrass species and their epiphytic algae. Marine
Ecological Progress Series 179: 201-213.

van Wijnen, H.J., and J .P. Bakker. 1999. Nitrogen and phosphorus limitation in a coastal
barrier salt marsh: the implications for vegetation succession. Journal of Ecology 87:
265-272.

van Raalte, C. D., l. Valiela, and J .M. Teal. 1976. Production of epibenthic salt marsh
algae: light and nutrient limitation. Limnology and Oceanography 21 :862-872.

Zedler, J ., T. Winﬁeld, and D. Mauriello. 1978. Primary productivity in a So. California
estuary. p. 649-662. In constal Zone. ’78. Symposium on Technical, Environmental,,
Socioeconomic and regulatory aspects of coastal Zone Management, vol. 2, American
Society of Civil Engineers, NY, USA.

Zedler, J .B. 1980. Algal mat productivity: comparisons in salt marsh. Estuaries 3:
122-131.

Zedler, J .B. 1993. Canopy architecture of natural and planted Cordgrass marshes:
selecting habitat evaluation criteria. Ecological Application 3: 123-138.

85

Chapter 4
Relations between nutrient conditions, algal growth, and species

composition in the restored wetlands of southern Michigan

4.1 Abstract: Nutrient limitation is one of the main factors regulating periphyton growth
and species composition in wetlands. To study nutrient limitation of algae and their
response to nutrient addition in restored wetlands, I placed nutrient diffusing substrata in
22 restored wetlands for 3 weeks in southern Michigan during the summer of 2000.
Physical and chemical variables, as well as epiphytic, sediment, and planktonic algal
assemblages were evaluated during the experimental period. Nitrogen (N) was the main
limiting nutrient for periphyton growth in restored Michigan wetlands. Algal N limitation
was positively correlated with phosphorus (P) concentration and negatively correlated
with N:P ratios in wetland water and sediment. Canonical correspondence analysis
(CCA) indicated that changes in diatom assemblages in natural habitats and control
artiﬁcial habitats were correlated mainly to pH and secondarily to nutrient, primarily total
phosphorus (TP) and dissolved inorganic nitrogen (DIN). Individual diatom taxa
responded to nutrient addition differently. Overall, the responses of indicator species in
nutrient bioassays were positively correlated with natural nutrient gradients among
natural wetlands indicating a good match of diatom indicators between natural and

experimental systems.

86

4.2 Introduction

Wetlands have increasingly been impacted by nutrient input from point and non-point
sources (Carpenter et al. 1998). A considerable body of evidence indicates nutrient
pollution not only alters the structural composition of a biological community in aquatic
systems, but also changes decomposition rates, productivity, and other ecosystem
functions (Craft et a1 1995, Newman et a1 1996, Bedford et al. 1999, Craft and
Richardson 1997, Qualls and Richardson 2000, Svengsouk and Mitsch 2001). In natural
communities, nutrient enrichment and shifts in resource ratios will lead to changes in
algal production and shifts in species composition (Tilman 1981,Tilman et al. 1986,
Sommer 1993). Thus, identifying the limiting nutrients in wetlands could help us
understand factors regulating community structure and function and better manage these
valuable ecosystems.

The primary limiting nutrients in aquatic systems are usually either nitrogen (N),
phosphorus (P), or both, depending on different systems and geological regions
(Downing et al. 1999, Bedford et a1. 1999). Early studies of nutrient lirrritation in
freshwater systems used multiple approaches to show P limitation of phytoplankton
ranging from algal cultures in incubators to phytoplankton response to whole lake
fertilizations (Emery et al. 1973, Hecky and Kilham 1988, Elser 1990, Sommer 1993,
Wurtsbaugh et a1 2001). Periphyton bioassays have also been used to identify nutrient
limitation in lotic systems (Bothwell and Stockner 1980, Elwood et a1. 1981, Peterson et
al. 1983, Grimm and Fisher 1986, Mulholland et al. 1991, Francoeur et al. 1999,
Chambers et al. 2000, Wurtsbaugh et al. 2001) and have shown relations among region,

geology, and N or P lirrritation. Boreal and temperate eastern and central streams are

87

generally P-lirrrited (Peterson et al. 1983), while temperate western and southwestern
streams in landscapes of volcanic or tectonic origin are generally N-limited (Grimm &
Fisher 1986).

Few studies have examined nutrient limitation and enrichment in wetlands (Hooper-
Reid and Robinson 1978, Bayley et a1. 1985, Rader and Richardson 1992, McDogal et al.
1994, McCormick et al. 1996, 2001, Pan et al. 2000), where plants and algae can be N or
P limited depending on wetland type. It is widely assumed that most freshwater wetlands
are P-lirrrited in north temperate zone of North America, but this conclusion is based on
little data (Goldsborough and Robinson 1996). A meta-analysis on nutrient limitation and
plant diversity suggested (Bedford et al. 1999) a large proportion of North American
wetlands, especially bogs and fens, were either P limited or co-limited by N and P. They
also used N: P ratios in plant tissues and surface soils to predict that marshes and the
majority of swamps were N limited. Aquatic systems in Southern Michigan were found
to be mostly P-lirrrited (Hart and Robinson 1990, Fahnenstiel et a1. 2000, Wetzel 2001) or
not limited (Hamilton 2000) by nutrients. But N: P ratios in ground water and
precipitation were generally very high, which also indicates P-lirrritation (Hamilton,
Personal communication).

Experimental determination of algal nutrient lirrritation at a regional scale has never
been conducted in wetlands. Nutrient diffusing substrata have been an effective way to
determine nutrient limitation in stream and lake systems at regional scales (Grimm et al.
1981, Lowe et al. 1986, Pringle et al. 1986, Genter et al. 1987, Hecky and Kilham 1988,

Carrick et al. 1993, Wurtsbaugh et al. 2001). They have also been used to determine

88

nutrient limitation and algal production in individual wetlands (McCormick et al. 1996,
2001, Pan et al. 2000).

The main purpose of this research was to examine the effects of nutrient enrichment
on periphyton production and algal species composition in restored wetlands. Field
experiments were conducted with nutrient diffusing substrata to determine the effects of
N and P availability on periphyton accrual in 22 restored wetlands in southern Michigan.
While previous studies indicate both N- and P-limitation were limited in aquatic
ecosystems of temperate Northern America, I hypothesized that P would be the primary
limiting factor for algal growth in restored wetlands of southern Michigan because of
high groundwater N:P ratios. I also tested the hypotheses that 1) algal nutrient limitation
are correlated with nutrient concentrations and the N :P ratio in the wetlands, and 2)
nutrient conditions regulate the natural variability in algal species among wetlands. The
latter hypothsis was tested by experimental nutrient addition to establish cause-effect
relations and with variation in species composition and nutrient conditions among

wetlands.

4.3 Methods

4.31 Sample collection

Twenty-two restored depressional wetlands were selected in the Maple and Upper Grand
watersheds of southern Michigan for this study (Figure 4.1). All these mitigated wetlands
were restored within the past 15 year and were located in an agricultural and urban

dominated landscape. Previous research showed that nutrient concentration and

89

 

Figure 4.4.1. Map of Lower

 

Peninsula of Michigan, showing the
Upper Grand River and Maple River
watersheds where sampling sites

were located.

 

algal assemblages were not related to age of these wetlands (Zheng and Stevenson,
unpublished data). The sampling sites were mainly Typha- or Nuphar- dominated
marshes intermingled with several swamps. Duckweed covered at least some proportion
of all wetlands during the experimental period. Physical, cherrrical, and biological
variables were assayed or collected during the mornings of July 2000. Water temperature
and dissolved oxygen (DO) was measured with a YSI® DO meter (YSI Inc., Yellow
Springs, Ohio, USA) at the beginning and at the end of the sampling period, allowing at
least one hour between measurements. Conductivity and pH were measured using a YSI®
model 33 SCT meter with conductivity and pH electrodes (Denver Instrument Company).
Canopy coverage by emergent vegetation was estimated with a spherical canopy
densiometer held at water level and by averaging canopy cover from 4 directions.
Duckweed coverage was estimated by averaging the coverage of the wetland surface

before and after the nutrient-enrichment experiments. Irradiance at the water surface (10)

90

and 30 cm below the water surface (12) were recorded with a LICOR light meter. Light
extinction coefﬁcients were calculated (ln(Io)-ln(Iz)/Z, Wetzel 2001). Water samples were
collected in 2 125-ml acid-washed polyethylene bottles, stored in ice, transported to the
laboratory, and frozen until chemical analysis. Phytoplankton samples were collected in 2
l-L polyethylene bottles for chlorophyll a and algal cell counts. The top 1 cm of sediment
was collected randomly from eight different locations in each wetland using plastic tubes
(internal diameter = 3.8 cm, length = 20 cm). All 8 sediment samples from a wetland
were combined in a Whirl-pak® bag. Epiphytic algae were collected by cutting 4-5
macrophyte stems, brushing epiphytes from each macrophyte surface, and putting them
in a Whirl-pak® bag. Metaphyton were collected for qualitative identiﬁcation. Algae were

preserved with M3(APHA 1998) after subsampling for chl a analysis.

4.3.2 Nutrient diffusion experiment

Nutrient enrichment experiments were conducted in each wetland to determine
whether nutrients limited periphyton growth. Although a variety of experimental
approaches can be used to study the effects of nutrient enrichment, nutrient diffusing
substrata are an effective and economical way to determine nutrient limitation of
periphyton (Fairchild et al. 1985, Hecky and Kilham 1988, Wurtsbaugh et al. 2001).
Nutrient-diffusing bioassays were condicted with clay ﬂowerpots (upper diameter =
4.1cm, lower diameter =5.8 cm, height = 4.1 cm) as described in Fairchild et al. (1985).
Eight pots for each wetland were sealed with plastic plates on the bottom and ﬁlled with
2% Bacto® agar and one of four different nutrient treatments: an untreated control, N -

enriched (0.5M NaN03,), P-enriched (0.5 M NazHPO4), and NP-enriched (0.5M NaNO3

91

and NazHPO4) treatments. Pots were capped with acid-washed rubber caps. Eight pots
were attached to the comers of two 25 X 25 cm tiles to ensure stability during the
experimental period.

Nutrient diffusion rates were monitored over a 3-week period in distilled water in the
laboratory. Water samples in the containers for all treatments were sampled and analyzed
every two days. Nutrient diffusion rates were constant over the 3-week period.

The pots were placed at a 20-40 cm depth in each wetland. After three weeks, the pots
were collected and immediately transported in a cooler to the laboratory. The pot surface
was cleaned with a razor blade and a hard-bristled toothbrush. Periphyton from the two
pots in the same treatment were combined and rinsed into a glass beaker with deionized
water. The periphyton samples were shaken well and then subsampled for assays of total
nitrogen (TN) and total phosphorus (TP), chlorophyll a (chl a), dry mass (DM), ash free
dry mass (AFDM), and species composition, cell density, and biovolume of non-diatom
algae and diatoms.

4.3.3 Laboratory analyses

One 250-ml water sample from each wetland was analyzed for TN and TP. Another
bottle was ﬁltered through Coleman® glass ﬁber ﬁlters (0.45um diameter) and the ﬁltrate
was analyzed for nitrate (NO3'-N) and soluble reactive phosphorus (SRP) in a
Spectronic® GenesysTM 2 Spectrophotometer. TP and TN were oxidized to SRP and
N03’ (APHA 1998). TP and SRP were measured using the ascorbic acid method, while
TN and nitrate and nitrite (N Ox-N) were measured using cadium reduction methods
(APHA 1998). Ammonium (NHI-N) was analyzed using a Wuick-Chem 8000

autoanalyzer in University of Michigan Biological Station (UMBS). Dissolved inorganic

92

nitrogen (DIN) was thus calculated as the sum of ammonia, nitrate, and nitrite
concentrations. Sediment samples were rinsed with deionized water to 200 ml and
homogenized using a homogenizer. Appropriate dilution was made for sediment samples
to measure sediment TN and TP concentration following the same methods as the water
samples. Periphyton samples from clay pots were ﬁltered on GF/C ﬁlters for the analysis
of algal cell nutrient contents.

Chl a samples were assayed by extraction in 90% acetone overnight at 4°C, reading
absorbance with a Spectronic® GenesysTM 2 Spectrophotometer and calculating
pheophytin-corrected chl a concentration (APHA 1998). DM and AFDM of periphyton
were determined according to standard methods (APHA 1998).

Phytoplankton, epiphytic algae, sediment algae, and clay pot algae were diluted or
condensed as if necessary before counting. Algal densities and non-diatom species
composition were determined using a Leica® microscope and a Palmer-Maloney (0.1 ml)
counting chamber under 4OOX and by counting 300 natural units. Algal taxa were
identiﬁed to the lowest possible taxonomic level. Diatom taxa and relative abundance
were determined from permanent Naphrax® mounts of acid-cleaned diatoms (Stosch and
Reimann 1970). At least 500 diatom valves were counted. Biovolume was estimated for at
least 15 cells of each taxon by assigning geometric shapes and measuring appropriate cell

dimensions (Stevenson et al. 1985, Hillebrand et al. 1999)
4.3.4 Data analysis

To evaluate the magnitude of N and P limitation, I deﬁned primary nutrient limitation

in the 1‘" wetland as algal biomass of either N or P treatment (N,- or P,-) minus biomass in

93

the control treatment (C,-) then divided by C. For example, primary N limitation=(N.- -
C,-)/C.- and primary P limitation =(P,- -C,- )/C,- . Algal secondary nutrient limitation in a
wetland was deﬁned as algal biomass in the NP treatment (NP,- ) minus either P,- or N,- ,
then divided by biomass in that nutrient treatment. Therefore, secondary N
limitation=(NP; -P,- )/P.- and secondary P limitation =(NP,- -P,- )IN;.

To calculate cellular nutrient content in periphyton mass, particulate N and P contents
of algal biomass on artiﬁcial substrata were divided by algal DM. Mass N:P ratios were
also calculated based on mass N and P contents.

To determine which treatment led to biomass accrual, algal cellular N and P contents,
chl a, algal biovolume, and algal density were compared among treatments. Because of
the data distribution and site-speciﬁc dependence, a non-metric paired comparison was
conducted using a Wilcoxon Signed Ranks Test in Systat® 10.

To determine if the type and magnitude of nutrient limitation was correlated with
nutrient concentrations or nutrient ratios, correlations were calculated using Systat @10. I
related nutrient limitation to nutrient concentrations and the following molar ratios:
TN/TP, DIN/SRP, and DIN/TP to determine the relative importance of dissolved and
total nutrient in regulating nutrient limitation.

Algal species composition of phytoplankton, and diatom species composition on
macrophytes, in sediments, and on control treatments of artiﬁcial substrata were related
to environmental factors using canonical correspondence analysis (CCA) (CANOCO 4.0,
ter Braak and Srrrilauer 1997). Diatom taxa with a relative abundance greater than 1%
and appearing in at least 3 wetlands, or with relative abundance greater than 5% and

appearing in at least 1 wetland were selected for CCA. Relative abundance values were

94

square-root transformed to reduce the importance of dominant taxa and attain normal data
distribution. Algal relative biovolume of phytoplankton genera were also selected and
transformed as the same rules as relative abundance of diatom assemblages. A forward-
selected CCA ordination was then used to determine which environmental variables
accounted for the greatest amount of variance in the distribution of the algal or diatom
taxa. The signiﬁcance of each variable added in this fashion was tested using a Monte
Carlo permutation test with 99 unrestricted permutations (P _<. 0.05) (ter Braak and
Smilauer 1997).

To compare the relationships between diatom species responses to nutrients among
wetlands and their responses to experimental N, P, and NP enrichment, I identiﬁed
indicator species that speciﬁcally responded to N or P additions on aritiﬁcial substrata.
Diatom species whose relative abundance in N treatments were signiﬁcantly higher than
control treatments, or in NP treatment signiﬁcantly higher than P treatments were deﬁned
as N indicator species (W ilcoxon Signed Ranks Test P<0.l). Diatom species whose
relative abundance in P treatments were higher than in control treatments and higher in
NP treatments than N treatments were deﬁned as P indicator species. The sum of the
relative abundances of indicator species for either N or P in each wetland was then

related to nutrient concentrations in wetlands.

4.4 Results

4. 4. I Wetland habitats

95

421:1— ~02}— ° 4:1}— . {I} °

1 l l l l

0 20 40 60 80 00 0.1 0.2 0.3 0‘0 0‘5 {.01} i0 {,5 :1 8 1.21.6
Canopy Coverage(%) DIN(mg/L) TN(mg/L) Sed. TN(ug/mg DM)

fDIl—JD—oo 9.43}. o

L

 

 

 

l l I I; l l l j

 

15 20 25 30 0.0 0.5 1.01.50123456 0 l 2 3 4

 

 

 

 

Temperature<°C) SRP(mg/L) TP(mg/L) Sed. TP (pg/mg DM)
—l:l:l—~ 13:1— ~n:I—— -- {1:1— - -
0123456789 0123456 0 5 1015 20 024681012

DO(mg/L) DIN:SRP N:P Sed. N:P

—El:l-—° '—Cl:l-—-'° r-CEI— *1]:—

l l l _l L l l I l ‘ l n I

1 2 3 o°§o°§§ o°°i°° 6 7 8 9 10 0.0 0.10.2 0.3 0.4
ADO(mg'1'l'h'l) Conductivityumﬁro 57cm) pH Sed.OMP

Figure 4.2. Box plot of environmental variables in the 22 wetlands.

Although most of the 22 wetlands were not forested, the canopy 27(i25% ) and
duckweed cover (and other ﬂoating plants) (45:1:12%) on the surface of most wetlands
reduced irradiance to extremely low levels. Light available for use by underwater algae
(30cm below water surface) averaged 21 (i 18) % of irradiance at the water surface, and
the light extinction coefﬁcients (LEC) averaged 1.20 (i012). Water of these wetlands
had a wide range of conductivity (490:1:251 umol/cm). pH was generally neutral
(7.86i0.66). Average DO was 4.58 (:1: 2.59) mg/L before 10 a.m. and ranged from
complete depletion (0.7 mg/L) to oversaturation (l 1mg/L) in different wetlands. The

productivities based on DO change rates were 0.76 (i0.15) g-m'z-h'l (Figure 4.2).

96

Table 4.1. Correlation matrix of ambient nutrients in 22 wetlands. A * indicates a
signiﬁcant correlation (Dunn-Sidak P<0.05).
NOx NH4+ DIN SRP TN TP SedimentTN

NOx 1.000

N114+ 0.432 1.000

DIN 0771* 0.758* 1.000

SRP -0.126 0.259 0.135 1.000

TN 0.070 0.517* 0.515* 0.009 1.000

TP -0.128 -0.146 -0.109 0532* -0.l37 1.000
Sediment TN 0.112 -0.058 0.065 -0.099 -0.141 -0.l61 1.000
Sediment TP 0.119 0.139 0.207 0.151 0.017 0.027 0.852*

 

 

DIN (0.047:t0.050 mg/L) and TN (0.680:t0.448 mg/L) concentrations were generally
low in the water column compared with high SRP (019510.286 mg/L) and TP (1.143
i1.353 mg/L) concentrations (Figure 4.2). N114+ was the main form of DIN while NO3'
was completely depleted (<0.010 mg/L) in 17 of the 22 wetlands. Both DIN: SRP ratios
and TN: TP ratios were less than 17 indicating potential N limitation in the water column.
The predominant substratum was generally sand overlain by very ﬁne organic material
and silt. Sediments were mineral soils with organic matter (OMP) composing less than
35%. Sediment N:P ratios showed extremely low values (<2), except in one site, which
also indicated an N-limited habitat.

Nutrient concentrations in the water column were not correlated with nutrient
concentrations or ratios in sediments (Table 4.1). TN in the water column was correlated
with inorganic nitrogen concentration (r=0.517 with DIN, 1:0.515 with NHI), and TP in
the water column was correlated with SRP(1=0.532). TN and TP in sediment (mg/kg

DM) were highly correlated with each other (r=0.852).

4.4.2 Effects of nutrient enrichment and nutrient concentrations

97

Nutrient enrichment through clay pots signiﬁcantly changed algal cellular N content,
but had little effect on P content (Figure 4.3). Nitrogen content (TN: DM) in both the N-
enriched and the NP-enriched treatments was much higher than in the control treatments
(P=0.003 and 0.014) and the P enriched treatments (P=0.001 and 0.004). P content of
periphyton (TP: DM) in the nutrient enrichment treatments was not signiﬁcantly different
from that in the control treatment (P > 0.10). P content in the P enriched treatments was

marginally higher than in the N-enriched treatments (P=0.062).

 

 

 

 

 

 

 

 

 

 

 

 

50 r i r 1 .1 r l 40 1 l 1
4O : 3O ..
37°28 -?0 ~92” '
if, 510.r 1910 '
=10 3° 5 29*-
:s 3 ‘2
E E : 1%
Z (5.21? 12
"‘ : :
C N P NP C N P NP C N P NP

Figure 4.3. Algal cellular nutrient contents and molar N:P ratios in control, N-enriched,
P-enriched, and NP-enriched treatments on artiﬁcial substrata.

Low cellular N:P ratios on clay pots indicated that algal growth would be N limited
(Figure 4.3). Molar cellular N:P ratio averaged 4.490 (21:0.686) in the control treatment,
11.148(:l:2.163) in the N-enriched treatment, 5.584(i1.614) in P-enriched treatment, and
9.971(i2.163) in NP treatment. Cellular N:P ratio in N- and NP-enriched treatments were

signiﬁcantly higher than control (P=0.004 and 0.016) and P-enriched (P=0.042 and

0.059) treatments. N:P ratios were not less in P-enriched than control treatments.

98

 

 

h

 

 

 

 

Chl a (ug/cmz)
5 c3, 9
Algal densrty (lemz)
9%. .22» ..3 3 c
- Z '——l:l:l—‘
AF‘DM (mg/erg?)
a 5 o
- z —1:1:1—-

 

 

 

 

 

 

 

 

 

 

 

 

: t r r I l 1 103 ‘ l I
C N P NP A C P NP C P NP
100 I r N I I r l I
NA F 1&0"! 1. .. ..
5101'F i 5101: :A10 E 3
>1 : : u "I I” :
310-2- ‘ E s 35
0 I I 3 —3 h
E : : O10 I IE 0- .
:1 _; > : : 10
310'! . 1. .2 - N
> 5 E mloir 15
.2104! E E 50
m : I O ' ‘
10 L r r r ‘ E105!- r r r 1 10-1 1 r r -
C N P NP 9 C N p NP C N P NP

Figure 4.4. Algal biomass/production in artiﬁcial substrata of control, N-enriched, P-
enriched, and NP—enriched treatments. They are represented by different attributes.
Differences in algal biomass on nutrient diffusing substrata also indicated that
nitrogen was overall the most important limiting nutrient in the 22 restored Michigan
wetlands (Figure 4.4). Chl a, algal cell density, and algal biovolume were all signiﬁcantly
higher on N- and NP-enriched substrata than on control and P-enriched substrata
(P<0.05). DM and AFDM responses to N enrichment were not as great as chl a, cell
density, and biovolume. DM with N and NP additions was signiﬁcantly greater over the
control treatment (P=0.001 and 0.026 respectively) but not among other treatments.
AFDM in the NP enriched treatments was higher than in control, P-enriched, and N-
enriched treatments (P<0.01). No differences between the P-enriched treatments and
control treatments or between NP- and N-enriched treatments were found (Figure 4.4).

Little evidence of secondary P lirrritation was found.

99

Chl a: DM and Chl a: AFDM ratios were different among different treatments

(Figure 4.5). N- and NP- enriched treatments had signiﬁcantly higher Chl a :DM ratios

than control and P-enriched treatments (P<0.05). N-enriched treatment also had higher

Chl a: AFDM ratio than the control, P-enriched, and NP-enriched treatments. chl a :

AFDM in the NP—enriched treatment did not differ from control treatment, but was higher

than the P-enriched treatment.

 

 

 

 

.. I i l I ‘
100F 1
’53 E =1: 5
E .
w .
t? 102'
2 E 1
a . .
52
a 1P i
E E * E
U i t . f -
C N P NP

 

 

 

 

l lllJ

 

 

..‘L
’67)
E 1.0? l 1
°° E — 5
§ : :
a T I .
26 * ‘
50.1:- '5
l I l -
C N P NP

Figure 4.5. Chl a: AFDM and Chl azDM ratios in control, N-enriched, P-enriched, and
NP-enriched treatments on artiﬁcial substrata.

Table 4.2. Algal nutrient limitation and corresponded water nutrients and sediment
nutrients. Value in the parentheses are correlations between log-transformed value and

algal nutrient limitation. A * indicates signiﬁcant correlation (Dunn-Sidak P<0.05)
Observed Correlation
water nutrients

sediment nutrients

 

N P DlNzTP N:P N P N:P
Primary N (N-C)/C 0.034 0612* 0.437 0.297 0.199 0.634- 0.378
lirrritation (0.122) (0592') (0546*) 00.601") (0.082) (0.150) (0.430)
Primary P (P-C)/C 0.121 0.239 0.203 0.300 0.141 -0.089 0.137
Limitation (-0.069) (0.253) (0.237) (-0.282) (0.138) (0.043) (0.089)
Secondary N (NP-P)/P 0.143 0.062 0.139 0.014 0.181 0.153 0.007
Linritation (0.028) (0.131) (0.049) (0.1 15) (-0019) (-0.106) (0.019)
Secondary P (NP-N)/N 0.116 0.049 0.168 0.180 -0.064 0.239 0.309
Limitation (0.141) @049) (0.011) (0.115) (0.206) (0567*) (0.382)

 

100

The magnitude of N - and P-lirrritation in individual wetlands also showed that N
limitation was consistently greater than any evidence of P limitation (Figure 4.6). Several
wetlands showed some signs of P-linritation. However, the magnitude of N limitation,

ranging from 0-13.5 units was much greater than the 0-2.5 units range in P-limitation.

 

 

 

15 I I I
O
0 Figure 4.6. Comparisons of algal N and P
O 10 ' . , ‘ limitation in all 22 wetlands. Primary N or P
G . limitation is deﬁned as algal biomass of that
é 5 . : . nutrient treatment minus control treatment then
8 ‘ . weighted by control treatment. Primary nutrient
3:3: 5:. linritation is deﬁned as algal biomass of that
'g 0 " 0‘ ' nutrient treatment minus control treatment then
'23 weighted by control treatment.
_5 I l I

 

-5 0 5 10 15
P limitation (P-C/C)

Primary limitation of algal accrual by N supply was related to nutrient concentrations
and ratios both in the water column and sediments (Table 4.2, Figure 4.7). Direct
correlation between N-lirrritation and DIN concentration was very weak. However, the
magnitude of N-limitation in a wetland was positively correlated with TP both in the
water and in the sediments, negatively correlated with DlNzTP and N: P ratios in the
water column, and negatively correlated with N:P ratio in the sediments (Table 4.2,
Figure 4.7). Secondary P limitation was negatively correlated with TP concentration in
the sediments (Table 4.2).

4.4.3 Algal species composition in natural and artiﬁcial substrata

Changes in species composition among wetlands indicated nutrients were an

important factor regulating algae in epiphytic habitats. A total of 115 diatom taxa out of

213 algal taxa were found in epiphytic samples. Epiphytic non-diatom algae were

101

predominated by Oedogonium, Mougeotia, Scenedesmus, Pediastrum, and Trachemonas

species, while Navicula minima, Achnanthes minutissima, and Navicula cryptocephala

were the dominant diatom species. CCA based on epiphytic diatom assemblages and

environmental variables showed that 31.4 % of the total variance of diatom assemblages

were explained by the ﬁrst three CCA axes (Table 4.3). Epiphytic diatom assemblages

were closely correlated with pH (P=0.005). Conductivity (P=0.07), DIN (P=0.075), and

TP (P=0.015) were the other important environmental factors with high eigenvalues

explaining diatom distribution (Table 4.3, Figure 4.8).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30 I f 1 y 20 I I I I I 20
Pearson r=-0.199 Pearson 1:0608
20- ° - . '
o 10 h. 0 I 10
z. 0 1- C‘ :‘ 0 .0 a O h T O
‘E
e
g _10 r r 1 1 r r 1 r r
E 0.0 0.5 1.0 1.5 2.0 25100 1 2 3 4 5 6'10
' TN 1n water(mg/L) TP in water(mg/L
Z 30 r ' ' ' 20 I I I I a 1 20
t: Pearson r=-0.034 Pearson 1:0.646
o 20- . d O .
. ' 10 - o '. - 10
10’. . '1 0.0 . o .
.0 ‘ 1? ' ' t . . "
0... J... 5 o .. O O
- o 10 20 30 4o 50'100 10 20 3O 40 50 60 7010
TN in sediment (pg/mg) TP in sediment(ug/ mg)

 

 

' 1' "I'." V ' "U'VI'

Pearson r: - 0.591

 

 

0.1

1.0 10.0
N :P in water

 

 

I FTTI

Pearson r=-0.454

 

1 LI]!

 

2 4 8 12
N:P in sediment

Figure 4.7. Comparisons of algal N and P limitation in all 22 wetlands. Primary N or
P limitation is deﬁned as algal biomass of that nutrient treatment minus control treatment
then weighted by control treatment.

Changes in algal assemblages of sediment also revealed that nutrients regulated

species composition. The majority of algal taxa (173) found in sediment samples were

102

diatom taxa (148). Non-diatom algae were mainly composed of Oscillatoria and
Scenedesmus, while diatom communities were dominated by N. minima, A. minutissima,
N. cryptocephala, and Achnanthes lanceolata var. frequentissima. CCA related changes
in diatom assemblages to pH and N gradients (Figure 4.8). The ﬁrst three ordination axes
of CCA explained 27.4% of diatom taxa data. The ﬁrst axis was correlated with pH

(P=0.01), and the second axis was positively correlated with DIN (P=0.04)(Table 4.3).

Table 4.3. Weighted correlation matrix from Canonical Correspondent Analysis (CCA).
Epiphytes Sediment phytoplankton Artiﬁcial substrata
Axisl AxisZ Axis3 Axisl Axis2 Axis3 Axisl AxisZ Axis3 Axisl AxisZ Axis3
Variance 14.3 9.2 7.9 11.5 8.2 7.7 9 8.5 7.9 14.3 9.0 8.0
pH 0.596 -0.453 0.128 -0.655 -0.026 -0.330 0.433 0.090 0.040 -0.509 -0.245 0.609
Conductivity -0.036 0.497 -0.208 -0.248 0.031 -0.482 -0.099 -0.076 -0.385 -0.244 0.746 0.261
Canopy Cov. -0.259 0.338 0.052 0.088 0.325 0.148 -0.075 0.338 -0.045 -0.073 -0.286 0.456

DIN -0.308 0.363 -0.520 -0.314 0.577 -0.511 -0.189 -0.044 0.084 -0.261 0.390 0.378
SRP -0.454 -0.094 0.135 0.123 -0.331 -0.156 -0.460 0.223 -0.250 -0.230 0.277 0.160
TN -0228 -0.201 -0.502 0.193 -0.250 -.0118 -0.353 -0.040 0.304 -0.230 0.277 0.160
TP -0.331 -0.008 -0.654 0.216 -0.034 0.296 -0.415 0.660 0.051 -0.115 0.104 0.492
TNzTP 0.289 0.374 -0.052 -0.066 -0.144 -0.115 -0.022 0.022 0.233 -0.125 -0.620 -0.421
Sed.N 0.138 -0.062 -0.117 0.013 -0.202 0.283 -0.456 0.005 0.304 -0.172 -0.132 0.250
Sed.P 0.218 0.335 0.152 -0.407 -0.l48 -0.416 -0.293 0.045 0.304 0.268 0.167 0.175

 

Phytoplankton assemblages were also affected by nutrient factors. Phytoplankton
samples were very low in algal species and cell density in most wetlands. Only 99 non-
diatom taxa were found. Swimming or ﬂoating genera, such as Scenedesmus,
Pediastrum, Euglena, Trachelomonas, and Oscillatoria appeared in some of these sites,
but none of them dominated in all wetlands. Diatom density was extremely low so they
were not used for further analysis. CCA of non-diatom assemblages related changes in
non-diatom species composition to TP (Figure 4.8). The ﬁrst three ordination axes of
CCA explained 25.4% of diatom taxa data. The only signiﬁcant environmental variable

was TP (P=0.03), which correlated with the second CCA axis (Table 4.3).

103

   
  
   
        
 

epiphytes DIN 3 sediment
Conductivity

DIN:SRP
. CMMICROC

   
    
   
   
 

. NAPUPREC
NACAPITA

 

 

 

 

 

DIN CABACILL GOCLAVAT . Sed. N
ACLANCEO ' .
ACMINUT ACHUNGAR
° ' ° ° Conductivity ° .
.' 0 Axis L NALIBONE ' °-. '. Axisl ‘
V r . U . r
ACEXIGUA - RPGIBBA ' NIPALEA
NAERIFUG . . '-
EUBILUNA o . . o NAVENE‘I‘A . EUBILUNA
oooRACII.NALIBO PH FRCAPGRA
O
FRCAPMES .SYRUMPEN CI'H'HCROC
' DO
NIPALDEB PH
N
o 24 C d “M 5 artificial substrata
'11) AK AB :5 phytoplankton 0“ “ 9'

CMMICROC

NIRADICU ‘ .NIPALDEB
NAVENETA . GOPARVUL

. .
TP' - ACHUNGARA“:l

NASUBMIN . .0 0. .

- NIPALEA

.

, ° ACLANCEO
RPGIBBA

CABACILL

 

 

 

 

 

 

N:P

Figure 4.8. CCA biplots of algal assemblages and environmental variables. Diatom taxa
on macrophytes, in sediments, and on artiﬁcial substrata in control treatments, and algal
genera of phytoplankton were used for the CCA. Codes for diatoms: ACEXIGUA-
Achnanthes exigua, ACI-IUNGAR- Achnanthes hungrarica, ACLANCEO- Achnanthes
lanceolata, ACMINUTI— Achnanthese minutissima, CABACILL- Caloneis bacillum,
CCPLALIN- Coconeis placentula var. linearis, CMMICROC-Cymbella microcephala,

E UBILUNA- Eunotia bilunaris, FRCAPGRA- F ragilaria capucina var. gracilis,
FRCAPMES-Fragilaria capucina var. mesolopta, GOCLA VAT-Gomphonema clavata,
GOPARVUL-Gomphonema parvulum, GOGRACIL—Gomphonema gracilis, NACAPITA-
Navicula capitata, NACRYPTO- Navicula cryptocephala, NACRYTEN- Navicula
cryptotenella, NAERIFUG- Navicula erifuga, NALIBONE- Navicula libonensis,
NAMINMA- Navicula minima, NAPUPREC-Navicula pupula var. rectangularis,

NA VENETA- Navicula veneta, NIFRUSTU-Nitzschiafrustulum, NIPALEA-Nitzschia
palea, NIPALDEB-Nitzschia palea var. debilis, NIRADICU- Nitzschia radicula,
RPGIBBA-Ropalodia gibba, S YRUMPEN-Synedra rumpens. Code for phytoplankton
genus: AB-Anabaena, AK- Ankiostrodesmus, DI-diatoms, CE—Coelastrum, CF-
Cladophora, CQ—Coelosphaerium, DB- Dinobryn, GL—Glenodinium, KR- Ceratium, MB
-Micrasterias, OF- Ophiocytium, OS-Oscillatoria, PA- Pediastrum, PV- Pandorina, SC-
Scenedesmus, XA- unknown algae.

104

Only 137 non-diatom taxa and 113 diatom taxa were found on the control artiﬁcial
substrata. About half of the wetlands were dominated by Stigeoclonium, Oedgonium, and
Mougeotia green macroalgae. The others were mostly diatom species, predominantly N.
minima, A. minutissima, and N. cryptocephala. CCA indicated that 31.3% of variance in
the diatom taxa data was explained by the ﬁrst three CCA axes. Diatom assemblages in
control substrata were correlated with pH (P=0.005), conductivity (P=0.015), N: P

(P=0.06), and TP (P=0.065) (Table 4.3).

 

 

 

 

 

 

 

 

100 I 1 I 1 1 I Figure 4.9. Box-plots of
90 _ _ similarities of diatom
assemblages between different
80 ' ‘ habitats and treatments. E-
70 __ __ epiphytes, S-sediment, C-
L i control, N-N treatment, P-P

g; 60 ' ‘ treatment, NP- NP treatment. X-
‘7,’ 50 _ _ axis shows the comparison of
(L __ two paired habitats. Y-axis

40 ' 1" “ shows the Percentage similarity

30 .. '1' .. between the two habitats.

:1:
I -
1 0 1 1 1 1 1 1

E28 C:E C:S C:N C:NP C:P
Substrate comparison

Different substrata might have different effects on diatom species assemblages,
however, the similarities in species composition on natural habitats, epiphyton (E) and
sediments (S), was not signiﬁcantly different than between natural habitats and control
treatments (Figure 4.9). The average similarities between epiphytic and sediment
diatoms, and epiphytic and control treatment diatoms, and sediment and control treatment

diatoms (0.444(:|:0.053), 0.442 (21:0.038), and 0.474 (i0.034) respectively) were not

105

signiﬁcantly different (P>0.05). Diatom species assemblages of all treatments on artiﬁcial
substrata were more similar (P<0.05) to each other than with those on natural substrata
(Figure 4.9).

4. 4.4 Algal assemblage response to nutrient additions

Both diatom and non-diatom algae increased in density and biovolume in N -enriched
treatments and NP-enriched treatments (P<0.01). The most dominant algal taxonomic
group in control treatments was Chlorophyta (45.2 :1: 0.077%), followed by diatoms
(21.9: 6.2%), Euglenophyta (15.7i 5.9%). Only a small proportion of the biovolume was
cyanobacteria (5.6 i- 2.7%). N-enrichment did not change the relative biovolume of
diatoms (29.6 i 6%) and ﬁlamentous green algae (48.51 7.6%) overall, but the relative
biovolume of cyanobacteria (0.9: 0.6%) was signiﬁcantly lower in the N-enriched
treatment than in the control treatment (P<0.05). P-enrichment did not change the relative
biovolume of any of the algal groups. NP-enrichment also decreased the relative
biovolume of cyanobacteria (1.9i 0.9%). The most dominant algal genera, Oedogonium,
Mougeotia, and Stigeoclonium, did not respond to nutrient addition based on relative
biovolume (P>0.1). although they all increased in total biovolume and density with
nitrate addition.

The most abundant diatom taxa responded to nutrient additions in a variety of ways
on artiﬁcial substrata (Table 4.4). The relative abundance of the most abundant taxa, such
as N. minima, A. minutissima, A. exigua, C. placentula var. lineata, and N. frustulum did
not change with nutrient addition. The relative abundance of G. parvulum, and N.

cryptocephala increased with nitrogen addition (P<0.1), while the relative abundance of

106

N. palea and N. palea var. debilis decreased. The relative abundance of A. lanceolata and
A. lanceolata v. frequentissima increased with P enrichment.

Table 4.4 Twenty-six most dominant taxa and their responses to nutrient additions. The
4-digit number represents comparisons of the relative abundance of a diatom taxa in a
treatment with control, N-enrichment, P-enrichment, and NP-enrichment treatments. A
signiﬁcant change (P<0.l) of relative abundance between treatments was presented by 1,
and insigniﬁcant effect was 0. * is empty when compared with itself.

Diatom taxa N-enriched P-enriched NP-enriched Indicator

 

Achnanthes exigua Grun. 0*00 00*0 000* -
Achnanthes hungarica (Grun.) I *00 00*0 000* N
Grun

Achnanthes lanceolata (Breb.) 0*00 10*0 100* P
Grun.

Achnanthes lanceolata v. 0*00 10*0 000* P
frequentissima Lange-Bert.

Achnanthes minutissima Ktltz. 0*00 00*0 000* -
Amphora veneta Kutz. 0*00 00*0 000* -
Cocconeis placentula v. lineate 0*00 00*0 000* -
(Ehr.) V. H

Cymbella microcephala Grun. 0*00 00*] 101 * N
Epithemia sorex Kutz. 0*00 00*0 000* -
Eunotia billularia (Ehr.) Mills 0*00 00*0 000* -
F ragillaria construens(Ehr.) 0*00 00*] 001 * N
Grun.

F ragillaria capucina v. gracilis 0*00 00*0 000* -
(08m) Hust.

Gomphonema clavatum Ehr. I *01 11 *1 001 * N and P
Gomphonema gracile Ehr. 0*00 00*0 000* -
Emend. V.H.

Gomphonema parvulum Kutz. 1*00 00* 1 001* N
Navicula capitata Kutz 0*00 00*0 000* -
Navicula cryptocephala Kutz I *00 00*0 000* N
Navicula minima Grun. 0*00 00*0 000* -
Navicula semilunum Grun 0*10 11 *1 001 * N
Navicula trivialis Kutz 0*00 00*0 000* -
Navicula veneta Kutz 0*00 00*0 000* -
Nitzschia amphibia Grun. 1 *00 10*0 000* N
Nitzschia frustulum (Kutz) 0*00 00*0 000* -
Grun.

Nitzschia palea v. debilis 1 *00 01 *0 000* P*
(Ktltz.) Grun.

Nitzschia palea (Kuetz.) Grun. I *00 00*0 000* P*
Nitzschia perminata (Grun.) 0*00 00*] 001 * P

Peragallo

107

A number of diatom species distributed along the nutrient gradients in CCA
ordinations on artiﬁcial substrata, macrophytes, and sediments (Figure 4.8). On
macrophytes, Cocconeis placentula dominated in high DIN environments and Nitzschia
palea var. debilis, Synedra rumpens. and Navicula libonensis dominated in low DIN
environments. Cymbella microcephala preferred high DIN: SRP environments but
F ragilaria capucina var. mesolopta dominated in low DIN:SRP environments.
Achnanthes exigua and Eunotia bilunaria were abundant in high TP environments. In
sediment habitats, Navicula capitata, Gomphonema clavatum, Navicula pupula var.
rectangularis appeared in high DIN environments while F ragilaria capucina var. gracilis
and E. bilunaria dominated in low N conditions. On artiﬁcial substrata, Navicula
cryptonella and Caloneis bacillum showed preference for high N: P environments while
N. palea var. debilis for low N: P environments.

Table 4.5. Correlations between relative abundance of nutrient indicator species and
nutrient concentration in different habitats. * indicates signiﬁcant correlation (P<0.05).

 

 

 

DIN SRP TN TP Sed. N Sed. P
N indicator on artiﬁcial substrata 0.310 0.236 0.034 -0.105 0.044 0.123
N indicator on macrophytes -0.004 0.041 0.339 -0.137 -0. 122 -0.350
N indicator in sediment 0.168 0.006 0.299 -0.068 -0.208 -0.079
P indicator on artiﬁcial substrata -0. 165 0.373 0137 0.562* -0.045 0.032
P indicator on macrophytes -0.241 0.330 -0.220 0.470* -0.163 -0.115
P indicator in sediments -0.019 0.282 -0.11 1 0.437 * -0.040 -0.076

 

Although individual taxa responded to a nutrient gradient differently, the indicator

species (Table 4.4) delineated from experimental nutrient additions correlated with

natural nutrient concentrations of wetlands (Table 4.5). Indicator diatom species for N

and P were listed in Table 4.4. Relative abundance of N indicator species on artiﬁcial

substrata weakly correlated with DIN concentration (r=0.331) in water column (Table

108

4.5). N indicator species on macrophytes and sediments weakly correlated with TN
concentrations in water column. P indicators on all substrata were consistently correlated

with TP concentrations in water column (Table 4.5).

4.5 Discussion

4.5.] Algal nutrient limitation

Four lines of evidence indicated that N was the primary limiting nutrient for
periphyton growth in restored wetlands in southern Michigan. First, ambient NO3-N
concentration and N :P ratios were low. Second, N and NP additions caused increased
periphyton cellular N content and cellular N:P ratios in the wetlands, which indicated a
physiological N limitation for algal cells. Third, the highest algal biomass on N and NP
enriched substrata revealed that N could stimulate algal biomass accrual in these
wetlands. Fourth, this N limitation was 5 times stronger than any indication of P
limitation in these systems.

The low DIN concentration, especially depleted NO3'-N concentration (<1011g N03'-
N/L), in most of the wetlands indicated strong N-lirrritation. Previous studies have found
that periphyton growth could be limited when ambient levels were 50-90 11g NO3-N/L
(Grimm and Fisher 1986, Lohman et al. 1992, Stelzer and Lamberti 2001), while
Maberly et al. (2002) found that nitrogen limitation was more likely than phosphorus
limitation where the DW was <65 pmol IL (7111g NIL) . The restored wetlands in
southern Michigan were generally depleted in N03'-N, but not N Hf. Several possible
biogeocherrrical processes might have led to the low NO3'-N concentrations in these
wetlands. One possible cause was sequestering of nitrogen by duckweed, which can

compete with phytoplankton for nutrients and deplete DIN supply (Pokomy and

109

Rejmankova 1983, Oscarson et al. 1989). Denitriﬁcation also can be a signiﬁcant path of
nitrogen loss, especially N031 In duckweed covered wetlands, low light irradiance and
high decomposition rate led to low DO concentration. The low dissolved 02 could have
limited nitriﬁcation rates and caused a relatively high NH]: N03 ratio.

Although nutrient limitation is presumably related to the concentration of that
limiting nutrient, the magnitude of nutrient limitation was not negatively correlated with
DIN or TN concentrations. It was negatively correlated with TP concentrations in
southern Michigan wetlands. Studies of nutrient addition experiments in New Zealand
streams (Francoeur et al. 1999) indicated that the degree of nutrient limitation (mostly N
limitation or NP co-limitation) was negatively correlated with NO3-N concentration when
NOg' ranged from 3.2 —367p.g/L. NO3'-N concentrations in our wetlands were extremely
low or depleted. The absence of a gradient of N concentrations could have caused the
lack of a possible relationship between N and the magnitude of nutrient limitation. The
increased TP concentration increased the difﬁculty for periphyton to absorb NO3-N from
the water column and led to greater N-lirrritation with increased P gradient. In high
ammonium wetlands, duckweeds covered the water surface and light may have been a
limiting factor that prevented NH4+ utilization. Algal biomass accrual was barely
detectable in these wetlands.

N :P ratios in the water column and in sediments were much lower compared with the
Redﬁeld ratio, the optimal N:P ratio for periphyton growth (Redﬁeld 1958, Tilman 1977,
Fairchild 1985, Sommer 1993, Vymaza 1995, Hillebrand and Sommer 1999, Guildford
and Hecky 2000, Stelzer and Lamberti 2001). In a review study on nutrient limitation of

phytoplankton in lakes and oceans, Guildfold and Hecky (2000) concluded that the

110

absolute value of TN might not be a strong predictor of N limitation, the TN:TP ratio
might be more useful determining the potentially limiting nutrient. However, Stelzer and
Lamberti (2001) found that the overall abundance of lotic periphyton responded
positively to increased total nitrogen concentration, but not to the N:P ratio. Apparently,
nutrient concentration and nutrient ratios could both play important roles in different
systems.

Increased cellular N content and N :P ratio after N addition is a strong evidence for N
limitation of algal accrual. Algal cellular N:P ratios in our control treatment were much
lower than the Redﬁeld Ratio. Increased N availability signiﬁcantly increased cellular N
content, but P addition did not change the P content which indicated that the ambient P
concentration in the water column was saturated for algal growth. Previous studies
showed that N and P enrichment would change N:P ratios and N and P contents in algal
tissues (Vymazal et al. 1994, Vymaza] and Richardson 1995, Hillibrand and Sommer
1999, McCormick et al. 2001, Hillibrand and Kahlert 2001, Stelzer and Lamberti 2001).
In the P-lirrrited Everglades, algal N:P ratio was higher than 100:1 (McCormick et a1.
2001) and increased P loading increased P content in the algal periphyton. Experimental
manipulation of stream periphyton growth under different nutrient concentrations and
N:P ratios revealed that periphyton N and P content increased with the N and P
concentration of stream water (Stelzer and Lamberti 2001), and the cellular N: P ratio
signiﬁcantly followed the N:P ratio in the stream water.

Algal chl a concentration, biovolume, and cell density increased with N and NP
enrichments in restored Michigan wetlands, but DM and AFDM showed different

response. This difference can be found by examining the chl a to AFDM and DM ratios.

111

My results showed that N-enriched periphyton was characterized by higher chl a to
AFDM ratios than N-limited periphyton, and NP-enriched periphyton had lower chl a :
AFDM than N enriched treatment. Previous studies also found the same results (Grirrun
and Fisher 1986). Triska et al. (1983) found no difference in AFDM between N enriched
and control treatments despite large differences in chl a because of higher chl a to AFDM
ratios of N-enriched periphyton. Healey and Hendzel (1979) proposed that chl a to
biomass ratio was a reasonably good indicator of N deﬁciency for several algal species
because chl a synthesis often was reduced or ceased in N -deﬁcient algal cells. Such cells
tend to accumulate carbon storage compounds (Healey 1973). Grimm and Fisher (1986)
attributed low chl a :AFDM ratio to earlier development of a thick layer of organic
material (detrital accumulation) on the substratum surface in N-limited substrata. Weber
(1973) proposed that the autotrophic index should be calculated based on estimates of
periphyton chl a and AFDM. Apparently, N enrichment could stimulate high algal
accumulation rather than high organic matter accumulation in P additions. Competition
for nutrients in N-lirrrited substrata may have stimulated heterotrophic microbial
processes (Tank and Webster 1998).

Nutrient limitation is a complicated process and could be easily confounded by other
factors. Algal nutrient limitation could be affected by many biotic factors (herbivory,
resource competition) and abiotic factors (light, hydrology, and substrata) (Hill and
Knight 1988, Munn et al. 1989, Hansson 1992, Cronk and Mitsch 1994, Toetz 1999,
Lamed and Santos 2000, Rosemond et al. 2000). These factors sometimes confound
change in periphyton growth with nutrient addition (Lowe et al. 1986, Hill and Knight

1988, Millard et al. 1996). The wetland systems in this study had high duckweed cover

112

and low phytoplankton density. Low algal accrual on control and nutrient enriched
treatments indicated that light limitation and grazing were probably important in these
wetlands. Despite the possibility that these confounding factors might have reduced the
signiﬁcance of algal nutrient limitation, nutrient diffusing bioassays still showed strong
responses of algal biomass and algal assemblages with N additions.
4.5.2 Algal species composition and nutrient concentrations

Although different habitats within wetlands and pH variation among wetlands
affected algal species composition, the response of algal species to nutrient and nutrient
ratios were consistently observed in all three CCA in our study. Previous studies in lakes
and streams have indicated that substrata type (Pringle 1990, Cronk and Mitsch 1994,
Tank and Webster 1998), pH (Dixit et al. 1992), and nutrients (Schindler 1987, Elser et
al. 1990, Stevenson 1998, Leland et al. 2001) are three important factors affecting algal
species composition. Algal communities on different substrata had different algal
assemblages. However, the similarities of diatom communities between artiﬁcial
substrata with sediment and epiphytic habitats were not different than the similarities of
algal assemblages between natural (sediment and plant) habitats in this study. Although
algal species composition in different habitats can respond to nutrients in different ways
(Pringle 1990), pH was the most important factor regulating diatom assemblages on all
three substrata in my study, and in a number of studies (Dixit et al. 1992, Lindstrom
1996, Stevenson and Pan 1999). Although algal response to nutrient gradients in wetlands
was not as strong as the response to pH and conductivity (Pan et al. 1996, Stevenson et al.
1999), TP and N:P ratio are also important factors regulating algal assemblages in both

regional surveys and experimental manipulation of nutrient additions (McCormick et al.

113

1996, 2001, Pan et al. 2000). Similarly, this studies indicated nutrient concentrations and
ratios could both be important in determining diatom species composition in restored
wetlands.

One of the most noticeable changes of algal groups on artiﬁcial substrata was that
cyanobacteria decreased in relative biovolume with N addition. The property of
cyanobacteria ﬁxing N and dominating in N deﬁcient ecosystems have been noted by
many studies (Rader and Richardson 1992, McCormick et al. 1996, 2001). Most studies
indicated that N-enrichment could decrease bluegreen biomass, and low N environment
stimulated cyanobacteria dominance. Marinho and Huszar (2002) reported that
cyanobacteria abundance was negatively associated with NO3' and N/P ratio, and
positively associated with temperature and light in a reservoir. My N- and NP-enriched
treatments did decrease the abundance of cyanobacteria. However, our wetland system
did not have high cyanobacteria biovolume. Green algae (primarily Stigeoclonium and
Oedogonium) and diatoms dominated in the wetlands. Even though cyanobacteria were
supposed to dominate to compensate N-linritation in N-limited systems (Smith 1986,
Vitousek et al. 2002), low light levels and low temperatures in northern (high latitude)
wetlands could have limited cyanobacteria dominance (Watermann et al. 1999, Sekar et
al. 2002). Cyanobacteria abundance and N ﬁxation could also have been constrained by
trace-element limitation and/or grazing even when N:P ratios were low (Howarth et al.
1999).

Individual diatom taxa responded to N and P additions on artiﬁcial substrata in a
similar or different manner to their response along natural gradients. Some species

showed exactly the same response to a nutrient addition as I observed on natural substrata

114

and controlled artiﬁcial substrata. For example, N. palea and N. palea var. debilis were
found in low N and high P conditions on natural substrata as well as previous studies
(Fairchild et al. 1985). This study proved that N enrichment decreased the relative
abundance of these taxa and P addition did not change their abundance in P-enriched
environments. G. clavatum and C. microcephala were also found in high DIN
environments in natural habitats, and N addition also increased their relative abundance
on artiﬁcial substrata. On the other hand, several species that appeared in the high or low
end of nutrient gradients did not change signiﬁcantly with nutrient addition. For example,
Caloneis bacillum appeared in high DIN: SRP environments but N addition did not
signiﬁcantly increase its relative abundance. A. minutissima, which was reported to
positively respond to N addition (Fairchild et al. 1985), did not respond to nutrient
addition in this study. Pringle (1990) reported that A. minutissima was signiﬁcantly less
for all nutrient treatments relative to the control on sand-agar. A lack of pristine P
environments may have caused insensitive response from species responding both to N
and P additions and thus may have confounded the effect of N addition. Complicated
wetland environments such as different pH among wetlands and grazing may have
confounded individual species response to nutrient additions (John 1993, Rosemond et al.
1993).

Little evidence has quantitatively related experimental response of diatom taxa to
their response in natural habitats. Diatom species responses to natural nutrient gradients
have been determined and indicator values have been assigned for many species (van
Dam 1994, EPA 2002). Multispecies indicators of trophic status based on specis nutrients

preferences proved to be applicable across geological regions (Kelly 1998). A few studies

115

have related natural distribution of diatom assemblages along nutrient gradients to
experimental nutrient gradients (McCormick et al. 1996, 2001, Underwood et al. 1998,
Pan et al. 2000). My studies indicated that despite variability among taxa, the overall
response of the sensitive diatom taxa was similar for natural and experimental nutrient
gradients. However, the response of N indicator species were not as strong as P indicator
species. This may have been due to the lack of immediate N supply and N gradients in
these wetlands resulting N indicator species not responding to N addition in a gradual
way. Although the correlations between experimental results and natural gradients were
weak, they provided a pioneering attempt to identify indicator species experimentally that

could be used for wetlands.

4.6 Conclusions

Algal growth in the restored wetlands of southern Michigan was limited by N supply.
Cellular nutrient contents and N:P ratios, algal biomass accrual response to a nutrient
addition, and the strong N limitation vs. weak P limitation proved that N was the primary
limiting nutrient. Although the magnitude of limitation was not directly correlated with N
concentration in the water sediments, its strong correlation with P concentration and N: P
ratio in both the water column and the sediments indicate a strong limitation caused by N
depletion. Algal species composition in natural habitats also responded to nutrient
gradients. Experimental enrichment changed relative abundance of many dominant
species. Overall, the relative abundance of species that increased with N or P enrichment

on experimental substrata also increased as N or P concentrations increased in wetlands.

116

4.7 Acknowledgments

Tom Eitniear from Michigan Department of Fisheries and Wildlife Service provided
valuable site information. Christian Parker participated in some of the ﬁeld work and sample
collection. Kalina Manoylov assisted in diatom identiﬁcations. Julie Heinlein and Christian

Parker provided valuable comments on an early draft of the manuscript.

117

4.8 References

American Public Health Association (APHA). 1998. Standard methods for the examination
of water and wastewater. 20‘h edition by LS. Clececeri, A.E. Greenberg, and AD. Eaton.
Washington D. C., USA.

Bayley, S. E., Zoltek, J ., Hermann, A. J ., Dolan, T. J., and Tortora, L.1985. Experimental
manipulation of nutrients and water in a fresh-water marsh - effects on biomass,
decomposition, and nutrient accumulation. Limnol. Oceanogr. 30:500-512.

Bedford, B., Walbridge, M. R., and Aldous, A. 1999. Patterns in nutrient availability and
plant diversity of temperate North American wetlands. Ecology 80: 2151-2169.

Bothwell, M. L., and Stockner,J.G. 1980. Inﬂuence of secondarily treated kraft mill
efﬂuent on the accumulation rate of attached algae in experimental continuous-ﬂow
troughs. Can. J. Fish. Aquat. Sci. 37: 248-254.

Carpenter, E. R., Caraco, N.F., Correll, D.L., Howarth, R. W., Sharpley, A. N., and
Smith, V. H. 1998. Nonpoint pollution of surface waters with phosphorus and
nitrogen. Ecol. Appl. 8: 559-568.

Carrick, H.J., Schelske, C.L., Aldridge, F.J., and Coveney, M.F. 1993. Assessment of
phytoplankton nutrient limitation in productive waters - Application of dilution
bioassays. Can. J. Fish. Aquat. Sci. 50: 2208-2221.

Chambers, P.A., Dale, A. R., Scrimgeour, G.J., and Bothwell, ML. 2000. Nutrient
enrichment of northern rivers in response to pulp mill and municipal discharges. J.
Aquat. Ecosyst. Stress and Recovery 8:53-66.

Craft, C. B. and C. J. Richardson. 1997. Relationships between soil nutrients and plant
species composition in everglades peatlands. J. Env. Qual. 26: 224-232.

Craft, C. B., Vymazal, J ., and C]. Richardson. 1995. Response of everglades plant
communities to nitrogen and phosphorus additions. Wetlands 15: 258-271.

Cronk, J. K., and Mitsch, W. J. 1994. Periphyton productivity on artiﬁcial and natural
surfaces in constructed fresh-water wetlands under different hydrologic regimes.
Aquat. Botany 48: 325—341.

Dixit, S. S., Smol, J .P., and Kinston, J .C., and Charles, D. F. 1992. Diatoms - powerful
indicators of environmental-change. Environ. Sci.Tech. 26: 22-33.

Downing, J .A., Osenberg, C.W., and Samelle, O. 1999. Meta-analysis of marine nutrient-

enrichment experiments: variation in the magnitude of nutrient limitation. Ecology
80: 1157-1167.

118

Elser, J. J ., Marzolf, E. R., and Goldman, C. R. 1990. Phosphorus and nitrogen limitation
of phytoplankton growth in the freshwaters of North America: a review and critique
of experimental enrichments. Can. J. Fish. Aquat. Sci. 47 : 1468-1477.

Elwood, J. W., Newbold, J. D., Trimble, A. F., and Stark, R. W. 1981. The limiting role
of phosphorus in a woodland stream ecosystem - effects of P-enrichment on leaf
decomposition and primary producers. Ecology 62:146-158.

Emery, R. M., Moon, C.B., Welch, E. B. 1973. Enriching effects of urban runoff on the
productivity of a mesotrophic lake. Water Res. 7: 1505-1516.

Fahnenstiel, G. L., Stone, R. A., McCormick, M. J ., Schelske, C. L., and Lohrenz, S. E.
2000. Spring isothermal mixing in the Great Lakes: Evidence of nutrient limitation

and nutrient-light interactions in a suboptimal light environment. Can. J. Fish. Aquat.
Sci. 57: 1901-1910.

Fairchild, G. W., Lowe, R.L., and Richardson, W. B. 1985. Algal periphyton growth on
nutrient-diffusing substrata: An in situ bioassay. Ecology 66: 465-472.

Francoeur, S. N., Biggs, B. J. F., Smith, R. A., and Lowe, R. L. 1999. Nutrient limitation
of algal biomass accrual in streams: Seasonal patterns and a comparison of methods.
J. N. Am. Benthol. Soc. 18: 242-260.

Genter, R. B., Cherry, F. S., Smith, ER, and Cairns, J., Jr. 1987. Algal-periphyton
population and community changes from zinc stress in stream mesocosms.
Hydrobiologia 153: 261-275.

Goldsborough, G., and Robinson, G.G.C. 1996. Pattern in wetlands. In Algal Ecology:
Freshwater benthic ecosystems. Edited by R. J. Stevenson, M. L. Bothwell, and R. L.
Lowe. Academic Press, San Diego. pp. 77-117.

Grimm, N. B., Fisher, S. G., and Minckley, W. L. 1981. Nitrogen and Phosphorus
Dynamics in Hot Desert Streams of Southwestern USA. Hydrobiologia 83:303-312.

Grimm, N. B., and Fisher, S. G. 1986. Nitrogen limitation in a Sonoran Desert stream. J.
N. Am. Benthol. Soc. 5: 2-15.

Guildford, S. J ., and Hecky, R. E. 2000. Total nitrogen, total phosphorus, and nutrient
limitation in lakes and oceans: Is there a common relationship? Limnol.
Oceanogr.45:1213-1223.

Hamilton, S. 2001. Nitrogen uptake and transformation in a Midwestern U.S. Stream: A
stable isotope enrichment study. Biogeocherrristry 54: 297-340.

119

Hansson, LA. 1992. Factors regulating periphytic algal biomass. Limnol. Oceanogr. 37 :
322-328.

Hart, D. D., and Robinson, C. T. 1990. Resource limitation in a stream community -
phosphorus enrichment effects on periphyton and grazers. Ecology 71:1494-1502.

Healey, F. P. 1973. Inorganic nutrient uptake and deﬁciency in algae. Crit. Rev.
Microbiol. 3:69-113.

Healey, F .P., and Hendzel, L. L. 1979. Indicators of phosphorus and nitrogen deﬁciency
in ﬁve algae in culture. J. Fish. Res. Board Canada. 36: 1364-1369.

Hecky, RE, and P. Kilham. 1988. Nutrient limitation of phytoplankton in freshwater and
marine environments: A review of recent evidence on the effects of enrichment.
Limnol. Oceanogr. 33: 796-822.

Hill, W. R., and Knight, A.W. 1988. Nutrient and light limitation of algae in two northern
California streams. J. Phycol. 24:125-132.

Hillebrand, H., and Kahlert, M. 2001. Effect of grazing and nutrient supply on periphyton
biomass and nutrient stoichiometry in habitats of different productivity. Limnol.
Oceanogr. 46: 1881-1898.

Hillebrand, H., Durselen, C. D., Kirschtel, D., Pollingher, U., and Zohary, T. 1999.
Biovolume calculation for pelagic and benthic rrricroalgae. J. Phycol. 35:403-424.

Hillebrand, H., and Sommer, U. 1999. The nutrient stoichiometry of benthic rrricroalgal
growth: Redﬁeld proportions are optimal. Limnol. Oceanogr. 44: 440-446.

Hooper-Reid, N. M., and Robinson, G. G. C. 1978. Seasonal dynamics of epiphytic algal
growth in a marsh pond - Productivity, standing crop, and community composition.
Can. J. Bot. 56: 2434-2440.

Howarth, R. W., Chan, F., and Marino, R. 1999. Do top-down and bottom-up controls
interact to exclude nitrogen-ﬁxing cyanobacteria from the plankton of estuaries? An
exploration with a simulation model. Biogeochemistry 46: 203-231.

Kelly, M. G. 1998. Use of the trophic diatom index to monitor eutrophication in rivers.
Water Res.32:236-242.

Larned, S.T., and Santos, SR. 2000. Light- and nutrient-limited periphyton in low order
streams of Oahu, Hawaii. Hydrobiologia 432: 101-111.

Leland, H. V., Brown, L. R., and Mueller, D. K. 2001. Distribution of algae in the San

Joaquin River, California, in relation to nutrient supply, salinity and other
environmental factors. Freshw. Biol. 46: 1139-1167.

120

Lindstrom, E-A. 1996. The hurrric lake acidiﬁcation experiment (HUMEX): Impacts of
acid treatment on periphyton growth and nutrient availability in Lake Skjervatjem,
Norway. Environment International 22: 629-642.

Lohman, K., Jones, J. R., and Perkins, B. D. 1992. Effects of Nutrient Enrichment and
Flood Frequency on Periphyton Biomass in Northern Ozark Streams. Can. J. Fish.
Aquat. Sci.49:l 198-1205.

Lowe, R. L., Golladay, S.W., and Webster, J. R. 1986. Periphyton response to nutrient
manipulation in stream cleaning clearcut and forest watersheds. J. N. Am. Benthol.
Soc. 5: 221-229.

Maberly, S. C., King, L., Dent, M. M., Jones, R. I., and Gibson, C. E. 2002. Nutrient
limitation of phytoplankton and periphyton growth in upland lakes. Freshw. Biol.47:
2136-2152.

Marinho, M. M., and Huszar, V. L. D. 2002. Nutrient availability and physical conditions
as controlling factors of phytoplankton composition and biomass in a tropical
reservoir (Southeastern Brazil). Arch. Hydrobiol.153: 443-468.

McCormick, P. V., Rawlik, P.S., Lurding, K, Srrrith, E. P., and Sklar, F. H. 1996.
Periphyton-water quality relationships along a nutrient gradient in the northern
Florida Everglades. J. N. Am. Benthol. Soc. 15: 433-449.

McCormick, P.B., O’Dell, M.B., Shuford III, R.B.E., Backus, J. G., and Kennedy, WC.
2001. Periphyton responses to experimental phosphorus enrichment in a subtropical
wetland. Aquat. Bot. 71: 1 19-139.

McDougal, R.L., Goldsborough, L. G., and Hann, B. J. 1997. Responses of a prairie
wetland to press and pulse additions of inorganic nitrogen and phosphorus:
Production by planktonic and benthic algae. Archiv Fur Hydrobiologie 140: 145-167.

Millard, E.S., Myles, D.D., Johannsson, OE, and Ralph, KM. 1996. Seasonal
phosphorus deﬁciency of Lake Ontario phytoplankton at two index stations: Light
versus phosphorus limitation of growth. Can. J. Fish. Aquat. Sci. 53: 1112-1124.

Mulholland, P. J ., Steinman, A. D., Palumbo, A. V., Elwood, J. W., and Kirschtel, D. B.

1991. Role of Nutrient Cycling and Herbivory in Regulating Periphyton Communities
in Laboratory Streams. Ecology 72:966-982.

Munn, M. D, Osborne, L. L., and Wiley, M. J. 1989. Factors inﬂuencing periphyton
growth in agricultural streams of central Illinois. Hydrobiologia 174: 89-97.

121

Newman, 8., Grace, J. B., and Koebel, J. W. 1996. Effects of nutrients and hydroperiod
on Typha, Cladium, and Eleocharis: Implications for Everglades restoration. Ecol.
Appl. 6: 774-783.

Oscarson, P., Larsson, C.M., and Ingemarsson, B. 1989. Growth and nitrate uptake
properties of plants grown at different relative rates of nitrogen supply. 2. Activity
and afﬁnity of the nitrate uptake system in Pisum and Lemna in relation to nitrogen
availability and nitrogen demand. Plant Cell Environ. 12:787-794.

Pan, Y., Stevenson, R.J., Vaithiyanathan, P., Slate, J ., and Richardson, C]. 2000.
Changes in algal assemblages along observed and experimental phosphorus gradients
in a subtropical wetland, USA. Freshw. Biol. 44: 339-353.

Perrin, CI, and Richardson, J S. 1998. N and P limitation of benthos abundance in the
Nechako River, British Columbia. Can. J. Fish. Aquat. Sci. 54: 2574-2583.

Peterson, B. J ., Hobbie, J. E., and Corliss, T. L. 1983. A continuous-ﬂow periphyton
bioassay: Tests of nutrient limitation in a tundra stream. Limnol. Oceanogr. 28: 583-
591.

Pokomy, J., and Rejmankova, E. 1983. Oxygen regime in a ﬁshpond with duckweeds
(Lemnaceae) and Ceratophyllum. Aquatic Botany 17:125-137.

Pringle, C.M., Paaby-Hansen, P., Vaux, P. D., and Goldman, CR. 1986. In situ nutrient
assays of periphyton growth in a lowland Costa Rican stream. Hydrobiologia 134:
207-213.

Pringle. C. M. 1990. Effects of water and substratum nutrient supplies on lotic periphyton
growth: An integrated bioassay. Can.J.Fish.Aquat.Sci.4l:1247-1251.

Qualls, R. G., and Richardson, C. J. 2000. Phosphorus enrichment affects litter
decomposition, immobilization and soil microbial phosphorus in wetland mesocosms.
Soil Sci. Soc. Am. J. 64: 799-808.

Rader, R. B. and Richardson, CJ. 1992. The effects of nutrient enrichment on algae and
macroinvertebrates in the everglades: A review. Wetlands 12: 121-135.

Redﬁeld, AC. 1958. The biological control of chemical factors in the environment.
American Scientist 46: 205-221.

Rosemond, A. D., Mulholland, P. J., and Elwood, J. W. 1993. Top-Down and Bottom-up
Control of Stream Periphyton - Effects of Nutrients and Herbivores. Ecology
74: 1264- 1280.

Rosemond, A. D., Mulholland, P. J ., and Brawley, SH. 2000. Seasonally shifting
limitation of stream periphyton: Response of algal populations and assemblage

122

biomass and productivity to variation in light, nutrients, and herbivores. Can. J. Fish.
Aquat. Sci. 57: 66-75.

Schindler, D. W. 1987. Detecting ecosystem responses to anthropogenic stress. Can. J.
Fish. Aquat. Sci.44: 6-25.

Sekar, R., Nair, K. V. K., Rao, V. N. R., and Venugopalan, V. P. 2002. Nutrient
dynamics and successional changes in a lentic freshwater bioﬁlm. Freshw.
Biol.47: 1893-1907.

Smith, V. H. 1986. Light and nutrient effects on the relative biomass of blue- green-algae
in lake phytoplankton. Can. J. Fish. Aquat. Sci.43: 148-153.

Sommer, U. 1993. Phytoplankton competition in plusssee - A ﬁeld-test of the resource-
ratio hypothesis. Limnol. Oceanogr. 38: 838-845.

Stelzer, R. S., and Lamberti, G. A. 2001. Effects of N : P ratio and total nutrient
concentration on stream periphyton community structure, biomass, and elemental
composition. Limnol. Oceanogr. 46: 356-367.

Stevenson, R. J ., Singer, R., Roberts, D. A., and Boylen, C. W. 1985. Patterns of Epipelic
Algal Abundance with Depth, Trophic Status, and Acidity in Poorly Buffered New-
Hampshire Lakes. Can. J. Fish. Aquat. Sci.42:1501-1512.

Stevenson, R. J. 1998. Diatom indicators of stream and wetland stressors in a risk
management framework. Environ. Monit. Assess.51: 107-118.

Stevenson, R. J ., Sweets, P.R., Pan, P., and Schultz, R. E. 1999. Diatom community
patterns in wetlands of Michigan (USA) and their use as indicators of ecological
conditions. In Proceedings of INTECOL’s Vth International Wetland Conference.
Edited by A. J. McComb and J. A. Davis. Gleneagles Press, Adelaaide, Australia. pp.
5 17-527 .

Stevenson, R. J ., and Pan, Y. 1999. Assessing environmental conditions in rivers and
streams with diatoms. In The diatoms: applications for the environmental and earth
sciences. Edited by .E. F. Stoermer and J. P. Smol. Cambridge University Press,
Cambridge, UK. pp 11-40.

Stosch, H. A. von, and Reimann, B.E.F. 1970. Subsilicea fragilarioide gen. Et spec. nov.,
eine Diatomee (Fragilariaceae) rrrit vorwiegend organischer membran Nova Hedwigia
Beih. 31: 1-36.

Svengsouk, L. J., and Mitsch, W. J. 2001. Dynamics of Mixtures of Typha latifolia and

Schoenoplectus tabemaemontani in nutrient-enrichment wetland experiments. Am.
Midl. Nat. 145: 309-324.

123

Tank J - L., and Webster, J. R. 1998. Interaction of substratum and nutrient availability on
wood bioﬁlm processes in streams. Ecology 79: 2168-2179.

ter Braak, C.J.F., and Srrrilauer, P. 1997. Update Note: CAN OCO v 4.02. Ceter for
biometry Wageningen, CPRO-DLO, Wageningen, The Netherlands.

Tilman, D. 1981. Tests of resource competition theory using four species of lake
Michigan algae. Ecology 62: 802-815.

Tilman, D., Kiesling, R., Sterner, R., Kilham, SS, and Johnson, RA. 1986. Green,
b1 uegreen and diatom algae: Taxonomic differences in competitive ability for
phosphorus, silicon and nitrogen, Arch. Hydrobiol. 106: 473-485.

Toetz, D. W. 1999. Multiple limiting nutrients in a subalpine stream, Colorado Front
Range. J. Freshw. Ecol. 14: 349-355. '

Tl‘i Ska, F. J ., Kennedy, V. C., Avanzino, R.J., and Reilly, B.N. 1983. Effect of simulated
canopy cover on regulation of nitrate uptake and primary production by natural
periphyton communities. In Dynamics of lotic ecosystems. Edited by TD. Fontaine
and S. M. Bartell. Ann Arbor Sci Publ., Ann Arbor, MI. pp 129-159.

U - S - EPA. 2002. Methods for evaluating wetland condition, #11 Using algae to assess
environmental conditions in wetlands. United States Environmental Protection

Agency, Washington, DC, USA. EPA-822-R-02-021.

Underwood, G. J. C., Phillips, J ., and Saunders, K. 1998. Distribution of estuarine benthic
diatom species along salinity and nutrient gradients. Eur. J. Phycol. 33:173-183.

Van Dam, H., Mertens, A., and Sinkeldam, I. 1994. A coded checklist and ecological
indicator values of freshwater diatoms from the Netherlands. Neth. J. Aquat. Ecol.

28:117-33.

Vitousek, P. M., Cassman, K., Cleveland, C., Crews, T., Field, C. B., Grimm, N. B.,
Howarth, R. W., Marino, R., Martinelli, L., Rastetter, E. B., and Sprent, J. I. 2002.
Towards an ecological understanding of biological nitrogen ﬁxation.
Biogeochemistry 57:1-45.

Vymazal, J, Craft, C. B., and Richardson, C]. 1994. Periphyton response to nitrogen and
phosphorus additions in Florida Everglades. Archiv fuer Hydrobiologie 103: 75-97.

VYmazal, J. 1995. Algae and element cycling in wetlands. Lewis Publisher, Boca Raton,
FL.

VYmazal, J ., and Richardson, C. J. 1995. Species composition, biomass, and nutrient
content of periphyton in the Florida Everglades. J. Phycol. 31: 343-354.

124

Watermann, F., Hillebrand, H., Gerdes, G., Krumbein, W. E., and Sommer, U. 1999.
Competition between benthic cyanobacteria and diatoms as inﬂuenced by different
grain sizes and temperatures. Mar. Ecol-Frog. Ser.187:77-87

Weber. 1973. Recent developments in the measurement of the response of plankton and
periphyton to changes in their environment. In Bioassay Techniques and
Environmental chemistry. Edited by G. Glass. Ann Arbor Science publishers, Ann
Arbor, Michigan. pp. 119-138.

Wetzel, R.G. 2001. Limnology: Lake and River Ecosystems. Third Edition, Academic
Press, San Diego. p 61.

Wurtsbaugh, W. A., Gross, H.P., Budy, P., and Luecke, C. 2001. Effects of epilimnetic

versus metalimnetic fertilization on the phytoplankton and periphyton of a mountain
lake with a deep chlorophyll maxima. Can. J. Fish. Aquat. Sci. 58: 2156-2166.

125

Chapter 5
Assessing wetland response to land use using diatoms and paleoecological

characterization of reference condition

5.1 Abstract. Diatom-based biological indicators were used to establish biological and
stressor criteria for wetland protection and restoration and to indicate wetland ecosystem
health. Physico-chemical characteristics and surface sediment diatom assemblages were
studied in 35 palustrine wetlands in the lower peninsula of Michigan across a wide range
of human disturbance gradients. Paleoecological diatom assemblages from cores and pre-
settlement sediments were used to represent natural, reference conditions in wetlands.
Stressor-response relations were used to delineate stressor criteria and the land use
causing problems when biological conditions deviated from the "normal" range of natural
conditions and when deviations were associated with stressors and land use. Canonical
correspondence analysis (CCA) revealed that diatom species assemblages responded
mainly to conductivity and total phosphorus (TP). Weighted average predictive models
for these variables were generally successful. The similarity in species composition and
number of native species in current wetlands decreased with increasing nutrient
concentrations. van Dam's diatom trophic index was linearly correlated with the TP
gradient. Both the weighted average TP model and van Dam's trophic index positively
responded to agricultural land use around wetlands. These indices deviated from the
"normal" natural condition when TP exceeded 46.6 ug/L. A threshold change in TP and

biotic indices was detected when agricultural land exceeded 60% total land use.

126

5.2 Introduction

Assessment of wetland systems has been increasingly considered important with the
growing impact of human disturbance. About 53% wetlands have disappeared or been
impaired by increasing human activities since European settlement, especially due to
agricultural and urban development (Mitsch and Wilson 1996, Carpenter et al. 1998,
Detenbeck et al. 1999, Freeland et al. 1999, USEPA 2000). Under most circumstances,
wetlands are sinks and transformers of chemicals and nutrients. They buffer the impact of
terrestrial nutrient pollution on aquatic systems (Richardson 2000) and maintain
biodiversity. With the better understanding of large river, stream, and lake systems in the
past decades, the importance of wetland systems has received increasing attention. Many
states and environmental agencies have started to incorporate wetlands into their water
quality program. The US. Environmental Protection Agency (USEPA) recently
published a series of methods for evaluating the biological integrity of wetlands (e.g., van
der Valk 2001). These methods recommend using a series of ecological parameters
including landscape characterization, nutrient load estimates, hydrology, algae,
vegetation, and soil and water chemistry.

Ecological assessment should characterize condition of valued ecological attributes
(VEA) by comparison to pre-deﬁned criteria and diagnosis of stressors and human
activities that threaten or impair valued attributes. Appropriate classiﬁcation and
deﬁnition of reference conditions could help to deﬁne the normal range of natural
conditions and reduce natural variability among different systems. Stressor-response
relation help to identify causes of impairment to VEA, and relate the stressors to human

disturbance (land use). Combining the reference approach with stressor-response

127

relationships, appropriate biocriteria can be deﬁned and related to stressors causing
problems. Although both reference approaches and stressor-response relationships have
been used for assessment of ecological conditions (Somlyody 1998, Stevenson et al.
2002), no studies have combined both reference and stressor-response relationship.
Establishment of stressor criteria based on these approaches has never been reported.

Reference conditions are usually deﬁned by characterizing chemical and biological
condition at least impacted sites, which are often selected based on surrounding land use.
Unfortunatedly, most aquatic systems have suffered from eutrOphication by different
sources, such as agriculture fertilizer and urban runoff, and ﬁnding pristine sites is
unlikely (Hughes 2000). Historical records of aquatic ecosystems are well preserved in
the paleoecological sediments. Shifts in species composition between historical and
extant assemblages could reﬂect the change due to anthropogenic stresses (Wilson and
Muhler 1983, Wilcox 1995). Models based on weighted average calibrations have been
used by paleolimnologists to reconstruct the limnological history of lakes (Dixit et al.
1992, Cumming and Smol 1993, Hughes et al. 1993, Dixit et al. 1996, Smol et al. 1998).
Thus, paleoecological records from species assemblages before European settlement
could be used to establish reference conditions to indicate environmental changes before
and after human disturbance. Comparison between assemblages in appropriate reference
condition and impaired wetlands can help identify impaired conditions and degree of
impairment in these systems.

Stressor-response relationships are the foundation for evaluating ecosystems at risk
and directing ecological management (Gordon and Majumder 2000). This model is based

on an understanding of relationships between human disturbance, stressors, and VEAs of

128

biological communities. Assessment of aquatic communities identiﬁes degradation of
VEAs. By exploring the relationships between aquatic community and possible causes of
degradation based on ecological response from indicator organisms, I can relate
ecological responses of VEAs can be related to stressor gradients, and used to further
reﬁne human disturbance gradients (Lopez and Fennessy 2002). A variety of ecological
attributes and indices could be used to deﬁne biological criteria when comparing with
reference conditions (Hill et al. 2003). Based on pairing “stressor” variables with
“response” variables, ecologically meaningful criteria can be established to protect and
restore VEAs.

Diatoms have been used as indicators of ecological conditions in streams and lakes
(Charles 1985, Hall and Smol 1992, Dixie et al. 1996, Leland and Porter 2000), but
relatively few studies have used diatoms as indicators of wetland conditions (Pan and
Stevenson 1996, McCormick et al. 1996, Stevenson et al. 1999, Pan et al. 2000,
Stevenson et al. 2001). Because of their great diversity and species-speciﬁc responses to
varied chemical and physical conditions, diatoms should serve as valuable indicators of
the heterogeneous environments in wetlands (Stevenson 1998, Stevenson et al. 2001).
Diatom valves are usually well preserved in sediments because their cell walls are
composed of resistant opaline silica. Therefore, diatom paleoecological assemblages
represent past environmental conditions (Dixit et al. 1992). In recent years, diatom-based
bioassessment has been improved and numerous indices have been developed in stream
and lake habitat assessment (van Dam et al. 1994, Whitton and Kelly 1995, Hill et al.
2000, 2003). van Dam’s trophic index based on diatom autoecology (van Dam et al.

1994) is one of the most successful indices proven to be valid in inferring conditions of

129

aquatic systems in Europe. Although the diatom paleolimnological approach has rarely
been used in wetlands (Gasse 1987), diatom-based inference models by weighted average
calibration have been developed for conductivity and total phosphorus in Kentucky
wetlands (Pan and Stevenson 1996). Diatoms thus have all the potentials for indicating
wetland health.

The goals of this research were to use diatoms to establish biocriteria and related
nutrient criteria for wetland management. I used paleoecological information from pre-
European settlement sediments to represent reference conditions in wetlands, and thus
deﬁne biological criteria. Stressors were delineated by relating water chemistry to land
use around wetlands. Stressor-response relations were used to identify nutrient criteria to
protect the "normal" range of natural biological conditions.

5.3 Methods
5.3.1 Materials and methods

Preliminary analysis (Stevenson et al. 1999) of diatom assemblages in Michigan
wetlands indicated low species-environmental correlation due to the great variability
among wetlands. I selected 35 palustrine wetlands (as deﬁned by Cowardin et al 1979,
Detenbeck 2001) in the lower peninsula of Michigan along a north-south and an east-
west transect (Figure 5.1). Although there was seasonal and annual variation, water depth

of all wetlands was less than 2 m, and all but one site had water throughout the year.

130

 

Figure 5.1. Distribution of

the 35 wetlands sampled in

 

 

 

fall 1994 and summer 1995

 

 

in the lower peninsula of

 

 

 

 

 

 

Michigan along a north-

 

 

 

 

 

 

 

south and an east-west

 

 

 

 

 

transect.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80 O 80 160 Miles
55

All wetlands were visited once in fall 1994 and again in summer 1995. A sediment
core was taken from a central location in each wetland. The coring methods used
depended on water depth: a modiﬁed piston corer was used in waters deeper than 70 cm,
and push corers (Dixit et al. 1996) were used in waters less than 70 cm. The cores,
ranging in length from 20-50 cm, were cut into variable segments up to 4 cm in length.
All the segments were preserved with M3 solution (APHA 1998). The top 1 cm of
sediment depth was later used for analysis of surface diatom, and only fall sediment
samples were included. One segment from a sediment core aged older than 200 years
were used as presettlement wetland assemblages. Core samples from 14 wetlands were
analyzed, and all taxa appearing in these samples were considered as native taxa.

The dissected sediment samples from fall 1994 were digested for diatom

enumeration using concentrated sulfuric acid and KCrO3 (Stosch and Reimann 1970). All

131

surface sediments and 14 core samples from fall 1994 were selected for analysis. Diatom
samples were washed with distilled water and then mounted in Naphrax®. A minimum
of 500 diatom valves were identiﬁed at lOOOx using a Nikon® Labophot-2 microscope.
Diatom taxa were identiﬁed using the taxonomic monographs of Hustedt (1927-66),
Kramer and Lange-Bertalot (1986-91), and Patrick and Reimer (1975).

All water samples were collected from the middle of the water column. A number of
water chemistry parameters were monitored in the ﬁeld, including pH, speciﬁc
conductance, total dissolved solids (TDS), apparent and true color, and temperature.
Aerated pH was measured within 24 hours after sample collection. Water samples were
preserved and analyzed on a Skalar® Sans-Plus automated water chemistry system
according to APHA methods (1998) for total phosphorus (TP) and total nitrogen (TN),
soluble reactive phosphorus (SRP), nitrate, ammonium, and chloride. In addition, a
known volume of water was ﬁltered through a 0.45-um Millipore® membrane ﬁlter for
chlorophyll a (chl a) analysis on a Turner Designs® TOC-SOSOA ﬂuorometer. Another
water sample was also ﬁltered for dissolved organic carbon (DOC) analysis on a
Shimadzu® TOC-SOSOA.

5.3.2 Land Use coverage

Land use coverage and digital soil data were obtained from the Spatial Data Library
maintained by the Michigan Department of Natural Resource (MIDNR). These data are
stored as Georef projections and as 1:250,000 scale maps from the US Geological Survey
(USGS). Land use was grouped as urban, agriculture, rangeland, forest, water, wetland,
and barren land. Michigan major land resource and soil types were categorized according

to Michigan resource inventory as Indiana and Ohio Till Plain (1), Northern Michigan and

132

Wisconsin Sandy Drift (11), Southern Michigan and Northern Indiana Drift Plain (111),
Southern Michigan Fruit and Truck Belt (IV), and Western Michigan and Northeastern
Wisconsin Fruit belt (V) (1982). They were categorized into three regions as low erosive,
medium erosive, and high erosive regions. The Lower Pennisula of Michigan was
classiﬁed into two sub-ecoregion according to ecoregion map. Arcview 3.2 with the
Spatial Analyst 1.1 extension, along with various extensions including Michigan
projection extension and MIDNR Geometry Calculator Extension, were used to perform
all spatial analyses. The watershed boundaries of wetlands were determined by selecting
l to 2 km buffer areas around palustrine wetlands (according to size of wetlands) to

determine surrounding land use (Berven and Grudzien 1990, Findlay and Lenton 2001).

5.3.3 Statistical analyses

Diatom species relative abundances, species richness, and Shannon species diversity
(H ’) were calculated from the species composition and enumeration data. van Dam
diatom indices (van Dam et al. 1994) were calculated as the sum of products of species
indicator values and proportional relative abundance (converted for proportion of species
with indicator values) and used to indicate diatoms sensitive to nutrient enrichment
change. All variables were standardized before analysis. Descriptive statistics, Pearson
correlations, and regression analyses were conducted in Systat® v10. The Dunn-Sidak test
was used to determine the signiﬁcance of the correlation.

Principal components analysis (PCA) was used to reduce and summarize variation of
environmental variables in the region. The PCA (eigenvalue>l, 6 axes) sample scores
were used as composite variables of chemistry to assess the differences among

ecoregions or soil types.

133

Only diatom species with a relative abundance 2 1% in a minimum of three sites or _>.
5% in at least one site were included for comparison and gradient analysis.

Nonmetric multidimentional scaling (NMDS) was used to identify diatom
distribution patterns with relation to presettlement and current conditions. Euclidean
distances were calculated for all samples based on diatom species composition, and then
projected to 2-dimensional plots using NMDS (PCORD 4.0). Multiple response
permutation procedures (MRPP), a nonmetric comparison method to discriminate
differences among groups of assemblages, was used to determine the difference between
reference and current diatom assemblages.

Canonical correspondence analysis (CCA) was conducted to explore the relationship
between diatom species distribution and environmental conditions (ter Braak 1990). All
analyses were performed with the computer program CAN OCO v4.02 and PCORD v4.10
(McCune and Mefford 1999). Relative abundance data were square-root transformed and
environmental variables were log transformed as necessary to obtain normal
distributions. Whenever a variance inﬂation factor (VIF) was greater than 5, the least
signiﬁcant, collinear variables were dropped from the analyses .The environmental
variables included all measured variables and the TN:TP ratio, DIN:SRP ratio, and
wetland coordinates. Forward selection and Monte Carlo permutation (199 random
permutations) were performed to determine statistical signiﬁcance of a subset of
environmental variables. The program CALIBRATE v0.6 (Juggins and ter Braak 1992)
was used to perform weighted average regression and calibration. All models were based
on square-root transformed diatom relative abundance data. J ackkniﬁng was used to test

the accuracy of the predictive models. Regression tree analysis (De'ath and Fabricius

134

 

2000) was performed using Systat® v10 to detect appropriate threshold values of
environmental variables and community compositions with increasing agriculture land
use.

The Bray-Curtis similarity index (SI) was calculated as similarity of the relative
abundance of surface diatom assemblage at one wetland with centroid of paeleological

diatom assemblages to represent similarity with reference conditions.
5.4 Results:

5.4.1 Environmental variables

The annual average of environmental variables varied among wetlands (Figure 5.2).
Conductivity of these wetland waters ranged from 32.2 to 495.5 umol/cm (average :1:
standard error (SE), 189.8i16.2 umol/cm). The pH was generally neutral (6.45:0.75).
DOC averaged 17.9 mg/L (il.2 mg/L), DIN (137.2:t25.9 jig/L), and TN (1456.6:l:95.5
ug/L) concentrations were much higher than SRP (22.015.8ug/L) and TP (82.81: 16.3
ug/L) concentrations. Si and Cl averaged 3.9(i0.7) mg/L and 5.2(il.2) mg/L
respectively. N :P (36.01 3.2) and DIN :SRP (26.0:I: 9.0) ratios were higher than 17
indicating potential P limitation.

Environmental variables, especially nutrient variables were closely related to each
other (Table 5.1). CI correlated with conductivity(r=0.668, P<0.001) and TDS (0.593,
P<0.01). Si also correlated with conductivity and TDS (r=0.737, P<0.001), as well as
NH4 (r=0.618,P<0.05), DIN (120.551, P<0.05), and SRP (r=0.547, P<0.05). Nutrient
concentrations were tightly correlated with each other, and related to chl a concentration
in the water column. TP correlated with TN(r=0.711, P<0.05), SRP(r=0.717, P<0.05),

and N H4 concentrations (r=0.753, P<0.001). Chl a concentration in water column was

135

highly signiﬁcantly (P<0.001) correlated with TP (m0.793), TN (r=0.805), and NH4

 

 

 

 

 

 

 

 

 

 

 

(r=0.850).

O 100 200 300 400 500 5 6 7 8 9 0 100 200 300
Conductivitymmol/cm) pH Ch] a(mg/L)

4:13—— »« —Cl:l—— *4 1 l—*

l l l ' W l l I l I l
0 10 20 30 1,0 10,0 \0 \‘3 '19 $53035
SI(mg/L) CL(mg/L) DOC(mg/L)

l 2 3456 4,9 \QQ\59,5® 01020304050607080
N:P
TN(mg/L) TP<mgIL>
W l ‘ 'iJ l J l '1
100 1000 b.9 $9 (69 \9 '19 39W)“
DIN(mg/L) SRP(mg/L) DIN:SRP

Figure 5.2. Box plot of environmental variables in 1994-1995 in the Michigan wetlands.

The ﬁrst three PCA axes accounted for 61.4% of variance in the environmental data
set. The ﬁrst axis was correlated with ionic concentrations (conductivity and pH) and the
second axis correlated with total nutrients (TN and TP). Multiviate test of the ﬁrst 3 axes
revealed that environmental variables in wetlands were not different among sub-
ecoregions (MANOVA, Wilks’s 2» F627 =l.784, P = 0.140) and among soil types
(MANOVA, Wilks’s 7t F1154 =1.054, P = 0.386).

Agricultural land use around wetlands correlated with average TP concentration in
1994 to 1995 (Table 5.3, Figure 5.3). Regression tree analysis was used to explore the

relationship between agricultural land use and TP concentration. A threshold effect

136

occurred when agricultural land use exceed 58% and average TP increased from 46.7 to

207.3 ug/L. The regression tree model explained signiﬁcantly more variance than a linear

model (F test, P<0.05).

 

 

 

 

 

 

 

400 I I I I I f I I
300 l— . _
E: o
100 - O . . O -
CI ’4. .. L...l. r ”m: r r
0.0 0.2 0.4 0.6 0.8
%Agriculture

Figure 5.3. Relationship between average TP and % agricultural land use in sampled
wetlands. Regression tree analysis was used to explore the relationship between TP and
land use. A threshold effect occurred when agriculture occupied 58% land use around the
wetland watershed.

137

 

 

 

...de- 848- $2 $2... $8 and $3... 848 88 :3 83- 83 83 88.8 Ed- 233%?
83 :88- Pod 3.8- 48.8 58. ~88- 888- 5.8- 88.8 28.8 33- 23. M88 83- $8qu
83 $3 38- 83 ES 89c... 83 83- 88 33 08.8- 48.8- :3 ES ﬂip
83 888- was 84.8 83- ~88- $8 :28 M88 088- 83 Ed 83- 25
83 83 888- 83... 838 388 3.8 Rod- 88 83 23 88.8 “am
88.3 :3 83 $8... 488 one? too 588 3.8 80¢ 88 £2
83 38.8- Rod- 83 3.8 98.8 was- 88 :3 88- ..oz
83 82 ~88 888- 888 88 83 W88 83- E.
o8._ :3 58 $8 $3 :8 ﬁne 28 E
.83 3d :3 33 88.9.. $2 83- 6
83 83 $2.. 83... 82 89¢ a
88.3 83... 88.8- 83. 58 won
83 8.8 RS- 88 8.8 25
o8._ £3... 83- £82.28
83 38. an
83 deep
$5qu $2 a :6 “am £2 52 E. E 6 a com 5ch £368.50 mm .85

 

.... .3 8565 a 8.3m 83385 032288 832:8 358%?
<.W®O1VQQ— Cm Cam—Sou: MO N—Dmﬂmconm hogs mag—ON avg—~03 WM MO mD—DNE> §€OE€OHM>GD MO KEG—H— angman—0:00 COW—NOAH ..mm D~£Nr~t

138

5.4.2 Current diatom assemblages and environmental conditions

A total of 269 diatom taxa were found in the surface sediment in the 35 palustrine
wetlands during fall. Seventy-one diatom taxa (abundance > 1%, Table 5.2) from
palustrine wetlands from surface sediment in fall were used in the CCA to explore the
relationship between diatom assemblage and environmental variables. The ﬁrst two axes
explained 28.7% of cumulative variance in the taxonomical data. Stepwise permutation
tests indicated that, of the 24 environmental variables, conductivity, TP, pH, and DIN:
SRP ratio had signiﬁcant canonical coefﬁcients (P<0.05) with CCA axes and thus were
selected in the ﬁnal CCA. The four environmental variables alone in the ﬁrst two axes
captured 19.1% cumulative percentage variance of taxonomical data (Figure 5.4a, 4b).
Conductivity and TP were the most signiﬁcant environmental variables (P<0.005). The
position of each taxon in the bioplot (Figure 5.4b) indicated its weighted-average
optimum relative to other taxa and environmental variables. Epithemia adnata, Cyclotella
nana, and Gomphonema mexicanum, among other taxa, indicated high TP concentration
in the ordination space; while Cymebella microcephala, and Navicula leptostriata

appeared most in low TP concentration.

139

A

TN:TP
NOx

 

 

Ox

N
TN:TP EUPRAERU

HAAMPHI o
CMCESATI ,
NALEP’I‘OSO

0°

 

00.
CMMICROCO °

 

O
NASUBTIL ° 0‘50
oo° o M
PIVIRIDI o

ANSERIAN o

PISTREP'I‘ :

PIINTERR 0°

PIMAIOR PIOSUBCAP

° EUPARALL

O

o
FSRHOSAX

 

EUPALUDO

i

.41

DIN:SRP

39
O

7033

tl I
.. L GTN Conduc vty

OCYGAMMA

NAPRAETE
DIN:SRP
o P"
FRCONBIN
FRPINNOA . CSUBSAL
O % o

LGTN Conductivity

 

a": NAGYSINLL

0 ° 0 ACEXIGUA
° ° . GOMEXICA
o 0 0° 0
O
CYNANA LG‘I'P
oFRNITZSC

. 0 O
m , . .
o EPADNATA
GOCLAVAT

0.“

ﬂ

Figure 5.4. CCA biplot of
sites and fall
environmental variables
from 35 palustrine
wetlands in Michigan. (a)
Dots show site scores in
ordination space, (b) dots
shows species scores in
ordination space, and
arrows indicate the
strength and correlation
between environmental
factors and the ordination
axes, The dotted circle
shows sites with above
58% agricultural land
cover. The dotted circle
shows sites with above
58% agricultural land
cover. Diatom species are
indicated by eight-digits
code (see Table 5.2 for
complete taxon names).

Table 5.2. The most dominant diatom taxa used in CCA and the correspondent taxa code

 

 

Code Taxa name Code Taxa name

Gomphonema gracile Ehr. emend.
ACCONSPI Achnanthes conspicua A. Mayer GOGRACIL V. H.

Gomphonema parvulum (Kiltz.)
ACEXIGUA Achnanthes exigua Grun. GOPARVUL Kutz.

ACHUNGAR Achnanthes hungarica (Grun.) Grun GOTRUNCA
ACLANCEO Achnanthes lanceolata (Breb.) Grun. NACRYP'TC

Gomphonema truncatum Ehr.

Navicula cryptocephaloides Hust.

140

 

ACMINUTI
ALPELLUC

ANSERBRA
ANVITREA

AUCRENUL
CCPLACEN

CCPLALIN

Achnanthes minutissima Kiltz.

Amphipleura pellucida (Kutz.) Kiltz.

Anomoeoneis serians var.
brachysira (Bréb. ex Kiltz.)
Hust.

Anomoeoneis vitrea (Grun.) Ross
Aulacoseira crenulata (Ehr.)
Thwaites

Cocconeis placentula Ehr.

Cocconeis placentula var. lineata
(Ehr.) V. H.

CMLUNATA Cymbella lunata W. Sm.

CMMICROC
CMMINUTA
CMSILESI
DNKUETZI
EPI'URGID
EUARCUS

EUBILMUC

EUCURVAT
EUEXIGUA

EUFLEXUO
EUNAEGEL
EUPRAERU

Cymbella microcephala Grun.
C ymbella minuta Hilse ex Rabh
Cymbella silesiaca Bleisch
Denticula kuetzingii Grun.
Epithemia turgida (Ehr.) Kiltz.

Eunotia arcus Ehr.

Eunotia bilunaris var. mucophila
Lange-Bert. & Norpel

Eunotia curvata (Kiltz.) Lagerst.

NACRYPTO
NACRYTEN

NALEPTOS
NAHARDII

NAMEDIOC
NAMINIMA

NAPUPREC
NARADIOS
NASEMLUM
N ASUBTIL
NATANTUL
NEIRIDIS
NIAMPHIB
NIGRACIL

NIPALEA

NIPERMIN

Eunotia exigua (Bréb. ex Kutz.) Rabh NITROPIC

Eunotiaﬂexuosa Bréb. ex Kiltz.
Eunotia naegelii Migula

Eunotia praerupta Ehr.

EURHOMBO Eunotia rhomboidea Hust.

FRBREVIS

FRCAPGRA

FRCAPMES

FRCONSTU

FRCONVEN
FRCROTON
FRPINNAT

FSRHOCRA
FSRHOMBO
GOACUMIN

GOAN GUST

F ragilaria brevism'ata Grun.
F ragilan'a capucina var. gracilis
(Q8111) Hust.

F ragilaria capucinavar. mesolepta
Rabh.

F ragilaria construens (Ehr.) Grun.

F ragilaria construens var. venter
(Ehr). Grun.

F ragilan'a crotonensis Kitton

F ragilaria pinnata Ehr.
F rustulia rhomboides var.

crassinervia (Bréb. ex W. Sm.) Ross

F rustulia rhomboides (Ehr.) DeT.

Gomphonema acuminatum Ehr.

Gomphonema angustatum (Kutz.)
Rabh.

PIGIBBA
PIINTERR
PIMAIOR
PIMICROS
PISUBCAP

PIVIRIDI
SSANCEPS
SSPHOENI

SYNAN A
SYRUMPEN
SYTENERA

SYULNA
TAFLOCCU
TAQUADRI

141

Navicula cryptocephala Klltz.
Navicula cryptotenella Lange-Bert.

Navicula leptostriata
Jorgensen

Navicula hardii Hust.

Navicula mediocris Krasske

Navicula minima Grun.

Navicula papula var. rectangularis
(Greg.) Grun.

Navicula radiosa Kiltz.

Navicula seminulum Grun.
Navicula subtilissima Cl.
Navicula tantula Hust.

Neidium iridis (Ehr.) Cl.
Nitzschia amphibia Grun.
Nitzschia gracilis Hantz. ex Rabh.

Nitzschia palea (Kiltz.) W. Sm.
Nituchia perminata (Grun.)
Peragallo

Nitzschia tropica Hust.

Pinnularia gibba Ehr.

Pinnularia interrupta W. Sm.
Pinnularia maior (Kutz.) Rabh.
Pinnularia microstauron (Ehr.) Cl.

Pinnularia subcapitata Greg.
Pinnularia viridis (Nitz.) Ehr.

Stauroneis anceps Ehr.
Stauroneis phoenicenteron (Nitz.)

Ehr.
Synedra nana Meist.
Synedra rumpens Kiltz.

Synedra tenera W. Sm.

Synedra ulna (Nitz.) Ehr.
Tabellaria ﬂocculosa (Roth) Kiltz
Tabellaria quadriseptata Knud.

Table 5.3 Regression of observed environmental variables against predicted values based
on diatom species composition. Simple regression and jackkniﬁng cross-validation
indicators based on weighted average (WA) and tolerance down-weighted WA used to
compare the precision. MSE represents mean square error and RMSE represent residual
of MSE.

 

 

 

 

 

 

Simple Cross-Validation
MSE RMSE R2 MSE RMSE R2
Conductivity WA 4585 67.713 0.741 7132 86.693 0.578
T01 d/w WA 3672 60.598 0.792 1.23E+04 110.879 0.425
LGTP WA 0.0492 0.222 0.704 0.0958 0.310 0.435
T61 d/w WA 0.0496 0.223 0.702 0.123 0.351 0.334
pH WA 0.407 0.638 0.679 0.807 0.898 0.381
T61 d/w WA 0.299 0.547 0.764 0.757 0.870 0.446
DIN:SRP WA 133.8 11.568 0.634 282.1 16.795 0.255
T61 d/w WA 131 11.447 0.641 291 17.060 0.250

 

Diatom-based WA inference models for conductivity, TP, pH, and DIN: SRP were
developed to predict environmental conditions for palustrine wetlands (Table 5.3). The
WA inference model for conductivity produced a lower R2 (0.741) than the tolerance
down-weighted WA model (WA(tol)); however, the simple WA model produced a much
better model (R2=0.578, Figure 5.5a) after cross-validation with jackkniﬁng than the
tolerance down- weighted WA model (R2=0.425). ’Ihe WA TP inference models also
showed better regression (R2=0.435) after cross-validation than the WA(tol) model
(R2=O.334). The analysis of pH data indicated that tolerance down-weighted WA models
were much better than simple WA models both before and after cross-validation. R2
values for WA models for DIN: SRP were low, especially after cross-validation

jackkniﬁng (1120.250).

142

8

I IIIIIII

Predicted TP(ug/L)
'5

 

I I IIIIII

 

I IIIII‘

A LLJII.

.31

 

 

lO
Observed TP(ug/L)

100

300

 

 

 

 

I I I I I I I
C
I
200 . . . 0'
O
O
100 o '7
' o
C . ... . ..O‘
1 L. 1 1 1 1 1
0102030405060708090
% Agricultural land

Figure 5.5. Regression models of diatom inferred environmental variables against
observed environmental variables. 5b shows the relationship between diatom inferred TP
and agricultural land use.

Based on diatom autecological data, predicted TP concentrations of wetlands along an

agricultural land use were used to infer wetland trophic level changes (Figure 5.5a). In

this model, diatom inferred TP showed a linear increase with increase of agricultural land

use reached (R2=0.246, P=0.003) (Figure 5.5b).

5.4.3 Comparison between reference and current conditions

A total of 169 diatom taxa were identiﬁed in 14 sediment core paleoecological

samples, While 189 taxa were found in the corresponding surface sediment samples. An

average of 55.4% (i2.4%) of the taxa in the surface samples were not found in the core

samples in the same wetland, while 50% (i 4%) of the core taxa were not present in

surface samples in the same site. Forty-one diatom taxa in paleoecological samples were

not found in all 35 surface sediment. Diatom species compositions were signiﬁcantly

different between current and paleoecological samples. (MRPP, P<0.001, Figure 5.6).

143

2 1 Figure 5.6. Nonmetric
o Paleoecological multidimensional scaling ordination

‘ diagrams of paleological diatom
1 ‘ Current assembalges and current diatom
assemblages. Reference(paleological
diatom assemblages) and current
(surface diatom assemblages) were
indicated by different symbols. The
centroids with 68.5% conﬁdent ellipse
were plotted around each region.

 

N MDSZ
o

 

 

 

 

 

 

 

Native Species

 

 

 

 

 

 

 

0 l llJ 00 LII!

100 200400800 ' 100 200300400

TP(ug/L) TP(Mg/L)
Figure 5.7. Relationship between richness of diatom native taxa with nutrient
concentration (TP) (a), and relationship between similarity index (SI) of current diatom
assemblage with pressettlement assemblages with nutrient concentration (TP) (b) in
wetlands. Native taxa were deﬁned as taxa found in the paleoecological samples. SI is the
Bray-Curtis similarity of diatom relative abundance in surface sediment and the centroid
of paeloecological samples.

144

Table 5.4. Correlation among community characteristics and environmental variables

 

LGTN LGTP DIN SRP Chl a % agricultural

 

land
Species Richness -0471 -0513 -0434 -0427 0.359 0.130
Shannon H’ -0.534 -0420 -0.514 -0357 0.265 0.093
NO' 9f Name -0291 -0452 -0439 -0.302 0.195 0.015
Specres
Similarity Index 0453 -0.357 -0395 -0.258 0.297 -0.062

 

Both the number of native taxa in surface sediments and similarity of diatom species
composition with reference conditions decreased With increasing nutrient concentrations
(Table 5.4, Figure 5.7). The number of native species in surface sediments was negatively
correlated with increased nutrient concentrations, especially TP in wetlands. Diatom
species composition in reference conditions showed large variation among different sites.
The similarities ranged from 0.03 to 0.38 with a median similarity of 0.13. The
similarities of current diatom relative abundances with paleoecological diatom species
composition were also negatively correlated with TN and TP concentrations (r: —0.453
and —0.357 respectively). These similarities were generally low and varied Within the
range of similarities among reference conditions. Thus a deﬁnition of biological criterion
based on similarity change was unlikely.

van Dam’s trophic indices deﬁnes the importance of trophic sensitive diatom species
and trophic conditions in wetlands. I deﬁned the 75th percentile of paleoecological diatom
trophic indices in wetlands as the biological criterion of pristine conditions. van Dam’s
trophic indices in reference conditions varied from 1.7 to 4.5 with a median value of 3.4.
van Dam trophic indices in current diatom species composition was more viriable,
especially at the higher value which represented higher relative abundance of trophic

species in some wetlands now than in historic times. The values of indices showed a

145

linear increase with increase of TP concentration (R2=0.409, P=0.001, Figure 5.8a). The
nutrient concentration at the intercept of the biological criterion with regression lines
represented the nutrient criterion (stressor criterion=45.6 um/L). Change of van Dam
trophic indices was also linearly correlated with agricultural gradient (R2=0.517, Figure
5.8b). However, a regression tree analysis also showed a changepoint at 56% agricultural

land use where the van Dam trophic index increased from 3.46 to 4.85.

 

 

 

 

:36 I I '0' I I 6 r l l r l . I r r 1
E "D‘ Paleoecological .
:5 _ . . ... . .- 5- ' Current .. . °". ..
E o ) g 0 o ...-”‘0 9
D4 1 o 9' ~-"'
04 - . . a ' . ............................ ,9-
b 0 ° 0". 3 . 0'...“ 9. o
33 3. ----- "s '3 ' .r" :0
- ..."..v C -' Q -1
Q .0 o
5 2 - ' o - -
> :1:
1 ﬂ 1 1 1 11 I 1 I A 1 l l 1 l
U

 

 

 

 

 

 

«3% \90 $3090 0 10 20 30 40 50 60 70 80 90
log 10TP(ug/L) Agricultural land use (%)

C

Figure 5.8a. Relationship between van Dam trophic index and palustrine wetland nutrient
concentration (TP). Box-plot shows paleological diatom trophic index from 14 palustrine
wetlands to represent reference wetland condition. Surface diatom trophic index (dots)
respond linearly to TP concentration in wetlands (R2=0.407, P=0.001). The intercept of
biological criteria and regression line represented nutrient criteria (stressor criteria=46
ug/L). 8b. Relationship between van Dam trophic index and palustrine wetland land use.
Box-plot shows paleological diatom trophic index from 14 palustrine wetlands
representing reference wetland condition; dots represent change of trophic index with
agricultural land use; a threshold effect was observed when agriculture exceeds 56% total
land use, with van Dam trophic index increasing from 2.46 to 4.85.

5.5 Discussion

146

The goal of ecological assessment is to assess condition of ecosystems, often relative
to an expected or desired condition, to determine the stressors that threaten or impair
valued ecological attributes, and to determine the human activities producing those
stressors. This has been challenging in wetlands due to spatial and temporal variability. I
found that changes in algal species composition in wetlands were related to human
activities and then developed biological and nutrient criteria that would protect the
natural diatom assemblages of wetlands. I found that development of assessment criteria
required classifying wetlands by type and restricting biological metrics to those most
sensitive and directly related to stressors. Characterizing reference condition with
paleoecological assemblages reduced doubt of the biological potential of the wetlands.

Nutrient concentrations, especially total phosphorus concentration in the Michigan
wetlands studied, were directly related to human activities, especially agricultural land
use. Human activities are well-known causes of nutrient enrichment of aquatic systems
(Vitousek et al. 1997, Carpenter et al. 1998). Land cover/land use indicators of human
activities and related water quality models are commonly correlated with nutrient
concentrations in surface waters (Soranno et al. 1996, Tufford et al. 1998, Freeland et al.
1999). As the receiver of storm runoff, overland ﬂow, and ground water inﬂux, palustrine
wetlands received high nutrient input from agriculture (Thompson and Polet 2000). My
study not only found that high phosphorus concentration in Michigan wetlands were
correlated with agricultural land use surrounding wetlands, but also discovered a
threshold effect when agricultural land use exceeded 57.9% total land cover. Similarly,

Waltman et al. (unpublished) also found a threshold effect on nutrient concentrations

147

when the % agriculture in Kentucky stream basin exceeded 60%. This effect may
represent saturation of the assimilative capacity of soils within watersheds.

I found that changes in diatom species composition, number and relative abundance
of individuals belonging to native species, and other valued ecological attributes changed
with nutrient enrichment associated with human activities. TP, as a common cause of
change in species composition, was an environmental stressor directly related to land use
in Michigan wetlands. Examination of diatom taxa found in wetlands with 57.9 %
agricultural land use showed that those wetlands were dominated by taxa requiring high
nutrients, such as Epithemia adnata and Cyclotella nana, and had the lowest number of
native taxa. TP enrichment was the possible cause of loss of native taxa. Both increased
species diversity and decreased species diversity have been found in aquatic systems with
increased nutrient concentrations (J utter et al. 1996, Hillebrand and Sommer 2000, Pan et
al. 2000). I expected a decreasing native taxa richness because native taxa were less
competitive in nutrient enriched environments. Competitive exclusion for other limiting
resources, increased predation, or alteration of other abiotic factors through indirect
effects of nutrients (e. g., dissolved oxygen concentration, microbial interaction, or habitat
structure) may all decrease native taxa richness.

The similarity of species composition to reference condition should be more sensitive
than loss of native taxa (presence/absence of taxa), because relative abundance of taxa
should change before they are lost from a habitat. However, my results only found a
weak linear decrease in the similarity index (S1) with increasing nutrient concentration.
The weak response of S1 is possibly because SI was affected by multiple natural

variables, such as pH and conductivity, as well as nutrient enrichment. Therefore, S1

148

responses are complexly relate to a stressor and its interaction with other stressors
(Lansid and McLaughlin 2000). Alternatively, van Dam’s trophic index represents
sensitive response of diatom relative abundance speciﬁc to one environmental stressor, in
this case, nutrients. Although it was developed in Europe, the strong relationship between
diatom trophic indices and TP in Michigan wetlands suggests that this index could have
broader geographic application.

Reference condition is ideally based on the central tendency and variability in
conditions in habitats with the least human disturbance. Paleoecological assessments
derived from sediments deposited before European alterations of landscapes provide the
ideal characterization of reference condition. When using S1 to describe reference
condition, little difference was observed between reference condition and present
condition in wetlands. SI was very low among paleoecological conditions in MI
wetlands (usually 0.1-0.2) compared to present day variability among reference sites in
the Everglades (0.7, Stevenson et al. 2002). This low similarity among reference wetlands
was probably related to the great natural variability in pH, conductivity, and even
development stage among wetlands. Greater differences between reference condition and
present day condition of wetlands was observed when diatom assemblages were
characterized using van Dam’s trophic indices, probably because van Dam's trophic
index based on diatom autoecological information of individual taxa characterizes
biological accumulation of trophic status in the system. Nutrient conditions were
probably more consistently low before European alteration of landscapes than pH,
conductivity, and other naturally varying factors that affect diatoms. Compared with S1,

autoecological indices are usually more precise for assessment and diagnosis of aquatic

149

ecosystem problems (Stevenson et al. 2001). Thus, biocriteria for wetlands were based on
van Dam's trophic index indicating the relative abundance of diatoms that were sensitive
to a single stressor, nutrient enrichment.

The biocriterion based on Van Dam's trophic index was established at the 75th
percentile of its frequency distribution, which represents an upper bound of nutrient
tolerant diatoms in the natural biological assemblages. The 75‘h percentile of reference
conditions has been deﬁned to indicate deterministic variation of algal assemblages and it
has been used in other studies (Hill et a1. 2003). A variety of percentiles in reference
conditions have been used by different agencies for different purposes. An ideal approach
for deﬁning biological criteria would be based on a threshold response along a stressor
gradient.

With the determination of a criterion for a valued ecological attribute, stressor criteria
can be determined based on the VEA criterion and a stressor-response relationship. A
cause-effect relationship has been established between nutrient enrichment and algal
response in Michigan wetlands (Chapter 4). Stressor criteria should be no higher than the
stressor level predicted to cause VEAs to be at the level of the biocriterion (Wang and
Stevenson 2002, Stevenson et al. accepted). An alternative approach is to observe
thresholds in VEA response along the stressor gradient. The TP criterion based on
stressor-response relations and the biological criterion (van Dam trophic index) was 46
ug/L. This value was similar to the TP levels below the threshold response in TP with
increasing agricultural land use. These complementary lines of evidence provide sound

conceptual foundations for deﬁning nutrient criterion.

150

The ultimate goal of assessing and managing aquatic ecosystems is to restore these
systems to natural conditions. Thus, I ﬁrst should estimate the natural condition should be
estimated by direct assessment of reference condition with minimum human activity,
contamination, and alteration. Second, cause-effect relationship should be determined
between algal indicators and stressors by manipulated experimental approaches. Third,
stressor criteria could be determined by integrating results from physical and chemical
analyses, experimental and observational relation of stressor- response gradients, and bio-
criteria based on reference approach. A suite of VEAs, such as native taxa richness, shifts
in relative abundance, or sensitive taxa should be measured in relation to stressor
gradients. Future research should explore ecological response of algae to nutrient
concentrations in wetlands, and a nationwide database of individual diatom
autoecological indices should be established. An understanding of benthic algal ecology
is a prerequisite to implementing a biological approach to ecosystem management using

algae.

151

5.6 References

American Public Health Association (APHA). 1998. Standard Methods for the Examination
of water and wastewater. 20th edition by LS. Clececeri, A.E. Greenberg, and AD. Eaton.
Washington D. C., USA.Anderson, D. J ., and B. Vondracek. 1999. Insects as

indicators of land use in three ecoregions in the prairie pothole region. Wetlands 19:
648-664.

Anderson, D. J. and B. Vondracek. 1999. Insects as indicators of land use in three
ecoregions in the prairie pothole region. Wetlands 19: 648-664.

Berven, K. A., and T. A. Grudzien. 1990. Dispersal in the Wood Frog (Rana-Sylvatica) -
Implications for Genetic Population-Structure. Evolution 44:2047—2056.

Byron, ER, and C. R. Goldman, 1989. Land-use and water quality in tributary streams
of lake tahoe. Califomia-Nevada. Journal of Environmental Quality 18:84-88.

Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N. Sharpley, and V. H.
Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen.
Ecological Applications 8:559-568.

Carpenter, K. D., and I. R. Waite. 2000. Relations of habitat-speciﬁc algal assemblages to
land use and water chemistry in the Willamette Basin, Oregon. Environmental
Monitoring and Assessment 64: 247-257.

Charles, D. F. 1985. Relationships between surface sediment diatom assemblages and
Lake-water characteristics in Adirondack lakes. Ecology 66: 994-1011.

Cowarding, L.M., V. Carter, F. C. Golet, and E. T. LaROe. 1979. Classiﬁcation of
wetlands and deepwater habitats of the United States. FWS/OBS-79/31, US. Fish and
Wildlife Service, Washington, DC. 103 pp.

De'ath, G., and K. E. Fabricius. 2000. Classiﬁcation and regression trees: A powerful yet
simple technique for ecological data analysis. Ecology 81:3178-3192.

Detenbeck, N. E., S. M. Galatowitsch, J. Atkinson, and H. Ball. 1999. Evaluating
perturbations and developing restoration strategies for inland wetlands In the Great
Lakes Basin. Wetlands 19:789-820.

Detenbeck, N. E. 2001. Methods for evaluating wetland condition: Wetlands
classiﬁcation. EPA 822-R-01-007g. US. Environmental Protection Agency; Ofﬁce of
Water; Washington, DC.

Dixit, S.S., J. P. Smol, J. C. Kinston, and D. F. Charles. 1992. Diatoms: powerful

indicators of environmental change. Environmental Science and Technology 26:23-
33.

152

Dixit, A. S., S. S. Dixit, and J. P. Smol. 1996. Long-term water quality changes in
Ramsey Lake (Sudbury, Canada) as revealed through paleolimnology. Journal of
Environmental Science and Health Part a- Environmental Science and Engineering &
Toxic and Hazardous Substance Control 31:941-956.

Findlay, C. S., J. Lenton, and L. G. Zheng. 2001. Land-use correlates of anuran

community richness and composition in southeastern Ontario wetlands. Ecoscience
8:336-343.

Freeland, J. A., J. L. Richardson, and L. A. Foss. 1999. Soil indicators of agricultural
impacts on northern prairie wetlands: cottonwood lake research area, North Dakota,
USA. Wetlands 10: 56-64.

Gasse, F. 1987. Diatoms for reconstructing paleoenvironments and paleohydrology in
tropical serrri-arid zones: examples of some lakes in Niger since 1200bp. Hydrologia
154:127-163.

Gordon, S. I., and S. Majumder. 2000. Empirical stressor-response relationships for
prospective risk analysis. Environmental Toxicology and Chemistry 19: 1106-1112.

Hall, R.I., and J.P. Smol. 1992. A weighted averaging regression and calibration model
for inferring total phosphorus concentration from diatoms in British-Columbia
(Canada) lakes. Freshwater biology 27: 417-434.

Heaney, S. 1., R. H. Foy, G. J. A. Kennedy, W. W. Crozier, and W. C. K. O'Connor.
2001. Impacts of agriculture on aquatic systems: lessons learnt and new unknowns in
Northern Ireland. Marine and Freshwater Research 52:151-163.

Hill, B. H., A. T. Herlihy, P. R. Kaufmann, R. J. Stevenson, F. H. McCormick, and C. B.
Johnson. 2000. Use of periphyton assemblage data as an index of biotic integrity.
Journal of the North American Benthological Society 19:50-67.

Hill, B. H., A. T. Herlihy, P. R. Kaufmann, S.J. DeCelles, M. A. Vander Borgh. 2003.
Assessment of streams of the eastern United States using a periphyton index of biotic
integrity. Ecological Indicators 2: 325-338.

Hillebrand, H., and U. Sommer. 2000. Diversity of benthic microalgae in response to
colonization time and eutrophication. Aquatic Botany 67:221-236.

Hughes, R. M., C.B. Johnson, S. S. Dixit, A.T. Herlihy, P.R. Jaufmann, W.L.Kinney,
D.P.Larsen, P.A. Lewis, D.M. McMullen, A.K. Moors, R. J. O’Connor, S.G. Paulsen,
R.S. Stemberger, S.A. Thiele, T.R. Whittier, and D. L. Kugler. 1993. Development of
lake condition indicators for EMAP — l991n pilot. Pages 7-90 in D. P. Larsen and SJ.
Christie. (editors). EMAP- surface waters 1991 pilot report. EPA-620-R-93-003.
U.S.Environmental Protection Agency, Ofﬁce of Research and Development,
Corvallis, Oregon.

153

Hughes, R. M., S. G. Paulsen, and J. L. Stoddard. 2000. EMAP-Surface Waters: a
multiassemblage, probability survey of ecological integrity in the USA.
Hydrobiologia 422:429-443.

McCune, B., and J. Grace. 2002. Analysis of Ecological community. MJM Software
Design. Pp304. Gleneden Beach, Oregon.

Hunsaker, C. T., D. A. Levine, S.P. Tmins, B. L. Jackson, and R. V. O’Neill. 1992.
Landscape characterization for assessing regional water quality. Pages 997-1006 in D.
H. McKenzie. D. E. Hyatt, and V. K .McDonald. (editors). Ecological indicators.
Volume two. Elsevier Applied Science, New York, New York, USA.

Hustedt, F. 1930-1966. Die kieselalgen Deutschlands, Osterreichs und der Schweiz unter
Berucksichtigung der urtigen Lander Europas sowie der andrenzenden
Meeresgebiete, Pages920pp,845 pp and 816 pp. in L. Rabenhorst editor.
Kryptogamen-Flora von Deutschland, Osterreich und der Schweiz.Academische
Verlagsgesellschaft Leipzig.

Juggins, S., and C. J. F. ter Braak. 1992. CALIBRATE—a program for species-
environment calibration by weighted averaging partial least squares regression.
Environmental Change Research Center, University College, London.

Juttner, I., H. Rothfritz, and S. J. Ormerod. 1996. Diatoms as indicators of river quality in
the Nepalese Middle Hills with consideration of the effects of habitat-speciﬁc
sampling. Freshwater Biology 36:475-486.

Krammer, K., and H. Lange-Bertalot. 1986. Bacillariophyceae. 1. Teil:
Naviculaceae.Pages 1-876 in H. Ettl, Gerloff, J ., Heynig, H., & Mollenhauer, D.
editor. Sitsswasserﬂora von Mitteleuropa.Gustav Fisher Verlag, Germany.

Krammer, K., and H. Lange-Bertalot. 1991. Bacillariophyceae.VEB Gustav Fischer
Verlag Jena.

Krammer, K., and H. Lange-Bertalot. 1997. Bacillariophyceae.VEB Gustav Fisher
Verlag Jena.

Landis, W.G., and J .F. McLaughlin. 2000. Design criteria and derivation of indicators for
ecological position, derecetion, and risk. Environmental Toxicology and Chemistry:
19: 1059-1065.

Leland, H. V. 1995. Distribution of Phytobenthos in the Yakima River Basin,

Washington, in Relation to Geology, Land-Use, and Other Environmental-Factors.
Canadian Journal of Fisheries and Aquatic Sciences 52: 1108-1129.

154

Leland, H. V., and S. D. Porter 2000. Distribution of benthic algae in the upper Illinois
River basin in relation to geology and land use. Freshwater Biology 44: 279-301.

Lopez, R. D., and M. S. Fennessy. 2002. Testing the ﬂoristic quality assessment index as
an indicator of wetland condition. Ecological Applications 12:487-497.

McCormick, P. V., P. S. Rawlik, K. Lurding, E. P. Smith, and F. H. Sklar. 1996.
Periphyton-water quality relationships along a nutrient gradient in the northern Florida
Everglades. Journal of the North American Benthological Society 15:433-449.

McCune, B., and J. Mefford. 1997. PC-ORD. Multivariate Analysis Of Ecological Data.
Version 3.0 For Windows. Mjm Software Design, Gleneden Beach, OR, USA.

Mitsch, W. J ., and R. F. Wilson. 1996. Improving the success of wetland creation and
restoration with know-how, time, and self-design. Ecological Applications?6:77-83.

Osborne, L. L., and M. J. Wiley. 1988. Empirical relationships between land use/cover
and stream water quality in an agricultural watershed. Journal of Environmental
management 26:9-27.

Pan, Y. D., and R. J. Stevenson. 1996. Gradient analysis of diatom assemblages in
western Kentucky wetlands. Journal of Phycology 32: 222-232.

Pan, Y., R. J. Stevenson, P. Vaithiyanathan, J. Slate, and C. J. Richardson. 2000. Changes
in algal assemblages along observed and experimental phosphorus gradients in a
subtropical wetland, USA. Freshwater Biology 44:339-353.

Patrick, R., and C. W. Reimer. 1975. The Diatoms of the United States, exclusive of
Alaska and Hawaii. The Academy of Natural Sciences of Philadelphia, Philadelphia.

Richardson, C. J. 2000. Freshwater Wetlands. Pages 448-499 in MG. Barbour, and W.D.
Billings (Editors). North American Terrestrial Vegetation. Cambridge University
Press, New York.

Smol, J. P., B. F. Cumming, A. S. Dixit, and S. S. Dixit. 1998. Tracking recovery patterns
in acidiﬁed lakes: A paleolimnological perspective. Restoration Ecology 61318-326.

Somlyody, L. 1998. Eutrophication modeling, management and decision making: The
Kis-Balaton case. Water Science and Technology 37 : 165-175 .

Stevenson, R. J. 1998. Diatom indicators of stream and wetland stressors in a risk
management framework. Environmental Monitoring and Assessment 51: 107-118.

Stevenson, R. J ., R. P. Sweets, Pan Y., and R. E. Schultz. 1999. Diatom community

patterns in wetlands of Michigan (USA) and their use as indicators of ecological
conditions. Pages 517-527 in A. J. McComb, and J. A. Davis (editors). Proceedings of

155

INTECOL’s Vth International Wetland Conference. Gleneagles Press, Adelaaide,
Australia.

Stevenson, R. J ., P V. McCormick, and R. Frydenborg. 2001. Methods for evaluating
Wetland Condition: Using Algae to assess environmental conditions in wetlands.
EPA 843-B-00-002k. US. Environmental Protection Agency; Ofﬁce of Water,
Washington, DC.

Stevenson, R. J ., Y. Pan, and P. Vaithiyahathan. 2002. Ecological assessment and
indicator development in wetlands: the case of algae in the Everglades, USA. Verh.
Intemat. Verein. Limnol. 28: 1248-1252.

Stosch, H. A. von and B.E.F. Reimann. 1970. Subsilicea fragilarioide gen. Et spec. nov.,
eine Diatomee (Fragilariaceae) mit vorwiegend organischer membran. Nova
Hedwigia Beih. 3121-36.

ter Braak, C.J.F. 1990. Update Note: CANOCO v 3.1. Agricultural Mathmatics Group,
Wageningen, The Netherlands.

Thompson, J. R., and G. Polet. 2000. Hydrology and land use in a Sahelian ﬂoodplain
wetland. Wetlands 20: 639-659.

Tufford, D. L., H. N. McKellar, Jr., and J. R. Hussey.l998. In-stream Nonpoint Source
Nutrient Prediction with Land-use proximity and seasonality. Journal of
Environmental Quality 27:100-111.

United States Department of Agriculture, 1982. Michigan Data 1982 National Resource
Inventory.

United States Environment Protection Agency (USEPA). 2000. Principles for the
Ecological Restoration of Aquatic Resources. EPA841-F-00-003. Ofﬁce of Water
(4501F), United States Environmental Protection Agency, Washington, DC. 4 pp.

Van Dam, H., A. Mertens, and J. Sinkeldam. 1994. A coded checklist and ecological
indicator values of freshwater diatoms from the Netherlands. Netherland Journal of
Aquatic Ecology 28:117-133.

Van der Valk, A. 2001. Methods for evaluating wetland condition: land-use
characterization for nutrient and sediment risk assessment. EPA 822-R-01-007q. US.
Environmental Protection Agency; Ofﬁce of Water, Washington, DC.

Vitousek, P. M., J. D. Aber, R. W. Howarth, Ge.E. Liken, P. A. Matson, D. W. Schindler,

W. H. Schlesinger, and D. G. Tilman. 1997. Human alternation of the global nitrogen
cycle: sources and consequences. Ecological Applications 7 2737-750.

156

Wang, Y-K., and R. J. Stevenson. 2002. An ecological assessment framework for
watershed management. Journal of the Chinese Institute of Environmental
Engineering 12: 209-215.

Whitton, B. A., and M. G. Kelly. 1995. Use of algae and other plants for monitoring
rivers. Australia. Journal of Ecology 20:45-56.

Wilcox, D. A. 1995. Wetland and Aquatic Macrophytes as Indicators of Anthropogenic
Hydrologic Disturbance. Natural Areas Journal 15:240-248.

Wilson, M. V., and C.L.Mohler. 1983. Measuring compositional change along gradients.
Vegetatio 54:129-141.

157

Chapter 6
Conclusions

This dissertation develops a better understanding of relationship among a variety of
ecological factors within wetlands, evaluates a number of valuable ecological attributes to
deve10p indicators for wetland bioassessment and wetland restoration, and establishes
biological and environmental criteria to protect wetland ecosystems. Because of the
spatial and temporal variability within and among wetlands, ecological assessment in
wetlands has to include development of habitat or class speciﬁc attributes for effective
evaluation. My studies in Michigan wetlands indicated that the similarity of algal
assemblages among habitats within wetlands and between restored and extant wetlands
were related to a number of environmental characteristics. Local pairwise comparison
based on reference conditions could reduce regional variation among wetlands and
identify the successional pattern of ecological development after restoration with restored
wetland age. Algal attributes in both freshwater and salt marshes were closely related to
nutrients. Experimental determination indicated that algal biomass and algal assemblages
were directly related to nutrient concentrations and nutrient ratio, and a cause-effect
relationship was established between algal assemblages and nutrient gradients. Based on
the stressor-response relationship between nutrient stressors and algal attributes, and by
choosing paleoecological diatom assemblages as reference condition, I developed bio-
criteria and nutrient criteria for wetland protection.

I ﬁrst evaluated spatial and temporal variation of environmental and algal attributes
among habitats within wetlands and between restored and extant wetlands. Restored

wetlands differed from extant wetlands in a number of ways in Southern Michigan.

158

Physical and chemical variables, especially plant shading and nutrient concentrations,
were distinctly greater in extant wetlands and contributed to differences in algal
communities among habitats within wetlands and between restored and extant wetlands.
Whether algal assemblages in these two types of wetlands were similar or distinct
depended on different habitats. Epiphytic assemblages in the restored and extant wetlands
were relatively distinct, while neither sediment nor phytoplankton assemblages were as
different between restored and extant wetlands. On the other hand, non-diatom algal
assemblages from different habitats were mostly different, but sediment and epiphytic
diatom assemblages were not signiﬁcantly different. Ecological restoration is a long-term
process and does not occur in a short period of time. I found no direct evidence for
increasing similarity in algal assemblages between restored and extant wetlands with
increasing age of restored wetlands. However, response of algal assemblages to nutrient
concentration and light, two factors distinguishing restored and extant wetlands, indicated
indirectly that succession processes in restored wetlands were regulating algal
assemblages and restored wetlands < 10 years old may need more time to develop mature
qualities.

In the next chapter, I evaluated a suite of algal attributes of restored salt marshes
based on paired reference and restored marsh approach and developed indicators for
wetland development. I assessed algal biomass (dry mass (DM), ash free dry mass
(AFDM), chl a content, algal biovolume), algal species composition and diversity, and
gross primary production on both Spartina alterm'ﬂora and sediments (sediment algae).
Controlling for regional variation in reference marshes substantially increased precision

in relations between attributes and the increase in age of restored marshes. The organic

159

matter restoration ratio of sediments increased with age of restored marshes in both
spring and summer. The algal biomass restoration ratios of epiphytes, calculated with
algal biovolume and chl a, increased with restored marsh age in summer but not during
spring. Biomass of sediment algae was not related to marsh age. The similarity of diatom
species composition between paired restored and reference sites increased with age of
restored marshes during spring and summer. Primary production by epiphytic and
sediment algae in summer showed site-speciﬁc changes and did not change consistently
with marsh age. Algal biomass, algal diversity, and diatom species assemblages during
summer were positively correlated with sediment nitrogen and phosphorus concentration.
I concluded that overall structural and functional development of restored wetlands,
especially nutrient storage in sediments, regulated algal community structure and
function, which can be used to evaluate marsh restoration.

I then determined the limiting nutrient in wetlands and established the cause-effect
relationship between nutrient and algal assemblages because both the previous two
chapters indicated that nutrients were one of the most important factors regulating
periphyton growth and species composition in wetlands. Algal bioassays using nutrient
diffusing substrata indicated that nitrogen (N) was the main limiting nutrient for
periphyton growth in restored wetlands in southern Michigan. Algal N limitation
positively correlated with phosphorus (P) concentration and negatively correlated with
N:P ratios in wetland water and sediment. Canonical correspondence analysis (CCA)
indicated that diatom species composition in natural habitats and control artiﬁcial habitats
changed mainly along a pH gradient and secondarily along nutrient gradients indicated by

total phosphorus (TP) and dissolved inorganic nitrogen (DIN). Individual diatom taxa

160

responded to nutrient addition differently. Overall, the responses of indicator species in
nutrient bioassays were positively correlated with natural nutrient gradients among
natural wetlands indicating a good match of diatom indicators between natural and
experimental systems.

Chapter 5 illustrated how the stressor-response relationship and reference approach
could be used to develop stressor and biological criteria to protect wetland health.
Diatom-based biological indicators were used to establish biological and stressor criteria
for wetland protection and restoration and to indicate wetland ecosystem health.
Paleoecological diatom assemblages from pre-settlement sediments were used to
represent natural, reference conditions in wetlands. Stressor-response relations were used
to delineate stressor criteria and human disturbance. Canonical correspondence analysis
(CCA) revealed that diatom species assemblages responded mainly to conductivity and
total phosphorus (TP). Similarity in species composition to reference condition and
number of native species in current wetlands decreased with increase of nutrient
concentrations in these wetlands. van Dam's diatom trophic index was imported to
characterize wetland trophic status in North American wetlands and was linearly
correlated with the TP gradient. Both the weighted average TP model and van Dam's
trophic index positively responded to agricultural land use around wetlands. These
indices deviated from the "normal" natural condition when TP exceeded 46 ug/L. A
threshold change in TP and biotic indices was detected when agricultural land exceeded
57.9 % total land use.

In summary, most extant wetlands have suffered a certain degree of nutrient impact

from human disturbance. Wetland restoration has signiﬁcantly decreased nutrient

161

concentrations and compensated for wetland loss. However, wetland restoration is a long-
term process and it takes a long time for different attributes to develop to the status of
reference condition. Continuous monitoring and assessment would lead to better
understanding of ecosystem processes. My studies developed algal assemblages as
indicators for diagnosing and quantifying nutrient pollution in wetlands and they
identiﬁed ecosystems at risk due to excess nutrients in the environment. Based on the
establishment of cause-effect relationship and stressor-response relationship between
nutrient and algal attributes, ecological thresholds, algal biological criteria, and stressor
criteria were developed for watershed management. The results of this research could be
used to support ecologically sensitive decision-making in land planning and wetland
restoration by providing targets for decreasing anthropogenic pollutions and thereby

increasing the rate of successful restoration projects.

162

Appendices

163

Appendix A.

Wetland sites sampled in Southern Michigan. Land use around wetland 100m radius were

deﬁned based on observation. RES: resident; AG: agricultural; CIT: constantly

inﬂuenced by road; UND: undisturbed.

 

 

Site Code Latitude Longitude Land County Sample Artificial Age Site
code use Date substrata of property
Res.
1 Dansla 42.56604 84.3968 RES INGHAM 8-18-2000 YES 6 Res.
2 JoIIy18 42.68161 84.4154 RES INGHAM 8-24-2000 YES 11 Res.
3 Jolly19 42.682 84.4182 RES INGHAM 8-24-2000 NO N/A Ext.
4 Green20 42.90778 84.5202 RES CLINTON 8-23-2000 NO N/A Ext.
5 Green25 42.81349 84.4527 RES CLINTON 8-23-2000 NO 8 Res.
6 Herbison2 42.8341 84.3962 AG CLINTON 8-1-2000 NO 7 Res.
7 27N 42.931 84.515 CIT CLINTON 8-23-2000 NO 8 Res.
8 278 42.9262 84.5159 CIT CLINTON 8-23-2000 NO 8 Res.
9 DansSl 42.55581 84.3642 RES INGHAM 8-18-2000 YES 6 Res.
10 Airport542 42.7905 84.6067 RES CLINTON 8-21-2000 YES 2 Res.
11 Alrport539 42.79212 84.6027 RES CLINTON 8—21-2000 NO N/A Ext.
12 Airport615 42.79212 84.6026 RES CLINTON 8-21-2000 NO N/A Ext.
13 Herbison6 42.8342 84.4007 AG CLINTON 8—1-2000 NO N/A Ext.
14 Alward 42.89237 84.4581 RES CLINTON 8-25-2000 NO N/A Ext.
15 Barnes 42.52287 84.3594 AG INGHAM 8-16-2000 NO N/A Ext.
16 CampusB 42.72074 84.4776 CIT INGHAM 8-25-2000 YES 9 Res.
17 CampusS 42.72079 84.4766 CIT INGHAM 8-25-2000 YES 11 Res.
18 Centerline 42.90436 84.4738 RES CLINTON 8-25-2000 NO N/A Ext.
19 Centeer 42.82354 84.4178 UND CLINTON 9-1-2000 NO N/A Ext.
20 CentraIPk 42.72789 84.4218 CIT INGHAM 8-22-2000 YES 2 Res.
21 Clark55 42.81374 84.4528 RES CLINTON 8-25-2000 NO N/A Ext.
22 CIark81 42.81422 84.4013 RES CLINTON 8-25-2000 NO N/A Ext.
23 Close 42.7564 84.5242 CIT INGHAM 8-21-2000 YES 11 Res.
24 College 42.63054 84.4852 CIT INGHAM 8-24-2000 YES 5 Res.
25 Cutler 42.8625 84.6893 UND CLINTON 8-19-2000 NO N/A Ext.
26 FarmL 42.70518 84.4804 CIT INGHAM 8-28-2000 YES 9 Res.
27 Forest 42.8621 84.6759 RES CLINTON 8-19-2000 NO N/A Ext.
28 Hawley 42.5187 84.3824 UND INGHAM 8-16-2000 NO N/A Ext.
29 Howe 42.84207 84.4137 RES CLINTON 8-19-2000 NO N/A Ext.
30 Krepps 42.9124 84.5073 AG CLINTON 8—23-2000 NO N/A Ext.
31 Loren 42.5207 84.3541 AG INGHAM 8-16—2000 YES 10 Res.
32 Marsh 42.7741 84.4124 CIT CLINTON 9-1-2000 YES 5 Res.
33 MasonPk 42.57729 84.4316 RES INGHAM 8-18-2000 NO N/A Ext.
34 Meijer 42.5674 84.3443 CIT INGHAM 8-18-2000 YES 12 Res.
35 Open 42.7564 84.5242 CIT INGHAM 8-21-2000 YES 7 Res.
36 Orchart 42.5838 84.4587 UND INGHAM 8-18-2000 YES 3 Res.
37 Pratt 42.90436 84.524 AG CLINTON 8-23-2000 YES 10 Res.
38 Price 42.9286 84.4418 RES CLINTON 8-25-2000 YES 5 Res.
39 Railway 42.72044 84.4788 CIT INGHAM 9-1-2000 NO N/A Ext.
40 Rosendom 42.77521 84.405 RES Clinton 9-1-2000 YES 2 Res.
41 Stall 42.79659 84.6052 RES CLINTON 8-21-2000 NO N/A Ext.
42 Tom 42.8543 84.7094 RES CLINTON 8—19—2000 YES 3 Res.
43 Upton 42.81598 84.4024 RES CLINTON 8-1-2000 NO N/A Ext.
44 VaughnE 42.5003 84.363 AG INGHAM 8-16-2000 YES 2 Res.

 

164

 

45 VaughnW 42.5003 84.363 AG INGHAM 8-16-2000 YES 2 Res.

Appendix B.
EMAP sampling location and their site property. Michigan subecoregion were classiﬁed
into 6 types. Soil type are classiﬁed as: 1, Northern Michigan and Wisconsin Sandy Drift.
2, Western Michigan and Northeastern Wisconsin Fruit belt; 3, Southern Michigan and
Northern Indiana Drift Plain; 4, Indiana and Ohio Till Plain; 5, Southern Michigan Fruit
and Truck Belt. Agricultural land use around wetlands are based on 1 km radius around

 

 

Site Code latitude longitude Ecoregion Soil Region 96 AGR
MIWET03 44.39583 85.0242 2 1 0.166347
MIWETO4 44.43513 84.91 15 2 2 0
MIWE105 44.625 85.0552 2 1 0
MlWETl 0 45.36889 84.4081 1 1 0.049657
MIWETT 2 44.72083 84.4283 2 3 0.810269
MlWETl 8 43.26359 85.2996 3 3 0.776972
MIWET20 42.6521 7 85.5165 4 3 0
MIWET25 42.31386 84.1 737 5 4 0.842661
MIWET26 42.36821 84.2022 5 3 0.498948
MIWET31 42.57201 83.5506 5 4 0.518641
MIWET33 42.53601 83.6563 5 4 0.568018
MIWET35 42.33016 84.1287 5 3 0.646123
MIWET36 42.44701 83.9982 5 4 0.53634
MIWET39 42.31658 85.3603 4 3 0.276599
MIWET41 42.25543 85.3603 4 3 0.330213
MIWET42 41 .92935 85.8401 4 3 0.256192
MIWET43 41 .85054 85.75 4 3 0.383817
MIWET48 42.4021 7 85.3971 4 5 0.284899
MIWET49 41 .91848 85.7813 4 3 0.162268
MIWET50 42.14946 85.601 1 4 3 0.556
MIWET51 42.46603 85.3695 4 3 0.482928
MIWET52 42.49796 85.2206 4 3 0.605755
MIWET53 42.53533 86.0404 6 3 0.455534
MIWET54 42.64674 85.9228 6 5 0.56166
MIWET58 42.52446 86.0055 6 5 0.479886
MIWE159 42.54755 86.0037 6 5 0.338
MIWET60 42.46603 85.2243 4 3 0.144218
MIWET62 42.3913 84.977 4 3 0.671064
MIWET68 42.45788 83.9982 5 4 0.203087
MIWET69 42.45245 84.0055 5 4 0.002193
MIWET70 42.44973 84.0147 5 4 0.096715
MIWET71 43.77446 85.2132 2 3 0.224564
MIWET72 44.18278 84.8183 2 3 0
MIWET77 45.22147 84.3824 1 3 0.051222
MIWET78 45.55707 84.6531 1 3 0
MIWE180 45.71603 84.6725 1 2 0.519926

 

165