 

 

 

n1 ,

a.

1. ;

 

..

10.?
1!... .
‘ If .t...

. . I
www.mrmﬂs... g...

:3 _‘ hulk?!

4:...

l

a.

£4».qu
A.

A9.

. 4: I
I . ROAM-W A
\tﬁ
. .. v.11-

mums

3%....

$5. p . s.

35.2%”?!

93d“: #1".
aII:\

0.3.

:4. 1.» Y
v0.3 .1 It

s

.vn.
312‘!

.171;

It ’n’
. 21.9....

2:!

. r":
‘49:... d i

.9! any)“.
..

r

i. ii .
.
3.: .nnWanr
9.51.1. {Ansﬁtnouit
{vulva .5 .. 5.2.
1|.Utx

:. a».

1.... J
Q: biymhﬂumrluiihé

uh.

M

l.

a».

9

.x
#4....

O.
.0...

.
t o :5 5.5.!
:39. I...

.1 .1}. Eli...‘ .
J4: . 04.

 

3-!
52
a)
.2
c:
:3

.93
CO
>25
CC
c:
Em
[11.9
3f;
2

 

This is to certify that the
dissertation entitled

THE USE OF MACROINVERTEBRATES AS INDICATORS
OF WETLAND QUALITY IN THE MUSKEGON RIVER
WATERSHED, MICHIGAN

presented by

MOLLIE DAY MCINTOSH

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Entomology

 

ﬂi/yz/aﬁ’i/

Major Professor’s/Signature
/Z/»7/o ?

Date

MSU is an affirmative-action, equal-opportunity employer

.__._..-.—.----c--—.———.—.—-_.-

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DaleDue.indd—p.1

 

THE USE OF MACROINVERTEBRATES AS INDICATORS OF WETLAND
QUALITY IN THE MUSKEGON RIVER WATERSHED, MICHIGAN

By

Mollie Day McIntosh

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Entomology

2007

ABSTRACT

THE USE OF MACROINVERTEBRATES AS INDICATORS OF WETLAND
QUALITY IN THE MUSKEGON RIVER WATERSHED, MICHIGAN

By
Mollie D. McIntosh

Wetlands provide many functions, both ecological and economical;
however, they are continuously threatened by numerous human activities. These
activities (e.g., agriculture, development) subsequently generate environmental
stressors (e.g., nutrients, road salt) that can impact the physical, chemical and
biological integrity of wetland ecosystems. It is hypothesized that
macroinvertebrate communities will vary with the presence and/or intensity of an
environmental stressor; however, identifying these relationships can be difﬁcult
due to complex interactions among multiple human activities and stressors and
wetland type and size. The main purpose of this study was to assess and
characterize the macroinvertebrate community of inland wetlands of the
Muskegon River Watershed, Michigan, and to understand how these
communities respond to natural and human-induced changes in the environment,
Ultimately the information from this research can add to a growing database on
the use of macroinvertebrates as biological indicators of wetland ecosystem
health, the identiﬁcation of biological criteria used for water quality standards,
predicting change in wetland structure and function, and in the regulation and

protection of Michigan wetland ecosystems.

This dissertation is dedicated to my parents, Candace and Michael McIntosh,

who have always supported me throughout this journey.

iii

ACKNOWLEDGEMENTS

I would ﬁrst like to thank my advisor, Dr. Richard Merritt. Throughout my
project he provided endless support, encouragement, and new opportunities that
have made me a better scientist, person and future mentor. I must also thank Dr.
Vanessa Lougheed, who not only served on my committee, but also helped with
ﬁeld logistics, data management and analyses for this project; her involvement
made my experiences at MSU really fun and enjoyable. I am also grateful to all
my committee members, including Drs. R. Jan Stevenson, Tom Burton and Mike
Kaufman, who provided signiﬁcant advice and support throughout this study.

I would also like to thank the numerous individuals at MSU who played a
signiﬁcant role during my graduate years. Eric lBenbow and Ryan Klmbirauskas,
my good friends, provided invaluable support and encouragement; without them,
this journey would not have been so enjoyable. Todd White, Leia Watkins,
Jennifer Schmitz and Rebecca Kolar, who worked numerous, long hours to
process endless wetland samples for my project. Vanessa Lougheed, Chris
Parker and Sarah LeSage should be commended for their tireless efforts in the
ﬁeld and at happy hour. Gary Parsons should also be recognized for his help
and guidance in identifying insect specimens. Finally I would like to thank the
members of the Merritt Lab, including JoAnna Lessard, Kelly Wessell, Jaree
Johnson, Kristi Zurwaski, Osvaldo Hernandez, Ryan Kimbirauskas, Eric Benbow,
Christian Lesage, Todd White and Matt Wessner; for making the lab a wonderful

and always entertaining place to work.

My family and friends should also be recognized for supporting me during
my years at MSU. My parents, Candace McIntosh and Michael McIntosh, my
sisters, Melissa Mitiska and Marielle McIntosh, Diana McIntosh and David
Hulefeld, were there for me during all the good and stressful times and provided
continuous encouragement; without them I would never have made it to this far —

I am truly thankful and love them all very much.

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................... viii
LIST OF FIGURES ....................................................................................... xi
CHAPTER 1
INTRODUCTION TO THE USE OF MACROINVERTEBRATES AS
INDICATORS OF WATER QUALITY .......................................................... 1
CHAPTER 2
INFLUENCE OF SPATIAL SCALE ON WETLAND MACROINVERTEBRATE
COMMUNITY RESPONSE TO ENVIRONMENTAL STRESSORS ............. 10
Abstract ................................................................................................. 1 1
Introduction...............; ............................................................................ 12
Methods ................................................................................................. 15
Study Location and Scale ....................................................... 15
Sample Collection ................................................................... 17
Statistical Analyses ................................................................. 1 9
Results ................................................................................................... 22
Environmental Variables ......................................................... 22
Macroinvertebrates ................................................................. 24
Environmental-Macroinvertebrate Response .......................... 27
Discussion ............................................................................................. 30
Ecosystem Scale .................................................................... 30
Class Scale ............................................................................. 33
Habitat Scale .......................................................................... 38
Tables .................................................................................................... 41
Figures ................................................................................................... 50
CHAPTER 3
UTILITY OF PRE-EXISTING MACROINVERTEBRATE METRICS IN MRW
WETLANDS AT MULTIPLE SPATIAL SCALES .......................................... 59
Abstract ................................................................................................. 60
Introduction ............................................................................................ 61
Methods ................................................................................................. 64
Study Location and Scale ....................................................... 64
Sample Collection ................................................................... 67
Metric Calculation and Statistical Analyses ............................ 68
Results ................................................................................................... 70
Discussion ............................................................................................. 74
Conclusion ............................................................................................. 83
Tables .................................................................................................... 85
Figures ................................................................................................... 90

vi

CHAPTER 4
INFLUENCE OF WETLAND SIZE AND WATER QUALITY ON ISOLATED

MACROINVERTEBRATE COMMUNITIES .................................................. 92
Abstract ................................................................................................. 93
Introduction ............................................................................................ 94
Methods ................................................................................................. 97

Study Location and Scale ....................................................... 97
Sample Collection ................................................................... 98
Statistical Analyses ................................................................. 1 00
Results ................................................................................................... 102
Water Chemistry and Land-Use Land Cover .......................... 102
Macroinvertebrate Abundance ................................................ 103
Macroinvertebrate Diversity .................................................... 104
Macroinvertebrate Functional-Feeding Groups ...................... 105
Discussion ............................................................................................. 106
Conclusion ............................................................................................. 1 11
Tables .................................................................................................... 1 13
Figures ................................................................................................... 1 16

APPENDIX A:

LIST OF COLLECTED MACROINVERTEBRATE TAXA FROM MUSKEGON

RIVER WATERSHED INLAND MARSHES .................................................. 118

APPENDIX 8:

RECORD OF DEPOSITION OF VOUCHER SPECIMENS .......................... 122

LITERATURE CITED ................................................................................... 133

vii

LIST OF TABLES

CHAPTER 2

Table 2.1: Principal componentanalysis (PCA) of 9 water chemistry
variables at the ecosystem, class (lacustrine, palustrine, riverine)
and habitat scale (emergent (E), submergent/ﬂoating (SIF)) in
marsh wetlands. The percent variance explained by the ﬁrst two
principal components (PCs) of each PCA are listed for each scale.
Water chemistry variables signiﬁcantly correlated (p < 0.05) to each
PC axis are listed (** = p < 0.001, * = p < 0.01). Water chemistry
variables include pH, conductivity (cond), dissolved oxygen (DO),
turbidity, total phosphorus (TP), ammonia (NH3), soluble reactive
phosphorus (SRP), chloride (Cl), and mean depth (Depth) ............... 41

Table 2.2: Principal component analysis (PCA) of 6 land-use land-cover
(LULC) variables at the ecosystem scale, class scale (lacustrine,
palustrine, riverine) and habitat scale (emergent (E),
submergent/ﬂoating (SIF)) in marsh wetlands. The percent
variance explained by the ﬁrst two principal components (PCs) of
each PCA are listed for each scale. Land—use land-cover variables
signiﬁcantly correlated (p < 0.05) to each PC axis are listed (** = p
< 0.001, * = p < 0.01). Land-use land-cover variables were relative
proportions of urban, rangeland, agriculture, forest, water and
wetland .............................................................................................. 42

Table 2.3: Mean(SE) values for all environmental variables at the
ecosystem scale, class scale (lacustrine (L), palustrine (P), riverine
(R)) and habitat scale (emergent (E), submergent/ﬂoating (SIF) in
marsh wetlands. Water chemistry variables include, pH,
conductivity (uS/cm), dissolved oxygen (DO mg/L), turbidity, total
phosphorus (TP ug/L), nitrate (NH3 ug/L), soluble reactive
phosphorus (SRP ug/L), chloride (Cl mg/L), and mean depth (cm).
Land-use land-cover (LULC) variables were relative proportions of
urban, rangeland, agriculture, forest, water and wetland. ANOVA
results were presented for the class scale within each type of
habitat; signiﬁcant values (p < 0.05) and non-signiﬁcant (ns) values
are reported. Tukey HSD (p < 0.05) mean comparison results were
shown among classes in each scale; different letters represent
signiﬁcant difference between wetland class ..................................... 43

Table 2.4: Mean (SE) values for 12 dominant macroinvertebrate taxa

within each habitat (emergent (E), submergent/ﬂoating (SIF)) at the
ecosystem scale and class scale (lacustrine (L), palustrine (P),

viii

riverine (R)) in marsh wetlands. ANOVA results were presented

for the class scale within each type of habitat, p values are reported
with signiﬁcant values (p < 0.05, gray highlighted). Tukey HSD (p <
0.05) mean comparison results were shown among classes in each
scale; different letters represent a signiﬁcant difference between
wetland class ..................................................................................... 44

Table 2.5: Macroinvertebrate taxa unique to the class or habitat scale in
this study. Included are taxa class, order, family and generic level
information, functional feeding group and habitat classiﬁcation ......... 45

Table 2.6: Correspondence analysis (CA) of macroinvertebrate community
at the ecosystem scale, class scale (lacustrine, palustrine, riverine)
and habitat scale (emergent (E), submergent/ﬂoating (SIF)) in
marsh wetlands. The percent variance explained by the ﬁrst two
CA axes are listed for each scale. Macroinvertebrates signiﬁcantly
correlated (p < 0.05) to each CA axis are listed (** = p < 0.001, * =
p < 0.01) ............................................................................................ 46

Table 2.7: Pearson correlations (p < 0.05) between the ﬁrst two axes of
each correspondence analysis (CA 1 and CA 2) and the ﬁrst two
principal components (PC1 and P02) calculated from the principal
component analysis of water chemistry values and from land-use
land-cover (LULC) variables. Signiﬁcant correlations (p value) at
the ecosystem scale, class scale (lacustrine, palustrine, riverine)
and habitat scale (emergent (E), submergent/ﬂoating (SIF)) in
marsh wetlands ................................................................................. 47

Table 2.8: Signiﬁcant relationships (correlations p < 0.05) between the
principal component analysis (PCs) scores and correspondence
analysis (CA) scores and the direction of each response (positive or
negative) in each wetland class and habitat. The main chemical
variables correlated to each PC axis (and the associated direction)
are listed and macroinvertebrate taxa that respond to these
changes in the PC are listed, taxa have either a positive (+) or
negative (-) response ......................................................................... 48

Table 2.9: Canonical Correspondence analysis (CCA) of
macroinvertebrate community at the ecosystem scale, class scale
(lacustrine, palustrine, riverine) and habitat scale (emergent (E),
submergent/floating (SIF)) in marsh wetlands. The CCA was
constrained by water chemistry variables. The percent variance
explained by the ﬁrst two CCA and CA axes are listed for each
scale .................................................................................................. 49

ix

CHAPTER 3

Table 3.1: Various macroinvertebrate metrics used in the assessment of
wetlands. The citation column lists studies that have utilized the
metric. Each number represents the following citation: (1) Burton et
al. 1999, (2) Helgen 2001, (3) Helgen and Gernes 2002), (4) Merritt
et al. 2002 and 2006, (5) Apfelbeck 2001, (6) Uzarski et al. 2006
and (7) Kashian and Burton 2000 ...................................................... 85

Table 3.2: Ecosystem parameters and corresponding functional ratio
ecosystem surrogates and expected ratios. (Modiﬁed from
Cummins and Merritt, 1999 and Merritt et al. 2002) .......................... 87

Table 3.3: Macroinvertebrate metrics signiﬁcantly different between
reference and impacted wetlands of various class (lacustrine,
palustrine, and riverine) and habitat (emergent, submergent).

Median values are listed with Vlﬁlcoxon test scores and signiﬁcance
values of either p < 0.1 (*) or p < 0.05 (**). Those values left blank
were non-signiﬁcant according to these tests .................................... 88

CHAPTER 4

Table 4.1: Ecosystem parameters and corresponding functional ratio
ecosystem surrogates and expected ratios utilized in this study.
(Modiﬁed from Cummins and Merritt, 1999 and Merritt et al. 2002) .. 113

Table 4.2: Mean (SE) Shannon Index, Simpson Index, and Evenness
Index values for each wetland treatment (impacted large, impacted
small, reference large and reference small) for total wetland (T),
emergent (E) and submergent (S) habitats. The only signiﬁcant
result (p < 0.05, *) was an interaction of wetland quality and size
observed in the Evenness Index of the submergent habitat .............. 114

Table 4.3: Functional feeding group ratios representing ecosystem
surrogates. Mean(SE) values for each wetland treatment
(impacted large, impacted small, reference large and reference
small) for total wetland (T), emergent (E) and submergent (S)
habitats. The only signiﬁcant effect (p < 0.05, *) of wetland size
was observed in the top down ratio of the emergent habitat ............. 115

APPEDNIX A

Table A1: List of macroinvertebrate taxon identiﬁed from inland marsh
wetlands in the Muskegon River Watershed, Michigan ..................... 118

LIST OF FIGURES

CHAPTER 2

Figure 2.1: Schematic illustration of three spatial scales (ecosystem, class
and habitat) and levels within each scale that were analyzed in this
study. The class scale is composed of three groups: lacustrine,
palustrine, and riverine wetlands. The habitat scale is composed of
two groups: emergent and submergent/ﬂoating habitats ................... 50

Figure 2.2: Principal component analysis of nine water chemistry
variables and six land-use land-cover variables at the ecosystem
level (n = 50 emergent sites). Each site is denoted by an open
circle. (A) Water chemistry variables include pH, conductivity
(Cond), dissolved oxygen (DO), turbidity (Turb), total phosphorus
(TP), ammonia (NH3), soluble reactive phosphorus (SRP), chloride
(Cl) and mean depth (Depth). (B) Land-use land-cover variables
include relative proportions of urban, rangeland (Range),
agriculture (Agricul), forest, water and wetland ................................. 51

Figure 2.3: Principal component analysis of (A) nine water chemistry
variables and (8) six land-use land-cover variables at the
ecosystem level (n = 50 emergent sites), with each site denoted by
a letter representing the marsh class. (A) Water chemistry
variables include pH, conductivity (Cond), dissolved oxygen (DO),
turbidity (Turb), total phosphorus (TP), ammonia (NH3), soluble
reactive phosphorus (SRP), chloride (Cl) and mean depth (Depth).
(B) Land-use land-cover variables include relative proportions of
urban, rangeland (Range), agriculture (Agricul), forest, water and
wetland .............................................................................................. 53

Figure 2.4: Mean(SE) values for (A) macroinvertebrate abundance and
(B) taxa richness for each habitat (emergent and
submergent/ﬂoating) from marsh wetlands at the ecosystem and
class scales (lacustrine, palustrine, and riverine). Signiﬁcant
differences in habitat (p < 0.05, denoted by *) were calculated by
ANOVA at the ecosystem and each class scale. Tukey HSD mean
comparison results were shown among classes at each habitat;
different letters represent a signiﬁcant difference (p < 0.05) between
wetland class ..................................................................................... 56

Figure 2.5: Correspondence analysis of 57 macroinvertebrate taxa at the

eCOSystem level in the emergent habitat (n = 50 sites), Each site is
denoted by a letter representing the marsh class. Sites separate

xi

into a palustrine dominated group (long- dashed circle) and
lacustrine and riverine dominated group (small-dashed circle) .......... 57

Figure 2.6: Canonical correspondence analysis of 57 macroinvertebrate
taxa from 50 sites in the emergent habitat at the ecosystem level.
Each site is denoted by aletter representing the marsh class. Sites
separate into a palustrine dominated group (long- dashed circle)
and lacustrine and riverine dominated group (small-dashed circle).
The graph has been constrained by water chemistry variables that
include pH, conductivity (Cond), dissolved oxygen (DO), turbidity
(Turb), total phosphorus (TP), ammonia (NH3), soluble reactive
phosphorus (SRP), chloride (Cl) and mean depth (Depth) ................ 58

CHAPTER 3

Figure 3.1: The relative composition (°/o) Gastropoda in emergent habitats
of reference (R) and impacted (I) wetlands at the class scale
(lacustrine, palustrine, and riverine). Solid lines within the box
indicate are median values, the box values represent the inter-
quartile ranges (25% and 75%), and the range bars indicate the
maximum and minimum values. Signiﬁcant differences (p < 0.05,
denoted by ** or p < 0.1, denoted by *) were calculated by non-
parametric ercoxon tests ................................................................. 90

Figure 3.2: A comparison of functional or ecosystem attribute ratios
between reference (R) and impacted (I) emergent habitats of
palustrine wetlands in the MRW. Solid lines within the box indicate
are median values, the box values represent the inter-quartile
ranges (25% and 75%), and the range bars indicate the maximum
and minimum values. Signiﬁcant differences (p < 0.05, denoted by
** or p < 0.1, denoted by *) were calculated by non-parametric
Wilcoxon tests ................................................................................... 91

CHAPTER 4

Figure 4.1: Mean (SE) macroinvertebrate abundance (individuals per
sample) for (A) total wetland, (B) emergent and (C) submergent
habitats in isolated palustrine wetlands. All four wetland treatments
are represented: small reference, large reference, small impacted
and large impacted. A signiﬁcant effect (*) of wetland quality (p <
0.05) and an interaction effect (p < 0.01) were observed in only the
submergent habitat (C) ...................................................................... 1 16

xii

Figure 4.2: Mean (SE) taxa richness (taxa per sample) for (A) total
wetland, (B) emergent and (C) submergent habitats in isolated
palustrine wetlands. All four wetland treatments are represented:
small reference, large reference, small impacted and large
impacted. No signiﬁcant effects of wetland quality, wetland size or
an interaction effect were observed in this study ............................... 117

xiii

Chapter 1: Introduction to the Use of Macroinvertebrates as Indicators of

Water Quality

Chapter 1: Introduction to the Use of Macroinvertebrates as Indicators of

Water Quality

The Clean Water Act (CWA) was established to restore and maintain the
chemical, physical and biological integrity of water bodies in the United States
(Adamus 1996). In the past, the condition of a water body was determined
mainly through physical and chemical measurements. These data were then
compared to previous data from that site or a reference location, and subsequent
management action was taken to improve the water body if necessary. Although
many of these physical and chemical criteria can be easily measured, when
collected in a rapid assessment (e.g., once a year) they provide only limited
information on the current conditions of the water body (Danielson 1998).
Typically when human activities degrade an aquatic ecosystem, effects can be
observed through a change in the biota (Danielson 1998). Biological organisms,
and thus biological measurements, can respond to short and long-term changes
in the environment, providing temporal information useful in evaluating the
integrity of a water body in a rapid assessment (Rosenberg and Resh 1993).
The biological community present at the time of assessment is the cumulative
result of the physical and chemical conditions over time. Organisms that
respond to these changes in a predictable manner can act as biological
indicators and are used to establish biological thresholds or criteria, and are
monitored in rapid assessments. As a result, the EPA and numerous states have

developedbiological criteria useful in assessing the condition of aquatic habitats,

hence the term bioassessment. The use of all three criteria, physical, chemical
and biological combined, provide for the most thorough assessments (Apfelbeck
2001).

Biological assessments have successfully been created for stream, lake,
and river systems; but few states have developed biological criteria and
assessments for wetland ecosystems (Danielson 1998). Compared to these
other aquatic habitats, less information is known about wetland biological
communities. The presence/absence of biota, the interactions of these
organisms with the biotic and abiotic environment, and individual organism traits
(e.g., life histories, tolerance levels) are still unknown in wetlands and can make
the identiﬁcation of good biological indicators difﬁcult. Natural temporal variability
common to the chemical and physical wetland environment can also inﬂuence
the biological community in both the short and long term; and the existence of
numerous wetland types (e.g., coastal marshes, fens, swamps) make the
establishment of standard biological criteria difﬁcult for wetlands (Rader 2001 ).
Despite the complexity, effort is still being given to develop biological criteria that
can aid in the regulation and protection of wetland quality.

Following the same trend, most aquatic research has focused on lake or
stream habitats; however, within the past twenty-ﬁve years an increasing amount
of research has been conducted on wetland ecosystems. Wetlands provide both
economic and social beneﬁts (e.g., ﬂood prevention, groundwater recharge,
water puriﬁcation) and ecological signiﬁcance (e.g., species diversity, unique

wildlife habitats) (Wrssinger 1999, Mitsch and Gosselink 2000). However, over

the past 200 years, approximately 50% of wetlands within the continental United
States have been lost or converted to other types of land uses (e.g., agriculture,
development) (Mitsch and Gosselink 2000). In Michigan, where this study was
conducted, 50% of wetlands have also been lost, with the threat rising due to
increased development (Mitsch and Gosselink 2000). Of those remaining
wetlands, many are still endangered due to increases in human activity and lack
of government protection. Increased human activity could lead to a decline in
water quality from nutrient enrichment, sedimentation, and other types of
pollution (Lemly and King 2000). A recent ruling in 2001, by the United State
Supreme Court (Solid Waste Agency of Northern Cook County [SWANCC] vs.
US. Army Corp of Engineers) has reduced the amount of protection for some
wetlands. The court ruled that isolated waters were no longer, under certain
circumstances, under the protection of the CWA; this decision makes isolated
waters, especially isolated depression wetlands, potentially vulnerable to human
disturbance (Tiner 2003a). Wetland managers, researchers, government
ofﬁcials and the general public have recognized this need to protect wetlands,
and have become more active as a result.

Biological assessments have been developed using ﬁsh,
macroinvertebrates, zooplankton, plant and algal communities for assessing
water quality; however, macroinvertebrates are the most commonly used group
(Hellawell 1986, Rosenberg and Resh 1993). Macroinvertebrates play a
fundamental role in the biological community of aquatic ecosystems. They

facilitate the transport of energy by connecting primary producers and higher

trophic levels in fobd webs. Therefore, shifts in macroinvertebrate structure
could have cascading effects on other levels of the food web, and in return, shifts
at other levels could have a direct inﬂuence on macroinvertebrate communities
(Wrssinger 1999, Helgen 2001). Because of this connection, macroinvertebrates
have been useful as management tools in the evaluation and monitoring of
aquatic habitats. Additional reasons for using macroinvertebrates are (1) their
ubiquitous nature, (2) range of response to environmental stressors by different
species, (3) relative stationary nature, (4) their relatively long life cycles, (5) the
ease in which macroinvertebrates are collected and the (6) relatively low costs
involved with collections (Rosenberg and Resh 1993).

All of these advantages apply to most aquatic habitats (e.g., streams and
rivers); however, some do not yet apply to wetland ecosystems. For example,
Batzer et al. (2001) maintained that many wetland macroinvertebrates are
mobile, some life cycles can be quite short (1-2 months), and that much
information is still unknown for certain taxa (e.g., response to pollution and
taxonomy), thereby, limiting the use of macroinvertebrates in wetland
assessments. For example, a study by King and Richardson (2002) found that
family-level identiﬁcation of wetland macroinvertebrate communities were not
capable of detecting impairment and suggested that genus or species-level
identiﬁcation be utilized, especially within the dominant Chironomidae family.
However, this type of identiﬁcation, especially with the Chironomidae or Annelida
can be difﬁcult, require expertise, time consuming and more expensive (King and

Richardson 2002), and for some taxa in some regions identiﬁcation keys may not

be available. In general, more research is needed on the general biology and
ecology of macroinvertebrate communities in various types of wetlands over
different spatial and temporal scales for us to better understand their use as
assessment tools.

Several studies, however, have successfully used macroinvertebrates in
the assessment of wetland quality. Burton et al. (1999) developed a preliminary
index of biotic integrity for coastal Great Lake wetlands using 14 invertebrate
metrics. Additional evidence supporting the use of certain invertebrate indicators
within costal freshwater wetlands were identiﬁed by Kashian and Burton (2000)
and later validated by Uzarski et al. (2004). Cooper et al. (2006) found certain
macroinvertebrate taxa to respond to land-use and water quality parameters in a
drowned river mouth wetland on Lake Michigan, supporting the use of
macroinvertebrates as indicators of human disturbance. In forested bottomland
wetlands, macroinvertebrate communities were also found to be sensitive to
human disturbance (e.g., highway roadways) and potentially useful in biological
assessment (King et al. 2000). Helgen and Gernes (2001, 2002) developed two
indices of biological integrity for vegetation and invertebrates to be used in
monitoring the water quality of depressional wetlands in Minnesota, and in
Montana, Apfelbeck (2001) found macroinvertebrates and diatoms to be useful in
evaluating the biological integrity of perennial wetlands. In Australia, Chessman
et al. (2002) successfully derived and tested a new biotic index for the Swan
Coastal Plain using macroinvertebrate taxa. Due to taxonomic limitations,

functional-group analyses can also provide insight to the assessment of aquatic

ecosystems (Cummins 1974, Cummins and Merritt 2001). Merritt et al. (1996,
1999, 2002b) found that macroinvertebrate metrics, including functional group
analyses, could be used as surrogates for ecosystem attributes in river oxbows
and ﬂoodplain ecosystems in Florida. However, other studies have found the
use of macroinvertebrates as biological indicators to be questionable in wetlands.
Studies in the Prairie Pothole Region, have found weak relationships between
macroinvertebrate communities and surrounding land use, suggesting that
macroinvertebrate indicators might not be useful in this region (Tangen et al.
2003). Also, Vlﬁlcox et al. (2002) found that biological indicators
(macroinvertebrates in addition to ﬁsh and plants) were inaccurate due to
ﬂuctuations in hydrology in drowned-river mouth wetlands. In general, all of
these studies suggest that macroinvertebrates might be useful in developing
biological criteria and bioassessment protocols, however, again more temporal
and spatial research is needed in wetlands of various regional, class and water-
level histories (Wilcox et al. 2002).

Common trends in wetland ecology have long been desired; however, due
to the large amount of variation that can be found in wetlands (e.g., hydrologic
condition, type of vegetation, location) these general trends have been difficult to
uncover and separating the natural variation from the human-induced variation
makes this task even more complex (Rader 2001). As a result, the
Environmental Protection Agency (EPA) has been working on conceptual models
to illustrate the relationship between biological condition (e.g., macroinvertebrate

attributes) and human disturbance in wetlands; thus, providing a framework to

help states fulﬁll the biological requirements of the CWA by supporting aquatic
life in wetlands. However, more research is needed to test this model by
assessing actual relationships between biota (e.g., macroinvertebrate
communities) and human disturbance (stressors) in the ﬁeld. Ultimately, if
useful relationships are found and tested, these guidelines could be a useful tool
in the development of macroinvertebrate/wetland assessments for many states.
This model will also provide consistency across states, scientiﬁcally defensible
benchmarks, a common framework for evaluation, and protection of wetland
habitats.

To understand if relationships do exist between the wetland
macroinvertebrate community and environmental stressors, more detailed
investigations, at various spatial and temporal scales, are needed. The main
purpose of this study was to assess and characterize the macroinvertebrate
community of inland wetlands of the Muskegon River Watershed (MRW),
Michigan, USA. To identify biological indicators, potential relationships were
examined at several spatial scales between MRW environmental stressors and
macroinvertebrate attributes (response indicators). i also investigated the
variability of the macroinvertebrate community within and between wetlands and
wetland habitats. This study has provided a better understanding of the
macroinvertebrate community present in inland Michigan wetlands and how
these communities respond to changes in the environment- information that can
then be utilized in the development of wetland biocriteria for the MRW,

assessment of MRW wetland integrity, predicting change in MRW wetland

structure and function, and in the regulation and protection of MRW wetland
ecosystems. The speciﬁc objectives of this study were divided into the following
three chapters:
1.) To determine macroinvertebrate communities and stressor response
relationships at multiple spatial scales
2.) To test the utility of pre-existing macroinvertebrate assessment metrics
in MRW wetlands
3.) To assess stressor impact and wetland size on macroinvertebrate
community structure and function in palustrine wetlands.
Ultimately the information from this research can add to a growing database on
the use of macroinvertebrates as biological indicators of wetland ecosystem
health and integrity, the identiﬁcation of biological criteria used for water quality

standards in Michigan, and provide a model for other states.

Chapter 2: Inﬂuence of Spatial Scale on Wetland Macroinvertebrate

Community Response to Environmental Stressors

10

Chapter 2: Inﬂuence of Spatial Scale on Wetland Macroinvertebrate

Community Response to Environmental Stressors

Abstract

Wetlands provide many functions, both ecological and economical;
however, they are continuously threatened by numerous human activities. These
activities (e.g., agriculture, development) subsequently generate environmental
stressors (e.g., nutrients, road salt) that can impact the physical, chemical and
biological integrity of wetland ecosystems. It is hypothesized that
macroinvertebrate communities will vary with the presence and/or intensity of an
environmental stressor; however, identifying these relationships can be difﬁcult
due to complex interactions among multiple human activities and stressors and
wetland type and size. In an effort to understand this complexity, we identiﬁed
relationships between macroinvertebrate communities and environmental
stressors at three spatial scales: the ecosystem (marsh wetlands), class (riverine,
palustrine, and lacustrine) and habitat (emergent and submergent vegetation).
During the summer of 2002 and 2003, macroinvertebrates were collected from
57 inland marshes within the Muskegon River Watershed (Michigan, USA).
Relationships between macroinvertebrate community abundance and
environmental stressors were explored using several multivariate ordination and
correlation techniques. Analyses showed spatially explicit variation among

macroinvertebrate communities, with most variation explained at the class scale,

11

but dependent on habitat type. In general, differences among the wetland
classes in emergent habitats indicated that palustrine marshes supported
different macroinvertebrate communities compared to riverine and lacustrine
marsh wetlands; the submergent/ﬂoating habitat did not differ among classes.
Because of the large variation of macroinvertebrate assemblages among all
spatial scales, it was not possible to identify useful biotic water quality indicators
for general wetland bioassessment in this watershed. Speciﬁc assessments
using macroinvertebrate communities may be needed for each class and habitat
scale. Results from this multi-scale analysis will allow for the identiﬁcation of
more useful indicators of wetland health that provide a guideline for scale-deﬁned

wetland bioassessments.

Introduction

Over the past twenty-ﬁve years increasing research has been conducted
on wetland ecosystems, due to their rapid disappearance, economic and social
beneﬁts (e.g., ﬂood prevention, groundwater recharge, water puriﬁcation) and
ecological signiﬁcance (e.g., species diversity, unique habitats) (Wrssinger 1999,
Mitsch and Gosselink 2000). In the United States alone, approximately 50% of
wetlands have been lost to development (e.g., urban, agricultural) leaving the
quantity and quality of those remaining wetlands at risk and vulnerable to human
activities (e.g., agriculture, mining, development) and subsequent environmental

stressors, both contributing to wetland degradation (Mitsch and Gosselink 2000).

12

The Clean Water Act (CWA) was established to restore and maintain the
chemical, physical and biological integrity of water bodies in the United States
(Adamus 1996), however, most emphasis has focused on lake, stream, and river
systems. Recent attention has shifted to developing similar CWA standards in
wetlands by improving methods and programs that protect wetland quality.

Biological indicators and their response to environmental stressors have
long been identiﬁed and successfully used in biological assessment protocols in
lotic ecosystems (King et al. 2000). Certain biota respond to both short and long-
term changes in the physical, chemical and biological environment, and are thus
good overall indicators of the environmental quality (Rosenberg and Resh 1993).
Macroinvertebrate communities are most frequently used in these assessment
protocols due to the following: (1) their role in facilitating energy transport by
connecting primary producers and higher trophic levels of the food web, (2) their
ubiquitous nature, (3) range of response to environmental stressors by different
species, (4) relative stationary nature, (5) their relatively long life cycles, and (6)
simple collection methods (Hellawell 1986, Rosenberg and Resh 1993,
Wissinger 1999, Helgen and Gernes 2001). However, Batzer et al. (1999) point
out that while these advantages apply to most aquatic habitats (e.g., streams and
rivers), some do not yet apply to wetland ecosystems. For example, many
wetland macroinvertebrates are mobile, some life cycles can be quite short (1-2
months), and that much information is still unknown (e.g., response to pollution
and taxonomy), thereby, limiting the use of macroinvertebrates in wetland

biological assessments (Batzer et al. 2001).

13

Environmental stressors on the macroinvertebrate community can be
abiotic or biotic factors and are commonly induced by human activities; however,
identiﬁcation of these relationships can be difﬁcult due to the complexity of
wetlands and interactions among multiple human activities and stressor. Several
studies, however, have identiﬁed these relationships and found that
macroinvertebrates can be useful in wetland bioassessments (Burton et al. 1999,
Kashian and Burton 2000, King et al. 2000, Apfelbeck 2001, Helgen and Gernes
2001, Chessman et al. 2002, Merritt et al. 2002b, Uzarski et al. 2004, Cooper et
al. 2006) while others have found little response between macroinvertebrate
communities and human activities (Tangen et al. 2003). Differences between
these studies question the utility of macroinvertebrates as general indicators of
wetland quality, and this may be due to high variability from independent or
interactive effects of hydrology, region, seasonality, and wetland type, among
others (Wilcox et al. 2002).

In order to understand this complexity, these relationships must be
addressed at multiple spatial scales. Macroinvertebrate communities and
environmental variables can vary among scales; thus, the stressor-biotic
response relationship should also be spatially explicit (e.g., wetland class, and
habitat) (Downes et al. 1993, Boyero and Bosch 2004, Rios and Bailey 2006).
An analysis that incorporates the role and importance of nested scales would
allow for the better evaluation of stressor -macroinvertebrate community
relationships (Kotliar and Wrens 1990), providing potential indicators of wetland

health and sampling procedures at the appropriate scale for assessment.

14

The ﬁrst objective of this study was to identify macroinvertebrate
community response to environmental stressors at multiple spatial scales among
inland Michigan wetlands. The following questions were addressed: (1) what
types and how do environmental stressors inﬂuence wetland macroinvertebrate
communities; (2) are macroinvertebrate communities good indicators of wetland
health; (3) what macroinvertebrate metrics can be developed for inland wetland
bioassessments. My second objective was to determine if macroinvertebrate
communities and their relationships with environmental stressors differ among
wetlands at each spatial scale. For each spatial scale of analysis the following
questions were addressed: (1) which macroinvertebrate taxa are ubiquitous vs
rare among wetlands at a particular spatial scale; (2) are relationships between
macroinvertebrate communities and stressors spatially explicit; and (3) should
each scale be evaluated separately in a bioassessment program for inland
marshes in Michigan? I hypothesized that macroinvertebrate communities would
differ among inland wetlands at each spatial scale and that relationships between
environmental stressors and macroinvertebrate attributes (e.g., metrics) would be

dependent on the scale of analysis.

Methods

Study Location and ﬁcale
The Muskegon River Watershed (MRW) begins in the north-central region

of Michigan’s lower peninsula, and drains southwest into Lake Michigan. The

15

watershed is approximately 7000 km2 and includes 94 tributaries, 183 stream
segments, hundreds of lakes and wetlands within 11 different counties (Torbick
et al. 2006). Dominant land use varies from forest to agriculture/urban
dominated areas within the Upper and lower regions of the watershed,
respectively (Lougheed et al. 2007). The continuum of the MRW has been
altered by 95 dams, has been impacted by human inﬂuence through logging and
agriculture, and is predicted to have a 50% increase in urban land use within the
next 35 years (Pijanowski et al. 2006). Inland wetlands (n = 57) were selected
from within and near the MRW; most sites,(~ 42) were randomly selected,
whereas remaining sites were purposely selected as described by Lougheed et
al. (2007) to ensure that a gradient of sites with variable water quality were
included in the study. Although the term wetland has been used to describe
many types of aquatic habitats, hereafter, in this study the term wetland referred
to only inland marsh ecosystems. All sites were surveyed prior to the ﬁeld
season to conﬁrm wetland existence, accessibility, and to obtain landowner
permission if necessary.

To achieve our objectives, three spatial scales were designated for this
study: the ecosystem scale, class scale, and habitat scale (Figure 1). The
ecosystem (watershed) scale classiﬁed all sites into a single group for analyses.
Within the ecosystem scale, all wetlands were further divided into the next scale
based on wetland class: palustrine (depression, n=25), riverine (n=18), and
lacustrine (n=14) wetlands (Figure 1). Within each class, habitats were nested

as distinct Vegetation zones (habitats) (Figure 1): when present, two zones were

16

sampled in each marsh, the emergent (E) and submergent/ﬂoating (S/F;
hereafter referred to as submergent) vegetation zones. The ﬂoating and
submergent zones were combined due to little variation in macroinvertebrate
composition (T. Burton, personal communication), and the inability to clearly

distinguish and separate these zones in many wetlands.

Sample Collection

Inland marsh wetlands were sampled once between July and early August
in 2002 and 2003. At each marsh site, three evenly spaced transects were
established perpendicular to one side of the wetland, thus extending from the
wet-meadow, or shore, into deep-water habitats. Along each transect, multiple
random sampling points (marked by a 1m2 quadrant) were established and
macroinvertebrates were collected from the ﬁrst random point in each habitat
zone. To ensure no disturbance in the macroinvertebrate community due to
other sampling activities (e.g., algae or zooplankton), macroinvertebrates were
collected at a random angle, 1—2 m from the main sampling quadrant. One
macroinvertebrate sample consisted of two sets of three sweeps (~1m2) using a
standard D-net (500 um) that was rinsed through a 500 um sieve to remove large
pieces of vegetation, and then combined into a single composite sample and
preserved in 100 °/o ethanol for laboratory identiﬁcation. Each wetland was
sampled only once giving a maximum of two composite samples per site (Halse

et al. 2002). Although 57 sites were sampled, some had only one type of habitat;

l7

therefore, a total of only 50 emergent composite samples and 50 submergent
composite samples were collected and analyzed in this study.

In the laboratory, macroinvertebrate composite samples were sieved and
sub-sampled to reduce processing time. In the sub-sampling protocol, each
sample was homogenized and divided into two equal proportions or sub-
samples. All macroinvertebrates from one sub-sample were sorted and identiﬁed
to the lowest practical taxonomic level using Merritt and Cummins (1996), Thorp
and Covich (1991), Pennack (1989) and Larson et al. (2000). Most
macroinvertebrate taxa were identiﬁed to the generic level, except for the
Chironomidae, Oligocheata, and Hirudinea. For other taxa, large numbers of
immature specimens were collected and could only be accurately identiﬁed to the
family level; all taxa from these groups (Planorbidae, Libellulidae,
Coenagrionidae, Corixidae, Notonectidae) were grouped at the family level.
Macroinvertebrate counts were adjusted (doubled) to account for the sub-
sampling protocol prior to statistical analyses.

Environmental factors measured at each wetland included chemical and
physical water variables. For water chemistry, a single 250ml sample was
collected from an open habitat area and analyzed in the laboratory for total
phosphorous (TP), soluble reactive phosphorus (SRP), ammonia (NH3), turbidity
(Turb) and chloride (CL) using methods described by Lougheed et al. (2007).
Field measurements of dissolved oxygen (DO), pH, conductivity and mean water
depth were collected at every site with a YSI 556 multiprobe meter. Land-use

land-cover‘(LULC) data for a 500 m buffer surrounding each wetland was

18

calculated using ArcGlS (Lougheed et al. 2007), categories determined were the
percentage of agriculture, urban, rangeland, forested, water, and wetland within

each buffer.

Statistical Analysis

We employed several multivariate statistical analyses that followed the
methods outlined by Cooper et al. (2006) to identify macroinvertebrate response
to environmental stressors at three spatial scales. In order to identify scale-
dependent relationships, the habitat zones were analyzed separately. For the
ecosystem scale analyses, 50 sites were used for each habitat. For class scale
analyses, the number of sites included depended on the habitat type: lacustrine
(E n=12; S/F n = 14), palustrine (E n=23; S/F n = 21), and riverine (E n=15; S/F n
= 15). Differences between the two habitats were compared for both the
ecosystem and class scale.

Principal component analysis (PCA) was conducted to identify stressors
among MRW sites at the ecosystem and class scale. For each scale, PCA were
conducted on two environmental datasets: (1) nine water chemistry variables log-
transformed and (2) six land-use land-cover (LULC) variables arcsine-square
root-transformed. The PCA will identify gradients in each environmental dataset
among the wetland sites. The ﬁrst two gradients or principal components (PCs)

from each PCA were correlated with each individual environmental variable from

19

the corresponding dataset in an effort to decompose the PCs as described by
Cooper et al. (2006) and Uzarski et al. (2004). Pearson correlations were
considered signiﬁcant at p < 0.05; those environmental variables signiﬁcant were
considered as potential stressors in MRW wetlands.

Correspondence analysis (CA) was conducted to identify patterns in the
macroinvertebrate community at the ecosystem and class scale. For each scale,
a CA was conducted on log-transfon'ned macroinvertebrate abundances, with 51
taxa groups in the emergent zone and 45 taxa groups in the submergent zone.
These taxa represent only organisms that were present in more than 10% of the
survey sites and 0.05% of the total macroinvertebrates collected (Cooper et al.
2006, Kratzer and Batzer 2007). The CA will identify gradients in
macroinvertebrate abundance among the wetland sites. Pearson correlations
(signiﬁcance level p < 0.05) were then conducted between the ﬁrst two gradients
or CA axis scores and corresponding macroinvertebrate abundances in order to
determine speciﬁc taxa abundance patterns in relation to the CA ordination at
each scale.

To relate macroinvertebrate community response to environmental
factors, two methods were used, Pearson correlations and canonical
correspondence analysis (CCA). Pearson correlations (signiﬁcance level < 0.05),
as described by Cooper et al. (2006), were conducted between the scores of the
ﬁrst and second CA of each scale and the corresponding scale site scores of the
ﬁrst two PCs. This method allowed relationships between biotic and abiotic

variables to be determined in an unconstrained method (Cooper et al. 2006).

20

Canonical correspondence analysis (CCA) was then utilized to relate the
macroinvertebrate community to the water chemistry variables, speciﬁcally to
environmental stressors. The same data sets used in the PCA and CA analyses
were used in the constrained CCA method for the ecosystem and class scale of
each habitat.

To identify speciﬁc differences at the ecosystem scale between habitat
types (between emergent and submergent) and at the class scale (among
lacustrine, palustrine and riverine wetlands) for each habitat type, one-way
ANOVAs for both environmental datasets and macroinvertebrate abundances of
the twelve dominant taxa were conducted, along with post-hoc multiple
comparison Tukey HSD tests at the class scale (signiﬁcance level of p < 0.05).
For macroinvertebrates, one-way ANOVAs for total macroinvertebrate
abundance and total taxa richness were conducted at the ecosystem and class
scale, along with post-hoc multiple comparison Tukey HSD tests to identify
speciﬁc differences among class scale for each habitat (signiﬁcance level of p <
0.05). Macroinvertebrate abundance data was log-transformed to meet statistical
requirements for these tests; taxa richness did not need transformation.
Dominant taxa were identiﬁed as the 12 most abundant individuals within each
habitat at the ecosystem scale. Ubiquitous and unique taxa were identiﬁed
among the wetland scales. For this study, taxa were considered unique to a
scale if it were absent in all but one level at the same scale and present in more
than two sites within that level (e.g., absent from lacustrine and riverine wetlands

but in more than two palustrine sites).

21

Principal component, correspondence, and canonical correspondence
analyses were conducted using R version 2.5.0. Pearson correlations, ANOVA

and Tukey HSD analyses were conducted using JMP IN version 5.1.2.

Results

Environmental Variables
Ecosystem Scale

At the ecosystem scale, the ﬁrst two principal components (PCs) of the
water chemistry and LULC ordination explained from 54-57% of the total
variance among sites (Table 2.1 and 2.2). There were signiﬁcant correlations
between the ﬁrst two PCs (PC1 and P02) and certain water chemistry variables
(Table 2.1). This suggested that each PC axis represented the association of
these variables and the signiﬁcance indicated that it was a main exploratory
variable in the PC. At the ecosystem scale, in both habitats, the water chemistry
PC1 was negatively correlated with turbidity, total phosphorus, ammonia, and
chloride and positively correlated with dissolved oxygen (Table 2.1, Figure 2.2a).
The water chemistry PCZ was negatively correlated with pH, conductivity,
chloride and positively correlated with dissolved oxygen (submergent zone only)
(Table 2.1, Figure 2.2a). The ﬁrst LULC PC was positively correlated with
%forest and %wetland and negatively correlated with %rangeland and agriculture
(Table 2.2, Figure 2.2b). The second LULC PC was positively correlated with

%water and %urban (Table 2.2, Figure 2.2b). In the ecosystem PCA ordination

22

with site labeled by wetland class, there was no clear separation of classes
based on water chemistry (Figure 2.3A) or LULC (Figure 2.38). No signiﬁcant

differences were observed in any environmental variables between habitats.

Class Scale

Individual PCA analyses at the class scale found from 60—72% of the
variation was explained by the ﬁrst two PCs of the water chemistry and LULC
ordinations in lacustrine and palustrine wetlands, whereas only 51-53% of the
variation was explained in riverine wetlands (Table 2.1 and 2.2). The water
chemistry PC1 for both habitats in the lacustrine and palustrine wetlands were
negatively correlated with chloride and conductivity, whereas in riverine wetlands,
PC1s were positively correlated with total phosphorus and turbidity and
negatively correlated with dissolved oxygen (Table 2.1). The second PC in all
three classes were negatively correlated with pH; however, other signiﬁcant
variables differed among classes (Table 2.1). The PCs of the LULC ordination at
the class scale identiﬁed different correlated variables for each class. Common
trends in all three classes have either %wetland or %water correlated with only
the ﬁrst PC and %agriculture correlated with the second PC axis; however, the
strength of these correlations varied (Table 2.2). In general, % agriculture and %
urban seem to be the main factors of disturbance in lacustrine and palustrine
wetlands.

Scale comparisons among environmental variables were conducted

(Table 2.3). Signiﬁcant differences were observed among classes in both

23

habitats for pH, TP and % water (Table 2.3). However, % agriculture and %
wetland were only signiﬁcantly different among classes for emergent habitats.
Post-hoc Tukey HSD tests showed submergent habitats in palustrine wetlands to
have signiﬁcantly lower pH values compared to lacustrine and riverine wetlands;
no signiﬁcant post-hoc differences were found between classes for the emergent
zone (Table 2.3). Total phosphorus was signiﬁcantly greater in palustrine
marshes compared to riverine marshes in both habitats; however, % agriculture
was only signiﬁcantly greater in palustrine marshes compared to lacustrine
marshes in the emergent habitats (Table 2.3). The % water was greater for
lacustrine wetlands in both habitats; this was expected since this class of

wetlands were deﬁned by their location adjacent to lakes.

Macroinvertebrates
Ecosystem Scale

A total of 182,688 macroinvertebrates were collected from three phyla, the
dominant ones being Arthropoda (75%), Mollusca (33%) and Annelida (7%).
The most dominant taxa were the Chironomidae (22.2%), Planorbidae (14%) and
Amphipoda (Hyalella, 12%). At the ecosystem scale, taxa richness was greater
in emergent than submergent habitats (Figure 2.4; df =99, F =34.6, p <0.001)
and macroinvertebrate did not differ among habitats (df = 99, F =1.49, p = 0.23).
The 12 most dominant taxa in the emergent and submergent/ﬂoating habitats are
listed in Table 5. At the ecosystem level, Physidae (df = 99, F = 6.7, p < 0.01)

and Sphaeridae (df = 99, F = 10.9, p < 0.001) were the only ubiquitous taxa to

24

have signiﬁcantly greater abundances in the emergent zone (Table 2.4). A total
of eleven taxa were found exclusively in emergent habitats and three only in
submergent zones (Table 2.5).

Correspondence analysis was utilized to identify patterns or gradients in
the macroinvertebrate community. At the ecosystem scale, the ﬁrst CA axis
(CA1) in both habitats explained 13-15% of the variance in the macroinvertebrate
community and combined with the second CA axis (CA2) explained only 23-25%
of the total variation (Table 2.6); however, macroinvertebrates signiﬁcantly
correlated to these axes varied between habitats. Thus, each CA was a
representation of the associated macroinvertebrates and the signiﬁcance
indicated that it was an important factor in determining the gradient of the CA
axis. Those signiﬁcant taxa may be responding to environmental changes in
MRW wetlands and may act as good biological indicators in future assessments.
In the emergent zone, the ﬁrst CA was negatively correlated with taxa from the
Odonata (Libellulidae, Coenagrionidae), Diptera (Dashylea, Bezzia) and
Trichoptera (Oecetis, Oxytheira) and the second CA was negatively correlated
with Amphipoda (Hyalella), Ephemeroptera (Caenis) and Mollusca (Viviparidae,
Hydrobiidae) (Table 2.6). Whereas, in the submergent/ﬂoating zone CA1 was
positively correlated with Mollusca (Viviparidae, Valavata) and Ephemeroptera
(Caenis) and negatively correlated with Coleoptera (Haliplus) and Odonata
(Lestidae). The second CA was positively correlated with lsopoda (Caecidotea)
and Amphipoda (Gammarus) (Table 2.6). The ecosystem CA ordination, with

sites labeled as class illustrated separation of marsh classes based on

25

macroinvertebrate composition (Figure 2.58). This separation of classes was
visually observed in the emergent ordination (Figure 2.5B), but not the
submergent ordination. This suggests that analyses at the class scale may

better explain macroinvertebratecommunity structure in MRW wetlands.

Class Scale

Individual CA analyses at the class scale were able to explain more
variation compared to the ecosystem scale, with the lacustrine macroinvertebrate
community explaining over 40% of the variation, followed by the palustrine and
riverine communities with 28-31% (Table 2.6). Several macroinvertebrate taxa
were signiﬁcantly correlated with the CA axes for each class, with no common
trends among classes or between habitats (Table 2.6).

At the class scale, a signiﬁcant difference in macroinvertebrate abundance
(df = 49, F = 3.41, p < 0.05) was observed in the emergent zone, with the highest
mean abundance in palustrine followed by riverine and lacustrine wetlands
(Figure 2.4A). However, post-hoc Tukey tests showed no signiﬁcant difference
among class means (Figure 2.4A). In addition, macroinvertebrate abundance in
the submergent habitat was not signiﬁcantly different between classes. Within
each class no signiﬁcant differences were observed between habitats (Figure
2.4A; lacustrine: df=24, F = 0.01, p = 0.92; palustrine: df=43, F = 2.86, p = 0.10;
riverine: df=30, F = 0.001, p = 0.97).

Total taxa richness of emergent habitats was signiﬁcantly different (df=49,

F = 5.41, p < 0.01) among classes, but not for submergent habitats (df=49, F =

26

1.16, p = 0.32) (Figure 2.4B). Tukey HSD tests found that taxa richness was
signiﬁcantly greater in palustrine compared to riverine wetlands in emergent
habitats (Figure 2.5B). Within each marsh class, emergent habitat taxa richness
was signiﬁcantly greater than the submergent habitat (Figure 2.43; lacustrine:
df=24, F = 9.57, p < 0.01; palustrine: df=43, F = 23.8, p < 0.001; riverine: df=30,
F=494p<00$.

Results from among class ANOVA analyses indicated signiﬁcant
abundance differences for the Chironomidae, Planorbidae, Hyalella, Copepoda,
and Lymnaeidae in emergent habitats, and Hyalella, Caenis, Neoplea,
Hyrdrachnida and Bezzia in submergent habitats (Table 2.4). In general, mean
comparisons with Tukey HSD tests indicated that palustrine wetlands were
signiﬁcantly different from riverine and lacustrine wetlands in both habitats (Table
2.4). Copepoda in palustrine wetlands (df = 43, F = 4.31, p < 0.05) and
Sphaeridae in palustrine and riverine wetlands (df = 30, F = 5.1, p < 0.05) were
the only ubiquitous taxa to be signiﬁcantly more abundant in the emergent zone
compared to the other habitats (Table 2.4). Speciﬁc taxa were exclusively found

within a single wetland class (Table 2.5).

Environmental -Macroinvertebrate Response
Ecosystem Scaﬁ

Signiﬁcant Pearson correlations between the PCA and CA axes for each
habitat at the ecosystem scale are presented in Table 2.7. The ﬁrst CA axis for

the emergent zone was negatively correlated with both water chemistry PC1 and

27

PCZ, whereas the ﬁrst CA axis (CA1) in the submergent habitat was not
signiﬁcantly correlated. At the ecosystem scale all the water chemistry PC
relationships had correlations below 0.5 (Table 2.7). For the LULC dataset, only
the emergent CA2 axis was correlated with PC1; all other relationships were not
signiﬁcant. The CCA analyses at the ecosystem level explained only 15% and
8% of the variation in the relationship between the marcoinvertebrate community
and environmental variables within the emergent and submergent habitat,
respectively. The same patterns found in the ecosystem level PCAs and CAs

were also found in CCAs for both habitats.

Class Scale

Signiﬁcant correlations between either CA1 or CA2 and the water
chemistry PC1 were observed at least once in each marsh class; all of these
relationships had a signiﬁcant correlation of +/- 0.59 or higher (Table 2.7). No
signiﬁcant correlations were observed at the class scale between the CA axes
and water chemistry PC2. Signiﬁcant relationships identiﬁed between gradients
at the class scale in the environmental variable data (PCAs) and
macroinvertebrate community (CAs) are listed in Table 2.8.

In emergent habitats of both lacustrine and palustrine classes, increases
in chloride and conductivity corresponded to in a shift from Corixidae, Cladocera,
and Hydrophilidae (Berosus) dominated communities to Hyalella, Haliplidae,
Stratiomyiidae (Odontomyia), Gastropoda (Valvatidae, Lymnaeidae,

Planorbidae), and Lepidoptera (Munroessa) communities (Table 2.8). In the

28

submergent zone of the lacustrine and palustrine marshes, macroinvertebrate
taxa had both positive and negative responses to increases in chloride,
conductivity and ammonia; however, the responsive taxa were different from
those in the emergent habitats. More speciﬁcally, in lacustrine marshes, a shift
occurred from Libellulidae, Corixidae, Cladocera, Oxythein'a, and Baetidae to
Hirudinea, Lymneaidae, and Hydrophilidae (Tropisternus) communities and in
palustrine marshes, the macroinvertebrate community shifted from Haliplidae and
Lepidoptera (Munroessa) to Hydrarchnida, Libellulidae, Copepoda, Caenis, and
Coenagrionidae (Table 2.8). Riverine marsh macroinvertebrate communities
signiﬁcantly responded to different water chemistry variables, as total
phosphorus and turbidity increased and dissolved oxygen decreased, a shift in
taxa from Caenis and Trichoptera (Oecitis) to Copepoda, Oligocheata, Physidae,
Hydroporinae, and Gammarus communities (Table 2.8).

At the class scale, the most signiﬁcant correlations between the LULC PC
and CA axes were observed in the lacustrine and palustrine classes. All
relationships between the CA axes and LULC PC1 had a signiﬁcant correlation of
+/- 0.5 or higher at the class scale (Table 2.7). Again no signiﬁcant correlations
were observed for the LULC P02. The ecosystem CCA ordination illustrated
separation of the marsh class based on macroinvertebrate communities and
water chemistry variables (Figure 2.6). This separation of the marsh class was
found in the emergent zones (Figure 2.6) but was not as deﬁned in for
submergent habitats. The CCA analyses at the class scale explained more of

the macroin‘vertebrate-environmental variable variance (24-37%) compared to

29

the ecosystem scale (8-15%) (Table 2.9), and most variation was explained in
lacustrine classes (36-37%) (Table 2.9). The same patterns observed in the
ecosystem level PCAs and CAs were also observed in the CCAs for both

habitats

Discussion

Ecosystem Scale

Principal components analysis was utilized to identify potential
environmental stressors that may inﬂuence macroinvertebrate communities in
MRW marshes. At the ecosystem scale, six of the nine water chemistry
variables included in the analyses were signiﬁcantly correlated to the ﬁrst PC
axes of both habitats, thus making it difﬁcult to identify speciﬁc stressors. These
results suggest that a single environmental variable was probably not responsible
for site separation observed in ordinations at this scale. All water chemistry
variables correlated to the ﬁrst PC axes had negative relationships except for
positive correlations with dissolved oxygen. These signiﬁcant correlations
indicated that environmental gradients exist among these sites and that these
correlated variables could be potential stressors on macroinvertebrate
communities in MRW inland marsh wetlands. Furthermore, these signiﬁcant
relationships were with variables that we would predict to change based on
previous wetland studies. Studies have found impacted wetland areas to have

higher chloride (Kashian and Burton 2000, Helgen and Gernes 2001, Cooper et

30

al. 2006), total phosphorus (Helgen and Gernes 2001), nitrogen/nitrate (Kashian
and Burton 2000, Cooper et al. 2006), turbidity (Helgen and Gernes 2001), and
SRP (Kashian and Burton 2000) values. Non-impacted wetland areas have been
reported to have higher dissolved oxygen (Kashian and Burton 2000, Helgen and
Gernes 2001, Cooper et al. 2006) and pH values (Cooper et al. 2006). Uzarski
et al. (2004) found that wetland water chemistry was dependent on surrounding
land-use; wetlands surrounded by agriculture had increased pH, temperature,
turbidity, alkalinity, sulfate and dissolved oxygen (in the daytime), and wetlands
surrounded by urban/road development had increased chloride, conductivity,
nitrate and ammonium. At the ecosystem scale in this study, the main
environmental variables indicated of a mixture of human disturbances, with the
water chemistry PC1 representing an agricultural disturbance gradient and PC2
representing an urban/road disturbance axis.

In the LULC analyses, four of the six variables were signiﬁcantly
correlated with PC1, again making the identiﬁcation of exact environmental
stressors difﬁcult. We would predict that % developed land (including %
rangeland and % agriculture), will increase with human disturbance similar to that
found by Soandso et al. (2002) and Cooper et al. (2006); whereas % forest and
% wetland- have been found to decrease with human disturbance (Cooper et al.
2006). However, response to surrounding land-use land-cover may not occur in
all wetlands and could be region speciﬁc (Tangen et al. 2003). At the ecosystem
scale in this study, the LULC PC1 and P02 may be indicative of a two different

disturbance axes, one based on agricultural differences and another on

31

urban/road differences (Uzarski 2004). When marsh class, however, was
superimposed on the ecosystem scale PCA ordinations (Figure 2.3), no clear
separation of sites was observed in either habitat. This indicated that observed
disturbance gradients may be affecting all three marsh classes in the MRW. It is
noteworthy that some separation of lacustrine sites in the LULC ecosystem
ordination was observed in relation to %water. This would be expected because
this class of wetlands are deﬁned by their position next to lake ecosystems.
Correspondence analysis (CA) was utilized to identify patterns or
gradients in the macroinvertebrate community based on abundance values.
Multiple taxa, in both habitats, were correlated with the CA axes, indicating
gradients in macroinvertebrate community abundance. The presence of these
gradients indicated that these taxa are changing across sites; however, the exact
reason for these changes are unknown from these analyses (and discussed
further below). Macroinvertebrate taxa represented by the CA axes differed
between habitats, indicated that community composition also differed between
habitats or that stressors on these communities differed between habitats.
Between habitat comparisons of the environmental variables and of dominant
taxa/unique taxa suggested that these CA gradients could be due to both factors.
The ecosystem CA ordination, with sites superimposed as marsh class (Figure
2.4B), illustrated separation of marsh class based on macroinvertebrate
composition. This indicated that the gradients observed in the macroinvertebrate
taxa (CA axes) at the ecosystem scale might be due to class instead of potential

environmental stressors. This separation of marsh classes was visually

32

observed in the emergent ordination, but not the submergent ordination, which
implied that submergent habitats are more similar among wetland class.

At the ecosystem scale, gradients were identiﬁed in the environmental
data and macroinvertebrate community; however, I was interested in determining
if relationships existed between these two factors. Signiﬁcant correlations in the
emergent habitat between the PCA and CA axes identiﬁed relationships that may
exist at the ecosystem scale; however the low value correlations (values < 0.5) of
these relationships was probably due to the inﬂuence of marsh class observed in
the CA emergent ordination. This was further supported by CCA analyses in the
emergent zone (Figure 2.6); ordination clearly separated the palustrine class
from the lacustrine and riverine class, indicating that main gradients at this scale
may not be due to environmental stressors. Differences in the macroinvertebrate
community and response may be a result of natural differences (e.g., physical,
hydrological) between lacustrine, palustrine, and riverine marshes. In the
submergent zone, no clear separation of class was observed in the CA or CCA
ordinations and no signiﬁcant correlations were observed between the PCAs and
CAs, revealing that macroinvertebrate community composition in this zone was
similar across class. Low correlations at the ecosystem level between the PCA
and CA analyses also indicated that speciﬁc relationships may be better

identiﬁed at the class scale and will be discussed in more detail below.

Class scale

33

Principal component analyses (of both water chemistry and land-use land-
cover data (LULC)) explained more variation among sites at the class scale,
speciﬁcally in lacustrine and palustrine marshes, with gradients observed in
mainly chloride, conductivity, ammonia and pH as seen in other wetland studies
(Helgen and Gernes 2001, Gernes and Helgen 2002). Riverine marshes had the
least variation explained of the three classes and lower variation compared to the
ecosystem scale; gradients in this class were observed in total phosphorus,
turbidity, dissolved oxygen and conductivity. According to Uzarski et al. (2004),
these same environmental gradients may be indicative of more urban and road
disturbance in lacustrine and palustrine habitats and agricultural disturbance in
riverine wetlands. Cooper et al. (2006) found water chemistry in drowned river-
mouth wetlands in Michigan to be extremely variable due to a range of
hydrological and biological conditions. These same dynamic conditions could
also have inﬂuenced riverine wetlands in the MRW. Among class comparisons in
the emergent zone had more signiﬁcant differences in water chemistry and LULC
values compared to the submergent zone, with palustrine marshes having
statistically different values from either lacustrine, riverine or both marshes. This
indicates that environmental variables affecting palustrine wetlands may be more
distinct, compared to more similar environmental variables in lacustrine and
riverine marshes.

Correspondence analysis at the class scale explained more variation in
the marcoinvertebrate community compared to the ecosystem scale. A large

variety of macroinvertebrate taxa were signiﬁcantly correlated with the CA axes

34

among the three classes and among habitats in the three classes, therefore,
making no clear indicator taxa evident across all marsh classes or habitats. The
signiﬁcant relationships identiﬁed between gradients in the environmental
variable data (PCAs) and macroinvertebrate community (CAs) were stronger
(higher) and thus more signiﬁcant than relationships at the ecosystem scale
which allowed for important variables in our stressor-response relationships to be
identiﬁed in these MRW inland marshes.

A variety of macroinvertebrate taxa had a negative response to increased
environmental stressors, suggesting their potential use as indicators of marsh
water quality. The Ephemeroptera (Baetidae, Caenis) in the submergent zones
of three marsh classes and the Trichoptera (Oecitis, Oxytheiria) in the
submergent zones of two marsh classes both responded negatively to increased
values in chloride, conductivity, total phosphorus and decreased dissolved
oxygen values. Both Ephemeroptera and Trichoptera have been considered
good indicators of poor water quality in lotic ecosystems, and although fewer taxa
are found in wetland ecosystems, other studies have found these taxa to be
useful as wetland indicators of water quality in coastal and depression wetlands
(Kashian and Burton 2000, Gernes and Helgen 2002, Uzarski et al. 2004). The
Odonata, including Libellulidae and Coenagrionidae, also had a negative
response to increased environmental stressors in only submergent zones of
palustrine and lacustrine marshes. Although not used as indicators in lotic
ecosystems, the Odonata are more diverse in wetlands and have been found to

be sensitive to low water quality (Kashian and Burton 2000, Gernes and Helgen

35

2002, Uzarski et al. 2004, Cooper et al. 2006). The Corixidae have often been
considered indicators of poor water quality and thus common in impacted
wetlands (Gernes and Helgen 2002, Cooper et al. 2006); however, my study
found this group to decrease in response to increased conductivity, chloride and
ammonia in MRW wetlands.

A variety of macroinvertebrate taxa had a positive response to increased
environmental stressors. Amphipoda had a positive response to increased
stressors, speciﬁcally Hyalella to increased conductivity and chloride in the
emergent zone and Gammarus to increased total phosphorus and turbidity in the
submergent zone. Other studies have also found Amphipoda to be ubiquitous in
wetlands, however, their response to stressors varied from no change in Hyalella
and Gammarus with increased impact (Cooper et al. 2006), to genus
replacement (from Hyalella to Gammarus) with increased impact (Kashian and
Burton 2000), and a decrease in abundance in Hyalella with increased impact
(Uzarski et al. 2004). Uzarski (2004) found that although the Amphipoda
ultimately decreased, its ﬁrst response was an increase in abundance with
intermediate levels of disturbance. Based on this, perhaps in our study the most
disturbed sites were only at this intermediate level. ln this study and in others
(Burton et al. 1999, Kashian and Burton 2000), the Gastropoda , including
Valvatidae, Planorbidae, Lymnaeidae and Physa positively responded to
increases in all signiﬁcant environmental stressors in at least one habitat and one

class.

36

Numerous macroinvertebrate taxa were ubiquitous across all spatial
scales. Chironomidae were the most abundant taxa collected among all three
marsh classes, as observed in other wetland studies (Kashian and Burton 2000,
King and Richardson 2002), andat the family level did not appear to respond to
environmental gradients in our wetlands. Similar results were found in costal
wetlands by Burton (1999) in which Chironomidae showed no response to
human disturbance. However, some wetland studies that have identiﬁed
Chironomidae to further taxonomic levels (e.g., sub-family, genus, species) have
found taxa that respond differently to environmental stressors (Kashian and
Burton 2000, King and Richardson 2002). Further identiﬁcation of the
Chironomidae in this study may have lead to signiﬁcant differences between and
among wetlands at various spatial scales.

Differences among wetland class were observed throughout this study,
with the palustrine marsh often being separated from the other two marsh types.
This separation was most evident in ordination techniques using the
macroinvertebrate communities and thus palustrine macroinvertebrate
communities are different compared to the riverine and lacustrine marsh
communities. Similar communities in the lacustrine and riverine marshes could
result from similar exposure to moving water through ﬂooding events and wave
exposure, whereas palustrine marshes tend to be more stagnant. Ordination
techniques based on water chemistry and LULC did not separate marsh classes,
indicating that other factors, such as these natural physical differences

mentioned, may be driving the distribution of macroinvertebrates. This would

37

support the importance of hydrology in the development of bioassessment
protocols (Wilcox et al. 2002) and the need to develop these protocols at the

class scale in inland marshes of the MRW.

Habitat Scale

Aquatic vegetation can inﬂuence biotic communities within wetlands by
providing a variety of habitat and food resources, shaping community interactions
(e.g., predation), and altering abiotic conditions (Voights 1976, De Szalay and
Resh 2000). In this study, the submergent zone appeared to be a more
homogenous and stable habitat compared to the emergent zone. Mean
macroinvertebrate abundance and total taxa richness were lower in submergent
habitats, with no statistical differences observed among marsh class. In addition,
only three unique taxa were identiﬁed from the submergent zone compared to 10
unique taxa in the emergent zone. More sensitive taxa, such as the
Ephemeroptera, Trichoptera, and Odonata were only observed to be responsive
to environmental stressors in this submergent habitat perhaps due to a more
stable location, with deeper more permanent water during the year. The
emergent habitat of marshes are usually more shallow with ﬂuctuating water
levels making them a less stable habitat (Murkin and Kadlec 1986, Burton et
al.1999). This increased disturbance and high habitat heterogeneity available in
emergent zones could be reasons for greater abundance, taxa diversity and
more unique taxa (Murkin and Kadlec 1986). Voights (1976) found similar results

with greater abundance in emergent compared to submergent zones, however,

38

they found the zone where emergent begins to mix with submergent zones to be
the most abundant and diverse habitats. Cardinale et al. (1997) also found
macroinvertebrate attributes (epiphytic Chironomidae abundance, biomass, and
diversity) to be greater in the mixed zone of emergent (littoral) vegetation and
open water habitats. They also found that plant growth and water level
inﬂuenced their results over time, suggesting that seasonal inﬂuences could have
an affect on wetland macroinvertebrate communities (Cardinale et al.1997). Our
results indicate that macroinvertebrate communities in emergent habitats were
more variable than submergent habitats; yet, both habitats had distinct taxa that
responded to environmental stressors. This data supported work by Burton et al.
(1999, 2004) in coastal wetlands, that vegetation type was an important factor in
macroinvertebrate community structure. More research is needed on
macroinvertebrate response to vegetation in inland wetlands, and studies over a
gradient of vegetation types and in mixed-habitat zones may be helpful for future
assessment protocols.

Other potential inﬂuences on the macroinvertebrate community that were
not taken into account for this study include both abiotic and biotic factors.
Abiotic factors such as ﬂuctuations in wetland hydrology and hydroperiod could
inﬂuence plant communities and habitat availability thus modifying the
composition, abundance and life histories of certain macroinvertebrate taxa
(Wilcox et al. 2002). Additional chemical or physical factors, could also have
inﬂuenced macroinvertebrate responses. The biotic community within each

Wetland coUId also have direct and indirect affects on macroinvertebrate

39

communities. For example, the presence or absence of ﬁsh could impact the
number and composition of macroinvertebrates due to increased predation
(Tangen et al. 2003). Future studies should address the relationship of
macroinvertebrates with other biological parameters such as plant, zooplankton

and algal communities among wetland classes and habitats.

 

2.64.300

 

 

 

Lapses» 2.53% 2.8036
.8031a L338 L338 Lowoimm 53-3580 L536 2.833.080
2.89328 2.5328 L8.o-vza 2.33% L333 L533 L83ucoo L531“. Non. 53>
L936 L398
med-Emw L33¢mw
2.80.53-33.22 L582» L39 on L836
2.63300 L238 2.83-512 2.63612 .63328 2.60332 2.5.39.2
LEQEE 2.3.393 2.3.328 Lmheucoo L336 L83ucoo 2.5.3.“: Lmo3n:
Lament Lamar: 2.8936 L396 L2.3mxz L8.o-v_o 2.82.393 2.83-3.3 son. 5?
gown—2.30
mm... 5... 8... 3... o... 5... 3.: mm... 459
Ed mg no mud mg 86 mud «No Non.
8.0 8.0 moo one 3.0 one 5.0 5.0 Ba
Quasar—«>3.
Em m ".5 m Em m Em m
octozm 2.2.329“. octagon.—
320 Eounamoom

 

.Eﬁmov Econ :92: can .203 3:25 .3me 355305 9:82 2328 .8123 mEoEEm .Eb mEOcamoca

.32 563.2 .69 5236 828% 228138328 .1a 8205 852.5, $2520 Lea; .33 v a u .. .306 v a u L 8%:

9m $8 on. :03 9 God v 3 83.9.8 bEmoEcQw mo_nmtm> QﬂEmco .053 .28» some .2 62m: 9m «do some 30 Amoav 9:88:50
.3655 92 “9....— oE E 3:598 8:99 2:088 on... .3552, :29: E 2&9 mczmocbccmLmEnawxwv EmmLmev diam “SEN:

ucm 65.02. 6553.3 .2598; $30 $9383 9: am $399 EgEmco Loam; m B Eon: mazmcm EccOano 69055 ”mm w_nm._.

41

 

 

 

 

.8..m< _8_.m< .8..m< .8..m< 8ch 608m. .53.: :85
IAwmdiv 2L vmdiv .5353 22853 2.95.8 1.3de 2285.9 2235.3
.88". .88. :25 :85 8.9. 32.9.. 555 .235 N o. 5.;
2.2de 2.398
~m>> uw>>
Cm.0i mm.Oi cA®®.Oiv ¢«AN®.OIV taAOWAViV ckAwmiv ELVGAUIV
hmum>> 50am; “G; «0; ummbOl _30_2_ O< _30_._ m<
2:98 L {.9 3.85.0. .5550; .2.?de 22599 «255.0; 2255.9
:59: C33 00cm“ 00cm”.— CNED Ewe: mar—mm “@050“—
u0>> u0>> «mthu. uwQLOl hmum>> L0~W>> uwmbou Gar—Gm P On— St;
GOSH—9:00
and nmé rod Ned bod th 3.9 and i_<._.0._.
mNd mud NNd Fwd mNd Fmd VNd NNd NOQ
mNd QNd and :10 mmd Yo Fwd vmd _.On_
. 0°CN_.—5> o\o
Em m Em m Em ll . m Em m
O=t0>mm OCEﬂ—gﬂa OCT—HQSOGJ
36—0 Eﬂwm>80m

 

68.83 9.6 .963 .822 6.3.35.8 68.88. .89: *o 85:80.... 62.56. 902, 8.95.9
.9868. 836:3 .Cod v a u .. .Sod v a
.9868. 836...“... .o.8m some .2 63m: 05 (on. 88 .o Amon: mucoc8E8 .8655 9.3 8.... 65 .5 85598 88:?
E698 oz... .m6cm_.c.s 8.9.. 5 2".va mazmogchmEnam .Amv «898.3 0.88 .359. 95 .852. .mcEmzﬁa 65.838;
0.8m mmm.o 6.8m E2938 05 .m 835.? 6.5.: .8868. 836cm. 0 .0 20a: m_m>_mcm 2.9.an8 .8655 Nd 6.8..

:v 69m: 0.6 m3... on. some 9 God v 3 6952.8 228586 8.95.?

42

 

 

 

 

 

 

 

 

< < < m: < < .388... .8....@.18.89...§...25 .588... .388... .888... .889... 280.5
m m < 5... v. m. m. .588... .588... .885... .885... .888... .888... .588... .588... .202.
< < < m: < < .888... .888... .0988... .888... .888... .388... .888... .888... .85.
< < < we 9:. . .888... .885... .388... .383... .888... .888... .588... .588... 30......
< < < m: < < .888... .888... .888... .888... .888... .888... .888.... .883... 8:00
< < < w: < < .888.... .0988... .888... .888... .888... .5.88... .588... .588... :85
.83....
< < < m: < < < m: .8. 8 .18 .88 .88 .8... E8 .8... .8... 5%..
m s. 8.3 8.8 8.8. :8 8.8 8.... 8....
< < < m: < < < w: .8... .88 .88 .08 .8... .8... .88 .8... _o
8.8 8.8 0.5 8.5 0.: R... 5.8 £8
< < < 2 < < < m: .0880 .0880 .888.... .03.... .388... .888... .3888 .8880 amm
< < < m: < < < m: .585... .585... .888... .888... .585... .588... .588... .588... 8.2
m < 9. 8...v m <m< 8...,v :8 .88 .2... .8... .8... .8... £838 .88 n:
H. 8.8. 8.8 8.8 8.... 8.8 8...... 8....
< < < 2 < < < 2 .8888 .888... .888... .888... .88.... .888... .583 .58.... £3
< < < m: < < < 2. £880 .588... .888... .888... .585: .385... .88.... .3888 0..
< < < 8 < < < m: .88.. .808 .38. .308 .58. .88. .85.. .8... 88
8.8.. 8.8.. 8.88 8.88 8.8 0.08 08 .8
< m < 5...v < < < 8....v. 82...... .288... .2850 .2880 .388. .288. .88.... .882. T...
E.0_E0:o
m n. . Em m n. .. Em m Em m . Em m Em m 29...;
438%.: 902.. 05.9.0 0:58.... 9.588.. $933
hmmnu E0wm>moom

 

 

 

.mm0.o 6:0..02. c0050.. 02.0.05... 50053.0 .c0m0...0. 92.0. 50.056 0.08

5000 5 380... 9.050 Esocm 0.02. m..:m0. co_m_.0ano :00... .8... v 3 am... >955 68.80. 0.0 003.9 .m... .c0oc_..m_m-coc 6:0 30... v
q. m0....0> .c0o......m.m ..0....0.. .o 0...... 500 5......) 0.00m 8.00 0... .o. 60.c0m0... 0.03 8.30. <>Oz< 6:0..03 6:0 .0.0>> ..m0.o. 0.3.8.50
6.5.09.0. 60...: .o 9.5.500... 02.0.0. 0.0; 855.9 6.5... .0>oo6..0. 0365.. ..E0. 5006 :00... 6:0 ...=mE .0. 05.2....

.38: .mm. 8.088.... 02.000. 058.8 .38: 2.2. 0.0.... .88 .8 8.9.89... .08. 5.65.... .88. on. $on 828”... .528.
3.2.0365... .10 .0620... m0...0..0> 8.0.5020 .0.0>> .m6:0_.0.s cm.0E 5 .06. m5.0oE.c0m.0En:m..m. E09050. 0.00m .0500: 6:0 .9...
05.9.. ..n.. 05.53.05 ..... 05:860.. 0.00m mm0... 0.08 50.9800 0... .0 m0.n0..0> .0...0E0co..>:0 =0 .0. m0:.0> .mw.c00s. .mN 0.90..

43

 

 

   

 

 

 

 

 

 

 

f ,v m: 2 a: a: 28:6 \66P68v\ 658 \6588 \ﬁmm
< < < to m: a: a: a: 6883. \ 66 \6.86~.~w\ 666.8 \33835
m m< < modv m: a: m: a: 6.886.: \E.E6.o¢\ 6,666.8 \ 6.98.5 \mecmoeuf
«55:5
8 m: m: m: < < < no \ A866 \ 9.86.8 \ 646.6 \6.:86.8 .cotmmcmmwm
m: m: a: a: < < < mod \ 6.68.6 \ 6,366.86 \ 6368.8 \6wé88 82.2.3:
2 m: m: m: ﬁ< < m ..&wo..v_\ 6.52.8 \ 2.86.8 \ 6.6.6 \68588 $2625.:
can—ED
< < < 8.0 < < < who 6.88.3 6.8668 6.86666 3.06869 6.66.: 2.65.8 6.96.3, 6.86.: “as am
m < m Sov < < < Ed 66:88 p.168 6.38.63 6,868.3? 66.858 2.86.6 688.8 686.8 8682
< < < 3.0 < < < No 6.368? 6663928688 6.9666: 3,866.6966326263688 6.8:.an 98830
< < < 2 < < < 8.0 6.588 886636.365; 66666.88 6.286.? 6.86.5 6566 6.86.6: 82628
m<m < 8.3 < < < 6.0 6.660: 6.86.8 6.8686 6.868 6863:5556?6.3:.6... 6.862 25mm
< < < 95 < < < .mowov 2695.82 6.86: £566.: 6.86.86 6.26:3. 6866.6 6.86.3 @668 «6888
,.< m < 506v < m < 6.66.16.62659 6.86.8 566.8 6886.8 28:39 6.6666: 3.8668 6.368 2.2%:
< < < NS m < 9. 3,6,» 6.86.8 6.86.8 6.5638 c.3868 66.8. 6.85.62696.686.96.38 365555
< < < 8.0 < < < modv 6.63963 6.86668 6:696 6635.86 39:882.98:86.66.686663636258920
. I £352.. 58
m n. .. Em m a 4 m Em m Em m Em m Em m 88h
as» <>oz< a: <>oz< 2:321 233.3 2.5283 . .. ; SEER.
«:mahoEnam «ccmumEm mam—o E3m>moom
.mmm_o

2953 5053 855:6 Eonch m Emmmam: 92$. 6.90:6 ”28m comm E mmmmﬂo 9095 £505 995 mzamm: comtmano :85
80.0 v S ow: >33 3059.59... 35 .86 v S m02m> E8556 53> umtoao. 2m mm:_m> Q 4.859.. “6 83 :08 £53, 28m 390
m5 .8 noEmmma 2m; $.39 <>Oz< .mucmzwz, cmﬁE E 25 0552.. An: mcEmama .Sv chmaom: 0.8m mmﬂo new 286 Ew~m>moom
m5 5 5:9 mczmogcomEEnam .Amv EQOEmv Etna: comm 55:3 mxS SmﬁmtggeomE EmEEou NF .8 mm:_m> $9:me EN 2an

Table 2.5: Macroinvertebrate taxa unique to the class or habitat scale in this study.
Included are taxa class, order, family and generic level information, functional
feeding group and habitat classiﬁcation.

 

 

Scale Class Order ﬁFamily Genus

Class

Lacustrine Insecta Diptera Ceratopogonidae Alluaudomyia

Palustrine Insecta Coleoptera Dytiscidae Agabetes
Insecta Coleoptera Dytiscidae Coptotomus
Insecta Coleoptera Dytiscidae Dytiscus
Insecta Coleoptera Dytiscidae Hydroporus
Insecta Collembola Poduridae
Insecta Diptera Chaoboridae Chaoborus
Insecta Diptera Corethrellidae Corethrella
Insecta Lepidoptera Pyralidae Acentria
Mollusca Gastropoda Viviparidae

Riverine Insecta Coleoptera Elmidae Dubiraphia
Insecta Odonata Calopterygidae Caloptelyx
Insecta Odonata Gomphidae Arigomphus

Habitat

Emergent Insecta Coleoptera Dytiscidae Celina
Insecta Coleoptera Dytiscidae Dytiscus
Insecta Coleoptera Dytiscidae Hydroporus
Insecta Coleoptera Dytiscidae Rhantus
Insecta Coleoptera Hydrophilidae Anacera
Insecta Diptera Corethrellidae Corethrella
Insecta Diptera Sciomyzidae Type 2
Insecta Diptera Sciomyzidae Type 3
Insecta Odonata Calopterygidae Calopterwr
Insecta Odonata Libellulidae Pachydiplax

Submergent Insecta Diptera Ephydridae Lemnaphila
Insecta Trichoptera Hydroptilidae Agraylea
Insecta Trichoptera Hydroptilidae Orthotn'chia

 

45

 

 

 

 

600-210 L606? £00600
.6380 60.00202 6030”:
60.3222 60.3.0.0 600.65 6.0620
60.30... 6531.6 60.352 .6350 60.0.66
20.0-.Iam L 53.3.. 803000 6600000 600050 L639:
30.0050 .65300m 60.3000 6003232 6006.00 60.06: ..60.0-0>_>
60000.... ..N50-.m_..n. 50.3mm: L 53000 5535.5 .6505). 5.5.0.63 560.020
.6595... ..E.0-Emm .6006: 800.05... .6506: .650600 :60.0§<0 L030<0
:E.06<m 5.65063 :60302 2.6000000 L 5.3%: L5.0-.<z< L00.000m_ :60.0-.<>I N 008. <0 5.;
L500002 60.350
L6.0.<..0 60.0030 16.3010
60.3.4.0 60.00%: :60602 L 6.58
63000 60.330 .6365 600020.). 600025 166.060.. L 6.0-050
60.32.00 603.206 600.50 60065 £0.00>_> 6003/; 5.600620 :60.0-Emm
600.66. 60.3235. 600005 :6000000 .60.0:<> 659000 «600.06. 3.603000
LE3>IQ 60.3020 600620 265060.). .6539... L 50:00 L603? ..6m.0-600
.60330 6000000 800050 850600 6500100 .653000 :600604. :603Nmm
1.650-600 60.065 L~59002 :600005 L 5.0602 653000 L00.0.._<> ..6m.0-vm<n_
L0506<0 600-00,}. .650-me $60.06: L00.30m_ L895: :60.0v>_> .6530: F win <0 5?
cozwﬁtoo
8.0 3.0 8.0 3.0 3.0 3.0 8.0 8.0 .25»
3.0 9.0 3.0 0.0 2.0 3.0 :0 00 ~<0
2.0 .20 3.0 8.0 8.0 8.0 3.0 2.0 20
wocata> .x.
Em m Em 0 Em 0 Em 0
05.531 0:_._um:_wn_ actumsoar.
meU EwumHmOom

 

.300 v a n 0180.0 v u w 3.0 030: 06 0x... <0 68 B 60.0 v 3

0000.00.60 5:005:06 0.000000003500002 0.000 0.000 08 000m: 000 00x0 <0 025 005 05 >0 0050.96 0000:0> E0000
0;... 6000.003 cm..0E E 2&9 00.000.0\0000._0E0:m Am; 0:090:03 0.00m 000.00,. 0:0 30:02.. 000.00.03.00 00:58.00:

0.000 000.0 .2000 E39680 00.0 00 BEsEEoo 000000t0>Eob0E 00 30v m_m>_0:0 00000000000080 .m.~ 0.005

 

 

 

 

m: 0: 0: 0: m: 0: m: m: N <0
0: 0: m: m: m: 0: m: 0: F <0
«la. NH and M0...- 010 Md «0: Mina.
0.. 00 .000 v.80 .000 v.00 .000 v. 00.0. 0: 0: .000 v.0~.0 0 <0
0: 0: 0: 30.0 v. 00.0 0: m: 0: m: .. <0
mm 0% 5n. am .0“. 5|... 01mm 5|“. 0.5..
0: 0c 0: 0: 0c 0: w: m: N <0
0: 0: m: m: 0: m: 0: 30.0 v. 3.0- .. <0
mmm 000 «00 mum «on «00 ~00 mum
0.. 0.. :00 v. 00.0- :00 v. 50- :00 v. 050 0.. 0.. 0.. 0 <0
:00 v. 00.0- . 0: 00 00 0.. :00 v. 050- 0: .000 v. 00.0 - 0 <0
0.0.“.- Gl... Flo... 5|“. ,0“. an viola. am 0.0.820 3.0.5
Em m Em 0 E0 0 Em 0 «Burg
0:...0>_m. 05.50200 05.303000
000.0 Eou0>000w

 

00:000.... 020:. :. 2&0. 05.00.350.253 Am. 0:090:00. 0.000
000.00: 0:0 0:002. 05:03.00. 0:50:00: 0.000 000.0 0.000 E000>0000 00. .0 .0205 a. 0:0..0.0..00 0:005:90 .00.00_.0> .003... .0>00
.05. 003-20. 80.. 0:0 m0:.0> 3.0.8000 .902, .0 0.m>_0:0 0:059:00 00.055 0:. Eo... 00.03200 30.. 0:0 000. 00:0:00E00 .0905...

0.2 00.: 0:. 0:0 AN <0 0:0 0 <0. 0.0205 00:00:800too 0000 .0 00x0 0.2 .0... 0... :00500 30.0 v Q. 0:200:00 :00.00n. KN 0.00.0

47

 

:0006 0020000

 

 

02050.00 :. 0000.000
0.0000 03.000 005500.02. .0030 0:0 3.0.0.0. - 0 0 3090.500
0.0000095 .00000000 0090000000
.0.0. :. 0000.05
05.030
. 0.:0EE0
000.500.0500 0.:000 .
. . . .000 0.2.8008
00000.50... 00.0.00... 000..00:0.0z 00000000 00:0.00 - N .. 300.0505
000__0__8_.. 00:60.31 0. 00.00.80
.000_0.0:0_0 00008.53. 0.:0EE0
0326.00 0002.00 0.500.300 .0000 at . 0:0 3.2.2.050 - N F .:0m.0:.m.
0000.00 :. 0000.05
00500.00
002.08 0.:0EE0
_ 00E0M0...00.... 0.3.90.0. 0.000006 0:0 .3_>..000:00 + N . .:0m.0E0aw
000.005: .. 00500.... 00000.00 00002.00... 00:0...0
. . . . :. 0000.000
000.005.}
. . 0.000 0 5.0: :00 :0
0005.200 0.0>.0> 0000200 m0...n000.0m 00.....0h0 :0. 000M000 - 0 P .:0m.0Em
03.00.0000 000.00.... . . . .
0530000..
00000000 .. 00... 005000... + 0x0... 00.00..0> 00.8000 <00 8 0.x< 0.x< 000.00... 000.0
00000000 <0 00
<0
00:0000.

A. 02.000: .0 .+. 02.08 0 05.0 0>00 0x0. 00.0: 0.0 00 0:. :. 000:0:0 000:. 0. 0:800. .00. 0x0. 0.0.00:0>:.0.00:.
0:0 00.0.. 0.0 .c0..00..0 00.0.8000 0:. 0:0. 00.0 00 0000 0. 00.0.9.8 00.00..0> .00.:.0:0 :.0.: 0: H .850:

0:0 000.0 0:000... :000 :. 02.000: .0 02.000. 00:0000. 0000 .0 5:00.... 0:. 0:0 00.000 20. 0.0005 00:00:0000..00
0:0 00.000 .000. m.m>.0:0 .:0:00:.00 .00.0:..0 0:. :00500 30.0 v 0 0:0..0_0..00. m0.:m:0..0_0. .:00...:m.w .mN 0.00..

48

 

 

 

 

 

9.0 0...: NNd 09.0 09.0 0.3 3.0 3.0 05.0...
mod mod ...o 00.0 nod mod 00.0 mod N <0
ﬁo ...o Nﬁc mod Nto to N_..o mod :0
«N... 0N... NNd 0N... on... 00.0 00.0 m v.0 ._<.PO._.
00.0 3.0 .0 00.0 0.0 Ned 3.0 mod N <00
m to 3.0 Nﬁo 9.0 No 0N... 3.0 00.0 .400
0000..0> 9..
":0 m “=0 m 0.0 m “:0 m
00..0>.m 00.50200 0050000..
000.0 E0u0>000m
W 0.000

0000 .0. 00.0.. 0.0 00x0 <0 0:0 <00 020 00.... 00. >0 00:.0.0x0 00:0..0> 0:00.00 00 0. .00.00..0> 0.0.0.000 .0002, >0 00:.0..0:00
00210.00 00 0. 0000.002, 00.0:. 0. 20.0. m:..0o=0:0m.0E0:0 0w. 0:00.900. 0.000 000.000 000 00002. 0:00:00 0:00:00:
0.000 000.0 0.000 0.0.0.0000 00. .0 >..:::.:.00 000.000.026.000. .0 200. 0.0>.0:0 00:00:0000000 .00.:0:00 .mN 0.00...

49

 

|_'_—

Scale

Ecosystem

AII MRW wetlands
n = 57

Class { i \

Lacustrine Palustrine Riverine
Wetlands Wetlands Wetlands

"=14 "=25 "=18

Habitat x F
Emergent Emergent
n = 12 V V n ' 15
Submergent! Submergent!
Floating Floating
“=14 n=15

Emergent Submergent!
n = 23 Floating

n=21

 

 

 

 

 

 

 

 

I

 

Figure 2.1: Schematic illustration of three spatial scales (ecosystem, class
and habitat) and levels within each scale that were analyzed in this study.
The class scale is composed of three groups: lacustrine, palustrine, and

riverine wetlands. The habitat scale is composed of two groups: emergent
and submergent/ﬂoating habitats.

50

0.2

 

 

 

 

 

 

 

 

-6 —4 -2 0 2
d) O

o .- 2

- -_ .. 0
—2
—4

COND ¢
PH _6
-0.6 :0.4 .02 0 (5.2
Comp. 11

Figure 2.2: Principal component analysis of nine water chemistry
variables and six land-use land-cover variables at the ecosystem level
(n = 50 emergent sites), Each site is denoted by an open circle. (A)
Water chemistry variables include pH, conductivity (Cond), dissolved
oxygen (DO), turbidity (Turb), total phosphorus (TP), ammonia (NH3),
soluble reactive phosphorus (SRP), chloride (CI) and mean depth
(Depth). (B) Land-use land-cover variables include relative
proportions of urban, rangeland (Range), agriculture (Agricul), forest,
water and wetland.

5|

Figure 2.2 (cont’d)

 

 

 

 

 

 

 

 

 

 

B
—4 -2 O 2 4
0.4
-4
0.2 .
2
0 O
Wetland
l —2
Forest
-O.2
-4
-0.4
$0 4 —0 2 0 0 2 -0 4

52

Figure 2.3: Principal component analysis of (A) nine water chemistry
variables and (B) six land-use land-cover variables at the ecosystem
level (n = 50 emergent sites), with each site denoted by a letter
representing the marsh class. (A) Water chemistry variables include
pH, conductivity (Cond), dissolved oxygen (DO), turbidity (Turb), total
phosphorus (TP), ammonia (NH3), soluble reactive phosphorus
(SRP), chloride (CI) and mean depth (Depth). (B) Land-use land-
cover variables include relative proportions of urban, rangeland
(Range), agriculture (Agricul), forest, water and wetland.

53

 

l

 

 

 

 

 

L

 

 

 

 

 

COND i
L = Lacustrine
P = Palustrine PH
R = Riverine
I T l i ﬁﬁ
-0.6 -O.4 -0.2 0 0.2
Comp. 1

54

 

Figure 2.3 (cont'd)

Comp. 2

B.

-4

 

0.4

0.2

 

 

 

 

 

 

 

 

 

 

p____
0 Agncul
P Wetland
P l
Range
_02 Forest
L = Lacustrine 1‘
P = Palustrine 5
R = Riverine l
—O.4 l
—0.4 —0.2 0 0.2 -0.4
Comp. 1

55

-4

Emergent I: Submergent

 
  

  

      

 

 

3500-
8
g 3000-
15 ‘6?
g a. 2500-
0 E
f, g 2000-
'5 w
5 5 1500-
e a
g s, 1000-
.5 a
e V 500-
g. A A
Ecosystem Lacustrine Palustrine Riverine

s: Emergent :lSubmergent

 

    

 

   

Ecosystem Lacustrine Palusrtine Riverine

Macroinvertebrate Taxa Richness

Figure 2.4: Mean(SE) values for (A) macroinvertebrate
abundance and (B) taxa richness for each habitat (emergent
and submergent/ﬂoating) from marsh wetlands at the
ecosystem and class scales (lacustrine, palustrine, and
riverine). Signiﬁcant differences in habitat (p < 0.05, denoted
by *) were calculated by ANOVA at the ecosystem and each
class scale. Tukey HSD mean comparison results were
shown among classes at each habitat; different letters

56

CA2

 

 

 

 

 

 

1 P

2 l

PP P‘9o"“'.'.ﬁ.-.'.O, '

P '3.“ O’o :

9 Z

P 0. pp R o

l o‘
O. P
RP ‘3 n R
O
i: P .~ P
P O. R
L: p L
0 P .
P 0. p
P P P: "a R
P .'
. L R U .
- i
-2 : L R ;
L = Lacustrine :L R L L .0. i
P = Palustrine . R 3 ;
R = Riverine '. R R L '- ,’ '.
‘0 L L o.
9 G.
‘4 ‘2 .0. 0 2 ....4
0 O
o. ) . o ’

CA1

Figure 2.5: Correspondence analysis of 57 macroinvertebrate
taxa at the ecosystem level in the emergent habitat (n = 50
sites), Each site is denoted by a letter representing the marsh
class. Sites separate into a palustrine dominated group (long-
dashed circle) and lacustrine and riverine dominated group

(small-dashed circle).

57

 

o""ln
.c n,.

9

 

 

 

 

 

 

 

 

 

 

O O.
': L = Lacustrine : R R '0,
' P = Palustrine : L L .9.
R = Riverine - - o.
'- R ‘ PH ’.
P
P
P L t
p :
i
P
P Pp
—3 -2 -1 O l 2 3
CA 11

Figure 2.6: Canonical correspondence analysis of 57
macroinvertebrate taxa from 50 sites in the emergent habitat
and at the ecosystem level. Each site is denoted by a letter
representing the marsh class. Sites separate into a palustrine
dominated group (long- dashed circle) and lacustrine and
riverine dominated group (small-dashed circle). The graph has
been constrained by water chemistry variables that include pH,
conductivity (Cond), dissolved oxygen (DO), turbidity (Turb),
total phosphorus (TP), ammonia (NH3), soluble reactive
phosphorus (SRP), chloride (Cl) and mean depth (Depth).

58

Chapter 3: Utility of Pro-existing Macroinvertebrate Metrics in MRW

Wetlands at Multiple Spatial Scales

59

Chapter 3: Utility of Pre-existing Macroinvertebrate Metrics in MRW

Wetlands at Multiple Spatial Scales

Abstract

Methods to monitor the ecological health of wetlands are needed to
protect these ecosystems from further human disturbances. One method of
interest has been the use of biological indicators in the assessment of water
quality. Macroinvertebrate communities are responsive to environmental change
and thus provide a good indication of water quality in other aquatic habitats
making their potential use in wetlands a logical and practical objective for wetland
management and protection. Wetland macroinvertebrate metrics sensitive to
disturbance have been determined for some wetland types and regions, however
the use of these methods or metrics in all wetlands is relatively unknown. The
objective of this study was to compare the utility of these pre-existing
macroinvertebrate metrics in wetlands, speciﬁcally inland marshes in the
Muskegon River Watershed (MRW) Michigan at multiple spatial scales, including
the class scale (lacustrine, palustrine and riverine marshes) and habitat scale
(emergent and submergent vegetation zones). Macroinvertebrates were
collected from impacted and reference wetlands of the MRW and 50
macroinvertebrate metrics were calculated and compared between groups at the
habitat scale Within each class. Of all metrics tested in this study, 27 had at least

one signiﬁcant difference (p < 0.10 or p < 0.05) in one class and habitat

60

combination, yet, those combinations that were signiﬁcant varied from metric to
metric. Few metrics were sensitive in all wetland classes or habitats, suggesting
that these pre-existing macroinvertebrate metrics did not work equally well in

MRW wetland assessments at different scales.

Introduction

Wetlands are threatened by various human activities (e.g., development,
agriculture) and as a result have declined in quantity and quality over the years.
Recent interest in protecting wetlands, including those still remaining, restored
and newly created wetlands, have resulted in the development of assessment
methods that can monitor the quality and identify degraded wetland ecosystems
(van Dam et al. 1998, Rader 2001). One method of interest is the use of biota as
indicators of wetland quality. Biological organisms, and thus biological
measurements, can respond to short and long-term changes in the environment,
providing temporal information useful in evaluating the integrity of a water body in
an assessment (Rosenberg and Resh 1993). Organisms that respond to these
changes in a predictable manner can act as biological indicators and are used to
establish biological thresholds or criteria and are monitored in assessments. The
success of these methods in other aquatic ecosystems (e.g., streams and lakes),
in addition to the rapid, reliable and cost efﬁcient manner in which these methods
are conducted, make the concept of biological assessments in wetland
ecosystems a logical and practical objective for wetland management and

protection.

61

Macroinvertebrates have been useful as biological indicators in aquatic
ecosystems due to their fundamental role in energy transport between primary
producers and higher trophic levels in food webs. Shifts in macroinvertebrate
structure could have cascading effects on other levels of the food web, and in
return, shifts at other levels could have a direct inﬂuence on macroinvertebrate
communities (Wrssinger 1999, Helgen and Gernes 2001). Because of this
signiﬁcant connection, macroinvertebrates have been useful as management
tools in the evaluation and monitoring of aquatic habitats. Additional reasons for
using macroinvertebrates are (1) their ubiquitous nature, (2) range of response to
environmental stressors by different species, (3) relative stationary nature, (4)
their relatively long life cycles and (5) the ease in which macroinvertebrates are
collected (Rosenberg and Resh 1993). For these reasons macroinvertebrates
are the most frequently used organisms in aquatic assessments (Hellawell 1986).
All of these advantages apply to most aquatic habitats; however, some do not yet
apply to wetland ecosystems. For example, Batzer et al. (2001) maintains that
many wetland macroinvertebrates are mobile, some life cycles can be quite short
(1-2 months), and that much information is still unknown for certain taxa (e.g.,
response to stressors and taxonomy), thereby, limiting the use of
macroinvertebrates in wetland assessments.

Several studies, however, have successfully used macroinvertebrates as
biological indicators of wetland quality. Burton et al. (1999) developed a
preliminary index of biotic integrity for coastal Great Lake wetlands based on

macroinvertebrate metrics. This study was later supported by work in Lake

62

Huron costal wetlands by Kashian and Burton (2000) and later validated by
Uzarski et al. (2004) in coastal wetlands of Lake Huron and Lake Michigan
wetlands. Helgen and Gernes (2001, 2002) developed an index of biological
integrity for macroinvertebrates in large depression wetlands in Minnesota, and in
Montana, Apfelbeck (2001) found macroinvertebrates to be useful in evaluating
the biological integrity of perennial wetlands. In Australia, Chessman et al.
(2002) successfully derived and tested a new biotic index for the Swan Costal
Plain using macroinvertebrate taxa diversity. Merritt et al. (1996, 1999, 2002b)
found that macroinvertebrate functional-group metrics, instead of taxonomic
based metrics, could be used in the assessment of wetlands, speciﬁcally in river
oxbows and ﬂoodplain ecosystems in Florida. Other studies, however, have
found macroinvertebrates to be non-responsive to human disturbances, making
their ubiquitous use as biological indicators questionable in wetlands (Wilcox et
al. 2002, Tangen et al. 2003).

Multiple methods or metrics that assess macroinvertebrate communities
have been developed and utilized in aquatic ecosystems; however, few
comparisons of these methods have been made (Ravera 2001), especially in
wetland ecosystems. Although many of these assessments have been
successful in certain types of wetlands (e.g., coastal, depressional) and in
speciﬁc regions (e.g., Great Lakes, Minnesota), the general use of these metrics
in all wetlands is relatively unclear. As a result, the main objective of this study
was to compare the utility of these pre-existing macroinvertebrate metrics in

wetlands, speciﬁcally inland marshes in the Muskegon River Watershed (MRW)

63

Michigan. We also tested the utility of these metrics at multiple spatial scales. In
general, the following questions can be answered from this study: (1) do existing
macroinvertebrate metrics work in MRW wetlands; (2) how do these metrics
compare; (3) do taxonomic or functional metrics work best in MRW wetlands; and
(4) do the utility of these metrics vary with different spatial scale? We
hypothesized that these pre-existing macroinvertebrate metrics work equally well
in MRW wetland assessments at all scales. The results from this study will
provide more information on macroinvertebrate communities in wetland
ecosystems, how they respond to changes in the environment, and in the

development of successful wetland assessment protocols in the MRW.

Methods

Mocation ancL Scale

The Muskegon River Watershed (MRW) begins in the north-central portion
of Michigan's lower peninsula, and runs in a southwest direction to the shoreline
of Lake Michigan. The watershed is approximately 7000 km2 in size and
includes 94 tributaries, 183 stream segments, hundreds of lakes and wetlands
within 11 different counties (Torbick et al. 2006). Dominant land use varies from
forest dominated to agriculture/urban dominated areas within the upper and
lower regions of the watershed, respectively (Lougheed et al. 2007). The
continuum of the MRW has been altered by 95 dams, and has been impacted by

human inﬂuence through logging and agricultural byproducts and is predicted to

have a 50% increase in urban land use within the next 35 years (Pijanowski et al.
2006). For all of our studies, inland marsh wetlands (n = 57) were selected from
in and near the MRW; most sites (~42) were randomly selected, whereas
remaining sites were purposely selected as described by Lougheed et al. (2007)
to ensure a gradient of sites with variable water quality in the study. Of these
sites, a subset (n=30) were selected for this study based on wetland class and
impact (see below). All sites were surveyed prior to the ﬁeld season to conﬁrm
wetland existence, accessibility, and to obtain landowner permission if
necessary.

To achieve our objectives, two spatial scales were designated for this
study: the class scale and habitat scale. Three wetland classes were sampled:
palustrine marshes (depression, n=12), riverine marshes (n=10), and lacustrine
marshes (n=8). Each class was then further divided into the next scale based on
habitat, which separated each marsh into distinct vegetation zones (habitats).
When present, two zones were sampled in each marsh, the emergent and
submergent/ﬂoating (now labeled as only submergent) vegetation zone. The
ﬂoating and submergent zones were combined due to little difference in
macroinvertebrate composition (T. Burton, personal communication), and the .
inability to clearly distinguish and separate these zones in many wetlands.

In order to test the utility of macroinvertebrate metrics, wetlands within
each class were divided into a reference and impacted group. These groups
were separated based on water chemistry measurements representing mainly

agricultural and urban disturbance. Turbidity and eutrophication were ranked as

65

two of the top three mechanisms of impact for wetlands in the Great Lakes
ecoregion (Detenbeck et al. 1999), and thus sites in this study were separated to
represent this. Statistical tests comparing reference and impacted sites (t-tests
for normal data and Wilcoxon tests for nonparametric data) conﬁrmed differences
in water chemistry. Lacustrine sites were signiﬁcantly different in total
phosphorus (ref = 11.9 ug/L, imp = 56 ug/L, t-test: t = 3.15, DF = 6, p < 0.01),
turbidity (ref = 1.43 ntu, imp = 4.62 ntu, t-test: t = 2.75, DF = 6, p < 0.05) and
soluble reactive phosphorous (ref = 3.23 ug/L, imp = 16.6 ug/L, Wilcoxon test: 2
= -2.16, S = 10, z < 0.05). Palustrine sites were signiﬁcantly different (p < 0.05)
in total phosphorus (ref = 15.6 ug/L, imp = 94.3 ng/L, t-test: t = 10.39, DF = 10, p
< 0.0001), chloride (ref = 30 mg/L, imp = 76.8 mg/L, Wilcoxon test: 2 = -2.16, S =
25, p < 0.05), soluble reactive phosphorous (ref = 1.2 ug/L, imp = 15.7 pg/L,
Wilcoxon test: 2 = -2.01, S = 26, z < 0.05) and % surrounding wetland land-use
(ref = 12%, imp = 4%, t-test: t = -3.08, DF = 10, p < 0.01). Conductivity, turbidity,
and % agriculture surrounding land-use were also signiﬁcantly greater at a more
relaxed level (p < 0.10), at impacted sites compared to reference palustrine sites.
Riverine sites were signiﬁcantly different in dissolved oxygen (ref = 7.73 mg/L,
imp = 2.9, t-test: t = -2.79, DF = 8, p < 0.05), turbidity (ref = 1.43 ntu, imp = 4.62
ntu, t-test: t = 2.82, DF = 8, p < 0.05) and total phosphorus (ref = 8.9 ug/L, imp =

48.17 ug/L, t-test: t = 6.46, DF = 8, p < 0.0001).

Sample Collection

Inland marshes were sampled once between July and early August in
2002 and 2003. At each marsh site, three evenly spaced transects were
established perpendicular to one side of the wetland, thus extending from the
wet—meadow or shore into deep-water habitats. Along each transect, multiple
random sampling points (marked by a quadrant) were established to facilitate
sample collections from each habitat zone. All macroinvertebrates were
collected from the ﬁrst random point in each habitat zone. To ensure no
disturbance in the macroinvertebrate community due to other sampling activities,
macroinvertebrates were collected at a random angle, 1—2 m from the main
sampling location (quadrant). One macroinvertebrate sample consisted of two
sets of three sweeps using a standard D-net (500 um) and subsequently rinsed
through a 500 um sieve to remove large pieces of vegetation. All three transect
samples from the same habitat were combined into a single composite sample
and preserved in 100 % alcohol for laboratory identiﬁcation. Because each
marsh was sampled only once, a maximum total of two composite samples were
collected from each marsh site (Halse et al. 2002). A total of 30 emergent
composite samples and 30 submergent composite samples were analyzed in this
study.

In the laboratory, macroinvertebrate composite samples were again
sieved and sub-sampled to reduce processing time. In the sub—sampling
protocol, each sample was homogenized and divided into two equal proportions

or sub-samples. All macroinvertebrates from one sub-sample were sorted and

67

identiﬁed to the lowest practical taxonomic level using Merritt and Cummins
(1996), Thorp and Covich (1991), Pennack (1989) and Larson et al. (2000).
Most macroinvertebrate taxa were identiﬁed to the generic level, except for
taxonomically difﬁcult organisms such as the Chironomidae, Oligocheata, and
Hirudinea. For other taxa, large numbers of immature specimens were collected
and could only be accurately identiﬁed to the family level. All taxa from these
groups (Planorbidae, Corixidae) were grouped at the family level for this study.
Microinvertebrates, including Cladocera, Copepoda and Ostracoda, were not
considered due to the sampling design in this study. Macroinvertebrate counts
were subsequently adjusted to account for subsampling procedures prior to

statistical analyses.

Metric Calculation and Statistical Analysis

Macroinvertebrate metrics that have been utilized in previous wetland
assessments can be divided into ﬁve general categories: macroinvertebrate taxa
richness, composition, tolerance, functional feeding, and diversity metrics (Table
3.1). All of these metrics (Table 3.1) have been found to respond predictably to
changes in some type of wetland environment. Richness measures assess the
number of taxonomically distinct individuals or groups of individuals within the
community. Often taxonomic groups known to be sensitive to anthropogenic
changes in aquatic environments are monitored through these measures (Burton

et al. 1999, Kashian and Burton 2000, Helgen and Gernes 2001, Gernes and

68

Helgen 2002, Uzarski et al. 2004). It should be noted, however, that taxonomic
resolution can affect the ability of these metrics to detect impairment, with more
detailed taxonomy (e.g., species level) providing more accurate detection (King
and Richardson 2002). Composition and tolerance metrics measure the relative
proportion of macroinvertebrate taxa abundance within the community and often
focus on tolerant and intolerant taxa to various stressors (Table 3.1) (Burton et al.
1999, Kashian and Burton 2000, Helgen and Gernes 2001, Gernes and Helgen
2002). Functional analyses rely on the separation of invertebrate taxa by
functional relationships (e.g., similar feeding morphology). These functional taxa
are then placed into ratios that provide surrogates of habitat assessment and
overall ecological condition (Table 3.2) (Merritt et al. 1996, Merritt et al. 1999,
Cummins and Merritt 2001, Merritt et al. 2002a). Another common category are
diversity measures, which assess the number of species (or lowest taxonomic
level) within a standard area (e.g., Simpson index and Shannon's index)
(McCune and Grace 2002). All metrics listed in Table 3.1 were calculated in
each class for each habitat (e.g., emergent lacustrine) of this study.

Statistical analyses included non-parametric Wilcoxon tests that were
conducted between impacted and reference sites (signiﬁcance level of p < 0.05
or p < 0.01) at each class and habitat combination. Non-parametric tests were
utilized due to smaller sample sizes, and the use of a majority of proportion data
that was not normally distributed (Kashian and Burton 2000, Quinn and Keough

2002). All statistical analyses were conducted with JMP IN 5.1.2.

69

Results

A total of 105,081 macroinvertebrates were collected from inland marsh
wetlands of the Muskegon River Watershed (MRW). At the class scale, the
majority of individuals were from palustrine sites (55%, n = 12), followed by 27%
from riverine sites (n = 10) and 18% from lacustrine sites (n = 8). At the habitat
scale, about 58,174 (55%) and 46,907 (45%) individuals were collected from the
emergent and submergent zones, respectfully. Of these individuals, the class
Insecta was the most abundant major taxonomic group, followed by the phylum
Mollusca and class Crustacea. Of the 50 metrics that were tested in this study
(Table 3.1), 27 tests had at least one signiﬁcant difference (p < 0.10 or p < 0.05)
in one class and habitat combination.

For the taxa richness metrics, 8 of the 12 measures detected impairment
in at least one class and habitat combination. For the total number of taxa, there
were signiﬁcantly greater taxa at both the generic and family level in the
emergent habitats of impacted riverine wetlands (Table 3.3); however, no
differences in total taxa numbers were observed in any other class or habitat.
Taxa richness measures including sensitive taxa only, such as the
Ephemeroptera, Odonata and Trichoptera, detected impairment in only riverine
and lacustrine wetlands. The number of Ephemeroptera taxa and the number of
Trichoptera taxa independently and combined were signiﬁcantly greater in the
submergent habitats of reference riverine wetlands (Table 3.3). In lacustrine

wetlands, the“ number of Odonata taxa was signiﬁcantly greater in both emergent

70

and submergent habitats at reference sites (Table 3.3). The number of Odonata
and Trichoptera Taxa combined and the number of Ephemeroptera, Trichoptera,
and Odonata (ETO) taxa combined were signiﬁcantly greater in the submergent
zone of only lacustrine wetlands (Table 3.3). It should also be noted that
differences in these taxa richness measures were observed in only lacustrine
and riverine wetlands; however, no differences were found in palustrine wetlands
(Table 3.3).

Of the 10 composition metrics, 7 measures detected impairment in at least
one type of wetland class and habitat. The only metric to have signiﬁcantly
different results between impacted and reference wetlands in all three classes
and habitat types was the metric measuring the relative proportion of Gastropoda
(Table 3.3). The lacustrine wetlands had a higher proportion of Gastropoda in
reference sites compared to impacted sites; however, the opposite was observed
in the palustrine and riverine wetlands, with the impacted sites having a higher
proportion of Gastropoda compared to reference (Figure 3.1). This same trend
was observed in both emergent and submergent habitats (Table 3.3). The %
Diptera and % Chironomidae also were signiﬁcantly greater in reference sites
compared to impacted in both habitats of palustrine and emergent only habitats
in riverine wetlands (Table 3.3). The % Trichoptera were greater in both habitats
of reference palustrine wetlands, whereas as % Mollusca + Crustacean were
signiﬁcantly lower in emergent zones of reference palustrine wetlands (Table

3.3).

71

Of the 13 tolerance metrics, 9 were signiﬁcantly different in at least one
type of wetland class and habitat. The lsopoda, generally tolerant taxa, were
found in greater abundance in impacted wetlands compared to reference riverine
wetlands, but only in the emergent habitats. Also, for the taxa Physa, the relative
proportion was signiﬁcantly greater in impacted sites of only submergent zones in
palustrine and emergent and submergent zones in riverine wetlands (Table 3.3).
Physa is also considered a tolerant taxa (Cooper et al. 2006). The dominant
metrics (% dominant taxa, % two most dominant taxa, % 3 most dominant taxa)
were greater in reference wetlands suggesting that dominant taxa were in greater
abundance at these sites; however, each signiﬁcant metric occurred in only one
class and habitat combination and each combination among wetland class was
different, indicating no clear pattern across the class or habitat scale (Table 3.3).

Of the 13 functional feeding metrics, 9 detected impairment in at least one
wetland class and habitat. A greater proportion of piercers, scrapers and total
shredders were observed in impacted wetlands compared to reference wetlands
(Table 3.3). For scrapers this difference was observed in both habitats of
palustrine wetlands and for piercers and shredders this difference was observed
in only emergent habitats of riverine wetlands. The photosynthesis to respiration
(P/R) ratio was signiﬁcantly greater in both habitats of palustrine impacted
wetlands compared to reference wetlands, indicating that impacted sites are
more autotrophic systems and reference sites are more heterotrophic systems
(Table 3.2 and 3.3; Figure 3.2A). Differences in the amount of particulate organic

matter, including the coarse particulate organic matter to ﬁne particulate organic

72

matter (CPOM/FPOM) ratio and the suspended particulate organic matter to
benthic particulate organic matter (SPOM/BPOM) ratio, were signiﬁcant in
emergent habitats only. The CPOM/FPOM ratio was signiﬁcantly greater in
impacted palustrine and riverine sites, indicating more shredder species at these
sites; however, all ratio values were still below the criteria (> 0.25) for normal
shredder riparian systems (Table 3.2 and 3.3; Figure 3.28). The SPOM/BPOM
ratio was also signiﬁcantly greater in emergent habitats of impacted palustrine
wetlands, indicating more suspended material in these sites; however, again, all
of these values were below the enrichment criteria (> 0.50) (Table 3.2 and 3.3,
Figure 3.2C). The ﬁrst habitat stability metric, based on functional feeding
groups, indicated that signiﬁcantly more stable habitat was available in impacted
sites compared to reference sites in both habitats of palustrine wetlands (Table
3.2 and 3.3; Figure 3.20). The second habitat stability metric, based on
macroinvertebrate habits, indicated the same results of more stable substrates in
impacted palustrine wetlands in addition to emergent riverine habitats; however,
it should be noted that a signiﬁcant opposite results was found in lacustrine
wetlands, with reference wetlands having more stable substrates (Table 3.2 and
3.3; Figure 3.2E). The benthic food ratio was signiﬁcantly greater in reference
wetlands compared to impacted wetlands in emergent palustrine wetlands only,
indicating these wetlands have better food availability for wading birds and
benthic ﬁsh, however, all values were still below the established criteria (> 0.60)

for good food supply (Table 3.2 and 3.3; Figure 3.2F).

73

Discussion

Numerous studies have begun to investigate the use of
macroinvertebrates as biological indicators of wetland ecosystem health; yet
many questions remain regarding the use of this approach in all wetlands. The
main objective of this study was to compare the utility of these pre-existing
macroinvertebrate metrics in detecting impacted verses reference wetlands,
speciﬁcally within inland marshes in the Muskegon River Watershed (MRW)
Michigan. In this study, impacted wetlands were inﬂuenced by agricultural
disturbance with increased total phosphorus and turbidity. We also tested these
metrics at two spatial scales, the class scale (lacustrine, palustrine and riverine
wetlands) and the habitat scale (emergent and submergent habitats). Although
this study has focused on only macroinvertebrate communities, other studies
have also begun to investigate the use of other biota as potential indicators of
wetland quality, including plants (Helgen and Gernes 2001, Wlssinger et al.
2001, Gernes and Helgen 2002, Lopez and Fennessy 2002, Lougheed et al.
2007), birds (Brown and Batzer 2001, Rivers and Cable 2003, Hierl et al. 2007),
fish and amphibians (Simon et al. 2000), zooplankton (Lougheed and Chow-
Fraser 2002, Lougheed et al. 2007), algae (Yangdong and Stevenson 1996,
Stevenson 2001, Lougheed et al. 2007), and bacteria (Lemly and King 2000,
McArthur 2001, Merkley et al. 2004).

In general, differences in the macroinvertebrate community were observed
between wetlands with dissimilar water quality. Of the 50 metrics that were

tested in this study (Table 3.1), 27 metrics had at least one signiﬁcant difference

74

in one class and habitat combination, yet, those combinations signiﬁcant varied
from metric to metric (Table 3.3). Few metrics were sensitive in all wetland
classes or habitats, suggesting that that these pre-existing macroinvertebrate
metrics did not work equally well in MRW wetland assessments at different
scales and many of these metrics used in other studies (e.g., diversity indices,
abundance values) were not sensitive at all in our study (Burton et al. 1999,
Kashian and Burton 2000, Apfelbeck 2001, Uzarski et al. 2004). However, prior
to a more thorough discussion of these metrics it should be noted that these
results could be a true representation of the differences and natural variability in
macroinvertebrate community response or other factors in the study design that
might have had a signiﬁcant role. For example, we cannot prove that observed
differences were caused by chemical and agricultural differences alone, many
other stressors not included in this study could also directly or indirectly have
inﬂuenced these communities. Also, some MRW wetlands (or classes and
habitats) classiﬁed as impacted in this study may not have stressor levels
necessary to affect wetland macroinvertebrate communities. In wetlands, natural
harsh conditions have made many wetland taxa tolerant to some degree of
variability in the surrounding environment (King and Richardson 2002). Wetlands
in more developed or disturbed areas outside the MRW may be at or above
stressor threshold levels and subsequently have changed macroinvertebrate
communities. If true, this would suggest that some macroinvertebrates metrics
are not good biological indicators of early wetland degradation or stress. This

also would suggest that signiﬁcant metrics in this study may be the most

75

sensitive macroinvertebrate metrics in the MRW and as more development
occurs in this watershed, these metrics may become more useful. Other factors
in our sampling design, such as single sampling dates, number of sites, time of
sampling, number of samples collected, etc. could also have inﬂuenced these
results. More research, especially on signiﬁcant metrics from this study will be
needed prior to the use of any metrics in biological assessments in the MRW.

In general, most wetland assessments have found total macroinvertebrate
taxa richness to decrease in impacted wetlands (Burton et al. 1999, Kashian and
Burton 2000, Helgen and Gernes 2001, Uzarski et al. 2004). In our study, no
differences in the number of taxa were observed between impacted and
reference sites except for an increase in taxa richness (at both the generic and
family level) in emergent habitats of riverine wetlands. No differences observed
in some wetlands may be an indication that impacted wetlands in the MRW are
not at levels that affect macroinvertebrate communities, as previously discussed.
However, it should be noted that although the number of taxa may not differ, the
taxa present might; sensitive taxa in reference wetlands could be replaced with
new tolerant taxa in impacted wetlands. Differences (and lack of differences) in
taxa number in our study and other wetland studies could be due to the level of
taxonomic resolution used. For example, a study by King and Richardson (2002)
found that family-level identiﬁcation of wetland macroinvertebrate communities
were not capable of detecting impairment and suggested that genus or species-
level identiﬁcation be utilized, especially within the dominant Chironomidae

family. Most wetland assessment studies have identiﬁed Chironomidae to

76

various levels including the family (Uzarski et al. 2004), sub-family or tribe
(Kashian and Burton 2000, Uzarski et al. 2004), generic levels (Apfelbeck 2001,
Helgen and Gernes 2001, Gernes and Helgen 2002) or species level (King and
Richardson 2002). Further identiﬁcation of the Chironomidae in this study may
provide for better detection between reference and impacted sites with taxa
richness metrics. However this type of identiﬁcation, especially with the
Chironomidae or other taxa (e.g., Annelida) can be difﬁcult, require expertise,
consume large quantities of time and be more expensive (King and Richardson
2002). Also, for some taxa identiﬁcation keys in certain regions may not be
available; making the utility of these metrics in rapid assessments questionable.
A few studies also have found taxa richness to increase in response to
disturbance. Rader and Richardson (1994) found increased taxa richness in
nutrient enriched wetlands in the Everglades and King et al. (2000) found
increased taxa richness in forested wetlands closer to highway disturbance. In
these studies, disturbance provided more variety of habitat (e.g., more plants,
structure); in the MRW riverine wetlands, the increased nutrient levels may
directly and indirectly cause macroinvertebrate taxa to increase due to more
habitat and food resource availability as suggested by the intermediate
disturbance concept (Townsend et al. 1997).

Certain taxa richness metrics detected signiﬁcant differences in the
macroinvertebrate community between impacted and reference sites in lacustrine
and riverine wetlands, but not palustrine wetlands. Lacustrine and riverine

wetlands are‘located adjacent to other aquatic habitats (lakes and rivers) and as

77

a result have the potential for waterﬂow through wave action or river movement
that can provide habitat and food resources for more types of taxa. The edge
habitats of these wetlands (e.g., the submergent zone) may support certain
sensitive taxa commonly used in the assessment of these other aquatic habitats.
These taxa may be more sensitive due to the dependence on waterﬁow for
movement of organic material, higher oxygen levels and potential removal of
pollutants. This might explain the reduction in taxa number and sensitive taxa,
including the number of Ephemeroptera, Trichoptera, and Odonata in mainly the
submergent zone of lacustrine and riverine habitats. Similar results in Great
Lake coastal wetlands found taxa richness of Ephemeroptera, Trichoptera and
Odonata to be lower in impacted wetlands (Burton et al. 1999, Kashian and
Burton 2000, Uzarski et al. 2004). Helgen and Gernes (2001, 2002) also found
the number of total taxa, number of Odonata taxa, and number of ETSD
(Ephemeroptera, Trichoptera, Sphaeriidae and dragonﬂies) to be reduced in
impacted palustrine wetlands. The absence of this result in palustrine wetlands
of this study could be due to the use of activity traps by Helgen and Gernes in
collecting movable predators and perhaps differences in taxonomic resolution.
Of all metrics tested in this study only one, the relative proportion of
Gastropoda, could detect impairment in all wetland classes and habitats;
however the response of this metric was different among wetland class. The
relative composition of Gastropoda decreased in impacted lacustrine wetlands,
whereas the relative composition increased in impacted palustrine and riverine

wetlands. At'impacted sites, increased nutrients may result in increased algal

78

growth, which is the main food source for Gastropoda (Kashian and Burton
2000); this subsequently may have increased the relative composition of these
taxa in palustrine and riverine wetlands of this study. The reduction of
Gastropoda observed in impacted lacustrine wetlands could be due to a
reduction in algal food resources, the presence of intolerant Gastropoda and the
inﬂuence of additional stressors in lacustrine wetlands not measured in this
study. However, other studies in coastal Great Lake wetlands, found opposite
results, with impacted wetlands having greater relative composition of
Gastropoda (Burton et al. 1999, Kashian and Burton 2000, Uzarski et al. 2004).
Although this metric was responsive in all wetlands, the variation in response
(positive and negative) based on wetland class suggests that more research is
necessary.

Of all the composition measures, the most signiﬁcant metrics were
observed in both habitats of palustrine wetlands, followed by emergent habitats
of riverine wetlands. The relative proportion of Diptera, Chironomidae, and
Trichoptera were greater in reference wetlands. Chironomidae, commonly the
most dominant taxa in wetlands and the most common Dipteran taxa in our
wetland samples, are generally thought of as tolerant taxa in aquatic
ecosystems, however, other studies have found different sub-families (Kashian
and Burton 2000), genera and species (King and Richardson 2002) within this
family to have variable tolerances to environmental stressors. Perhaps the
reduction observed in this study represents the loss of those sensitive

Chironomidae taxa. Helgen and Gernes utilized a metric assessing the number

79

of Chironomidae taxa (generic level) in palustrine wetlands for this reason;
however, this metric was not utilized in this study due to high taxonomic
resolution required for this family. Kashian and Burton (2000), however, found
the opposite result in Great Lake coastal wetlands, with approximately 20—40%
increase in Chironomidae relative composition. King et al. (2000) found no
differences in Chironomidae composition in reference compared to impacted
forested wetlands. Most Trichoptera are considered sensitive in all aquatic
habitats, similar reductions in the relative composition of this order have been
observed in other wetland studies (Kashian and Burton 2000), however it has not
been included in any assessment protocols. This could be due to the lower
number of Trichoptera generally found in wetland ecosystems (Helgen and
Gernes 2001). In inland lacustrine marshes we found no signiﬁcant differences
in these groups between impacted and reference wetlands.

Of the tolerance/intolerance taxa, the most sensitive metrics were
observed in palustrine and riverine wetlands. ln riverine wetlands, the relative
composition of lsopoda was greater in impacted wetlands. Uzarski et al. (2004)
and Burton et al. (1999) found opposite results with increased lsopoda
composition in reference coastal wetlands. However, Kashian and Burton
(Kashian and Burton 2000) found greater abundance of lsopoda in impacted
coastal wetlands, yet, signiﬁcant differences were small clue to high variability in
collected samples. Increased abundance in our study may be due to increased
nutrients and thus food resources in impacted wetlands. Differences between

our study and coastal wetlands could be due to taxonomic differences or other

80

stressors not included in this study. In both palustrine and riverine wetlands, the
relative composition of Physa was greater in impacted wetlands. Cooper et al.
(2006) found similar results in a drowned-river mouth wetland. The positive
response of these taxa in impactedwetlands could also explain trends observed
in the relative composition of Gastropoda.

The functional wetland metrics developed by Merritt and Cummins (Merritt
et al. 1996, Merritt et al. 1999, Cummins and Merritt 2001, Merritt et al. 2002b)
were made to assess the ability of wetlands to support bird and ﬁsh populations
in Florida. Criteria established by these authors in their assessment protocols
were developed speciﬁcally for that region and may not apply to all wetlands or
stressors in MRW wetlands. Although these metrics may not be sensitive to all
potential stressors, they can still provide an indication of changes in the
ecological condition between reference and impacted wetlands. In our study,
most impacted sites had higher functional group ratio values with overall greater
variability, whereas reference sites had lower functional group ratio values with
little variation among sites (Figure 3.2). This was true except for the benthic food
ratio, in which greater food availability was present in reference wetlands. The
low signiﬁcance value, and large amount of variability between reference and
impacted sites indicated that this metric might not clearly detect wetland
impairment. In addition, the criteria level for this ratio set by Merritt et al. (1996,
1999, 2002), developed and tested speciﬁcally for riverine wetlands in Florida,
was not relevant in this study. Similar results were observed with the

CPOM/F POM and SPOM/BPOM ratio criteria. This indicated that these metrics

81

and the corresponding criteria may not be useful in detecting impairment in
palustrine wetlands or all wetlands in the MRW. Further modiﬁcation and testing
of these metrics would be needed for their use in wetland assessment.

In general, greater variability in impacted sites could be due to
differences in the amount and number of stressors present at a site. The greater
P/R ratios and % scraper values in impacted wetlands suggest that increased
nutrient levels have stimulated more algal growth thus resulting in a more
autotrophic system that can support more organisms dependent on periphyton as
a food resource (Kashian and Burton 2000, Cummins and Merritt 2001, Merritt et
al. 2002b). Greater values of habitat stability observed in impacted wetlands
might again be the indirect result of increased nutrients stimulating plant and
algal growth, which can provide more habitats for macroinvertebrates. Greater
habitat stability was observed in impacted palustrine wetlands and impacted
emergent riverine wetlands; however, in the submergent zone of lacustrine
wetlands the opposite result was observed with greater stability in reference
wetlands. This opposite result was probably due to the lower number of
Gastropoda in impacted lacustrine wetlands, as discussed above. Kashian and
Burton (2000) also found greater autotrophy, stable habitat and proportion of
scrapers in impacted coastal wetlands. Merritt et al. (1996, 2002) found similar
types of ecological conditions in areas with more light penetration; impacted
wetlands in this study may have less canopy cover also providing for increased

habitat through plant and algal growth. Increased values of CPOM and SPOM

82

could also be a result of increased nutrient and plant growth in impacted

palustrine wetlands.

Conclusions

The results from this study suggest that macroinvertebrates may be useful
in the biological assessment of MRW inland wetlands; however, the lack in
consistency of most metrics among wetland class and habitats suggest that the
same metrics may not be useful for all wetlands. In general, this lack in
consistency is due to the high amount of variability observed in the
macroinvertebrate community. This variability is even evident among the
signiﬁcant metrics with the overlap of inter-quartile ranges and medians of
impacted and reference groups (Figure 3.1, Figure 3.2). The use of these
metrics in all wetlands could be misleading and lead to inappropriate
management decisions. The use of signiﬁcant metrics from this study in future
MRW tests may be possible but will require more research. Although no clear
group of metrics were identiﬁed in this study, some general patterns in
macroinvertebrate response among wetland class were observed from palustrine
to riverine to lacustrine wetlands. Speciﬁcally, palustrine and lacustrine metrics
were most dissimilar with wetland macroinvertebrates most sensitive to
composition and functional measures in palustrine wetlands and taxa richness
measures most sensitive in lacustrine wetlands. Riverine wetlands had an

overlap of measures, with emergent riverine macroinvertebrate response more

83

similar to palustrine wetlands and submergent riverine community response more
similar to lacustrine wetlands. These differences could be due to differences in
vegetation; visual observation of these ﬁeld sites would classify both habitats of
palustrine and emergent riverine wetlands as having more dense vegetation,
compared to both habitats of lacustrine and submergent riverine wetlands.
These results suggest that wetland assessment based on wetland class
(lacustrine, palustrine, and riverine) and habitat (emergent, submergent) may not
be the best method of grouping, perhaps general assessment categories based
on amount of vegetation or perhaps by more detailed types of vegetation may be
acceptable for assessment purposes in these inland wetlands (Burton et al.
1999). More research on the macroinvertebrate community and relationships to
the biological community (e.g., vegetation, algal communities) are needed to

validate these measures.

84

Table 3.1: Various macroinvertebrate metrics used in the assessment of wetlands. The
citation column lists studies that have utilized the metric. Each number represents the
following citation: (1) Burton et al. 1999, (2) Helgen 2001, (3) Helgen and Gernes 2002), (4)
Merritt et al. 2002 and 2006, (5) Apfelbeck 2001, (6) Uzarski et al. 2006 and (7) Kashian

 

 

and Burton 2000.
Metric I on
Richness Total taxa (g) 1,2,3,5,6,7
Total taxa (f) 1,6,7
Ephemeroptera (E) taxa (g) 2, 7
Odonata (O) taxa (g) 1,2,3,6
Trichoptera (T) taxa (9) 2,7
Diptera taxa (g) 7
Odonata + Trichoptera taxa 2
Ephemeroptera + Trichoptera Taxa 1,6,7
EOT Taxa 2
POET 2,5
ETSD (E+T+Sphaeriidae+Dragonﬂies) 2,3
Number of Individuals 5,7
Composition Odonata (%) 1,6,7
Trichoptera (%) 2,7
Chironomidae (%) 2,5,7
Diptera (%) 2,7
POET/(POET + Chironomidae) 2
Gastropoda (%) 1,6,7
Sphaeriidae (%) 1,6,7
Amphipoda (%) 1,6,7
Mollusca + Crustacea (%) 1,5,6,7
Corixidae (%) 3
Tolerance! Intolerance Odonata (%) 1,2
Trichoptera (%) 2
Ephemeroptera (%) 2
EOT (%) 2
Oligocheata (%) 2
Gastropoda (%) 1,2,6
Amphipoda (%) 1,2
lsopoda (%) 6
ETSD 2,3
Dominant Taxa(%), Dom. 2 Taxa(%), Dom. 3
Taxa(%) 2.3
Physa snail abundance (%) 2

 

85

Table 3.1 (cont'd).

 

 

Metric Citation
Functional Feeding Collector-gatherers (%) 7
Filterers (%) 7
Piercers (%) 7
Predators (%) 7
Scrapers (%) 7
Shredders (%) 7
Productivity/Respiration (P/R) 4,7
CPOM/F POM 4
SPOM/BPOM 4,7
Habitat Stability (based on feeding group and
habit) 4,7
Top-down control 4,7
Benthic Food 4
Diversity Evenness 1,6,7
Shannon Diversity 1,2,6,7
Simpson Index 1,6,7

 

86

 

0.0 v 8.. 0.0 A 8:
9.80. 25:8 9.80. 0......8
08 80.5 9.802. 0.8 99.0 9:83

.o. 2:388 .o. 9.888
8o. .8... 8o. .80
9...

A 9.0.8.88 3.0 v 9.0.8.88

.0800... .0 .0888 85.02
008088 .0>O
8.0 .
8 o
m :0.: o m
v.0“ 08.an A 9:0Ec88 .0.
0.88 09.2.... 08.808 0.005
8.0 .
on o
m :9: o m
v.9. 99“...“ A 9:08.098 .9
88 8...... 0.9.88 88
cm... A .208 5 cm... v .208 .3

00:0...8 E996 008......00 E996

mm... 1:85:99. 8.. A 8888

90EE.>>m
+ 9030.50 + 98.....0
+ 908:0 . 90.28.8m

8:05 Our...
.26 8 .92 888...

90EE..8
+ 90.8.8 + 90.50.50
.98....8 + 909...0

998:8 8.858 +
90000.8 .99 . 998:8
9.8:... + 980.8

998:8 9:858
828:8 9.5....

998:8
899 \ 90000.8 .90.

998:8 .99 + .2080.
90000.8 . €00.08.
8000.8 + 980.8

+ .98... 02.. 8000.8

8988 Sm
:9 0.5.8 .8 v.5 9.8;
5.2. .0589 80.0 9:088
0850902.. 0.5.80 0.88.2

2.888
009 >08 0... 9 90800...
.0 80.8.. 05 8.88.2

2.98
E08 .0 88 8 00.885.

8.500
.682. .98... .98. :90.
809.8 0.0.8 0.0.0.02...

E980 2...: .8
00.88... 208m 08 0E:_0>
:c: .8 00.88.: 208m

9..0E.00m ..0 88 :c: .8
9.8.8.888 5.08:2an

00:8 .3098 .0>0 88
:c: .8 9880.88... «S...

2000....
0.5.8. :9 968.0228 Em
8.5 8.003 .9 288 000.. 0006

€2.00 80.... 9.0.8.88
>0... .0 .088 .0888 950080...

3:58 8.8.... 08. 8.8.. .5
88.. 958 8.9.88. .98:

3.3.8
.881. 0.8. 9.80. .898... .6
.088 2.58 6.9.88. .98:

908.208. .08...

0.890 08.8...8 30.8800.
0.5.80 .0 8.2.0qu 0 8 .08E
0.890 08.8...8 009.896

.50....2208. .98.. 2.8.0
08.00.88 0:: .0 :0.qu..0q 0 8

.98... 0.890 08.8...8 09000

.12..
00.8.88. 3.8568 .0 :9.t0q08

0 8 8.82008 >888 80.0

 

9.0.8.88 8000.8
8000.5 88.0.. toaazw
mud A E088 mud v E998
02890.3( 0.898.901
«08¢ 0808.8

no.8". 0:20 88.8.5”.

0050! 2.0.8.88! «02.0

00.508! 98288.. 539308

 

.ANOON .8 .0 gum—2 v.8 mam: .3...0S_ 0cm mEEEso
So... 00:80.2. 80.8. 0908.8 0.8 888:8 E08880 0.8. 8.8.8.5.. m...0c88..8 08 99080.8 E0888m. Nd 0.8...

87

 

 

 

: 00- n ~ .00 u 0 0x0

 

00.0 00.0 8098.0 + 080:0.2
.méundruw .0...-u~.0..uw to...~-u~.m~uwtn.~-u~.kuw .0.:u~.mwuw .0...u~.m~uw
~00 00.0 3.0 3.0 00.0 v.0 50.0 3.0 9.0 93 N...0 N00 30. 000009000
:m.~u~.00uw 30¢.N"~.mmuw :0.Nu~<.mnw
9.0 00.0 2.0 0.0 :0 0.0 3... 0.90.0
:m.~uu.mmuw :0.~u~.~.muw :0.~u~.$um
00.0 8.0 3.0 9.0 00.0 00.0 .9. 000.800....0
:09 u~.:muw .0...n~.0vuw
0 00.0 0 3.0 .9. 9900000.: 00202800
. t u N .00 u 0
0.0 9 0x0... ...Ow
: tmuudmuw 0x0...
F 0 990000....
Leuuemuw . 0x0:
0.0 0.0 990000.; + 900000
.méuw‘dmum,
0 0 .9 0.9 990000....
..:.Nu~.0muw .0.:u~.v~um
P 0 0.0 m...» .9 0.9 900000
..0.~u~.00um
F N .9 0x9 90.00.00.003
:0..-n~.9u0
R S 8 0x9 .90...
8.780.0er
00 mm .9 0x9 .90 0. 80000.”.
_ m. _ m . m. . m . m. . m 0.502
.0090E00w E090Em. E09985 .0090Em. 900900.000 E0995
0:..0>.m 00.80.0m 0008000..

 

.98. 085 9 00.0.80 .08500890: 0.0.5 0.8.0 .00. 8:.0> 000:... A... 00.0 v 0 .0 C :0 v 0 85.0 .0
8:.0> 0080.090 000 8.80 .09 00x8._>> 5.3 09.0. 0.0 80.0> 8.00.2 000900.08 98900.0. .939. 000 80:02. 000 60.80.00
60.80000 08.0 9.0..0> .0 008.90; 090000.. 000 0000.90. 0003.00 .0998 2.800.090 0050:. 9900:0>0.0.00.2 ”0.0 0.00..

88

 

!m.N-uN.mFum

cm.—.UHN.—.mﬂw
v0.0 mrd

IN.N.uN.mNuwtmeJuNJuNnm .m. -uN.mNuw

coon. oEEmm

 

«.0 00.0 5.0 8.0 00.0 3.0 2.0 30 £03.: 35005 6:00...
.n._‘-un.wwum to.~.u~.-um
0:. 000 8+ 50 £520 .230:
.wo._.-uu.w~uw
00.0 00.0 EOQmSOQw
.m.r-u~.mruw .w.Tu~Kan
v0.0 .06 Nwd No.0 EOQH=EOQO
:0.N-un.omum 3N0.N-u~.v~um
v0 30 Rd 8.0 coagamoEasﬁauen.
tm.w-u~.o_. um
5.0 o manner—m .x.
LaﬁuNdNuw .~.F-u~.mwum
mud mod and 5.0 900mg o\o
t and- u N .9 u m @533“.
00.0 0 £8.05 o\o .3235".
t _..N-uN.\._.um .h._.-u~.mwuw :m.m-u~.mmnm
5.0 0 00.0 0 00.0 0 $00 85.“.
.mé-u~.wmum
00.0 3.0 3.0 0.me m EmEEoo
. 3.0 u N ._.m u m
$0 000 $90 05 N 0005800
.. me u N .00 u m
0.0 000 0x0 0x0... EwEEoQ
: 0.0.0- u N .3 n 0 3:80.85
00.0 0 $00 000002 30:80.0...
_ m _ a . m _ m _ m _ m 050:
0:090:55 EonEEm 00090805 0009005 «0090805 €09me
00:93 003028 00:380..

 

8.0080 m.m 20¢

89

% Gastropoda

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.9-

0.8- l

0.7-

0.6- * *

0.5-

0.4T H-

o.3~ T

0.2- _L - 512.

0.1- "L "' ”_,_'

o.c . . . . %ﬂ .
R I R l R I

-o.1- _ _ _ _
Lacustrine Palustrine Riverine

Figure 3.1: The relative composition (%) Gastropoda in
emergent habitats of reference (R) and impacted (I) wetlands
at the class scale (lacustrine, palustrine, and riverine). Solid

lines within the box indicate are median values. the box
values represent the inter-quartile ranges (25% and 75%).

and the range bars indicate the maximum and minimum

values. Signiﬁcant differences (p < 0.05, denoted by ** or p <

0.1, denoted by *) were calculated by non-parametric

Wilcoxon tests.

CPOM/FPOM Ratio Photosynthesis] Respiration Ratio

SPOM/FPOM Ratio

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1o.o- 1o:
1
2 a-
U
7.5 g ‘
g a
- .D
5.0 8 .
ﬂ 4-
25- g '
I 2-
0.0- o-
7.
o.4- E
is.
0.3- .9
5..
E
0.2- %a.
in
0.1- :1.
.D
i
0.0- o.
-o.1-
0.4-
3.5' O *
- a
I: 0.3-
2.5- §
IL
.2 0.2-
1.5- 5
C
:3 01- I
0.5-i
0.: . .
R I
-o.5-

 

Figure 3.2: A comparison of functional or ecosystem attribute ratios between
reference (R) and impacted (I) emergent habitats of palustrine wetlands in the MRW.
Solid lines within the box indicate are median values, the box values represent the
inter-quartile ranges(25% and 75%), and the range bars indicate the maximum and
minimum values. Signiﬁcant differences (p < 0.05. denoted by “ or p < 0.1. denoted
by ') were calculated by non-parametric Vchoxon tests.

91

Chapter 4: Inﬂuence of Wetland Size and Water Quality on Isolated

Macroinvertebrate Communities

92

Chapter 4: Inﬂuence of Wetland Size and Water Quality on Isolated

Macroinvertebrate Communities

Abstract

Wetlands are biologically diverse ecosystems that provide many beneﬁts
of both economic and ecological signiﬁcance; however, changes in the federal
wetland regulatory program have made isolated wetlands more vulnerable to
human disturbance. New regulations may be based on wetland size leaving
small isolated wetlands more at risk. In order to protect these wetlands more
information regarding the structure and function of biological communities within
isolated wetlands of all sizes and the response of these communities to human
activities is needed. The objective of this study was to compare
macroinvertebrate communities among isolated wetlands of variable quality
(impacted or non-impacted) and size (large or small). Four palustrine marshes
were sampled from each possible combination of treatments (large impacted,
large non-impacted, small impacted, and small non-impacted) from the
Muskegon River Watershed, Michigan, USA, during July 2002. In addition,
multiple habitats (emergent, submergent/ﬂoating) were sampled in each wetland,
if present. In this study, we found little differences in the macroinvertebrate
community between small and large wetlands and between impacted and

reference wetlands. However, we did ﬁnd different taxa exist in small and large

93

wetlands, and we did ﬁnd more unique taxa in small wetlands. These results
may be due to an interaction between wetland size and quality. This information
could provide wetland managers evidence for the protection of all wetlands, both

large and small.

Introduction

In 2001, the United States Supreme Court ruled that isolated, non-
navlgable waters of the United States were no longer protected through the
Clean Water Act based on the Migratory Bird Rule alone (Nadeau and Leibowitz
2003). The change in CWA jurisdiction and subsequent reduction in regulation
may result in once protected isolated waters and non-protected isolated waters
becoming more vulnerable to human disturbance (Christie and Hausmann 2003).
Falling under this category of isolated waters are isolated wetlands, generally
deﬁned as wetlands surrounded by upland areas (Tiner 2003b). The extent of
isolated wetlands through out the United States is unknown; however estimates
in the Great Lakes Region, where this study was conducted, suggest that ~40-
80% of wetlands may be considered isolated (T iner 2003a). These changes in
the federal wetland regulatory program have made isolated wetlands more
vulnerable to human disturbance and subsequently more research is needed to
understand the impact of disturbance on wetlands and to protect and better

manage these ecosystems.

94

Over the past 200 years, over half of the wetlands in the United States
have been destroyed due to human activities, mainly through the development of
agricultural, urban and industrial areas (van Dam et al. 1998, Mitsch and
Gosselink 2000). Loss of these wetland ecosystems may have resulted in
subsequent loss of many economic and social beneﬁts (e.g., ﬂood prevention,
groundwater recharge, water puriﬁcation) and in loss of areas of ecological
signiﬁcance (e.g., species diversity, unique habitats) (Wrssinger 1999, Mitsch and
Gosselink 2000). Methods to monitor and protect those remaining wetlands have
been of great research interest, especially methods using macroinvertebrates as
biological indicators of assessment (Burton et al. 1999, Merritt et al. 1999,
Kashian and Burton 2000, Helgen 200‘l, Helgen and Gernes 2001, Chessman et
al. 2002, Gernes and Helgen 2002, Merritt et al. 2002b, Tangen et al. 2003,
Uzarski et al. 2004). Most of these studies have addressed macroinvertebrate
communities in connected wetlands such as riverine and lacustrine wetlands;
however, few studies (Helgen and Gernes 2001, Gernes and Helgen 2002) have
addressed marcoinvertebrate community response to human disturbance in
isolated wetlands.

Often wetland regulation and protection are determined by wetland size,
with preference given to large wetland ecosystems (Babbitt 2005). This is based
on assumptions that smaller wetlands have shorter hydroperiods, identical
species and less species compared to large wetlands and are independent from
other wetlands (Snodgrass et al. 2000, Babbitt 2005). Whether these

assumptions hold true is still unknown and could vary dependent on the type of

95

biota. Studies have found an increase in bird species richness (Brown and
Dinsmore 1986), plant species richness (Lopez et al. 2002, Houlahan et al. 2006)
and a slight increase in amphibian and predatory macroinvertebrate richness
(Baber et al. 2004) with wetland size, yet other studies have found no
relationship between amphibian species richness and wetland size (Snodgrass et
al. 2000, Babbitt 2005). New regulations may be based on wetland size, leaving
small isolated wetlands more at risk, making further studies investigating the
relationship between wetland size and additional biological communities
necessary in order to understand the potential signiﬁcance of this new regulation.

The objective of this study was to address patterns in the biological
community of isolated wetlands, speciﬁcally, patterns in macroinvertebrate
community structure within a single wetland class (palustrine or depression
marshes) using an experimental approach. Macroinvertebrate attributes,
including abundance, diversity, and functional-group analyses, were compared in
wetlands and between habitats using two experimental factors: (1) stressor
impact (wetlands classiﬁed as impacted or reference); and (2) wetland size
(wetlands classiﬁed as large or small). I hypothesized that ( 1) in both large and
small palustrine wetlands, reference wetlands will differ in macroinvertebrate
community structure compared to impacted wetlands and that (2)
macroinvertebrate community structure will differ between palustrine wetlands of
different size. In general, I predicted that impacted wetlands, regardless of size,
would have lower macroinvertebrate abundance, diversity, and a shift in

invertebrate functional relationships compared to non-impacted wetlands. I also

96

predicted that large wetlands would have greater macroinvertebrate abundance
and diversity and a shift in invertebrate functional relationships compared to
small wetlands. The results of this study could provide valuable information for

future wetland regulation, protection and management.

Methods

Mocation andecale

The Muskegon River Watershed (MRW) begins in the north-central portion
of Michigan's lower peninsula, and runs in a southwest direction to the shoreline
of Lake Michigan. The watershed is approximately 7000 km2 in size and
includes 94 tributaries, 183 stream segments, hundreds of lakes and wetlands
within 11 different counties (Torbick et al. 2006). Dominant land use varies from
forest dominated to agriculture/urban dominated areas within the upper and
lower regions of the watershed, respectively (Lougheed et al. 2007). The
continuum of the MRW has been altered by 95 dams, and has been impacted by
human inﬂuence through logging and agricultural byproducts and is predicted to
have a 50% increase in urban land use within the next 35 years (Pijanowski et al.
2006).

Isolated wetlands, deﬁned in this study as depression marshes
surrounded by upland areas, were selected (n=16) from in and near the MRW
(Lougheed et al. 2006) and surveyed prior to the ﬁeld season to conﬁrm wetland
existence, accessibility, and to obtain landowner permission if necessary. Each

wetland was selected and classiﬁed based on two experimental factors: (1)

97

stressor impact (wetlands classiﬁed as impacted or reference); and (2) wetland
size (wetlands classiﬁed as large or small). Level of impact was determined by
the amount of human activity through visual surveys surrounding each wetland
and later conﬁrmed with chemical and GIS analyses. In general, those wetlands
with little human activity and forest-dominated surroundings were considered
reference, and those with high human activity and agriculture-dominated
surroundings were considered impacted. Wetlands < 0.5 hectares were
considered small, and wetlands >1.2 hectares were considered large as deﬁned
by Lougheed et al. (2006). Therefore, a total of four wetland treatments (large
impacted, large reference, small impacted and small reference) each with four

sites, were sampled in this study.

Sample Collection

All marshes were sampled once between July and early August 2002. At
each marsh site, three evenly spaced transects were established perpendicular
to one side of the wetland, thus extending from the wet-meadow or shore into
deep-water habitats. Along each transect, multiple random sampling points
(marked by a quadrant) were established to facilitate sample collections from
dominant vegetation habitats. For macroinvertebrate collections, two habitats
(the emergent and submergent/ﬂoating, now labeled as only submergent in this
paper, vegetation zones) were sampled in each marsh. The ﬂoating and
submergent zones were combined due to little difference in macroinvertebrate

composition, and the inability to clearly distinguish and separate these zones in

98

many wetlands (T. Burton, personal communication). In small wetlands, it was
difﬁcult to extend three parallel transects and not disrupt adjacent transects while
sampling. As a result three transects were extended from the center of the
wetland toward the bank. This allowed for three complete transects to be
sampled properly with no disturbance. In wetlands where walking was prohibited
or unsafe, boats were used to collect samples.

All macroinvertebrates were collected from the ﬁrst random quadrat in
each habitat. To ensure no disturbance in the macroinvertebrate community due
to other sampling activities, macroinvertebrates were collected at a random
angle, 1-2 m from the main sampling quadrant where other variables (e.g.,
physical, plant, zooplankton, algae) were collected. One macroinvertebrate
sample consisted of two sets of three sweeps using a standard D-net (500 pm)
and subsequently rinsed through a 500 pm sieve to remove large pieces of
vegetation. Since each marsh was sampled only once, a total of three samples
were collected per habitat and a total of 6 samples per wetland.

In the laboratory, macroinvertebrate samples were again sieved and sub-
sampled to reduce processing time (King and Richardson 2002). In the sub-
sampling protocol, each sample was homogenized and divided into two equal
proportions or sub-samples. All macroinvertebrates from one sub-sample were
sorted and identiﬁed to the lowest practical taxonomic level using Merritt and
Cummins (1996), Thorp and Covich (1991), Pennack (1989) and Larson et al.
(2000). Most macroinvertebrate taxa were identiﬁed to the generic level, except

for taxonomically difﬁcult organisms such as the Chironomidae, Oligocheata, and

99

Hirudinea. For other taxa, large numbers of immature specimens were collected
and could only be accurately identiﬁed to the family level (e.g., Planorbidae,
Corixidae, Notonectidae). Macroinvertebrate counts were subsequently adjusted
to account for subsampling procedures prior to statistical analyses.
Environmental variables collected for each wetland site included ﬁeld
(chemical and physical water variables, additional biological attributes) and
laboratory (chemistry, GIS) measurements. For water chemistry, a single 250ml
sample was collected from an open area of each marsh and analyzed in the
laboratory for total phosphorous (TP), total nitrogen(TN), and reactive
phosphorus (SRP), using methods described by Lougheed et al. (2007). Field
measurements of dissolved oxygen (DO), pH, conductivity, and mean water
depth were collected at every site with a YSI 556 multiprobe meter and
measuring stick, respectively. Biological measurements on plant, algal, and
zooplankton communities also were collected; however, these results are not
reported in this study and can be found in Lougheed et al. (2006, 2007). Land-
use, land-cover analyses were conducted with GIS maps as described by
Lougheed et al. (2006) to obtain land-use and wetland size information. These
data were collected to conﬁrm the classiﬁcation of wetland sites into correct

treatment groups.

Statistical Analysis

100

Common attributes used to characterize macroinvertebrate community structure
were calculated in this study, including macroinvertebrate abundance (number of
individuals per sample), and taxa richness (number of taxa per sample at the
lowest practical taxonomic unit). Numerous diversity indices were calculated to
characterize diversity within the macroinvertebrate community, including the
Shannon Diversity Index, Simpson Index and Evenness Index as described by
Burton et al. (1999). To better understand the ecological condition of these
isolated wetlands, macroinvertebrate taxa were assigned into functional feeding
and habit groups (Merritt and Cummins 1996, Cummins and Merritt 2001) and
subsequently calculated into the selected functional group ratios (Table 4.1) that
can serve as ecosystem surrogates for wetlands (Merritt et al. 1999, Cummins
and Merritt 2001, Merritt et al. 2002b). These functional metrics were developed
to assess the ability of wetlands to support bird and ﬁsh populations in Florida.
Criteria established by these authors (Table 4.1) in their assessment protocols
were developed speciﬁcally for that region and may not apply to our wetlands
(inland marshes) or the stressors in our wetlands. Although these metrics may
not be sensitive to some potential stressors, these metrics can still provide an
indication of changes in the ecological condition between reference and impacted
and small and large wetlands in this study. All attributes (abundance, diversity
and functional) were calculated for each habitat type, including (1) total wetland,
in which all samples from both habitats were included (6 samples per wetland);

(2) emergent habitat, in which only emergent samples were included (3 samples

10]

per wetland); and (3) submergent habitat, in which only submergent samples
were included (3 samples per wetland).

Comparisons for this study were calculated with a two-way ANOVA to test
for effects and interactions in the macroinvertebrate data (abundance, diversity,
functional-feeding analyses) (Quinn and Keough 2002). Non-normal data was
transformed to meet test assumptions; if necessary, abundance data was log or
fourth root transformed and proportion data was arcsin square-root transformed
(Quinn and Keough 2002). Again, analyses were conducted three times, for
each macroinvertebrate attribute: (1) total wetland (6 samples per wetland); (2)
emergent habitat (3 samples per wetland); and (3) submergent habitat (3

samples per wetland). All analyses were conducted on JMP IN 5.1.2.

Results

Water Chemistry and Land-(1139 Land Cover

Wetland sites had signiﬁcant differences in chloride between impacted
(59.46 +/- 17.6 mg/l) and reference sites (6.7 +/- 5.7 mg/l, t-test: n=16, t = 3.2, p
< 0.01). Similar trends were observed in conductivity, with 310.79 +/- 59.1 9
pS/cm and 109.91 +/- 30.9 pS [cm in impacted and reference wetlands,
respectively (t-test: n = 16, t = 2.33, p < 0.05). Impacted wetlands also had
higher amounts of total phosphorus (TP) (64.69 +/— 17.6 pg/L) and total nitrogen
(TN) (1.37 +/- .0-2 mg/L) compared to reference sites (TP: 31.52 +/- 6.2 pg IL, t =

1.47, p =0.16; TN: 1.12 +/- 0.11 mg/L, t = 1.11, p =0.28); however, these results

102

were not statistically signiﬁcant. Impacted wetlands had ~40% greater
developed (urban + agricultural) surrounding land use compared to reference
wetlands (t = 5.59, p< 0.001). Whereas, reference wetlands had signiﬁcantly
greater forest (71%) and wetland areas (17%) compared to impacted wetlands
(forest: 33%, t = -3.75, p < 0.001; wetland: 4%, = -2.91, p < 0.05). Wetland size
was found to be signiﬁcantly different (t—test: n = 16, t = 5.64, p < 0.0001), with
small wetlands 0.3 +/- 0.06 hectares in size compared to at large wetlands 1.66

+/- hectares. Overall, wetlands were correctly assigned to treatment groups.

Macroinverteliate Abundance

In this study, ~89,000 macroinvertebrates were collected from isolated
palustrine wetlands in the MRW. The community was dominated by Arthropoda
(74%), followed by Mollusca (23%) and Annelida (2%). Of the Arthropoda,
insects were most prevalent at 55%, followed by crustaceans (19%), including
both micro- (e.g., Copepoda) and macro-organisms (e.g., Amphipoda). At the
order level, the most dominant groups in all wetlands were the Diptera (27—35%),
Gastropoda (11-20%), Copepoda (9-10%), Hemiptera, (7-10%) and
Ephemeroptera (5-8%).

For the total wetland analyses, mean macroinvertebrate abundance was
greatest in impacted large wetlands and lowest in impacted small wetlands,
however no signiﬁcant effects of wetland quality (F = 1.02, p = 0.31), size (F =
0.01, p = 0.96) or interactions (F = 1.8, p = 0.19) were observed (Figure 4.1A).

Similar results'were observed in the emergent zone, with ~30-40% greater

103

macroinvertebrate abundance in large impacted wetlands compared to all other
treatments (Figure 4.1 B). In the submergent zone, we did observe a signiﬁcant
effect of wetland quality (F = 4.66, p < 0.05), along with a signiﬁcant interaction
between wetland quality and size (F #- 8.97, p < 0.01). This interaction effect is
evident in Figure 4.1C, with high abundances in small compared to large
reference wetlands and opposite results, of high abundances in large compared

to small impacted wetlands.

[Lacroin vertebrate ﬁversitv

Total mean taxa richness was consistent across all treatments, ranging
from 20-25 different taxa per sample with no signiﬁcant difference in the main
effects (quality: F = 0.16, p = 0.68; size: F = 0.02, p = 0.32) or an interaction (F =
1.01, p = 0.31) (Figure 4.2A). In the emergent zone even less variation was
present between treatments, means ranging from 27-29 different taxa (Figure
4.28). In the submergent zone no signiﬁcant difference was observed, however,
the same interaction observed with the abundance data was also apparent in the
taxa richness data (Figure 4.2C). It should be noted that the emergent habitats
across all treatments had greater abundance and taxa richness than the
submergent habitats (Figures 4.1 and 4.2).

Various community diversity indices were calculated to test for differences
among wetland treatments, with mean values reported in Table 4.2. We found
no signiﬁcant effects of wetland quality, size, or interactions in any of the three

habitats for the Shannon-Diversity Index (range for all three habitats: F = 0.02 —

104

2.51, p = 0.95 - 0.12) or Simpson Index (range for all three habitats: F = 0.02 -
3.27, p = 0.87 — 0.08). The only statistically signiﬁcant result was an interaction
between wetland quality and size in the Evenness Index for the submergent
habitat only (Table 4.2: F = 4.46, p s 0.05).

Because taxa richness and diversity were mainly the same across wetland
treatments, we examined each factor (quality and size) and each treatment (e.g.,
large impacted) for unique taxa. We considered unique taxa as organisms found
in a minimum of two sites within a single wetland factor or treatment. The most
unique taxa (6) were found in small wetlands compared to large wetlands (2
taxa); of these taxa in small wetlands, most were predators (5 taxa) belonging to
the family Dytiscidae or Hydrophilidae. In impacted wetlands, 4 unique taxa were
identiﬁed, one being lsospoda, a group often associated with increased
disturbance (Burton et al. 1999, Kashian and Burton 2000). In reference
wetlands, 3 unique taxa were identiﬁed, two of which are from insect orders
(Odonata and Trichoptera) considered as less tolerant and indicators of good
water quality (Burton et al. 1999, Kashian and Burton 2000, Helgen 2001, Helgen
and Gernes 2001). Of the total taxa identiﬁed in this study, approximately 11%
were found in either small or large and either impacted or reference wetlands. Of
this percentage, study treatments had even fewer unique taxa, with only 2 unique
macroinvertebrate taxa in small reference wetlands and one unique taxon in

small impacted, large reference and large impacted wetlands.

Macroinvertebrate Fynctional—Feeding Groups

I 05

Functional group ratio means are listed in Table 4.3. The ratio of
predators to available prey, or top—down control, was statistically greater in small
wetlands compared to large wetlands in emergent habitats (F = 4.41, p < 0.05).
This indicated that small isolated wetlands have a greater number of predators.
Although no statistical differences were observed in the rest of the data, some
general trends on the ecological condition among treatment wetlands were
evident. Differences between impacted and reference wetlands were observed
in the production/respiration (HR) and habitat stability (using macroinvertebrate
habits) ratios. In general, the P/R ratio in all reference wetlands were lower
compared to impacted wetlands, this indicated that reference wetlands may be
more heterotrophic systems and impacted wetlands may be more autotrophic
(Table 4.3). The exception was emergent habitats of impacted wetlands, which
also indicated heterotrophy. A similar pattern was observed with the habitat
stability ratio (based on macroinvertebrate habits), with less stable habitats in
reference and emergent impacted wetlands and more stable habitats in impacted
wetlands (Table 4.3). However, this ratio contradicted the additional habitat
stability ratio (based on functional feeding groups) in which all wetlands had

similar, higher values indicative of more stable substrates.

Discussion

Wetland ecosystems are threatened by various human activities and

recent changes in wetland regulation have made these systems, speciﬁcally

106

isolated wetlands, more vulnerable. Without protection, many of these isolated
wetlands, especially small sized systems, and the biological communities within
may be overlooked, despite the numerous beneﬁts these wetlands can provide.
In order to protect these systems, more information regarding the structure and
function of biological communities within isolated wetlands of all sizes and the
response of these communities to human activities are needed. The objective of
this study was to compare macroinvertebrate communities (including abundance,
diversity, and functional relationships) among isolated wetlands of variable
quality (impacted or reference) and size (large or small).

We predicted that impacted wetlands, regardless of size, would have
lower macroinvertebrate abundance, diversity, and a shift in functional
relationships compared to reference wetlands. However, most of our results
showed no statistical differences in abundance and diversity between impacted
and reference wetlands. Other wetland studies have found similar results with a
lack of macroinvertebrate response to environmental disturbances, such as
increased land-use, nutrients and development (King et al. 2000, Steinman et al.
2003, Tangen et al. 2003). Environmental conditions in these impacted wetlands
may not have reached the threshold limits for many naturally tolerant wetland
macroinvertebrates, thus explaining the lack of signiﬁcant change in abundance
or diversity in our study (King and Richardson 2002). However, other wetland
studies have found contrasting results. For example, both King and Brazner
(1999) and Kashian and Burton (2000) found lower macroinvertebrate

abundance and diversity at impacted wetlands; whereas, Rader and Richardson

107

(1994) found opposite results, with increased macroinvertebrate diversity in
response to increased nutrients in the Florida Everglades. This latter study
supports our signiﬁcant ﬁndings of increased macroinvertebrate abundance in
submergent habitats of large impacted wetlands compared to large reference
wetlands. Increased nutrient input, combined with a suitable amount of available
habitat (thus large impacted wetlands), may provide for increased plant and algal
growth, which subsequently can provide increased food and habitat availability
for more individuals (Wissinger 1999). Variation between all of these studies
could be due to numerous factors such as wetland hydrology, wetland type, the
amount and type of impact (e.g., type of nutrients), study design (e.g., taxonomic
resolution) or a combination the above (King and Richardson 2002, VWcox et al.
2002).

We also predicted that macroinvertebrate community structure would differ
between large and small wetlands; however, in general we found no
macroinvertebrate response due to wetland size. The only observed effect of
size was on top-down control in emergent habitats, due to a greater number of
predators in emergent habitats of small compared to large wetlands. Large
reference wetlands in general, had the lowest mean abundance and diversity of
treatments and thus might have fewer prey to support large predator populations.
A second reason for lower macroinvertebrate predator abundance could be due
to the presence of ﬁsh. We did not survey ﬁsh in these studies, however, the
likelihood of ﬁsh presence was increased with wetland permanence as would be

expected in large wetlands. Numerous studies have shown that ﬁsh can lower

108

the overall abundance, diversity, and biomass of macroinvertebrate populations
and compete with large macroinvertebrate predators for prey; thus limiting the
top-down control in these wetlands (Mallory et al. 1994, Wlssinger 1999, Zimmer
et al. 2000, Tangen et al. 2003). In general, large numbers of predators can be
typical in many wetland ecosystems and could be further supported by high prey
turn-over (possibly induced by high nutrient levels of impacted sites) and the
absence of ﬁsh (more common in small wetlands). The lack of ﬁsh common in
small wetlands can often lead to increased diversity in macroinvertebrate
predators (Vlﬁssinger 1999). This was evident in the 6 unique taxa, mainly
predators, found only in small wetlands of this study. These small wetlands may
provide important areas for insect diversity, not provided by large wetlands.
More studies with more intensive taxonomic resolution (e.g., identiﬁcation of
Chironomidae to species) are required to identify these differences (King and
Richardson 2002). Overall, a combination of biotic reasons (both prey levels and
ﬁsh presence/absence) may explain the overall lower abundance and diversity in
large reference wetlands and greater predator abundance and unique taxa in
small wetlands.

Additional functional analyses of macroinvertebrate communities can
provide information regarding the ecological condition of these wetland
ecosystems (Merritt et al. 1996, Merritt et al. 1999, Cummins and Merritt 2001,
Merritt et al. 2002b); however, it should be noted that criteria established by
these authors (Table 4.1) were based on a different region and type of wetland

and may not apply to wetlands in this study. Additional studies would need to be

109

done to quantify and test these criteria in inland marshes in Michigan; however,
the established criteria values of Merritt et al. (1996, 1999, 2002) in relation to
our study results (Table 4.3) are included for discussion and comparison
purposes. Increased levels of production to respiration or PIR ratios in our
impacted wetlands indicated these systems were more autotrophic compared to
reference wetlands. Increased urban and agricultural land-use around these
sites in addition to increased nutrient run-off might explain these differences.
However, the emergent zones in these impacted wetlands indicated more
heterotrophic systems; which could be due to high amounts of decomposition
and increased abundance of aquatic detritivores common in this type of habitat
(Wrssinger 1999). Kashian and Burton (2000) found similar results with lower
P/R ratios in reference compared to impacted costal wetlands in Lake Huron,
Michigan, and a study in restored riverine marshes of the Kissimmee River-
ﬂoodplain, Florida, found most vegetation habitats to be heterotrophic (Merritt et
al. 1996). Observed differences in the two habitat stability metrics were
probably due to the placement of the Bivalva, in either the denominator (habit
metric) or numerator (functional feeding metric) of the ratio calculations. Merritt
et al. (2002) suggested switching the Bivalva into a special case of burrower for
river-oxbows in Florida; if changed in this study, both habitat metrics would have
indicated similar amounts of stable substrate availability.

An interaction effect between wetland quality and size (Figure 4.1 and 4.2)
was observed numerous times in the abundance and diversity data. This

interaction indicated that large impacted wetlands had greater abundance and

110

diversity compared to large reference wetlands, whereas, small wetlands had an
opposite effect, in which reference wetlands had greater abundance and diversity
compared to impacted wetlands. As already mentioned in this discussion, many
abiotic and biotic factors may be playing a role in the treatments of this study.
Abundance of macroinvertebrate communities in large wetlands responded
positively with impact, probably due to available space for general biotic growth;
whereas, small wetlands respond negatively with impact. This negative impact
could be due to increased competition for habitat and food resources and
decreased recovery levels after carrying capacities have been reached. Also,
because smaller wetlands have a reduced area, less impact (e.g. accumulation
of N or P) may be necessary to reach the tolerant threshold levels in which
wetland macroinvertebrates respond. Macroinvertebrates in reference wetlands
seem to respond positively in small sized wetlands and negatively in large
wetlands; again, as mentioned above, this could be due to the absence and
presence of ﬁsh predators, respectively. Without the pressure of ﬁsh predators,
the macroinvertebrate community can diversify in small wetlands. Additional
examination of these data, that include further taxonomic identiﬁcation and
similarity analyses may provide more information on observed patterns in this

study.

Conclusions

111

Often wetland regulation and protection are determined by wetland size,
with preference given to large wetland ecosystems (Babbitt 2005). Assumptions
made regarding small and large wetlands were not supported by this study,
speciﬁcally, that macroinvertebrate communities have identical species in large
and small wetlands and that less species are found in small compared to large
wetlands (Snodgrass et al. 2000, Babbitt 2005). In this study, we found few
differences in the macroinvertebrate community between small and large
wetlands and between impacted and reference wetlands. However, we did ﬁnd
different taxa to exist in small and large wetlands, and we did ﬁnd more unique
taxa in small wetlands. These results may be due to an interaction between
wetland size and quality. These results provide support that small wetlands are
equally, if not more important, than large wetland ecosystems. This evidence, in
addition to potential greater diversity in small wetlands, could help support equal
protection, regulation and management of small and large isolated wetlands.
More research is needed on wetland ecosystems in general. However, studies
on small wetland diversity, biological interactions and connectivity to surrounding

wetlands are needed.

112

 

2.0 A 00000308 90
0900000 00 v 000003000
000000000 00>O 00.00000 .0852

omd 00.0
.........
0.000 000.00... 00000000 0.0000

00.0 .

00 0
0.000 00:00.... 00000000 0.030

00.0 A E0000 00.0 v .0000
0.0000050. 200000.000...

000000 00“.
0050 ..0 .002 0090000

0000.00.30
+ 20.30000 + 0002,0000
.200E_.o + 00005.0

00200.60
00000.00 + 000000000
.000. . 00200.60
00000.... + 00000000

0.2028 .92 + .2000.
000000000 \ 00000.00
000000 + 0.00080

+ 0000.0 02: 0000000

0:00:90
0001.000 00. 9 000000000
00 000050 0.... 0050005.

0.0000
.0000 00 00.0 00 005000.).

.00000
>000; 0000.0 .0000. .000
0000.000 0.000 0.0.0.024.

00000 500.00 .05 0000

0.0: 000 0.00E000000E En.

€300 00 .5 000005000 .000
.0 .0008 0200000 0260.00...

000.030 05.00....
0000 000.000 00 00000

0:090 0.00800 0.00:

3.0080 05.00:. 009
00.000.— 000.602 00 00000

0:590 00.0000. 0000...

ER. 809.009
0.020008 00 00.00000 0
00 00000090 0000000 00000

 

$0300. 8:00 03800

00.001
000.5 000.0000“.

.0050:
«00.09.0000: “00.00

0050005.
003080000 E000>000m

 

 

.3000 .0 .0 0.005. 000 000. .0005. 000 00.5050 69.. 00000.2. .630.
0.5 0. 000.0: 0000. 00.0000 000 000000000 000000000 0000 .0000002 0000000008 000 00.0.0000 000.0008. ”0.0 0.00 ._.

113

Table 4.2: Mean (SE) Shannon Index, Simpson Index, and Evenness Index
values for each wetland treatment (impacted large, impacted small, reference

large and reference small) for total wetland (T), emergent (E) and

submergent (S) habitats. The only signiﬁcant result (p < 0.05, *) was an
interaction of wetland quality and size observed in the Evenness Index of the

 

 

submergent habitat.
Diversity Impacted Impacted JReference Reference
lndices Lﬂge Small Large Small
Shannon
Index T 2.11(.11) 2.04(.07) 2.00(.09) 2.16(.06)
E 2.43(.07) 2.2(.07) 2.13(.12) 2.31(.05)
S 1.78(.18) 1.84(.09) 1.87(.12) 2.00(.10)
Simpson
Index T 0.79(.04) 0.81(.02) 0.79(.03) 0.84(.01)
E 0.87(.01) 0.81(.02) 0.79(.03) 0.84(.01)
S 0.69(.06) 0.76(.02) 0.77(.03) 0.78(.03)
Evenness T 0.67(.03) 0.68(.01) 0.69(.02) 0.68(.02)
E 0.74(.03) 0.67(.02) 0.65(.03) 0.70(.02)
S * 0.59(.05) 0.68(.02) 0.74(.02) .68(.03)

 

114

 

 

00.50 0.0000 .3020 60.5.0 0
00.0 A 08.0508 . . . . . . . .
50000 0 80000000 05 .00 .00 0 .00 .0. 0 .00 .00 0 .00 .0 0 .. 0
80000.0 30000.0 60000.0 80000.0 0 0260 000
.0902 80:00 @3000 .3000 6.0.00... 0
00000000 000008 00000000 0.0000 . . . . . . . .
00.. 000000.. .000 v 00056000 5.. .00 .00 0 .00 #0 0 E .00 0 .00 :0 0 m .
200000... 0000 02.5.. 02 8:90.00 00.000 £00000 0.5000 .0. 00.0 0 “HM“...
.0003 .0300 .0103 8.00000 0
00.0 . . . . . . . .
A 00056000 0. 2000000 0090 a .00 F 000 00. s .0100 0 0.0 000 F m 30.0.00
3...: F a 000 F :0 .0» F .0 00.0 0 e .530:
“SEN: acmemEm $093.0 Aowgwwd Cwqomd 3.350 m
.0800. 0.002900 00.0.0 000005. 3000000 $00000 .0300 80000.0 0
.000 v. 0080292 802200 £0.00 600000 .0503 6.0.000. F 0:.
:08» 090.. :00—0 l00000 0000000.:
000000 0001 .20000 0000.000”. 00000000”. 03000.0. 0300000 8000.00.00
E0~0>000m

 

000000 00000000 000 .6 0000
0260 000 0.0 0. 0020000 002, 0.00 000.003 00 A. .000 v 00 00000 00005090 0.00 00... 000000 Amy 0000000000 000
Ev 00000000 .Ab 000.002, .0000 .00 0.000 0000.000 000 00.0. 00000000. .__0Em 0000000. .0000. 00000000 00000000
000003 0000 0000020100000: 0000000000 000000000 000000000. 0000. 0090 00.0000 .00000000 ”0.0 0.000.

115

.>
.4
O
5.7.

D Reference - Impacted

n1

C.

2000-

E

g

E. ,

(indivdualslsample)
é .§

Mean Macroinvertebrate
Abundance

 

 

 

 

O

 

Small

—0— Reference +Impacted {l- Reference +Impacted

 

 

 

 

 

3 A
g 2 ,1
g g g- 1500- 1500‘
' 8
>

a
E 2% 1000‘ 1000‘ all
u 3 :3
a .9 E
2 < .2
g g sool 500‘

V Small LaFgo v Sula" LaFge

Figure 4.1: Mean (SE) macroinvertebrate abundance (individuals per
sample) for (A) total wetland, (B) emergent and (C) submergent
habitats in isolated palustrine wetlands. All four wetland treatments
are represented: small reference, large reference, small impacted
and large impacted. A signiﬁcant effect (*) of wetland quality (p <
0.05) and an interaction effect (p < 0.01) were observed in only the
submergent habitat (C).

116

Mean Taxa Rlchnless

A. Total
DReference -lmpacted

     

 

 

 

 

 

 

 

 

 

 

 

30.
(D
m
m l
E E
.2 zo- -I-
a:
(B
r:
5..
r: 10-
(U
Q)
2
l‘
v ' '
Small Large
35. LEI-Reference +lmpacted 3;. {Iv-Reference -o-Impacted
30- %- Id 30.
25- 25-
20- 20-
15- 15-
" Small Large .. Small Large

Figure 4.2: Mean (SE) taxa richness (taxa per sample) for (A)
total wetland, (B) emergent and (C) submergent habitats in
isolated palustrine wetlands. All four wetland treatments are
represented: small reference, large reference, small impacted
and large impacted. No signiﬁcant effects of wetland quality,
wetland size or an interaction effect were observed in this study.

117

APPENDIX A: LIST OF COLLECTED MACROINVERTEBRATE TAXA FROM

MUSKEGON RIVER WATERSHED INLAND MARSHES

118

Table A1: List of macroinvertebrate taxon identiﬁed from inland marsh
wetlands in the Muskegon River Watershed. Michigan.

 

 

Class Order Family Genera
Insecta Coleoptera Chrysomelidae immature
Insecta Coleoptera Curculionidae immature
Insecta Coleoptera Curculionidae Adult 1
Insecta Coleoptera Curculionidae Type 2
Insecta Coleoptera Dytiscidae Acilius
Insecta Coleoptera Dytiscidae Agabus
Insecta Coleoptera Dytiscidae Celina
Insecta Coleoptera Dytiscidae Celina
Insecta Coleoptera Dytiscidae Coptotomus
Insecta Coleoptera Dytiscidae Desmopachria
Insecta Coleoptera Dytiscidae Desmopachria
Insecta Coleoptera Dytiscidae Dytiscus
Insecta Coleoptera Dytiscidae Dytiscus
Insecta Coleoptera Dytiscidae Hydraticus
Insecta Coleoptera Dytiscidae Hydrovatus
Insecta Coleoptera Dytiscidae Hygrotus
Insecta Coleoptera Dytiscidae Hygrotus
Insecta Coleoptera Dytiscidae Laccophilus
Insecta Coleoptera Dytiscidae Liodessus
Insecta Coleoptera Dytiscidae Matus
Insecta Coleoptera Dytiscidae Neoporus
Insecta Coleoptera Dytiscidae Rhantus
Insecta Coleoptera Dytiscidae Uvarus
Insecta Coleoptera Elmidae Dubiraphia
Insecta Coleoptera Elmidae Macronychus
Insecta Coleoptera Elmidae Stenelmis
Insecta Coleoptera Gyrinidae Dineutus
Insecta Coleoptera Haliplidae Haliplus
Insecta Coleoptera Haliplidae Haliplus
Insecta Coleoptera Haliplidae Peltodytes
Insecta Coleoptera Haliplidae Peltodytes
Insecta Coleoptera Helophoridae Helophorus
Insecta Coleoptera Hydrophilidae Anacera
Insecta Coleoptera Hydrophilidae Berosus
Insecta Coleoptera Hydroptilidae Berosus
Insecta Coleoptera Hydrophilidae Enochrus
Insecta Coleoptera Hydrophilidae Enochrus
Insecta Coleoptera Hydrochidae Hydrochus
Insecta Coleoptera Hydrophilidae Hydrobius
Insecta Coleoptera Hydrophilidae Paracymus
Insecta Coleoptera Hydrophilidae Paracymus
Insecta Coleoptera Hydrophilidae Tropisternus
Insecta Coleoptera Hydrophilidae Tropisternus
Insecta Coleoptera Noteridae Hydrocanthus
Insecta Coleoptera Noteridae Hydrocanthus

 

119

Table A1 (cont'd)

 

 

Class Order Family Genera
Insecta Coleoptera Scirtidae Microcara
Insecta Coleoptera Scirtidae Scirtes
Insecta Coleoptera Staphylinidae Stenus
Insecta Coleoptera Staphylinidae Stenus
Insecta Coleoptera Curculionidae Tanysphyrus
Insecta Collembola Entomobryidae

Insecta Collembola Poduridae

Insecta Collembola Sminthuridae

Insecta Diptera Ceratopogonidae Pupae
Insecta Diptera Ceratopogonidae Atrichopogon
Insecta Diptera Ceratapogonidae Culicoides
Insecta Diptera Ceratapogonidae Dashylea
Insecta Diptera Ceratapogonidae lVlalIochohelea
Insecta Diptera Ceratapogonidae Sphaeromias
Insecta Diptera Chaoboridae Chaoborous
Insecta Diptera Chironomidae Immature
Insecta Diptera Chironomidae Pupae
Insecta Diptera Culicidae Aedes
Insecta Diptera Culicidae Anopheles
Insecta Diptera Culicidae Mansonia
Insecta Diptera Dixidae Dixella
Insecta Diptera Ephydridae Parydra
Insecta [Diptera Ephydridae Setacera
Insecta Diptera Larvae Type 1
Insecta Diptera Sciomyzidae Sepedon
Insecta Diptera Sciomyzidae Type 1
Insecta Diptera Sciomyzidae Type2
Insecta Diptera Sciomyzidae Type 3
Insecta Diptera Simuliidae Immature
Insecta Diptera Stratiomyidae Odontomyia
Insecta Diptera Stratiomyidae Stratiomys
Insecta Diptera Tabanidae Chrysops
Insecta Diptera Tabanidae Hybomitra
Insecta Diptera Tabanidae Merycomyia
Insecta Diptera Tipulidae Helius
Insecta Diptera Tipulidae Ormosia
Insecta Diptera Corethrellidae Corethrella
Insecta Ephemeroptera Caenidae Caenis
Insecta Ephemeroptera Ephemerellidae Eurylophella
Insecta Hemiptera Naucoridae Pelocoris
Insecta Hemiptera Belastomatidae Belastoma
Insecta Hemiptera Gerridae Neogerris
Insecta Hemiptera Hebridae Hebrus
Insecta Hemiptera Hebridae Merragata
Insecta Hemiptera Hydrometridae Hydrometra
Insecta Hemiptera Mesoveliidae Mesovelia
Insecta Hemiptera Nepidae Ranatra

 

120

Table A1 (cont'd)

 

 

Class Order Family Genera
Insecta Hemiptera Notonectidae Notonecta
Insecta Hemiptera Pleidae Neoplea
Insecta Hemiptera Veliidae Microvelia
Insecta Lepidoptera Pyralidae Munroessa
Insecta Lepidoptera Pyralidae Parapoynx
Insecta Megaloptera Corydalidae Chauloides
Insecta Megaloptera Sialidae Sialis
Insecta Neuroptera Sisyridae Sisyra
Insecta Odonata Aeshnidae Aeshna
Insecta Odonata Aeshnidae Anax
Insecta Odonata Calopterygidae Calopteryx
Insecta Odonata Calopterygidae Hetaerina
Insecta Odonata Coenagrionidae CoenagrionlEnallagma
Insecta Odonata Coenagrionidae Enallagma
Insecta Odonata Coenagrionidae Ischnura
Insecta Odonata Corduliidae Cordulia
Insecta Odonata Gomphidae Gomphus
Insecta Odonata Lestidae Lestes
Insecta Odonata Libellulidae Erythemis
Insecta Odonata Libellulidae Leucorrhinia
Insecta Odonata Libellulidae Libellula
Insecta Odonata Libellulidae Pachydiplax
Insecta Odonata Libellulidae Plathemis
Insecta Odonata Libellulidae Sympetrum
Insecta Trichoptera [Hydroptilidae Agraylea
Insecta Trichoptera Hydroptilidae Orthotrichia
Insecta Trichoptera Hydroptilidae Oxyethira
Insecta Trichoptera Leptoceridae Pupae
Insecta Trichoptera Leptoceridae Ceraclea
Insecta Trichoptera Leptoceridae Leptocerus
Insecta Trichoptera Leptoceridae Oecetis
Insecta Trichoptera Leptoceridae Triaenodes
Insecta Trichoptera Limnephilidae Grammotaulius
Insecta Trichoptera Limnephilidae Mystacides
Insecta Trichoptera Phryganeidae Fabria
Annelida Hirudinea
Crustacea Amphipoda Gammaridae Gammarus
Crustacea Amphipoda Pontoporeiidae Monoporeia
Crustacea lsopoda Asellidae Lirceus
Mollusca Bivalva Sphaeridae
Mollusca Gastropoda Hydrobiidae Amnicola
Mollusca Gastropoda Lymnaeidae Lymnea
Mollusca Gastropoda Lymnaeidae Stagnicola
Mollusca Gastropoda Physidae Physa
Mollusca Gastropoda Planorbidae Armiger
Mollusca Gastropoda Planorbidae Gyraulus
Mollusca Gastropoda Planorbidae Helisoma
Mollusca Gastropoda Valvatidae Valvata

. Mollusca Gastropoda Viviparidae Campeloma
Mollusca Gastropoda Viviparidae Viviparus

 

121

APPENDIX B: RECORD OF DEPOSITION OF VOUCHER SPECIMENS

122

Appendix B
Record of Deposition of Voucher Specimens*
The specimens listed on the following sheet(s) have been deposited in the named museum(s) as
samples of those species or other taxa, which were used in this research. Voucher recognition
labels bearing the Voucher No. have been attached or included in ﬂuid-preserved specimens.
Voucher No.: 2007-03

Title of thesis or dissertation (or other research projects):

THE USE OF MACROINVERTEBRATES AS INDICATORS OF WETLAND QUALITY IN THE
MUSKEGON RIVER WATERSHED, MICHIGAN
Museum(s) where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (MSU)

Other Museums:

lnvestigator’s Name(s) (typed)
M IlieM Intosh

 

 

 

Date 12/07I2007

*Reference: Yoshimoto, C. M. 1978. Voucher Specimens for Entomology in North America.
Bull. Entomol. Soc. Amer. 24: 141-42.

Deposit as follows:
Original: Include as Appendix B in ribbon copy of thesis or dissertation.

Copies: Include as Appendix B in copies of thesis or dissertation.
Museum(s) ﬁles.
Research project ﬁles.

This form is available from and the Voucher No. is assigned by the Curator, Michigan State
University Entomology Museum.

123

Appendix B

Voucher Specimen Data

of 9 Pages

 

 

3.0.02.5 0.9m. 00900:). 0:. E 000000
.0. 0008.800 00.0.. 0>000 0... 0020000.

 

No 10.00,“ .oz 0.009

 

 

 

not-NF 0.00
0090.05. 0:85.
.0003. 8.00.02 0..0.0m..00>0.

30000000 ._ 0.0000 .000...000 003.

 

 

Page 1

 

 

 

 

 

 

 

 

 

322 F 00001000012. 0.00002 .2 078 .00 00.8022 80.00020
30.2 0 00001000012. 0.0802 .2 078 .00 03902: 00000.06
30.2 F 008103.35 0009.002 .2 0.8 .00 00.03.02: 000.0008
30.2 F 0000108012. 0.0800 .2 3-8 .00 08.0.0022 00000.08
392 F 00001-000012. 5.00.002 .2 078 .00 080.020 00000.06
30.2 F 00001000012. 005288.. .2 078 .00 030.020 8000......
30.2 0 000010.001... 0.0800 .2 ..-8 .00 00000005000 80.00.30
30.2 F 00001000012. 3.8002 .2 078 .00 00000008000 80.8.0.0
392 N 00001000012. 0.0802 .2 00-8 .00 00.00.2000 000.0008
30.2 F 000015312. 8.0.0 .2 00-8 .00 00.000 00000.06
392 F 0081000012. 00.00.8000 .2 00-8 .00 00.000 00000.08
30.2 F 00001000012. 090588.. .2 00-8 .00 8000.0 80.00.20
30.2 F 008180012. 0.0.0.02. .2 00-8 .00 0:00... 00000.06
30.2 0 00001000012. 002.03% .2 8-8 .00 0 00? 000.02.00.00
30.2 F 00001000012. 0.0802 .2 00-8 .00 3.00... 000.02.00.00
30.2 0 00001000012. 30802 .2 00-8 .00 05.0.0.0. 000.00.30.00
30.2 F 008100.012. 0.0.52. .2 8-8 .00 0.0.0.0.... 80.00.0006

m m s .0000 00

m m M m Pm m m. m. m .000 0000.0 00.00.80 00088000 www.0.00 "00mg 09.0. 000.0 .0 00.00am

m .w. 0 0 .0 0 m m. 0 .0.

 

 

”.0 0000.02

 

 

124

Appendix B

Voucher Specimen Data

0.00 8.050

 

80000.2 200880.00
0.0.02.5 0.0.0 00000.2 0... 0. 000000
.0. 05.0.0000 00.0.. 0800 0... 002000”.

 

 

 

 

BENF 0.00
000.522 0:85.
8002.. 0.0802 08.09.0022

 

 

Page 2 of 9 Pages

 

 

 

 

 

 

 

 

 

 

 

 

 

.oz .0..o:o> 0.0000000: .. 0.0000 8:00.000 003.
30.2 F 008180012. 0.00.02. .2 00-8 .00 0000.00 000.....0000m
30.2 F 0081000012. 00.00.8000 .2 00-8 .00 0.00000. 000.0090»:
30.2 0.. 008108012. 0.00000 .2 00-8 .00 030200.01 00.009.00.00
30.2 F 0081000012. 0.0802 .2 .0-8 .00 00.000200 00000.00.
30.2 0 00001000212. 8000.002 .2 00-8 .00 00.02.00 00000.0:
30.2 0 0081000012. 00.00.8000 .2 00-8 .00 00.3.0: 000.0..0:
30.2 F 0000130012. 2.00082 .2 00-8 .00 00.0.0... 000.000..
30.2 F 0000100012. 0.00000 .2 00-8 .00 00.00.00 000.0000
30.2 F 00001000212. 0000.002 .2 00-8 .00 08.00000 000.....0
30.2 F 008320212. 8000.002 .2 00-8 .00 000000002 0000.0
30.2 F 00810300012. 0.0.0 .2 00-8 .00 0.0000000 0.00.5.0
30.2 F 0081000012. 0.0802 .2 00-8 .00 00003 00000000
30.2 F 000.010.000.12. 8000.002 .2 00-8 .00 030020 00000000
30.2 F 00001000012. 0.0802 .2 .0-8 .00 00000002 00000020
30.2 0 0000100212. 8000.002 .2 00-8 .00 00.0:. 00000000
30.2 0 0081000012. 00.00.8000 .2 078 .00 00000.00: 00000000
30.2 F 008108012. 0.0.0.02. .2 078 .00 00.308000 00000.0
m m m e 00000000 000
m Mr” M w m m m m m % 0000.0 00.00.80 00088000000000.0003 80.0.0508 00.0000
0 m 0 0 0 0 0 m. 0 00
00.00802

125

Appendix B

Voucher Specimen Data

mama

.9930

 

Enema: 30.0595
3.99.5 99w 55.6.2 9.. c. 58.8

 

 

 

.2 992.0me 35.. m>onm 9: 328mm

 

BENF 9mm.
59222 2:22
335 Amvamz 305959...

 

 

Page 3 of 9 Pages

 

 

 

 

 

 

 

 

 

.oz .m..o:o> 5338: F. £36 .9858 $3.
322 F 85:39.12. 5.8.82 .2 5-8 .am 85.38.36
:92 m 88:32:; 3.882 .2 8-5 .3 32:5me 08.5.8.5
322 F 85:39.2 985.2 .2 5.15 .am 85% 58.2.2..qu
:22 F 85:83:; 85588”. .2 3-5 .am 855 825.28%
:ws. 9 85:38:; 8.3822 .2 5-5 .am metam- 32:3
392 F 85-2532 3882 .2 3-5 .8 8885.: 82:3
322 N 88:38:... 38052 .2 3-5 .% mascmueuxm 82.802
392 F 85:88:; 53882 .2 3-5 .3 £5289an 03.882
:22 F 85155512. 286; .2 915 .3 85333 32.28.52:
:22 F 88:38:; 3882 .2 $25 .3. maEBmao: 32.28.52:
:92 N 85153.3... 558.32 .2 :15 .9". masxuemo. 03:28.3.
322 m 85:88:? SEES-8”. .2 3-5 .% 8:589. 82.28.52:
:22 F 85:89:; «.988 .2 3-5 .3 3593: 82.2862:
3w: F 85:88:; SEES-8”. .2 8-5 .3. 828.3: 8.26952:
392 F 88:39.12, 2.8.82 .2 5-5 .am 2585 32.2852:
:ws. F «80:38:; 3.82.2 .2 3.5 .3 95.85. 82.2852:
322 F 85.3312. SEES-v.3. .2 3.5 .9. 385m 82.592:
m m s .88 cm
mmm Mr. m m m m. m w 835 352.8 mcmEBmamwwwgmu “enema. 5535505 8.0on
m m w m m M m m m ,m.

 

 

go .3252

 

 

126

Appendix B

Voucher Specimen Data

060

.2230

 

.Eammas. 30.0225
3.20203 22w 08.50:... 0... E 200000
.2 808.00% 020.. m>00m 05 0020001

 

 

 

 

BENF 28
2020.05. 9:05.
8002.. .mvamz 01.282008.

 

 

Page 4 of 9 Pages

 

.0z .0...0:0> 380000: 2 202.0 .8223“. 00:.
:m: N 88:28:... 2830 .2 3-8 .3 88.8 82...»..&
3.0.... F 88:885.... 859.82 .2 3-8 .3 88“. 828.33
:22 F 88:88:... 2830 .2 2-8 .3 2.8.0 82.8
:22 N 88:88:... 3882 .2 ~78 .3 8382 822.8
:92 F 85:28:... 2830 .2 2-8 .3 8.2.8:... 822.30
:92 F 88:88:... .2882 .2 078 .3 88... 822.5
:22 N 88:88:... 3882 .2 8-8 .3 8288.20
:92 N 88:88:... .3882 .2 8-8 .3 8288.20
3m... F 88:88:... 8882 .2 8-8 .3 8882.0 82.885
:22 F 88882:... 88282 .2 8-8 .3 8.8.823- 82883.80
:22 F 88:89:; 283.0 .2 m0-m0 .3 8.2.2.0082 82888880
:92 m 88:88:; 22.8.... .2 8-8 .3 8.880 828508880
:22 F 85:28.2. 85833. .2 8-8 .3 88.8.80 82888880
392 N 88:058.... 828.. .2 8-8 .3 8888...... 82888880
:92 N 88:08.3... 85832. .2 5-8 .3 8&5 82888880
392 m NONOImmelg 0.0.28.5 .2 mo-No dm 80.55023
:9... F 88:38:... 5.8.82 .2 8-8 .3 82.88
m m m e 02.00000 0:0 ,
m m M Mr. m m m m m $ 008.0 0209.00 805.083228 .83 50.2.0508 8.00%
m m m. m M M m m. m a.

 

 

 

 

 

 

 

 

 

 

”.0 .8822

 

 

127

Appendix B

Voucher Specimen Data

0.00 .2050

 

830022 30.00.25
3.82.... 08.0 8.0.8.... 0... c. .88..
.2 802.6000 020.. 0800 0... 00>.000m.

 

 

 

 

BEN F 0.00
8250.2 0...0.2
.0098 30.002 0.209.008.

 

 

Page 5 of 9 Pages

 

 

 

 

 

 

 

 

 

 

 

 

 

.oz .0..0:o> 00000000.. .. 0.008 .0..0...000 003.
:0.... F 88:83:... 8332 .2 8-8 .3 3.88.8 82.882
30.2 F 88:88:... 882.32 .2 8-8 .3 2.2.8.80 82:28.83
Ems. N N08308:; 88808.. .2 F080 3.. 88.0 82:80
:0.... F 88:058.... 88... .2 8-8 .3 288.0 82.8:
:0.... F 88:88:... 3882 .2 8-8 .3 8...... 82.8:
:05. F 88:80.1... 8.882 .2 8-8 .3 28882 82:83
30.2 F 88:98:... 8.83.2 .2 8-8 .3 8.88. 823an
:0.... F 88:80.35 3.882 .2 8-8 .3 8380 82......ng
3m... F 88:83:... 88.88.. .2 8-8 .3 88.88 8228.88
30.... F 88:38:... 6.882 .2 8-8 .3 8.28.800 82828.3
:0.... F 88:83:... 880.92 .2 8-8 .3 232...... 82.2.3
:0.... F 88:28:... 8.83.2 .2 2-8 .3 M83 82.4-28.3
30.2 N 88:88:... «.830 .2 2-8 .3 89¢ 8288.8
30.2 N 88:88:... 8.880 .2 2-8 .3 .83 82028.8
:0.... F 88:28:... .88... .2 078 .3 8880 8238.8
m m s .0000 :0 -
mmm w m m W. m m w 008 .0 00.00.60 80E.0000F..M.200 “00mm. :98. .050 .0 8.080
m m m. m M M m m. a a.
no .0282

128

Appendix B

Voucher Specimen Data

2mm.

.9930

 

.8385. 30.0895
3.28.5 .28 896.2 8. c. 88.8
.8 mcmE_omam 3E. m>onm 9: 328mm

 

 

 

 

BEN r Ema
59522 2:22
€098 @952 {23:83.

 

 

Page 6 of 9 Pages

 

.oz .m...o:o> 3888 t 935 .8363 83
322 F 88:28-2. 8.088 .2 8-8 .8 85‘ 82.82
:ws. e 88:83:; 889.82 .2 5-8 .8 888- 82:88.
am: m 88:83:; 3.882 .2 5-8 .8. «5a 8.2.28
:92 F 88:88:; 88.82 .2 8.8 .8 £8- 82.8
:92 N 88:88:; 2882 .2 8-8 .8 8.20.86 82.880
:22 _. 888%qu SEES-8”. .2 8.8 .8 82088“. 8.2.8}.
:92 w 88:83:; 3882 .2 8-8 .8 888%: 8.2.85.
:22 F 88:83:; 85588”. .2 2-8 .8 £889.: 8o.._o>
:92 v 88:88:; 3.882 .2 9-8 .8 8.0.82 820.0.
:92 r 8810828. 889.82 .2 8-8 .8 282202 838882
:22 F 88:28:; 3882 .2 8-8 .8. 38:3. 8282
:22 .- 88:88:; 85588”. .2 8-8 .8 3.8082 828.682
:22 F 88:88:; 5.8882 .2 8-8 .8 E8883... 82.8692:
322 r 88182-2. 285; .2 8-8 .8 88852 82.8...
:92 v 88:28:; 8.88.2 .2 8.8 .8 8.8: 82.8:
:92 e 88:53:; 9.8.0 .2 8-8 .8 8882 89:8
:22 F 88:85:; SEES-0.3. .2 No-8 .3 «888.8 828E888
m m m e 3:830 ccm
mmm m m m W W m m. 88.0 882.8 mcmE.omam.£.m.mu .83 coxmtmseo 8.0on
M m «u. m M M ou- Nvu m cm.

 

 

 

 

 

 

 

 

 

 

no .8832

 

 

129

Appendix B

Voucher Specimen Data

0.00

.2050

 

.8308: 30.00.95
3.202.... 0.90 826.2 0... c. 888
.0. 808.0000 00.0.. 0>000 0c. 00>.000m

 

 

 

 

BEN F 0.00
58.0.0.2 050.2
600.5 @0802 0.0.0989...

 

 

Page 7 of 9 Pages

 

.02 .0..0:0> 308000: .. 0.00:0- .0:0...000 0.0.3.
:22 m 88808:... 5055080.. .2 8.3 .8 8.5.0820 82.08.02:
3w... F 88:225.]... 8.88.2 .2 8-0. .8 2.05.2.5 82.08.02:
:w... m 88.2.8.4... 5.8.82 .2 8.2 .8 8.8.2.. 82.08.02...
3w... N 88:80.1; 3.882 .2 078 .8 53.8.58- 82.2.8...
:05. F 88:88:; 5.8502 .2 2-8 .8 0.52.8... 82.2.8:
0w... F 881.211.... .8882 .2 1-8 .8 8.2.8.8. 82.2.8...
3w... F 88:88:; 505588.. .2 2-8 .8 22.00.... 82.2.8...
:22 N 88:85-... 505588.. .2 N 78 .8 22:50.80. 82.2.8...
:22 F 88i......n.s 30802 .2 F78 .8 0.5052... 82.2.00...
:ws. F 8848.-.... 20080 .2 078 .8 8.8.. 82.8..
00.2 F 88100813 8882 .2 8-8 .8 3.0500 825.500
30.5. N 88:53.... 2.8.0 .2 8-8 .8 2.8.00 82.3500
30.2 N 881082-... 588.82 .2 8-8 .8 5:558. 82:28:80
am... F 88:..sz.... 2880 .2 8-8 .8 858.85... 825058580
0w... F 88.23:... 5.8.502 .2 8-8 .8 858.82.888.80 82.2.8580
:92 F 88:82.3... 20080 .2 8-8 .8 8:08.01 82838.8
3.0.... F 88-83-; 589.82 .2 8-8 .8 80.8.80 82888.8
m m m e 00.50000 0:0 .
m m m m m m m m. m .50.. 008 .0 00.00.60 80E_0000 .0..0.00 .000. 00x0. .050 .0 00.00%
m m w m m m m m. a a.

 

 

 

 

 

 

 

 

 

 

”.0 .00E32

 

 

130

Appendix B

Voucher Specimen Data

0.00

.9080

 

 

8.0.02.5 0.0.0 000.00.... 0.... 0. 000000
.9 0006.080 00.0.. 0>000 0... 002000”.

 

 

BENF 0.00

83003.2 30.00.95
0090.05. 0505.
.08.... 8.0502 0.0.8.8....

 

 

 

Page 8 of 9 Pages

 

 

 

 

 

 

 

 

 

 

.02 .0..030> 2.000000: .. 0.00:0 .0..0...000 00:.
:0.... F 88:88:... 0.0082 .2 00-0. .8 80.0 82080
:0.... F 88:08:... 0.880 .2 8-0. .8 0.8.08.0 82805.:
30.2 8 88:88:... 8.880 .2 8-0. .8 0.00.0E< 82.00.02:
:0.... 0 88:80.1... 880.002 .2 8-0. .8 82.8.8
:0.... 0.. 88:...<2:... 88.002 .2 8-0. .8 0000.... 82.8..
:0.... 0 88882:... 0080.032 .2 8-0. .8 0.0.0080... 82.0.0038.
30.2 on 88:83:... 0080.002 .2 8.0. .8 00.05500 82.05500
00.... F 88:88:... 0055080.. .2 .0-.. .8 08.02....
am... N 88:28:... 0.00002 .2 .70. .00 0.80.. 8288......
:0.... F 08:22.24... 8880.2 .2 0.0. .8 80.880. 82.2085:
00.2 F 88:82.35 0.880 .2 8.0. .8 00.00.0550... 82.2085...
:0... N 88802:... 008.032 .2 8.0. .8 80008.... 82.80.08
20.... F 88:88:... 0.0.8.... .2 8.0. .8 0.00000 8280.8.
Ems. N Nomolmmzml... 0.00000 .2 00.0. .00 0300900.. 80008.00.
Dws. F 88:85:... 0.0080 .2 00.0. .8 00.00.00 82.80.00.
:0.... F 88:83:... 8.0082 .2 8.0. .8 80:0 82.80.08
m M: s .0000 :0
mmm m Pm m m m m w 0003.0 00.00.60 0..0....0000....0....0.00 “.208. 00x0. .050 .0 00.0000
0 m 0 0 M M m m. 0 0..

 

 

”.0 .8252

 

 

131

Appendix B

Voucher Specimen Data

mama

.2930

 

Sagas. 30.0E2cw

£292: 99w amazes. 9: 5 gmoamu
.2 2056on 09m: m>onm m5 umzmowm

 

 

 

 

BEN? Ema
F$950.2 2:22
68>: 3252 3039395

 

 

Page 9 of 9 Pages

 

 

 

 

 

 

 

 

 

.oz .mcoao> Emmmoooc : 93% .2363 33

DwE dm

3w: dm

DmE dm

3m: dm

DwE dm

DwE dm

3m: dm

3w: dm

3m: dm

Dmi dm

DmE dm

:ws. me Nowouﬁzmu; 288.0 :2 9-2 .3 333.3 «3:335

:92 N Nomouﬁxmu; 2880 =2 8.? .3 «E9856 3.2.835

:22 :V NooouE<za3 comoxmaz .2 8-2 .3 338, 825,;

am: N 88:32:; «.880 .2 3-2 .am «50%: $2922“.

:22 m 88:58:; 85888”. =2 8-? .9». 833% @8385:

:w: m 88:88:; 08332: :2 no-2 .Qm EQE< omgeocma
m m. s .m .
mmm m. Pm m w W m w 88.0 882.8 mcméomamwwwvmuoﬂummmm cox—35505 «90QO
m m w m m M m m. u. a.

 

 

«o 89:32

 

 

132

LITERATURE CITED

133

Literature Cited

Adamus, P. R. 1996. Bioindicators for assessing ecological integrity of prairie
wetlands. in U. S. E. P. Agency, editor.

Apfelbeck, R. S. 2001. Development of biocriteria for wetlands in Montana.
Pages 141-166 in R. B. Rader, D. P. Batzer, and S. A. Wlssinger, editors.
Bioassessment and Management of North American Freshwater
Wetlands. John Wiley and Sons, New York.

Babbitt, K. J. 2005. The relative importance of wetland size and hyrdoperiod for
amphibians in southern New Hampshire, USA. Wetlands Ecology and
Management 13:269-279.

Baber, M. J., E. Fleishman, K. J. Babbitt, and T. J. Tarr. 2004. The relationship
between wetland hydroperiod and nested patterns in assemblages of
larval amphibians and predatory macroinvertebrates. OIKOS 137:16-27.

Batzer, D. P., R. B. Rader, and S. A. Wrssinger. 1999. Invertebrates in
Freshwater Wetlands of North America‘. John Wiley and Sons, New York.

Batzer, D. P., A. S. Shurtleff, and R. B. Rader. 2001. Sampling Invertebrates in
Wetlands. Pages 339-354 in R. B. Rader, D. P. Batzer, and S. A.
Wlssinger, editors. Bioassessment and Management of North American
Freshwater Wetlands. John Vlﬁley and Sons, New York.

Brown, M., and J. J. Dinsmore. 1986. Implications of marsh size and isolation for
marsh bird management. Journal of VWdIife Management 50:392-397.

Brown, S., and D. P. Batzer. 2001. Birds, plants and macroinvertebrates as
indicators of restoration success in New York Marshes. Pages 469 in R. B.
Rader, D. P. Batzer, and S. A. Wlssinger, editors. Bioassessment and
Management of North American Freshwater Wetlands. John Wiley and
Sons, New York.

Burton, T. M., D. G. Uzarski, J. P. Gathman, J. A. Genet, B. E. Keas, and C. A.
Stricker. 1999. Development of a preliminary invertebrate index of biotic
integrity for Lake Huron coastal wetlands. Wetlands 19:869-882.

Chessman, B. C., K. M. Traylor, and J. A. Davis. 2002. Family— and species-level
biotic indices for macroinvertebrates of wetlands on the Swan Costal
Plain, Western Australia. Marine and Freshwater Research 53:9919-9930.

Christie, J., and S. Hausmann. 2003. Various state reactions to the SWANCC
decision. Wetlands 23:653-662.

134

Cooper, M. J., D. G. Uzarski, T. M. Burton, and R. R. Rediske. 2006.
Macroinvertebrate community composition relative to chemical/physical
variables, land use and cover, and vegetation types within a Lake
Michigan drowned river mouth wetland. Aquatic Ecosystem Health and
Management 92463-479.

Cummins, K. W. 1974. Structure and function of stream ecosystems. Bioscience
24:631-641. '

Cummins, K. W., and R. W. Merritt. 2001. Application of invertebrate functional
feeding groups to wetland ecosystem function and biomonitoring. Pages
85-111 in R. B. Rader, D. P. Batzer, and S. A. Wissinger, editors.
Bioassessment and Management of North American Freshwater
Wetlands. John Wiley and Sons, New York.

Danielson, T. J. 1998. Wetland Bioassessment Fact Sheets. in U. S. E. P.
Agency, editor.

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.

Downes, B. J., P. S. Lake, and E. S. G. Schreiber. 1993. Spatial variation in the
distribution of stream invertebrates: implications of patchiness for models
of community organization. Freshwater Biology 30:119-132.

Gernes, M. C., and J. C. Helgen. 2002. Indexes of Biological Integrity (IBI) for
Large Depressional Wetlands in Minnesota. CD-995525-01, Minnesota
Polloution Control Agency.

Halse, S. A., D. J. Cale, E. J. Jasinska, and R. J. Shiel. 2002. Monitoring change
in aquatic invertebrate biodiversity: sample size, faunal elements and
analytical methods. Aquatic Ecology 36:395-410.

Helgen, J. C. 2001. Methods for Evaluating Wetland Condition: Developing an
invertebrates index of biological integrity for wetlands. in U. S. E. P.
Agency, editor.

Helgen, J. C., and M. C. Gernes. 2001. Monitoring the condition of wetlands:
Indexes of biological integrity using invertebrates and vegetation. Pages
167-186 in R. B. Rader, D. P. Batzer, and S. A. Wissinger, editors.
Bioassessment and Management of North American Freshwater
Wetlands. John Wiley and Sons, New York.

Hellawell, J. M. 1986. Biological Indicators of Freshwater Pollution and
Environmental Management. Elsevier, London.

135

Hierl, L. A., C. Loftin, S., J. R. Longcore, D. G. McAuley, and D. L. Urban. 2007.
A multivariate assessment of changes in wetland habitat for waterbirds at
Moosehorn National Wildlife Refuge, Maine, USA. Wetlands 27.

Houlahan, J. E., P. A. Keddy, K. Makkay, and C. S. Findlay. 2006. The effects of
adjacent land use on wetland species richness and community
composition. Wetlands 26:79-96.

Kashian, D. R., and T. M. Burton. 2000. A comparison of macroinvertebrates of
two great lakes coastal wetlands: testing potential metrics for an index of
ecological integrity. Journal of Great Lakes Research 26:460-481.

King, R. S., and J. C. Brazner. 1999. Coastal wetland insect communities along a
trophic gradient in Green Bay, Lake Michigan. Wetlands 19:426—437.

King, R. S., K. T. Nunnery, and C. J. Richardson. 2000. Macroinvertebrate
assemblage response to highway crossings in forested
wetlandszimplications for biological assessment. Wetlands Ecology and
Management 82243-256.

King, R. S., and C. J. Richardson. 2002. Evaluating subsampling approaches
and macroinvertebrate taxonomic resolution for wetland bioassessment.
Journal of the North American Benthological Society 21:150-171.

Kotliar, N. B., and J. A. Wrens. 1990. Multiple scales of patchiness and patch
structure: A hierarchical framework for the study of heterogeneity. OIKOS
29:253-260.

Kratzer, E. B., and D. B. Batzer. 2007. Spatial and temporal variation in aquatic
macroinvertebrates in the Okefenokee Swamp, Georgia, USA. Wetlands
27:127-140.

Larson, D. J., Y. Alarie, and R. E. Rougjley. 2000. Predaceous Diving Beetles
(Coleoptera;Dytiscidae) of the Nearctic Region. National Research
Council of Canada, Ottawa.

Lemly, A. D., and R. S. King. 2000. An insect-bacteria bioindicator for assessing
detrimental nutrient enrichment in wetlands. Wetlands 20291-100.

Lopez, R. D., C. B. Davis, and S. Fennessy. 2002. Ecological relationships
between landscape change and plant guilds in depressional wetlands.
Landscape Ecology 00:1-14.

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. .

136

Lougheed, V. L., and P. Chow-Fraser. 2002. Development and use of a
zooplankton index of wetland quality in the Laurentian Great Lakes Basin.
Ecological Applications 12:474-486.

Lougheed, V. L., M. D. McIntosh, C. A. Parker, and R. J. Stevenson. 2006.
Contribution of small, isolated wetlands to local and landscape-scale
biodiversity. Biological Conservation in press.

Lougheed, V. L., C. A. Parker, and R. J. Stevenson. 2007. Using non-linear
responses of multiple taxonomic groups to establish criteria indicative of
wetland biological condition. Wetlands 27:96-109.

Mallory, M. L., P. J. Blancher, P. J. Weatherhead, and D. K. McNicol. 1994.
Presence or absence of ﬁsh as a cue to macroinvertebrate abundance in
boreal wetlands. Hydrobiologia 279/280z345-351.

McArthur, J. V. 2001. Bacteria as Biomonitors. Pages 469 in R. B. Rader, D. P.
Batzer, and S. A. Wissinger, editors. Bioassessment and Management of
North American Freshwater Wetlands. John Vlﬁley and Sons, New York.

McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MJM
Software Design, Gleneden Beach.

Merkley, M., R. B. Rader, J. V. McArthur, and D. Eggett. 2004. Bacteria as
bioindicators in wetlands: bioassessment in the Bonneville Basin of Utah,
USA. Wetlands 24:600-607.

Merritt, R. W., M. E. Benbow, and P. L. Hudson. 20023. Wetland
macroinvertebrates of Prentiss Bay, Lake Huron, Michigan: Diversity and
functional feeding group composition. The Great Lakes Entomologist
35:149-160.

Merritt, R. W., and K. W. Cummins. 1996. An Introduction to the Aquatic Insects
of North America, 3rd edition. Kendall/Hunt Publishing Co., Dubuque,
Iowa.

Merritt, R. W., K. W. Cummins, M. B. Berg, J. A. Novak, M. J. Higgins, K. J.
Wessell, and J. L. Lessard. 2002b. Development and application of a
macroinvertebrate functional-group approach in the bioassessment of
remnant river oxbows in southwest Florida. J. N. Am. Benthol. Soc.
21:290-310.

Merritt, R. W., M. J. Higgins, K. W. Cummins, and B. Vandeneeden. 1999. The
Kissimmee River-Riparian Marsh Ecosystem, Florida; Seasonal
Differences in Invertebrate Functional Feeding Group Relationships.
Pages 55 - 79 in D. P. Batzer, R. B. Rader, and S. A. Wissinger, editors.
Invertebrates in Freshwater Wetlands of North America: Ecology and
Management. John Wiley & Sons, Inc.

137

Merritt, R. W., J. B. Wallace, M. J. Higgins, M. K. Alexander, M. B. Berg, W. T.
Morgan, K. W. Cummins, and B. Vandeneeden. 1996. Procedures for the
functional analysis of invertebrate communities of the Kissimmee River-
ﬂoodplain ecosystem. Florida Scientist 59:216-274.

Mitsch, W. J., and J. G. Gosselink. 2000. Wetlands. John Vlﬁley and Sons, New
York.

Nadeau, T., and S. G. Leibowitz. 2003. Isolated wetlands: An introduction to the
special issue. Wetlands 23.

Pennack, R. W. 1989. Freshwater Invertebrates of the United States: Protozoa to
Mollusca, 3rd edition. John Wiley and Sons, New York, NY.

Pijanowski, B., B. Shellito, S. Pithadia, and K. Alexandridis. 2006. Forcasting and
assesing the impact of urban sprawl in coastal watersheds along eastern

Lake Michigan. Lakes and Reservoirs: Research and Management 7:271-
285.

Quinn, J. P., and M. J. Keough. 2002. Experimental Design and Data Analysis for
Biologists. Cambridge.

Rader, R. B. 2001. An introduction to wetland bioassessment and management.
Pages 1-9 in Bioassessment and Management of North American
Freshwater Wetlands. John Wiley and Sons, New York.

Rader, R. B., and C. J. Richardson. 1994. Response of macroinvertebrates and
small fish to nutrient enrichment in the northern Everglades. Wetlands
14:134-146.

Ravera, O. 2001. A comparison between diversity, similarity, and biotic indices
applied to the macroinvertebrate community of a small: the Ravella river
(Como Province, Northern Italy). Aquatic Ecology 35297-107.

Rivers, J. W., and T. T. Cable. 2003. Evaluation of farmed playa wetlands as
avian habitat using survey data and two rapid assessment techniques.
Transactions of the Kansas Academy of Science 106:155-165.

Rosenberg, D. M., and V. H. Resh. 1993. Introduction to Freshwater
Biomonitoring and Benthic Macroinvertebrates. Pages 1-9 in D. M.
Rosenberg and V. H. Resh, editors. Freshwater Biomonitoring and
Benthic Macroinvertebrates. Chapman and Hall, New York.

Simon, T. P., R. Jankowski, and C. Morris. 2000. Modification of an index of
biotic integrity for assessing vernal ponds and small palustrine wetlands
using fish, crayfish and amphibian assemblages along southern Lake
Michigan. Aquatic Ecosystem Health and Management 32407-418.

138

Snodgrass, J. W., M. J. Komoroski, A. L. Bryan, and J. Burger. 2000.
Relationships among isolated wetland size, hydroperiod, and amphibian
species richness: Implications for wetland regulations. Conservation
Biology 14:414-419.

Steinman, A. D., J. Conklin, P. J. Bohlen, and D. G. Uzarski. 2003. Inﬂuence of
cattle grazing and pasture land use on macroinvertebrate communities in
freshwater wetlands. Wetlands 23:877-889.

Stevenson, R. J. 2001. Using algae to assess wetlands with multivariate
statistics, multimetric indices and an ecological risk assessment
framework. Pages 469 in R. B. Rader, D. P. Batzer, and S. A. Wissinger,
editors. Bioassessment and Management of North American Freshwater
Wetlands. John Vlﬁley and Sons, New York.

Tangen, B. A., M. G. Butler, and M. J. Ell. 2003. Weak correspondence between
macroinvertebrate assemblages and land use in prairie pothole region
wetlands, USA. Wetlands 23: 1 04-1 15.

Thorp, J. H., and A. P. Covich. 1991. Ecology and classiﬁcation of North
American freshwater invertebrates. Academic Press Inc., San Diego, CA.

Tiner, R. W. 2003a. Estimated extent of geographically isolated wetlands in
selected areas of the United States. Wetlands 23:636-652.

Tiner, R. W. 2003b. Geographically isolated wetlands of the United States.
Wetlands 23:494-516.

Torbick, N. M., J. Qi, G. J. Roloff, and R. J. Stevenson. 2006. Investigating
impacts of land-use landcover change on wetlands in the Muskegon River
Watershed, Michigan, USA. Wetlands 26:1103-1113.

Townsend, C. R., M. R. Scarsbrook, and S. Doledec. 1997. The intermediate
disturbance hypothesis, refugia, and biodiversity in streams. Limnol.
Oceanogr. 42:938-949.

Uzarski, D. G., T. M. Burton, and J. A. Genet. 2004. Validation and performance
of an invertebrate index of biotic integrity for Lake Huron and Michigan
fringing wetlands during a period of lake level decline. Aquatic Ecosystem
Health and Management 71269-288.

van Dam, R. A., C. Camilleri, and C. M. Finlayson. 1998. The potential of rapid
assessment techniques as early warning Indicators of wetland
degredation.297-312.

Wilcox, D. A-., J. E. Meeker, P. L. Hudson, B. L. Armitage, M. G. Black, and D. G.
Uzarski. 2002. Hydrologic variability and the application of index of biotic

139

integrity metrics to wetlands: a Great Lakes evaluation. Wetlands 22:588-
615.

Wissinger, S. A. 1999. Ecology of Wetland Invertebrates: Synthesis and
applications for conservation and management. Pages 1043-1086 in D. P.
Batzer, R. B. Rader, and S. A. Wissinger, editors. Invertebrates in
Freshwater Wetlands of North America. John Wiley and Sons, New York.

Wissinger, S. A., S. G. Ingmire, and J. L. 8090. 2001. Plant and invertebrate
communities as indicators of success for wetlands restored for wildlife.
Pages 469 in R. B. Rader, D. P. Batzer, and S. A. Wissinger, editors.
Bioassessment and Management of North American Freshwater
Wetlands. John Wiley and Sons, New York.

Yangdong, P., and R. J. Stevenson. 1996. Gradient analysis of diatom
assemblages in western Kentucky wetlands. Journal of Phycology 32:222-
232.

Zimmer, K. D., M. A. Hanson, and M. G. Butler. 2000. Factors inﬂuencing
invertebrate communities in prairie wetlands: a multivariate approach.
Canadian Journal of Fisheries and Aquatic Science 57:76-85.

140

IllﬂllllﬂﬂlllllﬂljljﬂlllIlliilﬂljllll