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 LL DATE DUE DATE DUE
WI
NOV Q ‘3“

 

rm 5? 96155

 

 

4M 8%905
DE1C1116 €005

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuopes-p. t 5

THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON WATER CHEMISTRY
AND INVERTEBRATE COMMUNITY COMPOSITION IN A GREAT LAKES
COASTAL WETLAND
By

Craig Allen Stricker

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Zoology

2003

ABSTRACT
THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON WATER CHEMISTRY
AND INVERTEBRATE COMMUNITY COMPOSITION IN A GREAT LAKES
COASTAL WETLAND
By

Craig Allen Stricker

Fringing wetlands lie at the interface between terrestrial and lentic ecosystems.
The interaction between lake and wetland inﬂuences the hydrology and creates habitat
unique from other types of freshwater wetlands. Water level ﬂuctuations are common to
the Great Lakes and likely inﬂuence water quality and shape invertebrate communities,
yet potential relationships have not been explored. In this study, I documented these
changes in a Lake Huron coastal wetland over a four-year period that corresponded to a
1.04 m decline in water level.

A transect spanning a Scirpus-dominated wetland was established in 1997.
Sampling stations were spaced at 20-m intervals and extended 280 m shoreward of the
wetland/lake interface. Water chemistry (1997-2000) and invertebrate (1998-2000)
samples were collected at approximate monthly intervals during the active growing
season (June-September).

An increase in the total dissolved ion content of wetland surface water occurred in
conjunction with declines in lake level. Changes for all of the major anions and cations
were statistically signiﬁcant between years, particularly in 1999 and 2000 when lake
levels were below average. In 2000, water depth in the wetland averaged 0.09 m and

much of the near shore region of the wetland was devoid of standing water. As depth

declined, the combination of reduced lake/wetland mixing and a water table gradient of
10 cm per 100 m resulted in stronger interactions between wetland surface water and
sediments. Seiche-related water level ﬂuctuations increased connectivity between
interstitial and surface water. Higher concentrations of SO4'2, Mg”, and Si in 2000
relative to previous years supported tighter coupling between surface and interstitial
water.

A total of 60 invertebrate taxa were collected from the study wetland during the
three-year investigation. Richness and diversity increased as lake stage declined. Low
water levels stimulated the growth of benthic algae and likely excluded piscivorous
predators from the wetland. Scrapers increased 22% between 1998 and 2000, and
predators increased 6%. Invertebrate distributions within the wetland shifted in response
to declining lake levels, changes in hydrology, and food resources. Twenty-one taxa
were only collected during the low water year (2000) when water depth in the wetland
averaged 0.09 m. All of these taxa are common to temporary wetlands.

The documented changes in surface water chemistry and invertebrate community
composition illustrated the importance of water level ﬂuctuations in the ecology of Great
Lakes coastal wetlands. Climate change predictions for the Great Lakes basin suggest
that the frequency and duration of below average lake levels will increase.
Understanding how water level ﬂuctuations inﬂuence wetland water quality,
productivity, and habitat is critical for anticipating future climate-driven changes to the

Great Lakes ecosystem.

To my girls, Christine and Jordan

iv

ACKNOWLEDGEMENTS

I would like to thank my doctoral advisor, Dr. Thomas Burton, for the many
suggestions and critiques that greatly improved this manuscript. I am also very
appreciative for his willingness to listen to my ideas and for ensuring continuous
ﬁnancial support throughout my degree program. I also thank the members of my
advisory committee, Dr. Donald Hall, Dr. Stephen Hamilton, and Dr. Richard Merritt, for
their guidance and suggestions on this research project.

Financial support was provided by Michigan Department of Environmental
Quality and US. Environmental Protection Agency. I also acknowledge the ﬁnancial
support of the College of Natural Sciences for a Dissertation Completion Fellowship and
the Department of Zoology for making travel to meetings possible.

I especially would like to thank Dr. Don Uzarski, a long-time friend and
colleague. His encouragement and help over the years eased the stresses of graduate
school. John Genet was always eager to help with ﬁeldwork and offer his taxonomic
expertise. Donna Kashian provided a great deal of assistance during the second ﬁeld
season. Kari Divine, Kati Kiehl, and Rebecca Serbin helped launch the nutrient
limitation experiments. Thanks to all of you.

Finally, I thank my wife for pushing me on the days I needed to be pushed,
offering support and encouragement, and for making my life complete. My parents and
family understood why this was so important to me and kindly offered their unending

support.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES ........................................................................................................... xi
INTRODUCTION ............................................................................................................... 1
CHAPTER 1
THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON THE SURFACE
WATER CHEMISTRY OF A GREAT LAKES COASTAL WETLAND ......................... 4
Introduction .............................................................................................................. 4
Methods .................................................................................................................... 6
Saginaw Bay, Lake Huron ........................................................................... 6
Study Wetland .............................................................................................. 8
Surface Water Chemistry and Physical Attributes ..................................... 11
Ancillary Data ............................................................................................ 13
Interstitial Water Chemistry and Physical Attributes ................................ 14
Laboratory Analyses .................................................................................. 15
Hydrologic Connectivity between Lake and Wetland ............................... 17
Evaluation of Water Chemistry/Lake Level Relationships ....................... 18
Results .................................................................................................................... 18
Water Level Fluctuations and Climatic Inﬂuences .................................... 18
Speciﬁc Conductance and Total Dissolved Ions ....................................... 22
Major Ion Chemistry .................................................................................. 24
Major Ion Chemistry: spatial and seasonal variation ................................. 32
Nutrient Chemistry: spatial and temporal patterns .................................... 39
Chemistry of Potential Source Waters ....................................................... 44
Chemistry of Interstitial Water .................................................................. 48
Discussion .............................................................................................................. 52
CHAPTER 2
THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON INVERTEBRATE
COMMUNITY COMPOSITION IN A GREAT LAKES COASTAL WETLAND ........ 69
Introduction ............................................................................................................ 69
Methods .................................................................................................................. 71
Study Wetland ............................................................................................ 71
Habitat Measurements ............................................................................... 75
Hydrologic Connectivity between Lake and Wetland ............................... 75
Ancillary Data ............................................................................................ 76
Invertebrates ............................................................................................... 76
Temporal Patterns ...................................................................................... 80
Spatial Patterns ........................................................................................... 83
Results .................................................................................................................... 84
Habitat Characteristics ............................................................................... 84
Invertebrate Community Structure: 1998 vs. 2000 .................................... 92

vi

Invertebrate Community Composition in Late Spring ............................... 98

Invertebrate Community Composition in Early Summer ........................ 102
Invertebrate Community Composition in Mid-summer .......................... 106
Invertebrate Community Composition in Late Summer .......................... 111
Functional Feeding Group Composition: 1998-2000 .............................. 111
Spatial Trends of Select Invertebrate Taxa .............................................. 117
Discussion ............................................................................................................ 120
Invertebrate Community Composition: 1998-2000 ................................. 120
Water Level Fluctuations and Changes in Habitat .................................. 131
Water Level Fluctuations and Changes in Invertebrate Community
Composition ............................................................................................. 1 3 5
CONCLUSIONS .............................................................................................................. 142
APPENDIX A .................................................................................................................. 145
NUTRIENT LIMITATION PATTERNS IN A GREAT LAKES COASTAL
WETLAND ...................................................................................................................... 145
LITERATURE CITED .................................................................................................... 160

vii

Table 1.1.

Table 1.2.

Table 1.3.

Table 1.4.

Table 2.1.

Table 2.2.

LIST OF TABLES

Mean surface water chemical characteristics (i SE) of the Quanicassee
and Saginaw Rivers during spring/summer (June-September) of 1997 -
2000. QR corresponds to the Quanicassee River, SR to the Saginaw
River. Hydrochemical data for the Saginaw River during 1997-1998 and
2000 (N=2 for each respective year) were obtained ﬁom the USGS

http ://water.usgs. gov/mi/nwi s/qwdata?site_no=04 l 5 7000&agency_cd=US
GS&format=inventory_retrieval). Data for the Quanicassee River (1999,
N=6 and 2000, N=5) and some of the data for the Saginaw River (1999,
N=1) were collected as part of this study. BDL corresponds to below
detection limits (see methods). .................................................................. 45

Mean surface water chemical parameters collected from offshore stations
(see Figure 3), the study wetland, the Quanicassee River, and the Saginaw
River in 1999. Sampling stations are spatially related to one another in a
counter-clockwise direction beginning at the Saginaw River; Bay-2 was
located further offshore than the other two stations to provide information
of surface water chemistry of Saginaw Bay water removed from potential
long-shore currents ..................................................................................... 47

Mean (:SE) and range of surface water chemical characteristics measured
at two drainage ditches (south and canal) adjacent to the study wetland.
The ditch labeled canal was only sampled on one occasion (12 March
2000). The south ditch was sampled on three occasions (12 March, 18
May, and 12 June 2000), and following the last sample date contained no
water throughout the rest of the summer (through 12 September 2000).
BDL corresponds to below detection limits (see methods). ...................... 49

Proﬁle means (1|; SE) and ranges for interstitial water chemical parameters
collected during the spring/summer (June-September) of 2000 from mini-
piezometers located at each station along the study transect (20-280 m).
For purposes of resolving spatial variability, proﬁles were also divided
into three regions, outer wetland (Outer; 20-140 m from the wetland
interface), inner wetland (Inner; 160-280 111), and near shore wetland (Near
Shore; 340-460 m). BDL corresponds to below detection limits (see
methods) ..................................................................................................... 50

List of dates invertebrates were collected from the study marsh throughout
the spring and summer (June-September) of 1998-2000. Sampling
methods were identical for each collection date. Sample dates were
averaged by month to facilitate a concise analysis (see methods). ............ 77

Functional feeding group classiﬁcation for invertebrate taxa collected

during the three-year investigation (1998-2000). Functional feeding
groups were restricted to collector (C), shredder (S), piercer (Pi), scraper

viii

Table 2.3.

Table 2.4.

Table 2.5.

Table 2.6.

Table 2.7.

(Sc), and predator (P). Group membership was based on Merritt and
Cumrnins (1996) and Cardinale (1996). .................................................... 81

Mean annual percent abundance (: SE) of major invertebrate groups
(phyla, classes, orders) including contributions to total annual taxa
richness (S) measured throughout the spring and summer (J une-
September) of 1998-2000 (N=9, 1, and 4 respectively). Shannon’s
diversity index (H ’) and evenness (J ') were calculated based on mean
annual percent abundance values. .............................................................. 94

Mean annual percent abundance Q SE) of invertebrate taxa collected
throughout the spring and summer (June-September) of 1998 (N = 9) and
2000 (N = 4). Bold type indicates taxa that represented at least 2% of the
respective communities; only those taxa were assessed for statistically
signiﬁcant (*, p<0.05) changes in percent abundance between years; ns
represents not signiﬁcant. Chironomini and Tanytarsini (sub-tribes of
Chironomidae) were summed for 1998 (italicized; see methods). ............ 96

Mean percent abundance (t SE) of invertebrate taxa collected from the
study wetland in June 1998 (N = 28) and 2000 (N = 10). Bold type
indicates taxa that represented at least 2% of the respective invertebrate
communities; only those taxa were assessed for statistically signiﬁcant (*,
p<0.05) changes in percent abundance between years; ns represents not
signiﬁcant. Chironomini and Tanytarsini (sub-tribes of Chironomidae)
were summed for June 1998 (italicized, see methods); S designates taxa
richness, H’ and J’ correspond to Shannon’s diversity index and evenness
respectively. ............................................................................................... 99

Mean percent abundance (: SE) of invertebrate taxa collected from the
study wetland in July 1998 (N = 56) and 2000 (N = 14). Bold type
indicates taxa that represented at least 2% of the respective invertebrate
communities; only those taxa were assessed for statistically signiﬁcant (*,
p<0.05) changes in percent abundance between years; ns represents not
signiﬁcant. Chironomini and Tanytarsini (sub-tribes of Chironomidae)
were summed for July 1998 (italicized, see methods); S designates taxa
richness, H’ and J’ correspond to Shannon’s diversity index and evenness
respectively. ............................................................................................. 103

Mean percent abundance (1- SE) of invertebrate taxa collected from the
study wetland in August 1998 (N = 28), 1999 (N = 14), and 2000 (N =
14). Bold type indicates taxa that represented at least 2% of the respective
invertebrate communities; only those taxa were assessed for statistically
signiﬁcant (*, p<0.05) changes in percent abundance between years (1998
and 2000 only, 1999 was used to aid in the interpretation of community
shifts in relation to changes in water level); ns represents not signiﬁcant.

ix

Table 2.8.

Table 2.9.

Table 2.10.

Table 2.11.

Table 2.12.

Table 2.13.

S designates taxa richness; H’ and J ' correspond to Shannon’s diversity
index and evenness respectively. ............................................................. 107

Mean percent abundance (: SE) of invertebrate taxa collected from the
study wetland in September 1998 (N = 14) and 2000 (N = 14). Bold type
indicates taxa that represented at least 2% of the respective invertebrate
communities; only those taxa were assessed for statistically signiﬁcant (*,
p<0.05) changes in percent abundance between years; ns represents not
signiﬁcant. S designates taxa richness; H’ and J’ correspond to Shannon’s
diversity index and evenness respectively. .............................................. 112

Invertebrate taxa that had signiﬁcantly higher percent abundance values
associated with the outer marsh (20-140 m) region relative to the inner
marsh (160-280 m) region (left) and corresponding Pearson correlation
coefﬁcients (right). Statistically signiﬁcant (p<0.05) differences were
assessed using the non-parametric Mann-Whitney U test (see methods); p
values are noted, no p values are given for non-signiﬁcant tests.
Correlation analysis was used to further assess spatial patterns for only
those taxa with signiﬁcant U tests. Correlation coefﬁcients in bold were
signiﬁcant (p<0.05). ................................................................................. 119

Invertebrate taxa that had signiﬁcantly higher percent abundance values
associated with the inner marsh (160-280 m) region relative to the outer
marsh (20-140 m) region (leﬁ) and corresponding Pearson correlation
coefﬁcients (right). Statistically signiﬁcant (p<0.05) differences were
assessed using the non-parametric Mann-Whitney U test (see methods); p
values are noted, no p values are given for non-signiﬁcant tests.
Correlation analysis was used to further assess spatial patterns for only
those taxa with signiﬁcant U tests. Correlation coefﬁcients in bold were
signiﬁcant (p<0.05). ................................................................................. 121

Annual summary of abundant (>5%) and common (>2%, <5%)
invertebrate taxa collected from the study marsh during the spring and
summer of 1998-2000. Each taxon was abundant and/or common during
at least one month. ................................................................................... 123

List of newly collected invertebrate taxa from the study marsh throughout
the spring and summer of 1998-2000 and taxa not collected in this study,
but recorded in previous investigations. .................................................. 128

Classiﬁcation of invertebrate taxa based on presence (X) or absence (--)
relative to lake levels. Above average, intermediate, and below average
water levels refer to the years 1998-2000 respectively. Taxa noted as
general were collected during each respective year (or lake level). ........ 129

Figure1.1.

Figure 1.2.

Figure 1.3.

Figure 1.4.

Figure 1.5.

Figure 1.6.

Figure 1.7.

LIST OF FIGURES

Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The point bars (Point Lookout and Sand Point) delineating inner
Saginaw Bay, Lake Huron are connected by a dotted line. The
Quanicassee and Saginaw Rivers are referenced. ........................................ 7

Schematic of transect bisecting study marsh. The primary transect
extended 280 m shoreward of the wetland/open bay water interface (blue
line; O-m station). Stations (circles) were established at 20-m intervals
along the primary transect for collection of surface water samples. Mini-
piezometers were installed at each station along the primary transect in
2000 and three additional stations were added at 60-m intervals shoreward
of the 280-m station (squares) to collect interstitial water. The wetland
was arbitrarily divided into three regions and these are noted in the

graphic ........................................................................................................ 10

Enlarged view of the southern end of inner Saginaw Bay, Lake Huron
illustrating the proximity of offshore stations (B-l , B-2, and B-3; sampled
in 1999) and agricultural drains (sampled in 2000) to the study wetland (*)
and Saginaw and Quanicassee Rivers. ....................................................... 12

Mean annual stage (+ SE) record at Essexville, M1 for the years of study
and the 48-year average corresponding to the years 195 3-2001 (leﬁ pane).
Mean monthly stage record during the period of study including the 48-
year average (right pane). IGLD corresponds to International Great Lakes
Datum 1985. Vertical dotted lines partition the months corresponding to
ﬁeld data collection (June-September). ..................................................... 19

Mean annual water depth (+ SE) at the wetland/open bay water interface
(O-m station) relative to sediment surface proﬁle. Water depths and
sediment elevations are referenced to the O-m station (i.e. O-m station has
an elevation of 0.00 m; see methods). DF WI corresponds to distance from
wetland interface in meters. ....................................................................... 21

Box and whisker plot of water depth measurements collected within the
wetland (0-280 m) during each respective year. Box and center line
represent the 25th and 75th percentiles and the median respectively.
Whiskers correspond the 10th and 90th percentiles, circles represent
outliers. Numbers above each box represent the mean wetland depth for
each respective year; number of individual measurements are denoted
below. ......................................................................................................... 21

Mean annual estimates of evaporation, precipitation, and run-off for Lake
Huron during the period of study (1997-2000), including long-term means

xi

Figure 1.8.

Figure 1.9.

Figure 1.10.

Figure 1.11.

Figure 1.12.

Figure 1.13.

(48 years). Numbers over bars correspond to net basin supply (NBS),
where NBS = run-off + precipitation - evaporation, in mm. Run-off data
were not available for 2000 ........................................................................ 23

Annual precipitation totals (mm) corresponding to the period of study
(1997-2000) and relative to the 30-year average (1971-2000) measured at
Essexville, MI (station #202631). The dotted line corresponds to the 30-
year average. Data were acquired from the Michigan Climatological
Resources Program (see methods). ............................................................ 23

Box plots of speciﬁc conductance (uS cm") measurements collected from
the study wetland during the spring/summer (J une-September) of 1997-
2000 (N=60, 135, 90, and 73 respectively). Boxes represent 25th and 75th
percentiles, medians are indicated within boxes, whiskers correspond to
10th and 90th percentiles, and circles represent outliers. Mean values are
italicized below boxes. ............................................................................... 25

Relationship between speciﬁc conductance (uS cm'l) and total dissolved
ions (TDI; meq L'l) for wetland water samples analyzed for the full suite
of major anions and cations during the spring/summer (J une-September)
of 1998-2000 (N=3 7, 34, and 32 respectively). Cation concentrations
were not quantiﬁed in 1997. ...................................................................... 25

Box plots of major anion (A-C) concentrations (mg L") for marsh water
samples collected during the study period (1997-2000; N=60, 135, 90, and
73 respectively unless noted). Boxes represent 25th and 75th percentiles,
medians are within the boxes, whiskers correspond to 10th and 90th
percentiles, and circles are outliers. Note scale changes. Asterisks
indicate signiﬁcant (p<0.05) change relative to previous year. Numbers
above each box correspond to mean concentrations. ................................. 26

Box plots of the proportion of total anions comprised by T-alkalinity (A),
chloride (B), and sulfate (C) of wetland water samples collected
throughout the period of study (1997-2000; N=20, N=135, N=88, and
N=88 respectively). Boxes represent 25th and 75th percentiles, medians
are indicated within boxes, whiskers correspond to 10th and 90th
percentiles, and circles represent outliers. Mean proportions for each
respective year are indicated above boxes. ................................................ 28

Box plots of major cation (A-D) concentrations (mg L'l) for wetland water
samples collected during the study period (1998-2000; N=135, 90, and 73
for Ca and Mg, and N=37, 34, and 31 for Na and K). Boxes represent
25th and 7 5th percentiles, medians are within the boxes, whiskers
correspond to 10th and 90th percentiles, and circles are outliers. Asterisks
indicate signiﬁcant (p<0.05) change relative to previous year. Mean
concentrations are indicated near each box. .............................................. 30

xii

Figure 1.14.

Figure 1.15.

Figure 1.16.

Figure 1.17.

Figure 1.18.

Figure 1.19.

Box plots of the proportion of total cations comprised by calcium (A),
magnesium (B), sodium (C), and potassium (D) of marsh water samples
collected during 1998-2000 (N=3 7, N=34, N=32 respectively). Boxes
represent 25th and 75th percentiles, medians are indicated within boxes,
whiskers correspond to 10th and 90th percentiles, and circles are outliers.
Mean proportions for each respective year are indicated above boxes.
Note change in scale for potassium. .......................................................... 31

Mean speciﬁc conductance (uS cm'l; + SE) proﬁles along the study
transect during the spring/summer (June-September) of 1997-2000. The
number of observations for each station were N=4, N=9, N=6, and N=5
respectively. DFWI corresponds to distance from wetland interface
(wetland/open bay water interface) in meters. ........................................... 33

Mean mass loss (: SE) of alabaster substrata (AS) across the study
wetland during mid-summer (August) 1999 and 2000. Arrows indicate
subjective inﬂection points where mass loss appears to stabilize
throughout the remaining portion of the study transect. DFWI
corresponds to distance from wetland interface (wetland/open bay water
interface) in meters. ................................................................................... 35

Spatial trends for mean (+ SE) concentrations of major anions (T-
alkalinity, chloride, and sulfate) across the study wetland during the
spring/summer (June-September) of 1997-2000. Note changes in scale for
each anion. A common legend is shared for each graphic. DFWI
corresponds to distance from wetland interface (wetland/open bay water
interface) in meters. ................................................................................... 36

Spatial trends of mean concentrations (: SE) of major cations (calcium,
magnesium, sodium, and potassium) across the study wetland during the
spring/summer (June-September) of 1998-2000. Note changes in scale for
each cation. A common legend is shared for each graphic and is displayed
in the potassium panel. DFWI corresponds to distance from wetland
interface (wetland/ open bay water interface) in meters. ........................... 3 8

Box plots of wetland surface water inorganic nitrogen concentrations (left
panels) and mean (: SE) spatial trends for nitrate and ammonium
concentrations along the study transect (right panels). Nitrate observations
were N=16, N=135, N=89, and N=73 repectively; ammonium
observations were N=0, N=135, N=90, and N=73. Note logarithmic scale
for nitrate concentrations (left panel only). Means are indicated above
each box. DF WI corresponds to distance from wetland interface in
meters. ........................................................................................................ 40

xiii

Figure 1.20.

Figure 1.21.

Figure 1.22.

Figure 1.23.

Figure 1.24.

Figure 1.25.

Figure 1.26.

Box plots of wetland surface water phosphorus concentrations measured
during the period of study (1997-2000). SRP represents dissolved soluble
reactive phosphorus (N=0, N=135, N=90, and N=75 respectively) and T-P
is total phosphorus measured on unﬁltered samples (N =60, N=135, N=90,
and N=46 respectively). Means are indicated above each box. BDL
corresponds to below detection limits (0.005 mg P L'1 for SRP). Soluble
reactive phosphorus was not measured in 1997 ......................................... 42

Spatial trends for mean (3; SE) turbidity proﬁles across the study wetland
during the spring/summer (June-September) of 1998 (N= 135) and 1999
(N =90). Turbidity was not measured in 1997 or 2000. DP WI corresponds
to distance from wetland interface (wetland/ open bay water interface) in
meters. ........................................................................................................ 43

Box plot of dissolved silica concentrations measured during the
spring/summer (June-September) of 1999 (N=90) and 2000 (N=73).

Silica was not measured in 1997 and 1998. Boxes represent 25th and 75th
percentiles, medians are indicated within boxes, whiskers correspond to
10th and 90th percentiles, and circles represent outliers. Mean values are
italicized above boxes. ............................................................................... 43

Mean VHG (t SE) proﬁle across the study wetland in summer 2000.
Positive VHG values indicate upwelling or seepage zones and negative
values indicate downwelling or recharge zones. VHG is dimensionless;
see methods for calculation. DFWI corresponds to distance from wetland
interface ...................................................................................................... 5 3

Average (: SE) position of the water table (piezometric head) within the
study wetland in summer 2000. Piezometric head measurements were
relative to surface water levels and re-scaled to sediment elevation (see
methods). Values above sediment surface indicate localized seepage
areas; the converse is true of values below the sediment surface. Sediment
elevations were referenced to the wetland interface (O-m station). DFWI
corresponds to distance from wetland interface in meters. ........................ 53

Forty-eight year stage record for Saginaw Bay, Lake Huron (1953-2001;
Essexville, MI). Solid line represents long-term mean (176.56 1 0.05 m).
Markers correspond to studies conducted by Cardinale et al. 1997 (1),
Suzuki et al. 1995 (2), and EPA water quality surveys (Nalepa er al. 1996;
3) conducted within inner and outer Saginaw Bay; asterisk indicates the
onset of this study (1997-2000). IGLD corresponds to International Great
Lakes Datum. ............................................................................................. 55

Mean cumulative ion plots (meq L") of wetland surface water samples
collected during the spring/summer of 1997-2000 (A-D). Only anions
were quantiﬁed in 1997 (N=60 for T-alk and Cl' and N=16 for SO4'2);

xiv

Figure 1.27.

Figure 1.28.

Figure 1.29.

Figure 1.30.

Figure 2.1.

Figure 2.2.

N=135, 90, and 73 for all ions during 1998-2000 respectively, except
N=37, 34, and 31 for Na+ and K’). A common legend is shared by each
graphic; 804'2 concentrations were interpolated between 0, 80, 160, and
240 m stations. DF WI corresponds to distance ﬁom wetland interface in
meters. ........................................................................................................ 57

Mean annual proportions of major anions (A) and cations (B) of wetland
surface water samples collected during the period of study (1997-2000).
Numbers correspond to the proportion of total anions and cations
accounted for by Cl' and Mg+2 respectively. Proportions were

determined based on ion concentrations in meq L". Cations were not
quantiﬁed in 1997 (ND represents no data). Potassium accounted for less
than 1.5%. .................................................................................................. 58

Mean monthly discharge of the Saginaw River at Saginaw, MI. Daily
stream ﬂow data were obtained from the USGS National Streamﬂow
Information Program (station #04157000; see methods) and re-computed
as monthly averages. The break in hydrograph indicates no data were
available. A pro-longed period of drought corresponded with the years
1999-2001 (2001 data not available). ........................................................ 60

Box plots of speciﬁc conductance (1.18 cm") measurements from offshore
waters (left panel) of Saginaw Bay (see map inset) in 1991-92 (EPA 1-3
and 13; Nalepa er al. 1996) and measurements collected in this study
(1997-2000; denoted by * in map). Boxes represent 25th and 75th
percentiles, medians are indicated within the boxes, whiskers correspond
to 10th and 90‘h percentiles, and circles represent outliers. Numbers

above each box are sample means with the number of replicates indicated
below. ......................................................................................................... 66

Long-terrn stage record (1900-2002) for Lake Huron. Annual values (1;
SE) were derived from mean monthly water levels (GLERL, see methods).
Data for 2000-2002 were from the Essexville, MI station; deviations from
the previous years of record were less than 6 cm. The solid and dotted
horizontal lines represent the 102-year average (1- SE); 176.51 1 0.03 m.
The asterisk indicates the onset of this study (1997-2000). IGLD
corresponds to International Great Lakes Datum 1985. ............................ 68

Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The Quanicassee and Saginaw Rivers are also referenced. .......... 72

Schematic of transect spanning study wetland. Station locations denoted
along the study transect (circles) are relative to distance from the emergent
plant/Open bay water interface (O-m station) in meters and are spaced at 20
m intervals along the transect (280 m). Invertebrate sampling was
conducted at all stations except at the wetland interface. The shoreline

XV

Figure 2.3.

Figure 2.4.

Figure 2.5.

Figure 2.6.

Figure 2.7.

Figure 2.8.

was located approximately 500 m from the wetland interface. Naming
conventions used to designate regions of the wetland are noted in the
graphic; no invertebrate samples were collected from the near shore
region. ........................................................................................................ 74

Mean spring/summer (June-September) stage (1 SE) record at Essexville,
MI during the years of study and the 48-year average (i SE)
corresponding to the years 1953-2001 (left panel). Mean monthly stage
record during the period of study including the 48-year annual average (-_+-_
SE; solid and dotted lines), 176.55 : 0.053 m (right panel). Dashed lines
correspond to invertebrate collection periods (June-September). IGLD
corresponds to International Great Lakes Datum 1985. ............................ 85

Box and whisker plot of water depth measurements collected within the
wetland (0-280 m) during each respective year. Box and center line
represent the 25th and 75th percentiles and the median respectively.
Whiskers correspond to the 10th and 90th percentiles, circles represent
outliers. Numbers above each box represent the mean wetland depth for
each respective year; number of individual measurements are denoted
below. ......................................................................................................... 86

Mean (1: SE) cumulative stem density (no. stems m'z) along the study
transect during 1997-1999. Monthly (J une-September) data were
averaged. Stem densities were not signiﬁcantly different between years
within the outer wetland region (0-140 m); signiﬁcant (p<0.05) increases
were measured within the inner marsh (160-280 m) in 1999 relative to
1998. DFWI corresponds to distance from wetland interface in meters...86

Mass loss (: SE) of alabaster substrata (AS) across the study marsh during
mid-summer 1999 and 2000. Arrows indicate subjective inﬂection points
where mass loss appeared to become relatively uniform thereafter. DFWI
corresponds to distance from wetland interface in meters. ........................ 88

Mean (: SE) water temperature (° C) proﬁles measured across the study
marsh throughout the summers of 1997-2000 (A). DF WI corresponds to
distance from wetland interface in meters. Mean monthly modeled water
temperatures (B) and over lake air temperatures (C) for Lake Huron
(excluding Georgian Bay) during the years of study and relative to the
long-term average Q SE). Lake Huron data were obtained from USGS -
GLERL. The modeled Lake Huron water temperatures were limited to 0°
C or above. ................................................................................................. 90

Mean (: SE) dissolved oxygen (DO) percent saturation proﬁles across the
study wetland during June through September of 1997-2000 (N=4, 9, 6 per
station for each respective year). DFWI corresponds to distance from
wetland interface in meters. ....................................................................... 91

xvi

Figure 2.9.

Figure 2.10.

Figure 2.11.

Figure 2.12.

Figure 2.13.

Figure 2.14.

Figure 2.15.

Box plot of speciﬁc conductance (1.18 cm") measurements collected from
the study wetland during the summers of 1997-2000 (N=60, 135, 90,

and 73 respectively). Boxes represent 25th and 75th percentiles, medians
are indicated within boxes, whiskers correspond to 10th and 90th
percentiles, and circles represent outliers. Mean values are italicized
below boxes. .............................................................................................. 91

Mean (: SE) turbidity (NTUs) proﬁles across the study wetland during
June through September of 1998 and 1999 (N=9 and 6 per station for each
respective year). DFWI corresponds to distance from wetland interface in
meters. ........................................................................................................ 93

Average functional feeding group composition (percent abundance) of
invertebrate communities sampled during June through September of 1998
(A) and 2000 (B). Scrapers and piercers were summed since food
resources for these taxa were assumed to be similar (see methods). A
common legend is shared for all stacked bar graphics. ........................... 116

Average functional feeding group composition (percent abundance) of
invertebrate communities sampled in August 1998-2000. Scrapers and
piercers were summed since these taxa were assumed to have similar food
resources (see methods). .......................................................................... 118

Forty-eight year stage record for Saginaw Bay, Lake Huron (1953-2001;
Essexville, MI). Solid line represents long-term mean (176.56 1 0.05 m).
Markers correspond to invertebrate-based investigations conducted in the
study wetland (+ Brady (1992); I Burton et al. (2002), Cardinale (1996),
Stricker er al. (2001), and * Burton et al. (1999); arrow indicates the onset
of this study (1998-2000). IGLD corresponds to International Great Lakes
Datum ....................................................................................................... 132

Mean (t SE) spring/summer (June-September) water depth proﬁles across
the study wetland during the years of study (1998-2000) as inﬂuenced by a
hypothetical 0.10m seiche-induced increase (A) and decrease (B) in water
level. DFWI corresponds to distance from wetland interface. The zero
depth line (dotted) represents the sediment surface within the marsh.
Observations with a negative depth indicate the lack of standing surface
water. ........................................................................................................ 134

Average functional feeding group composition (percent abundance) of
invertebrate communities during the years of study (1998-2000). Scrapers
and piercers were summed since these taxa were assumed to have similar
food resources (see methods). .................................................................. 139

xvii

Figure 3.1.

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 3.5.

Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The Quanicassee and Saginaw Rivers are referenced. ............... 147

Schematic of transect bisecting study marsh. The primary transect
extended 280 m shoreward of the wetland/open bay water interface (O-m
station). Nutrient diffusing substratum (NDS) experiments were
conducted along three secondary transects established perpendicular to the
primary transect: 60 m, 120 m, and 180 m from the wetland interface.
Clay pots were evenly spaced along the perpendicular transects and
adjacent pots (treatments) were chosen at random. Circles and squares
correspond to stations in which surface water samples were collected for
the study of surface and interstitial water chemistry (see Chapter I). ...... 150

Mean (1 SE) periphyton chlorophyll 3 concentrations ﬁom the August
1999 NDS experiment (A). Mean (i SE) AFDM concentrations from the
August 1999 NDS experiment (B). DFWI corresponds to distance from
wetland interface in meters. N and P correspond to nitrogen and
phosphorous treatments. Probabilities from AN OVA are indicated above
each experiment; signiﬁcant pair-wise treatment comparisons relative to
the control (Dunnett's) are denoted by circles. ........................................ 154

Mean (: SE) periphyton chlorophyll a concentrations ﬁ'om the July 2000
NDS experiment (A). Mean (1: SE) periphyton chlorophyll a
concentrations from the August 2000 NDS experiment (B). DFWI
corresponds to distance from wetland interface in meters. N and P
correspond to nitrogen and phosphorous treatments. Probabilities from
AN OVA are indicated above each experiment; signiﬁcant pair-wise
treatment comparisons relative to the control (Dunnett's) are denoted by
circles. ...................................................................................................... 156

Mean mass loss (+ SE) of alabaster substrata (AS) across the study
wetland during mid-summer (August) 1999 and 2000. Arrows indicate
subjective inﬂection points where mass loss appears to stabilize
throughout the remaining portion of the study transect. DFWI
corresponds to distance from wetland interface (wetland/open bay water
interface) in meters. ................................................................................. 157

xviii

INTRODUCTION

There are approximately 65,547 ha of coastal wetlands remaining on the upper
Great Lakes (Prince er al. 1992). Coverage has been reduced by approximately 50%
since pre-settlement times, yet extensive complexes remain in shallow embayments that
offer some degree of protection from wave energy and storm surges (Brazner 1997).
Coastal wetlands lie at the interface between terrestrial and lentic ecosystems, and the
hydrology and habitat of these systems makes them unique from other classes of
freshwater wetlands. Lakeshore geomorphology, hydrologic connectivity between lake
and wetland, and water level ﬂuctuations characteristic of the Great Lakes interact to
create a diversity of habitats (Keough et al. 1999).

Great Lakes coastal wetlands have been recognized as important feeding and
nursery habitats for ﬁsh (Liston and Chubb 1986; French 1988; Jude and Pappas 1992;
Brazner 1997) and waterfowl (Prince et al. 1992). Despite the economic importance of
these systems, few studies have focused on the invertebrate communities that support ﬁsh
and waterfowl production (Duffy et al. 1987; McLaughlin and Harris 1990; Krieger and
Klarer 1991). Growing interest in assessment and monitoring of the ecological integrity
of these systems has culminated in a variety of intensive studies of coastal wetland
invertebrate communities (Brady and Burton 1995; Brady et al. 1995; Cardinale et a1.
1997; Gathman et al. 1999; Kashian and Burton 2000; Stricker er al. 2001; Burton et al.
2002).

Expanding the knowledge of coastal wetland ecology will require an
understanding of how abiotic and biotic factors interact to inﬂuence wetland productivity

and shape communities. Water levels of the Great Lakes ﬂuctuate naturally over the

short-terrn (hours, days) and long-term (years). Short-term ﬂuctuations generally result
in 10-20 cm changes in water depth and may go unnoticed during periods corresponding
to high water (Gathman et al. 1999). Long-term water level ﬂuctuations are more
dramatic. For example, mean annual stage of Lake Huron has historically ﬂuctuated 2.14
m (BishOp 1990). Water level ﬂuctuations are thought to alter habitat availability, water
quality, and aquatic plant, invertebrate, and ﬁsh communities (Gasith and Gafny 1990).
However, the inﬂuence of water level ﬂuctuations on the ecology of Great Lakes coastal
wetlands has not been studied (but see Gathman 2000).

In chapter 1, I described the surface water chemistry of a Saginaw Bay coastal
wetland in the context of a 1.04 m decline in lake level between the years 1997-2000.
The total dissolved ion content and ionic composition of wetland surface water changed
in conjunction with lake level declines. I argued that the reduction in water depth in near
shore regions of the wetland led to oxidizing conditions, exaggerated the water table
gradient across the wetland, and increased sediment/surface water interactions. The
implications for soil ﬂushing during periods of below average water levels were placed in
the context of drought and climate change scenarios proposed for the Great Lakes basin.

In Chapter 2, I describe the changes in invertebrate community composition
between 1998 and 2000 relative to lake level declines. A 0.71 m reduction in mean water
depth reduced colonizable habitat by nearly 90% and left most of the near shore region
devoid of standing water. These data represent the ﬁrst record of a Great Lakes coastal
wetland invertebrate assemblage at lake levels nearly 0.50 m below the long-term

average. Diversity and richness increased in response to lake level declines. I argue that

community composition changed in response to food resources, exclusion of piscivorous

predators, and a shift in hydrology that mimicked the hydroperiod of temporary wetlands.

Chapter 1

THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON THE SURFACE
WATER CHEMISTRY OF A GREAT LAKES COASTAL WETLAND

INTRODUCTION
The fringing wetlands of the Great Lakes lie at the interface between terrestrial

and lentic ecosystems and are hydrologically unique relative to other types of freshwater
wetlands (Wetzel 1990; Wetzel 1992). These systems provide diverse and complex
habitat because of the interaction between lakeshore geomorphology, surface wave
energy, and natural water level ﬂuctuations that occur over short and long duration
(Gathman et al. 1999; Keough er al. 1999).

Observations of surface water chemical disparities between offshore (or pelagic)
and littoral environments and among different plant communities within the littoral
environment are common (Dvorak 1970; Planter 1970; Howard-Williams and Lenton
1975; Smiley and Tessier 1998). In fact, a variety of investigations of fringing wetlands
have demonstrated that these ecotones represent an important mixing region whereby
pelagic water, terrestrial-derived run-off, and groundwater potentially interact (Howard-
Williams and Lenton 1975; Horsch and Stefan 1988; Wetzel 1990; James and Barko
1991a; James and Barko 1991b; Winter 1981; Huddart er al. 1999). Despite the potential
importance of fringing wetlands to large lake ecosystems, few studies have focused on
the surface water quality of Great Lakes coastal wetlands (Suzuki et al. 1995; Cardinale
1996; Cardinale et al. 1997).

Threats of cultural eutrophication in the Great Lakes basin have generated
considerable interest in water quality monitoring of offshore water. This is particularly

the case in areas with a long history of anthropogenic impacts such as Saginaw Bay, Lake

Huron (Bierman er al. 1984; Nalepa et al. 1996). Studies suggest that freshwater
wetlands function as a sink for many pollutants and buffer water quality impacts to
receiving waters (J ansson et al. 1994). Fringing wetlands along shoreline areas should
function in much the same way because of their position in the landscape. However, the
role of natural water level ﬂuctuations on lakeshore wetlands remains unknown. Studies
from peat wetlands suggest that drought-induced changes in hydrology result in the
release of a variety of inorganic ions; in particular mobilization of 804'2 and NH: ions
has been noted (Qiu and McComb 1996; Devito and Hill 1999). Water level ﬂuctuations
are common to the Great Lakes and long-term water level ﬂuctuations can vary up to 1.5
m. Climate change scenarios speciﬁc for the Great Lakes basin suggest an increase in the
frequency and duration of below average lake levels (Kling et al. 2003). Understanding
the impact of water level ﬂuctuations on coastal wetland surface water chemistry is
critical for anticipating the effects of climate change on the Great Lakes ecosystem.

The goal of this study was to characterize changes in the hydrology and surface
water chemistry of a Saginaw Bay coastal wetland relative to a 1.04 m decline in lake
stage. First, I characterized the major ion and nutrient chemistry of wetland surface
waters from June to September of 1997 through 2000. Second, the inﬂuences of local
source waters (riverine, agricultural drainage, and interstitial water) on wetland surface
water chemistry were evaluated. Third, changes in wetland hydrology that occurred in
conjunction with lake level declines were used to illustrate possible causes of changes in
wetland surface water chemistry. Lastly, the changes in wetland water quality were
placed in the context of extended periods of drought and climate change scenarios

forecasted for the Great Lakes basin.

METHODS
Saginaw Bay, Lake Huron -

Saginaw Bay is a shallow extension of the western shoreline of Lake Huron
covering an area of approximately 3,000 kmz. At the widest point, the bay is 42 km and
extends a distance of 82 km (Bierman et al. 1984; Budd er al. 1998). The bay can be
naturally divided into an inner and outer region along a natural constriction (~21 km
wide) stretching from Point Lookout to Sand Point (Figure 1.1). The inner bay is
warmer, shallower (mean depth 5.1 m), and has poor water quality relative to the offshore
waters of Lake Huron (Nalepa et al. 1996). The outer bay is cooler, deeper (mean depth
13.7 m), and more strongly inﬂuenced by the colder, nutrient-poor waters of Lake Huron
(Budd er al. 1998).

Because of the shallow nature of the inner bay and the low relief of the
surrounding landscape, extensive ﬁinging wetlands are characteristic features along the
shoreline areas. Historically, wetland coverage within this region was much greater
during pre-settlement times. Land drainage and conversion have greatly contributed to
the decline in present day coverage. Current land-use within the Saginaw Bay basin is
primarily row crop agriculture and industry. As a result, adjacent rivers and drainage
ditches carry signiﬁcant dissolved and particulate loads during the spring and fall of the
year when run-off is elevated. The Saginaw River is the most important point source to
the basin, contributing approximately 70% of total tributary discharge to the bay
(Bierman et al. 1984; Fahnenstiel et al. 1995; Budd er al. 1998). This constitutes a

signiﬁcant impact to the waters of Lake Huron because the Saginaw River system drains

an area of land approximately seven times greater than the Saginaw Bay watershed alone

 

 

Uri...

 

:‘hqm

F

 

inner &
Saginaw Bay 3 I '
r K
I
‘ Saginaw R. A?)
\ J-l
. 0 Jr 1
.5
Th - <>
, l S" 3 tudy Wetland
.13.

O uanicassee R.

 

A

u __

 

 

 

10

lGIomete 15
0

20 30

Figure 1.1. Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The point bars (Point Lookout and Sand Point) delineating inner Saginaw Bay,
Lake Huron are connected by a dotted line. The Quanicassee and Saginaw Rivers are

referenced.

(Bierman er al. 1984; Budd et al. 1998). Non-point source drainage ditches are common
throughout the basin and likely contribute to nutrient loading.

The circulation of surface waters in Saginaw Bay generally follows a counter-
clockwise direction (Budd et al. 1998). The ionically dilute offshore water of Lake
Huron ﬂows into the northwestern portion of the bay, mixes with solute-rich inner bay
water and tributary ﬂow, and subsequently discharges back into Lake Huron further up
the northeastern coastline of the bay. The convergence of offshore water with inner bay
surface water and tributary ﬂow results in the mixing of chemically distinct water masses.
Tributaries and drainage ditches have the most distinct hydrochemical signatures because
of high dissolved and particulate loads typical of agricultural run-off (Vanni et al. 2001).
Groundwater in the Saginaw Bay basin has a characteristically high dissolved solid load,
principally Ca, Mg, Na, Cl, Fe, and S04 (Long er al. 1988).

Study Wetland

The wetland complex selected for study was located along the southeastern shore
of Saginaw Bay, Lake Huron, U.S.A. (Figure 1.1). The portion of the wetland studied
encompassed only a small area of the total complex that extended approximately from the
Quanicassee River (Tuscola County) to the Sand Point/Wildfowl Bay area (Huron
County). The study area was located adjacent to Vanderbilt Park, Tuscola County,
Michigan (43° 37’N 83° 38’W). Predominant winds were out of the northwest, and the
wetland was unprotected from wind and wave exposure with a maximum fetch of 30 or
more km (Suzuki et al. 1995).

The emergent plant community of the study wetland consisted of a nearly mono-

dominant stand of three-square bulrush (Scirpus pungens Vahl) that extended

approximately 300 m shoreward from the wetland interface (the interface between the
outer edge of the emergent zone and open bay water). Less dominant species of bulrush,
S. acutus Muhl. and S. validus Vahl, were interspersed among S. pungens, primarily
within the inner (160-280 m shoreward of the wetland interface) and near shore (300-500
m shoreward of the wetland interface) regions of the wetland. A large cattail (Typha
angustifolia L.) complex bordered the northern edge of the study area adjacent to the
inner wetland region. At the onset of this study, vegetation within the near shore region
of the wetland (300-500 m from open water) included S. pungens, arrowhead (Sagittaria
3p), and a variety of submergent species (see Batterson et al. 1991). As lake stage
declined, the near shore region became dominated by species typical of wet
meadow/strand communities, such as sedges (Carex spp.), smartweeds (Polygonum spp.),
and cotton wood (Papulus deltoides Rydb.) seedlings. Several recent studies have been
conducted in this particular wetland and have provided critical background information
on solute chemistry, periphyton, vascular plants, and invertebrate communities (Brady
1992; Brady et al. 1995; Brady and Burton 1995; Suzuki et al. 1995; Cardinale 1996;
Cardinale et al. 1997; Cardinale er al. 1998; Burton et al. 1999; Stricker er al. 2001;
Vaara 2001; Burton er al. 2002).

A transect bisecting the S. pungens wetland was established in June 1997 and
used during ensuing ﬁeld seasons (1998-2000). The transect extended 280 m
perpendicular to the shoreline, extending shoreward (S 172°) from the outer edge of the
emergent plant zone/open bay water interface hereafter referred to as the wetland
interface (N 43°36’44” W 83°39’41 .3”). Sampling stations were established at 20-m

intervals along the study transect (Figure 1.2). During 2000, wetland interstitial water

.oEQEw 05 5 38: 2a 035 BE 32on out: 85 363% 35:38

was use—63 2:. .333 REESE 80:8 01833.5 ".266 :7ch 2: no 295205 35,52: 5-8 3 325 0.53 Sousa
3:02.63 3:: new coon E 88:93 SEE. 2: mac.“ 2236 :28 a 3:53.55 82> maQHoEonoEaem—z .3388 88>» ooﬁSm we
558:8 new 885: @553 2: wee—e 28935 S cm “a 352958 225 G225 225m @238 86 6:: 2.3V coatog 833
men coqoBaazoB one we engage: E owm 96:25 885: basin 2C. .58.: beam wcmaoomﬁ 885: we oﬁﬁoaom .~._ oSmE

\

\.

.1.

\ A e
\ y if» r »

E :3

A i: i x A
f» in”...

 

 

:2»?— oueﬁ .32. "E chA P P f
:3on .55: "E 38.3: P

Egan.— ..82. "E :3...

P aye»? .528
»

 

 

 

» > > P L» > ...w....w.._......

P ham eoaOzia—«og

10

was also sampled at 20-m intervals along the study transect, and at 60-m intervals
thereafter (280-460 m, near shore region of wetland).

To better resolve the source of surface waters contributing to wetland hydrology,
several offshore stations were established in 1999 (Figure 1.3). The ﬁrst station (Bay-1)
was approximately 1 km south of the study wetland (N43 35 50.1 W83 41 29.5), Bay-2
was located approximately 2 km north of the of the mouth of the Quanicassee River and
was the furthest station from the shoreline (N43 36 46.2 W83 41 13.6), and Bay-3 was
approximately 1 km northwest of the Quanicassee River mouth (N43 35 50.7 W83 40
13.8). These stations were chosen to evaluate the inﬂuences of Quanicassee and Saginaw
River discharge on wetland surface water chemistry. Sampling stations were also
established at the Quanicassee River (N43 35 12.1 W83 40 47.3) and the Saginaw River
(N43 38 29.6 W83 50 53.7). The Quanicassee River was sampled on the same schedule
as transect sampling in 1999 and 2000.

During spring 2000, two agricultural drainage ditches were sampled (Figure 1.3).
The ﬁrst ditch (south drain) was located approximately 1 km adjacent to the study
wetland and maintained ﬂow through mid-June after which it dried up completely. The
second ditch (canal) was located approximately 500 m further south and had sufﬁcient
depth to permit watercraft access into adjacent residential development. The canal did
not appear to ﬂow; yet irrigation pumps were observed to discharge into this water body.
Surface Water Chemistry and Physical Attributes

Surface water samples were collected monthly (1997, 1999, and 2000) and
approximately twice per month (1998) during the active growing season (J une-

Septcmber). Surface water grab samples were collected from mid-depth in acid washed,

ll

 

inner Saginaw Bay

B-2

 

 

 

Agricultural
B-l Drains

Saginaw .
River Quanicassee

River

 

 

 

Figure 1.3. Enlarged view of the southern end of inner Saginaw Bay, Lake Huron
illustrating the proximity of offshore stations (B-l , B-2, and B-3; sampled in
1999) and agricultural drains (sampled in 2000) to the study wetland (*) and
Saginaw and Quanicassee Rivers.

12

pre-leached, 5 00-mL polypropylene bottles after rinsing two times with water from each
respective station. Samples were placed on ice and transported to the laboratory for
chemical analyses. Depth, stem density (no. stems m'z), dissolved oxygen (YSI Model
51B; measured between 10 am and 12 noon), pH (Cole Parmer pH 5 Series), speciﬁc
conductance (25° C; YSI Model 30), and temperature (YSI Model 30) were measured in
the ﬁeld at each station along the transect. Stern density, pH, and dissolved oxygen
measurements were not collected in 2000. Riverine, pelagic, and drainage ditch stations
were sampled in an identical fashion, with the exception of depth and stem density.
Ancillary Data

Lake Huron stage data (International Great Lakes Datum 1985) were acquired
from the National Oceanic and Atmospheric Administration (N OAA real-time water
level data; http://co-ops.nos.noaa.gov/ data_res.html; Essexville, MI; station #9075035).
The historical record for this gauging station extended back to 1953, and therefore
provided a 48-year chronology of lake levels within Saginaw Bay. Longer historical
records were available for the Great Lakes (Environmental Protection Agency, Great
Lakes Environmental Research Laboratory, USEPA: www.glerl.noaa.gov), but this
station was most relevant to the study area. Net basin supply (NBS; run-off +
precipitation — evaporation, in mm) for Lake Huron was calculated using mean annual
run-off, precipitation, and evaporation data obtained from USEPA (www.glerl.noaa.gov).
Daily precipitation data for the period of May through September of 1997-2000 were
obtained from the nearest weather station, Essexville, MI (Michigan Climatological

Resources Program, Michigan State University). Mean daily wind velocity and

13

maximum daily wind velocity were retrieved from the National Weather Service (N WS;
www.nws.com) and correspond to the Essexville station.

Hydrochemical information for the Saginaw River was supplemented with data
available from the United States Geological Survey (U SGS; http://water.usgs.gov/mi/
nwis/qwdata, site #041 5 7000). The USGS sampling station was located several km
upstream of the Saginaw River station established in this study.

Interstitial Water Chemistry and Physical Attributes

In March 2000, mini-piezometers (m-p) constructed of 2.1-cm inner diameter
PVC were installed at each station along the transect bisecting the study wetland. The m-
p grid was extended toward shore to include stations at 340, 400, and 460 m from the
wetland/open bay water interface to characterize ﬂuctuations in the water table (see
Figure 1.2). Piezometers were 1.525 m in total length with 0.495 m buried in the soil.
The 0.10 m intake section of the piezometer extended from 0.150 to 0.250 m below the
soil surface and was perforated with uniform holes (0.7 cm diameter). The entire intake
section was encased in several layers of nylon mesh (0.5 mm). Extension of the
piezometer an additional 0.245 m below the intake section ensured collection of adequate
sample volume.

Mini-piezometers were sampled in conjunction with surface water in 2000. Prior
to sampling, head levels were measured in each m-p using a modiﬁed beep stick (battery
Operated buzzer that was activated when electrodes contacted water in the m-p) and each
m-p was pumped down prior to sample collection. To minimize disturbance, each m-p
was bailed down using a hand pump while collecting surface water samples and ﬁeld

measurements. After approximately 30 minutes of recharge time per m-p, a sample of

14

interstitial water was extracted and samples were immediately ﬁltered (Whatrnan GF/F)
in the ﬁeld. Temperature and speciﬁc conductance were measured on the ﬁltrate, which
was retained in a sample bottle, placed on ice, and transported to the laboratory for
chemical analyses.
The elevation of each m-p was determined using a laser level. The stage of Lake
Huron at the time of elevation measurements was used as a benchmark (N OAA real-time
water level data; http://co-ops.nos.noaa.gov/data_res.html; Essexville, MI; station
#9075035). Localized areas of groundwater discharge and recharge were determined
following the conversion of m-p head (piezometric head) levels to vertical hydraulic
gradient (VHG). Vertical hydraulic gradient is a dimensionless measure where positive
values indicate upwelling and negative values indicate downwelling areas; VHG is
calculated using the following equation (Baxter et al. 2003).
VHG = Ah / Al
where,
Ah = difference in head between m-p water level and water level of
wetland surface water in m
AI = depth from wetland sediment surface to the ﬁrst Opening in the m-p
sidewall in m; 0.15m

To better facilitate graphical interpretations of groundwater-surface water interactions
within the study wetland, piezometric head values were recalculated relative to wetland

surface water levels (Ah, see above equation) and re-scaled to wetland sediment

elevations (sediment elevation + Ah).

Laboratory Analyses
A 250-mL aliquot of each surface water sample (1997-2000) was initially passed

through a Whatrnan GF/F ﬁlter (0.7 pm nominal pore size) followed by a Millipore

membrane ﬁlter (0.45 pm). Interstitial water samples (2000) were passed through

15

membrane ﬁlters only. A 30-mL sub-sample of the ﬁltrate was preserved with 100 1.1L of
concentrated I-INO3 for dissolved cation analyses. Cation sub-samples and residual
ﬁltrate (220 mL) were collected into acid washed, pre-leached polypropylene bottles and
stored at 4°C until solute analyses could be performed.

Total alkalinity (T-alkalinity) was determined on unﬁltered surface water samples
by titration with 0.02 N H2804. Alkalinity attributed to the carbonate ion was quantiﬁed
using phenolphthalein indicator (pH 8.3 endpoint) and bicarbonate alkalinity by
bromcresol green-methyl red indicator (approximate pH 4.5 endpoint). Carbonate
alkalinity was observed on occasion (pH rarely exceeded 8.3), but generally accounted
for less than 10 mg CaC03 L". Interstitial Water (2000) T-alkalinity determinations were
based on ﬁeld-ﬁltered (GF/F) samples to ensure the removal of soil particles.

Turbidity was determined on unﬁltered surface water samples (1998-1999) after
equilibration to room temperature (Hach model 2100A).

Concentrations of chloride (Cl'), nitrate (NO3' as N), and sulfate (as SO4'2) were
determined by ion chromatography using chemically suppressed conductivity detection
and a Dionex AS4A column. The detection limit for all solutes determined by ion
chromatography was 0.030 mg L"; observations falling below detection limits were
assigned half the detection limit value (0.015 mg L"). Sodium (Na’) and potassium (K+)
were analyzed via atomic absorption spectroscopy (Perkin Elmer AAS 3100); calcium
(Ca+2) and magnesium (Mg+2) were also quantiﬁed via AAS (Perkin Elmer 5100 PC)
after spiking samples with 500 11L of lanthanum chloride (87g L") for suppression of

interfering ions.

16

Dissolved silica (Si) concentrations were determined colorimetrically (Bausch &
Lomb Spectronic-21) via the heteropoly blue method (Wetzel and Likens 2000).
Ammonium nitrogen (NI-14+) was quantiﬁed colorimetrically (Perkin-Elmer/Coleman
Model 124) using the modiﬁed phenol-hypochlorite method (EPA). Nitrate nitrogen
(N 03) was quantiﬁed via ion chromatography (see above). Nitrite (N02) was not
measured because concentrations were assumed to be low. However, occasionally NOz'
was observed in appreciable quantities in the ion chromatogram, though concentrations
were never quantiﬁed. Soluble reactive phosphorus (SRP) was determined on ﬁltered
samples following the ascorbic acid procedure (Perkin-Elmer/Coleman Model 124
Spectrophotometer) outlined by Wetzel and Likens (2000). Total phosphorus (TP) was
quantiﬁed on unﬁltered samples subjected to persulfate oxidation followed by
colorimetric analysis for SRP. Detection limits for silica, NH], NO3', and SRP were
0.050 mg L", 0.005 mg N L", 0.030 mg N L", and 0.005 mg P L" respectively.
Observations falling below detection limits were assigned one-half the respective
detection limit.

A charge balance was performed to verify the reliability of analytical
determinations. The analysis was restricted to samples in which the full suite of cations
and anions were determined. Cation and anion concentrations were converted to
equivalence units (meq L") and summed. An average deviation of i 7.3% (i 0.4) was

observed for a total of 159 samples. Deviations of 310% were assumed to represent an

acceptable ion balance.

Hydrologic Connectivity between Lake and Wetland

17

A relative measure of surface wave energy (or exposure) at each station along the
study transect was assessed by measuring the mass loss of an alabaster substratum (AS;
Muus 1968; Doty 1971). Alabaster substrata were constructed of commercially available
Plaster of Paris (CaSO4) mixed to the manufacturer’s speciﬁcations, and molded into
disposable plastic cups (90 mL total volume) with 1 cm diameter wooden dowels east
through the center. The substrata were allowed to air dry until masses stabilized,
deployed into the wetland (dowel driven into sediments), and retrieved after
approximately one week. In the laboratory, substrata were again air dried until masses
stabilized. The extent of dissolution, which was assumed to be proportional to turbulence
(Muus 1968; Doty 1971 ), was calculated by subtracting ﬁnal mass from initial mass.
Evaluation of Water Chemistry/Lake Level Relationships

Replicate water samples (N=15) collected on each sampling date were pooled
together to evaluate seasonal differences. All replicate samples collected each respective
year were pooled to evaluate trends across years and relative to lake level. Differences
between years were assessed by analysis of variance (AN OVA). Variables were
inspected for normality and, if necessary, were log-transforrned. Alpha was set at 5%
and all signiﬁcance tests were performed in SYSTAT (version 7.0).

RESULTS
Water Level Fluctuations and Climatic Inﬂuences

Mean annual water levels of Lake Huron declined consistently between 1997 and
2000 (Figure 1.4). The average for the 48-year period of record (1953-2001) was 176.55
(i 0.05) m, approximately 0.14 m greater than that calculated for the 127-year period
spanning 1860-1985 (Bishop 1990). In 1997 and 1998, lake levels were 0.44 and 0.18 m

greater than the 48-year mean and 0.30 and 0.55 m below in 1999 and 2000.

18

.Conanombczz

538:8 Se: 22.: 2 wEEOQmoCOo 2:58 2: 53:3: 8:: :88: Eomto> .93: 83mm moo—.3 $20 3:23an:— 9 3:88:09
Q49 .8ch 33¢ omega Sorrow 2: maﬁa—2: beam we vote: 2: mats: 988 09% >222: 532 A28: #33 58.32 Ewe»
2: 8 mcﬁcoqﬁtoa amp—02w an??? 2: :5 33m no menu» 2: 8.: =2 6:353“: “a 288 Amm H: omﬁm F558 :32 4e.— «Cami

no: 52 .90 new 9.4.. .3. E... .32 .:_< .32 3.: can. _a=:=< =32

 

 

 

 

 

OM.MF— _ p p L m p b _ . in p p r L p ; 6m.mb—
. r f
U 1
8e: . 25am r 23:
H 32 H S
. . . . m
8e: 1 m m - r . a
. :11“ . Senanzm . amen—int
. u p u . . r.
r . . r t n o . :2 m . m
J O > b v > O i .. ﬂ
8.:— - o m m c - 8.: H r 8.5 I
. . o o . o . ﬂ 06m
A o o A . G
. . :52”?— 111 . i
cm 5.6— U m 25" IAVI L I cmﬁh—
. u . 32 101 H .
. m m «a: 1.91 . H
. m 32 101 A
GG.QB— b p _ n _ p b _ m p _ p p - ﬁ soak.—

 

19

Maximum stage generally occurs in July with lowest water levels typical of
winter (Larson and Schaetzl 2001 ), yet this was variable during the years of study (Figure
1.4). Water levels ﬂuctuated on average by 0.08 m from June through September over
the 48-year period of record. Fluctuations during the years of study were higher, ranging
from 0.07-0.21 m. The latter occurred in 1998 and contributed to below average water
levels in 1999 and reﬂected the onset of drought (1998-2001) within the Great Lakes
basin (Larson and Schaetzl 2001; Hoerling and Kumar 2003). The stage difference
between the July 1997 and July 2000 peaks corresponded to 1.04 m and provided two
consecutive growing seasons at lake levels above and below the long-term average.

A signiﬁcant relationship between lake stage and depth measurements collected at
the wetland interface (O-m station) was observed (r=0.994, p=0.000). Water depth at this
station declined from a summer average of 1.31 (i 0.02) m in 1997 (N=4) to 0.27 (i
0.03) m in 2000 (N=7), corresponding to a change of 1.04 m (Figure 1.5). Maximum
(1.35 m) and minimum (0.15 m) water depths were observed in 1997 and 2000
respectively. The range in water depths in the study wetland (20-280 m) was greatest in
1998 (Figure 1.6), and consistent with the trend in monthly (June-September) lake stage
observed for this particular year (Figure 1.4). Water depths measured in the wetland
declined consistently during the period of study; mean depth was 1.09 (i- 0.02) m in 1997
and 0.08 (i 0.01) m in 2000. In 2000, portions of the wetland in the transect region were
observed to be devoid of standing water on three different sampling dates. In fact, upon
completion of sampling on 12 September (mid-moming) the entire wetland was observed
to rapidly de-water, a direct result of a moderate seiche. Shallow water depths in 2000

also left near shore regions of the wetland (>280 m from the wetland interface) dry,

20

 

1.40
1997; N=4
0 .................................................................................. E
l
1998; N=9 l A
Q .................................................................................. r- 1.” g
- =
.2
- 0.80 *5
. >
999 N- ' 060 g
l ; — ~ .
a .................................................................................. . I-
" e
F 0.40 5
2.999;.NT7. ..................... - 5"
. 3- 0.20 :
Wetland Interface Wetland Sediments I
//////////II " "-00

 

I V I ' I ' I ' I ' I ' I ' I ' I j I V I j I

0 40 80120160200240280320360400440
DFWI (m)
Figure 1.5. Mean annual water depth (1 SE) at the wetland/open bay water interface (O-m
station) relative to sediment surface proﬁle. Water depths and sediment elevations are
referenced to the 0-m station (i.e. O-m station has an elevation of 0.00 m; see methods).
DFWI corresponds to distance from wetland interface in meters.

 

 

 

 

 

 

 

 

 

 

1.06 I 1.40
O
0.80 31.20
+ E1.0o
.“ , - A. b A
-_ -0.80 E,
.. , : .5
’ G-
i :0.60 5
N=135 3
r 0.40
' 0.20
N=105
. . . 0.00
1997 1998 1999 2000

Figure 1.6. Box and whisker plot of water depth measurements collected within the
wetland (0-280 m) during each respective year. Box and center line represent the
25th and 7 5th percentiles and the median respectively. Whiskers correspond the
10th and 90th percentiles, circles represent outliers. Numbers above each box
represent the mean wetland depth for each respective year; number of individual
measurements are denoted below.

21

though shallow (<2 cm) remnant pools remained interspersed throughout this region
(Figure 1.5).

Over lake precipitation was approximately 75 mm greater in 1997 relative to
1998-2000 and the 48-year average. Net basin supply (NBS = precipitation + run-off —
evaporation) estimates paralleled water level trends for Lake Huron; NBS declined
consistently from 1997-1999 (Figure 1.7). Run-off and evaporation were more sensitive
measures of climatic inﬂuences on Great Lakes water levels. Run-off decreased
consistently between 1997 and 1999, whereas evaporation increased from 1997 tol 999.
Run-off data were not available for 2000; however, NBS was likely intermediate to 1998
and 1999 based on the similarity between precipitation and evaporation estimates for
1998 and 2000 (Figure 1.7). If this were true, the relationship between lake stage and
NBS would have likely diverged in 2000, yet would not have been realized immediately.
A lag time in the response of lake levels to changes in climatic conditions is common
(Larson and Schaetzl 2001), and although lake stage continued to decline through 2001
(mean annual stage = 176.96 3; 0.03 m), a 0.17 m rebound occurred in 2002.

Local (Essexville, MI) precipitation totals decreased consistently from 1997 to
1999 (Figure 1.8).. Precipitation was above average in 2000 relative to the 30-year mean
(1971-2000). These totals were much more sensitive indicators of year-to-year variation
compared to whole-lake estimates used in NBS calculations.

Specific Conductance and Total Dissolved Ions
A total of 358 speciﬁc conductance measurements were collected along the study

transect throughout the spring and summer (June-September) of 1997-2000. Values

ranged from 269.2 to 1120.0 uS cm"; the grand mean for all four years was 417.9 (: 5.4)

22

3000

 

  

2000 - 1266.2 ‘ L

4000 ‘ :1 Evaporation -
Precipitation

m Run-off

'2000 l l l I I

1997 1998 1999 2000 1953-2000

      

1000 -

mm
.\\

 

 

 

Figure 1.7. Mean annual estimates of evaporation, precipitation, and run-off for Lake
Huron during the period of study (1997-2000), including long-term means (48 years).
Numbers over bars correspond to net basin supply (NBS), where NBS = rim-off +
precipitation - evaporation, in mm. Run-off data were not available for 2000.

 

1000

 

 

800 _ ......................................................................

 

 

 

S
c

 

Precipitation (mm)

 

 

 

 

 

 

 

 

 

 

 

 

o , ' I _ , I]. _l ‘ ‘ ,l.
1997 1998 1999 2000 1971-2000

 

Figure 1.8. Annual precipitation totals (mm) corresponding to the period of study (1997-
2000) and relative to the 30-year average (1971-2000) measured at Essexville, MI
(station #202631). The dotted line corresponds to the 30-year average. Data were
acquired from the Michigan Climatological Resources Program (see methods).

23

11S cm". Within-year temporal variability was high and increased in conjunction with
declines in lake level. Mean speciﬁc conductance values increased systematically
between 1998 and 2000 (Figure 1.9). The range in observations was smallest in 1997 and
1998, 230.2 and 180.5 11.8 cm" respectively. Conductance measurements made in 2000
sparmed a range of 809.7 118 cm" and approximated the total range observed throughout
the four-year investigation (850.8 118 cm"). Several anomalously high measurements
were collected in 2000; comparable values were not measured in previous years.

The total dissolved ion (TDI; meq L") content of wetland surface water samples
was signiﬁcantly correlated (r=0.985; p=0.000) with speciﬁc conductance measurements
and concentrations increased consistently as lake levels declined (Figure 1.10). Cations
were not quantiﬁed in 1997, which precluded the calculation of TDI. Overlap in TDI
among 1998-2000 was evident, however, mean TDI values were 7.67, 9.41, and 11.52
meq L'1 respectively. The range in TDI values was similar for 1998 and 1999, but
increased in 2000 (Figure 1.10). Analysis of variance indicated that annual TDI
observations were signiﬁcantly different from one another (F 2,99 = 69.701; p=0.000).
Exclusion of the two outliers for 2000 did not alter the outcome of the ANOVA, but did
decrease the mean to 11.00 meq L".

Major Ion Chemistry

Mean T-alkalinity, Cl“, and 80.;2 concentrations increased throughout the
investigation (Figure 1.11). Carbonate alkalinity was not quantiﬁed in 1997 and 1998,
but based on pH measurements (1998 and 1999 means = 8.03 respectively) may have
contributed more to T-alkalinity during these years relative to 1999 (2.0 1; 0.2 mg CaCO;

L") and 2000 (1.0 i 0.2 mg CaCO3 L"). Total-alkalinity ranged from 58.0-540.0 mg

24

Speciﬁc Conductance (118 cm")

 

 

 

 

 

 

 

 

 

 

 

 

12% I 1 I I
I
1000 - -
O
800 - l
600 l i g -
40° ‘ 55-23 N ' V '
i o
200 d l-
384.9 361.2 431.3 533.3
0 I I I1 I
1997 1998 1999 2000

Figure 1.9. Box plots of speciﬁc conductance (1.18 cm") measurements collected from the
study wetland during the spring/summer (June-September) of 1997-2000 (N=60, 135,
90, and 73 respectively). Boxes represent 25th and 75th percentiles, medians are
indicated within boxes, whiskers correspond to 10th and 90th percentiles, and circles
represent outliers. Mean values are italicized below boxes.

 

 

 

 

254 ‘ IV V 1 r j v ' 1 11’
"A ‘ r
.4 l r
5.20, 0 -
o 1 >
a . O .
v 1 t
215: L
.3 j e .
e . <> .
s 4 ~
E101 _-
.2 j '
a , A .
'5 5: A 1998:
e . I 1999’
E- .

f o 2000.

o W .ﬁ-......-g,- ..

0 200 400 600 800 1000 1200

Speciﬁc Conductance (1.18 cm“)

Figure 1.10. Relationship between speciﬁc conductance (118 cm") and total dissolved
ions (TDI; meq L") for wetland water samples analyzed for the full suite of major anions
and cations during the spring/summer (June-September) of 1998-2000 (N=3 7, 34, and
32 respectively). Cation concentrations were not quantiﬁed in 1997.

25

.mcozmbcoocoo :35 o. becamotoo xon sumo gone. m..onE=Z .32» 9.5305 2 03.20.. amass”. Amodvnv

Ewoﬁcwﬁ 836:. 8.33.3. .mowcmso 28m 802 £2.50 0.... 8.88 E... £05582. Sea van :5. 3 9832.80 £3323 £889
05 55.3 2.. £582: £23582. 53. 28 52 E883. moxom 380: $2.... b02838. ms 98 .3 .mm. 6an SENSE.
v2.2— zuam 05 macaw 3.00:8 838mm .895 5.8:. .8 b-.. we. 82.88880 62$ :25 8.8:. mo 30:. xom ._ 2 233m

cgN aaa— @3— ram—
. ﬂu...

. ., 9% 93 ......m
. .

y. .98.
Van

 

ed

 

L
F
‘ ‘

 

 

 

 

. adm—

13m

:3 coal—f

. . 9.2."

m . cdmu
AU -«vOm .

46.

 

 

 

cécn

60c b p p p p P p p :0:

can . . . . WW WWW . .3...
. o . . .
$.GQ . p r 4 * .. GAuGN
an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o
a m
m . . . 3.2 k N2 . and
. o . e
3 . . m 5 .
.. 8 . . . . . e 8». D
J . . U a. n 3 . . o . O
... . 3‘ 3... . c
c 9% L . . Ndmé .. c 8? Fl...
. * v . v II
3.... . 3m . . . 2...».
. . . I
a .5 . 2 x. 35.3.... ..

 

 

adu— 0.25

26

CaCO3 L'1 with the lowest concentrations measured in 1998 and 1999, highest in 2000.
Chloride concentrations ranged from 24.2-82.5 mg L’1 throughout the four-year period
and on a sample-speciﬁc basis, generally exceeded SO4’2 concentrations (4.9-270.1 mg L'
1). Variation in the concentrations of major anions also increased systematically in
conjunction with declining lake levels (Figure 1.11). The most notable changes occurred
for C1' and 804'2 concentrations. Contrary to the other major anions, the range in 304'2
concentrations increased in both directions during 1999 and 2000. The greatest range in
804'2 concentrations occurred in 2000; low concentrations were also measured in 1999
(5.1 mg L").

Analysis of variance indicated that T-alkalinity and Cl' concentrations were
signiﬁcantly different (p<0.05) in 1998 relative to the previous year; T-alkalinity actually
decreased on average between 1997 and 1998 (Figure 1.11). Statistically signiﬁcant
increases in T-alkalinity also occurred in 1999 and 2000. Chloride concentrations were
similar between 1998 and 1999, but differed signiﬁcantly in 2000. Sulfate concentrations
signiﬁcantly increased in 1999 and 2000 (Figure 1.11).

Proportional shifts in major anion chemistry were evident in conjunction with
declines in lake stage. The proportion of total anions comprised by T-alkalinity
decreased from approximately 55% in 1997 to 43% in 2000 (Figure 1.12). A similar
trend for CI' occurred, though 1997 and 2000 were similar on average. The number of
observations for 1997 was comparatively low (N =20). The percent of total anions
accounted for by 804'2 increased consistently between 1998 and 2000 (Figure 1.12).

Sulfate represented 28% of the total anions in 2000. In addition to the proportional

27

.moxon 30% @2865 2a .80» 0282.3.— noao .8 322805 :32 @8225 .5852

8.26 98 .3552»... 53 can :5. 2 9.2.8on ﬂue—m2? £88.. 55.3 380%5 03 £8238 £25522. 5mm can 53
2.8292 moxom Abozwooamo. wwuZ v.8 .wwnZ .mmTZ .omuZ mooomgbav 25% mo @2qu 23 50.3.5.5 380:8 $383

.533 28:25 .«0 ADV 0328 use Amv 02820 Add €==§Eé .3 communaoo 82.3 139.3 c0930.... 05 mo 32.. xom .N: 95me

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.." a... .3... a...
. t r 8...
.d
. a M
. nuﬁuu a? .3... .m
. .2... m.
.. .t..... may. . 0
. «:6 N26 . 3.: u
. . m...
L
.. .. 99.: 106.
. o. . W
. a: .8... w.
0
. V v
O -u Om N
r 2..
8.." a... 23. as. 8.." 82 2a a...
$6.: - p p L h p p b so:
. . . . J
m
3.... . . .3... .m
i . .
I. n+- . . . w.
3.... . . u
. an... 2.3 o . . 5s a m.
. $§ .. .. I.
a... .T o . . m a; W
. “ENS . 34.6 W
3.... . . ....
. . . . 0
E .6 .< 5.5.5.34 m
2..— cc.—

 

 

 

 

 

 

28

changes in the major anions, variability increased across years, particularly in 1999 and
2000.

Concentrations of the major cations also increased during the investigation,
however the trends across years were not entirely consistent with those noted for the.
anions. Concentrations of Ca+2 were signiﬁcantly higher in 1999 relative to 1998, yet
slightly lower again in 2000 (Figure 1.13). Minimum values were similar among years
(30.5, 28.8, and 30.6 mg L'1 respectively), but a general increase in the upper range
occurred as lake levels declined; the greatest range occurred in 2000 (30.6—106.8 mg L“).
In contrast, mean Mg+2 concentrations increased consistently throughout the three years
in which cations were quantiﬁed (Figure 1.13). Statistically signiﬁcant differences were
evident in 1999 and 2000 relative to each respective previous year. Similar to Ca”, the
lower end of the range in Mg+2 concentrations was highly comparable among years, yet
progressively higher concentrations were measured each ensuing year, particularly in
2000 when wetland water depth was lowest. Sodium concentrations were variable,
though the range was similar between 1999 and 2000 (11.3-23.7 and 10.5-23.9 mg L'1
respectively). The concentration of Na+ was on average lowest in 1999, yet identical in
1998 and 2000 (Figure 1.13). Potassium concentrations were statistically different in
2000 relative to the previous year and ranged as high as 5.4 mg L'1 in 2000. Mean
concentrations and the range of concentrations (0.5-2.9 and 05-27 mg L'1 respectively)
were nearly identical in 1998 and 1999 (Figure 1.13).

The percent composition of the major cations was dominated by Ca”, which
accounted for approximately 40-5 7% throughout the study (Figure 1.14). The proportion

accounted for by Ca+2 was very similar in 1998 and 1999, yet decreased in 2000. In

29

.82 :28 .82. 3.86.: 2.. 828.838... E82 .80» 28.58 c. 03.22 owcaso Godqu
.cmuuﬁwmm 8.3%.: axmtoﬁ< 88:28 2.. 8.28 E... .8388...— ﬁoo 9... 5o. 3 283:8 80.7.23 .88.. on. 55.3 2.. 28:8...

£23583 £2. 2... £3 .5882 88m .0— 9... «Z .8 .m 9... .VM Kmuz 9... .32 can «U 8.. mu v.8 .8 .mﬂuz .88-”.83
“.28.. .323 on. mama... 828:8 «29:8 8...? 98—83 .8 A.-.“ we. 82.88880 5-5 .838 8.8:. .8 m8... xom .m _ ._ 08w...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

scan 33 «88— ga— ccen 28— 33 53
c.: b p P P n if p P Go:
. $ + . . S: .1... 3: .
...N. . . . w . . ....m
. O . . . .2
.2 . . . .
5.? A N N v . I. lip] . m
. . . . . 3
a... . o L a . ..m. J
a . , . t
r 2 . . - _ . lull-I s...
..... . . .. o m .....m~
. . .. ..
AG +v— AD +52 .
e...— t 0 .r . . . ¢ .T 8.3
..... . . i. . . . t . .....
. .5. 3. «.2 . . 3m New 3.. .
...... . . . . a..."
.....N L 1.0] . w . . & ........
. n a . . w . ‘ . m
.....n . G . . .. . .2... 8
. . * . 1P
. . o . . . .
...... . . o e i...
. a
.2... . o . . . ........
. 3 cm a»: . . A< 0 3:0 .
3... 3.2

30

8.3.08.2. 8.. 0.800. ... 08.2.0 082 80.8.. 0.6.... 8080...... 08 80.. 02.02.00. 2080

8.. 0.8828... .80.). 0.2.80 0... 00.0.8 88 82.8008. :50 .28 5o. 0. 2.2.3.80 0.8.0....» .88.. 52...... 8080...... 08 0880...
82.8008. 5mm .28 5mm 800.080. 008m. 6202.02.00. mmuZ .Vmuz .nmnzv oooméom. 8...... 8.02.00 00.9.8... .083 208:.

.8 a. 83088.. E... .6. 85...... Am. 89.00.88... .3... 85.280 .3 802.980 08.80 .88. .8 22.8.8... 0... .8 .8... 8m. .3 .. 0.3m...

 

 

 

 

 

 

 

 

 

 

GGGN a¢a~ waa— bam— GGON aaa— waa— baau
ago: p h P h b h n :60:
"Av AV 4 §N§o§ §N§o§ v ._ + v :N a m
. h~§$ . . MMNS ©V~$ wn~$ . .0
.56 . . 1 . I:
. . 3. c .I.
. . . . o
4.»
cc... . . . . :9... I.
3
. . B
. . . w.
.... .. . . . . 8 .. u
s
. 8 .0. G .2
:—.¢ 0 0 4f 0 u u u 0 cc.—
ccd . . h . . . . . $6.:
. wVMS ©~MS QQNS . VaVS QNhé ”3.6 .
. . . .d
an e . . 1 cu c M
+ d
. III! . .o
:9... r . . . an... I.
III I? r.
. . . I. . w.
a
co... . . . . co... .I
3
. . B
. . w.
ca 9 . . . 1 cm a u
2 S
Am +~w2 . Ad. 3&0
cc.— .5.—

 

 

 

 

 

 

31

contrast, the proportion of total cations accounted for by Mg+2 increased systematically
between 1998 and 2000; variation was similar during these years. There was no clear
trend in the proportion of total cations accounted for by Na+, though variation increased
across years. Potassium accounted for less than 2.3% of the major cations (Figure 1.14).
The range of K+ conCentrations was highly similar for 1998 and 1999, but increased in

' 2000.

Major Ion Chemistry: spatial and seasonal variation

Spatial patterns in speciﬁc conductance measurements collected along the study
transect were evident during 1998-2000 (Figure 1.15). Speciﬁc conductance
measurements were similar shoreward of the wetland interface in 1997; the mean
difference between the wetland interface and 280 m station was approximately 12 1.18 cm'
1. As lake levels declined, increases in speciﬁc conductance along the wetland transect
indicated progressively stronger chemical gradients. Mean differences between the
wetland interface and 280 m stations were 38.8, 141.6, and 165.1 uS cm'1 during the
years 1998-2000 respectively (Figure 1.15). The biggest changes occurred in 1999 and
2000 when lake levels were below the 48-year average. Variability at individual stations
increased systematically across years. In particular, high speciﬁc conductance
anomalies were repeatedly measured at several stations (120 and 260 m) in 2000.

The change in speciﬁc conductance proﬁles ﬁom no spatial trend to consistently
increasing values shoreward of the wetland interface implied that the extent of
lake/wetland mixing decreased as lake levels declined. The alabaster substrata (AS)
experiments were speciﬁcally designed to quantify the degree of lake/wetland mixing.

Mass loss observed for AS declined in a near linear fashion 100-120 m (outer region of

32

08.0... ... 300.82.. 8.03 .3 02.0.0.8..03. 80.8.... 9.0.83 80... 02.0.0...
0. 0.2.2.0280 SEQ 3.03.0098. ng 0:0 .cMZ .auZ .an 083 5:0... ..000 8.. 23.02030 .0 89.5... 0.... dooméba. ..o
..0.....0...0m-0..3.v .0E.....0\w......m 0... w........ 8080.. .30... 0... $8.0 00....0... .mm H . _....0 m3 00:08:28 0.200.... .803. .m . .. 0.9m...

 

 

 

 

 

 

 

.a. .3...
00. 0... 00. 00. 0.. 00 0.. 0

p . . . _ p — . . . p l r \- P 6:"

n .. - 00..
I I I .m.
. I 000 a.
m.
.A 1 a
_ 3
- r 000 m
. -- n.
4 I as. m
)
4 -r . m.
L - 000 m
4 v (bl

000. I9:
- -- a... IDI - 0...
03. It!
. 3.. IOI .
p . p p b . — p b F p p P r p :5:—

33

the wetland) shoreward of the wetland interface in 1999 (Figure 1.16). Mass loss was
greatly reduced beyond this region and suggested equilibrium dissolution was more
important than mass loss associated with physical abrasion (surface waves). A similar
pattern occurred during the 2000 experiment, except a substantial reduction in
lake/wetland mixing was suggested approximately 60-80 m from the wetland interface
(Figure 1.16). Mean daily wind velocities observed during each AS experiment were not
signiﬁcantly different (p>0.05; two-sample t test). However, mass loss during 1999 was
consistently greater than that observed during 2000.

Spatial trends for the major anions mirrored speciﬁc conductance proﬁles. Mean
T-alkalinity concentrations measured along the study transect increased shoreward of the
wetland interface during all four years (Figure 1.17). The difference between the wetland
interface and the 280 m stations was greatest in 1999 (57.7 mg CaCO; L'l), and similar in
magnitude during 1998 and 2000 (29.4 and 29.8 mg CaCO3 L'l respectively). However,
variation was high at the 120 and 260-m stations in 2000, consistent with speciﬁc
conductance measurements. Mean differences in Cl' concentrations along the wetland
transect were greatest in 1999 and 2000; distinct spatial patterns were not evident in 1997
or 1998. Variability in Cl' concentrations was greatest in 2000. Sulfate concentrations
decreased on average along the study transect in 1997 and 1998 by approximately 4.0 and
2.3 mg L'1 respectively (Figure 1.17). The patterns were nearly identical during these
years, but the number of replicates was low in 1997 (N=16). The pattern was reversed in
1999 and 2000, concentrations increased by 6.8 and 77.9 mg L“1 shoreward from the
wetland interface; variability increased in the same direction in 2000, particularly from

200-280 m shoreward of the wetland interface. Higher mean anion concentrations across

34

 

100.0 I I I I I T I ' I ' I I F ' I

—l— 1999
+ 2000

80411 -

as
9’
Q
l
j

Mass Loss (g)

40.0d ¢r _

20.0 « t

 

 

 

0.0 I r j ' I ‘ I T I j I I I f I

40 80 120 160 200 240 280
DFWI (m)

Figure 1.16. Mean mass loss E SE) of alabaster substrata (AS) across the study wetland
during mid-summer (August) 1999 and 2000. Arrows indicate subjective inﬂection
points where mass loss appears to stabilize throughout the remaining portion of the study
transect. DF WI corresponds to distance from wetland interface (wetland/open bay water
interface) in meters.

35

.282: E Aoomtouﬁ BBB >3 5233265 03.285 28:25 Set 353% 8 manage—co 3E0 .oEQEw some 8.. 335
mm Eamo— coEEoo < .2628 :26 8m 28m 5 moms—Eu BoZ .ooomAbE mo Conﬁoﬁomocsc SEEEEEEm 2: 9:26 28:03
338 2: $28 $8.28 98 62820 .bEm—nﬁaév 28:8 8.38 no 8385:0280 Amm 3 :38 com mp5: :3QO .5 _ ._ 83mm“

 

 

 

 

 

 

 

 

 

 

 

 

. . . an

A.
I
H
C
[-1 ‘0383 3m

 

a . . .
E. . . L % . 8n
. .6 . . Qésié .
g“ . p P L P p p . . u L p p r b p p p f b p h p F > p p u L b an”

 

 

 

 

 

 

36

years were consistent with the general increase in ionic strength of wetland surface water
as lake levels declined (Figure 1.17).

Spatial trends were observed for all of the major cations (Figure 1.18). Mean
Ca+2 concentrations increased shoreward of the wetland interface during each respective
year; the difference between the wetland interface and 280 m station increased from 6.6
mg L'1 in 1998 to 25.9 mg L'1 in 2000. Similar trends were evident for Mg2 and the
greatest mean difference occurred in 1999 (5.5 mg L‘l). Magnesium concentrations were
variable in 2000 relative to previous years; anomalously high Ca+2 and Mg2
concentrations were evident at the 120 and 260-m stations (Figure 1.18). Sodium
concentrations were spatially and temporally variable, yet lowest overall in 1999. The
mean difference between the wetland interface and the 280 m station was approximately
4.0 mg L'1 in 2000; no distinct spatial patterns occurred in 1998 or 1999. Spatial trends
for K+ were subtle, but generally declined by approximately 1.0 mg L'1 shoreward of the
wetland interface in 1998 and 1999. Spatial variability increased in 2000 and a general
trend in K+ concentrations across the study wetland was not evident (Figure 1.18).

Seasonal trends (i.e. spring versus summer) in speciﬁc conductance
measurements were difficult to assess, but higher values were more typical in June (late
spring) and September (late summer) 1997 and June 1998. Temporal variability in 1999
and 2000 could not be attributed directly to seasonality (i.e. periods of elevated run-off).
Concentrations of T-alkalinity, SO4'2, and perhaps Ca+2 followed the seasonality trends
noted for speciﬁc conductance. Total-alkalinity and Ca"2 concentrations were generally
elevated in September 1999 and 2000 relative to previous months, whereas SO4’2

concentrations measured in late spring 2000 were the highest observed during the four-

37

.882: :_ 38:88: 88>» b3 :28

85:25 08:88 8:283 :8.a 855% 2 8:88:80 SEQ 8:8 83380: 05 E wean—amﬁ ﬂ 8:: uEaﬁw some :8 38% ﬂ
Emmo— :oEEoo < .828 some :8 28m 5 83:8 082 :58-”me mo ConEoEombgc BEE:m\m:_8m 2: $5.8 8:283 82m
2: $28 AEBmmSo: 8:: .8568 .838:me 8:838 82:8 8.3:: ,8 Amm .3 80288880 :3:— .8 8:6: 38% .w _ ._ 08$.»

:5 .38 :5 :58
can :3 can 8— :2 cm 3 a an" ova SN 3: cu. cm as c

c h I I D P b I P P I I } P h h D D b P F b h I P D b b P I h m

 

 

 

If] 3m

 

 

 

 

:—

 

 

:—

 

 

34

an.

mu.

[TI 3m

an.

 

 

 

 

 

 

av

 

 

38

year study. By late summer 2000, however, 804'2 concentrations decreased to the lowest
(<5 mg L'l) measured throughout the investigation.
Nutrient Chemistry: spatial and temporal patterns

Concentrations of NO3' within the study wetland ranged from below detection
limits (<0.03 mg N L") to 24.8 mg N L". Maximum concentrations measured each year
(1997-2000) were 3.7, 24.8, 11.0, and 6.2 mg N L'1 respectively (Figure 1.19). In
general, mean NO3’ concentrations were similar among years. A total of 20 samples were
analyzed for N03’ in 1997 and despite low sample size, none were below the detection
limit. Nitrate concentrations were much more temporally variable during ensuing years
and often below detection limits by late summer (Figure 1.19). The frequency of
observations falling below the detection limit was slightly greater in 1999 (65%)
compared to 1998 (53%) and 2000 (53%). Concentrations were typically elevated in
June (late spring) and July (early summer) throughout the four-year investigation.

Concentrations of NH? were high within the study wetland; maximum
concentrations observed during 1998-2000 were 0.42, 0.13, and 0.38 mg N L'1
respectively. Measurements were never below the detection limit (<0.005 mg N L'l).
Temporal variability was higher in 1998 and 2000 relative to 1999 (Figure 1.19).
Ammonium concentrations did not increase in response to declining lake levels, yet
higher concentrations were more common in late spring relative to summer in 1999 and
2000.

Spatial trends in N03' concentrations were noted across the study wetland (Figure
1.19). Concentrations declined consistently from the wetland interface toward shore

during 1997, 1998, and 2000; concentrations were highest within the center of the

39

.8888 8 888:: 8:283 808 8:888 0. 8:88.80 SEQ .88 some 088. 88888 2.. 832 8:8 8:3 .8:
8088888.. 88:: 8.. 28m 385.8%2 SoZ .mnuZ 88. .ooHZ .mm TZ 6.1.2 083 808.888: 82888.. 503888 mnnz
8:: .meZ .mmTZ .EHZ 895 288.8830 88.2 A288 895 .888. 83m 2: m8... 82.88880 82:088.. 8:: 88::

:8 8:8. 88QO Em H5 :88 8:: $68.... to: 82.88880 :umoE: 28.90:. 883 8885 8:283 .8 82: 8m .2 A 88m...

:5 .38

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

can :3 gm 8: :2 on 88 a gen 3a 33 Sa—
....... L . . _ t t . . . . . 2....
I |
r z . S...
8.: E 8...
m
80
— .8... N
J
o . . r
8.5 2.... 8..
on... a 5...." ivl . em...
32 Ill
...=z
mu... . . . . .8...
8.... . . . . .
l- 1 : o z...
.2... .
5...". .2...
i, m
.m .....m. 3
N .2... N
1 8.... ﬂ
.1 o
....m . as a a; . 2...:
2.5. m: an:
8.: IOI .. a
2... oz - oz 8......

40

transect region of the wetland in 1999. In contrast, no spatial trends for NH;
concentrations across the study wetland were evident in 1998 or 1999, but generally
declined within the inner region (160-280 m) of the wetland in 2000 (Figure 1.19). The
spatial patterns for dissolved NO3' and NH4+ were in good agreement in 2000.

Soluble reactive phosphorus (SRP) was generally below detection limits (<0.005
mg P L4) in 1998 and 1999, although variability was greater in 1998. Concentrations
were much higher throughout 2000 (Figure 1.20). However, no distinct spatial patterns
were evident across the wetland during any of the years. Mean wetland surface water
concentrations of T-P (total phosphorous) were highest in 2000 and lowest in 1999
(Figure 1.20). Spawning carp were observed in high concentration up to 220 m from the
wetland interface during the ﬁrst sampling date of 2000 (12 June). Low water in
conjunction with excessive bioturbation caused by carp appeared to coincide with
elevated T-P concentrations. An approximate four-fold increase in concentration
occurred on this sampling date relative to the others, thereby skewing the mean higher
than previous years. The presence of carp also coincided with high NH4+ concentrations
during the same sampling date. Total-P concentrations generally declined across the
wetland in 1998, but the pattern was less distinct in 1999. No spatial pattern in T-P was
noted in 2000. Mean turbidity values across the study wetland in 1998 and 1999 agreed
well with spatial trends in T-P during each respective year (Figure 1.21).

Dissolved Si concentrations were similar on average in 1999 and 2000; Si was not
measured in 1997 or 1998 (Figure 1.22). Greater variability occurred in 2000 relative to

the previous year and concentrations exceeded 5 mg L". The concentration of dissolved

41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.04 . . a .
SRP
0'03 ‘ 0 005 0.013 '
O ’ r
.. l
=- 0.02 ~ -
u, o
E
o
o
0.01 l %
BDL '
0.00 i 4. : ;
0.40
T-P
0.090 L
I
0.30 - o
L]
n. 0.20 ‘
5D
E 0 060
-. 0%.?
. i
0.10 ‘ o
0.00 . . % .
1997 1998 1999 2000

Figure 1.20. Box plots of wetland surface water phosphorus concentrations measured
during the period of study (1997-2000). SRP represents dissolved soluble reactive
phosphorus (N=0, N=135, N=90, and N=75 respectively) and T-P is total phosphorus
measured on unﬁltered samples (N=60, N=135, N=90, and N=46 respectively). Means
are indicated above each box. BDL corresponds to below detection limits (0.005 mg

P L'1 for SRP). Soluble reactive phosphorus was not measured in 1997.

42

 

 

 

 

25.0 I ' ﬂ ' r I ' I
1 + 1998
—D— 1999
20.0 ‘1 '
5
H
a 15.0 - -
.8
8
:5 10.0 - +
h
:1 , ﬁ
[—
5.0 . W -
0.0 ' ' I v T v I v I ' I ' I ' U
0 40 80 120 160 200 240 280

DFWI (m)

Figure 1.21. Spatial trends for mean (: SE) turbidity proﬁles across the study wetland
during the spring/summer (June-September) of 1998 (N= 135) and 1999 (N=90).
Turbidity was not measured in 1997 or 2000. DFWI corresponds to distance ﬁom
wetland interface (wetland/open bay water interface) in meters.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6.0 . . -
J 1.4 '
5.0 a Q .
. b
4.0 a O r
1.3 1
1, ‘ o 0
an 3.0 " P
8 . .
2.0 - . . V . _
1.0 « f g‘ t. .I. ‘7 ..
' i t
0.0 . . . .
1997 1998 1999 2000

Figure 1.22. Box plot of dissolved silica concentrations measured during the spring/
summer (June-September) of 1999 (N=90) and 2000 (N=73). Silica was not
measured in 1997 and 1998. Boxes represent 25th and 75th percentiles, medians
are indicated within boxes, whiskers correspond to 10th and 90th percentiles, and
circles represent outliers. Mean values are italicized above boxes.

43

Si tended to increase shoreward of the wetland interface in 2000; no spatial trend was
evident in 1999.
Chemistry of Potential Source Waters

Source waters that potentially inﬂuenced wetland surface water chemistry
included the Quanicassee River, the Saginaw River, and adjacent drainage ditches (see
Figure 1.1 and 1.3). The Quanicassee River had the lowest speciﬁc conductance
measurements, but increased on average between 1999 and 2000 (Table 1.1). Speciﬁc
conductance measurements were similar to those collected from the wetland.
Additionally, the dominant ions were proportionally similar, yet concentrations were
generally elevated relative to the wetland during late spring; this was particularly true for
SO4'2, Ca”, NO3', and NH4+. The range in TDI for the Quanicassee River samples (6.54-
10.61 meq L") overlapped wetland surface water values during each respective year
(1999 and 2000). However, Quanicassee River values were generally near the lower
range of wetland TDI values, and in fact corresponded with samples collected at the
wetland interface (0 m) station. Although discharge was never measured for the
Quanicassee River, concentrations of NO3', NHK, and phosphorous were high and likely
constituted an important non-point source to the bay. In fact, T-P concentrations were
much higher on average compared to the study wetland; variability was also lower (Table
1.1). There was no obvious seasonal pattern for T-P, yet concentrations were consistently
higher in 2000 relative to 1999.

Speciﬁc conductance values were much higher for the Saginaw River compared
to the Quanicassee River; variability was much greater in 1999 and 2000 (Table 1.1).

Total alkalinity of the Saginaw River was much higher than any of the other Saginaw

44

..~02 82:05 83 3.95m 83.: Bacaam 2: 8.. .52.,

 

 

808 8x 808 «3 A28 2.. A88 8% 828 So 808 N3 9... as: a
25.8 23. #88 so... 388 23. 68.8 :3 $88 ~88 38.8 83. p: m 95 55
28.8 on; 88.8 33. 55.8 $3 38.8 E; 88.8 3 _ .c 88.8 2:; n-.. .— uev .E.
:28 8.3 28.8 a: $2.8 22 38.8 83 p-.. z “.5 z-..
88.8 858 88.8 8.3. :88 :38 28.8 E; 88.8 32 2.88 S; a-.. z «.5 .52
28.8 8: 83.8 33 3.8.8 :2 8:8 as... A5,: 8.; 823185 2.... z 95 .62

€48 03 A38 2% 5.8 93 A88 Ea :08 a? $8 .2 n... 95 .2
A38 3.: 5.3 3.3 3.3: 2.8 828 3:. :38 8...: A88 8.8 r: use .2
5.8 2.2 5. 8 of: 25.8 8.2 :58 8.8 5.: 3.8 Go. 8 an: n... ”.5 a. a:
a: c 8.8 5.8 8.8. 808 3% $08 8.8 83.8 2.8. 5. 8 8.? n: ”.5 :5
and 8.2. $8 2.8. 5.8 2.8. 828 2.3.. 8.8 3.? £28 3.8. n... 9.: «Sm
83: 8;: 3:8 3.. : 82.: 8.8. 5.8 no...» $188.3 5.8 2.8. n-.. 95 -6
3.5 m. z: 3.88 is. :28 3: $.88 on: 8.8 $8 3.8 :5 p... 505 «.5 .__<.._.
8:: : 3% A38 3: 8.88 e. .2 3.8 3% £68 2:. 8.3 38. £5 8: .25 3%
am I 2.3 mm - 32 mm - ama— zm I 2...: mo - econ mo - 98. .5852:

 

.3552: 83 £8: 5:836
328 9 magnumotoo 15m 35% $5 80 :3 mm 380:8 29$ :02 doc—V 53m Swimmm 2: do.“ 5% 2: mo 08% can 302

doom can 002 .328 821 0832530 05 do.“ Sam .egg—“Elbows;TEES,“QmOmDHvolxocowmﬁooonmGenealozmmﬁwgrc
\m_3=\_E\>ow.mwm:332539; mOmD 2: 80¢ 3:350 225 ﬁne» 3:832 :28 S.“ NHZV ooom ES woo—-33 mat—6 83M
BaEwam 05 not“ 3% 3382—86»: 33¢ 323$ 05 9 Mm 53.x 0333530 2: 9 muconmotoo MO 458-33 80 Conﬁoaom
-253 BEEEREEm wESw Egg BuEwmm use 03885.50 of no Amm HV 83288830 32823 $83 8&2; :82 .2 035.

45

Bay surface waters characterized in this study. The river was an important source of Cl',
Na+, N, and P to Saginaw Bay; high Na+ concentrations relative to the Quanicassee River
illustrated the differences in land use and drainage basin size (Table 1.1). Dissolved
inorganic N and P concentrations were generally elevated in spring. Several of the other
solutes (Ca+2, Mg”, and 1C) were less variable within years; however, no clear patterns
were evident across years and likely reﬂected low sample size. Total dissolved ion
content ranged from 10.69 to 16.62 meq L'l.

Several offshore stations were sampled around the bay in 1999 (Figure 1.3).
Comparison of solute concentrations from these stations with those measured at the
Quanicassee River and the Saginaw River made it possible to infer the inﬂuence of river
discharge on wetland surface water chemistry in 1999. In general, lower ion
concentrations at the Bay-l station indicated dilution of Saginaw River discharge, and the
concentrations of 804'2 and N03' measured at the Quanicassee River were similar to
those measured at Bay-1 and Bay-3 (Table 1.2). Concentrations observed at the mid bay
station (Bay-2) diverged from the surface water chemistry of the Quanicassee River,
indicating that discharge generally dispersed along the shoreline, rather than directly into
offshore areas (Table 1.2). Minimal dilution between Bay-3 and the study wetland was
noted; concentrations were similar.

The drainage ditch directly adjacent to the study wetland (south drain) was
sampled in an upstream-downstream fashion on three occasions in spring 2000. The
early sampling date (12 March) preceded the growth of emergent wetland vegetation.
However, concentrations for many of the surface water chemical parameters were similar

to later sampling dates (1 8 May and 12 June) and were included in the calculation of

46

 

 

$8 3. 3m : a: 3.8 Ba 3.8 mom :8 m. a A... 8 a: A.-.“ waa mi
38.8 23 88.8 23 $88 83 32.8 E: :28 3.2 32.8 $3 A.-.“ 2 may .52
cad 3.3 A88 8.8. 83.8 8.3 $08 8.2. 33.8 8.8 93.8 3.2 r: was ~-..Om

:8 EN 33 2 : G8 Sm 8.8 2.. 3.88 mg. :8 2:. 9-4 95 -5

N135 .m taﬁwmm m2»:— .8>_m .530 Tham— gmua—z 3.3m 88:83.—

 

 

.3558 20585: 85589 Bet @3058 883 mam 338mm 80 bEEozo 883 Suﬁsm 80 2388885 28303

2 8288 025 8:8 2: ~85 293:0 85.3 B882 33 ism ”82m 265me 05 8 wﬁzﬁwon 22888 336808
-8238 a E 8:88 28 9 688—8 338% 98 8288 wcszmm .32 5 83% 328mm 2: was 83% 838830
2: 6:383 .488 05 Am 23m:— oomv 2.288 econmto Eat 388:8 £88883 83:88 883 R588 :32 .N._ 035,—.

47

mean values due to limited sample size. This drain ceased to ﬂow following 12 June.
Speciﬁc conductance values from the south drain were the highest measurements made
for any of the Saginaw Bay surface waters characterized throughout the four-year
investigation (Table 1.3). Dissolved ion concentrations were generally very high for each
respective drain; TDI ranged from 8.75 to 24.87 meq L". In particular, these drains had
very high concentrations of T—alkalinity, cr, 80,", Ca", NO3', and NH4“; NO3' nearly
exceeded 70 mg N L'1 on 12 June 2000. In addition, Na+ concentrations were very high
in the canal. The overall importance of the agricultural drains as non-point sources to the
receiving waters of Saginaw Bay was unclear since discharge was not measured. The
drains were relatively small with considerable dispersion throughout the wetland,
particularly near the wetland interface. Yet, large differences in dissolved ion
concentrations were noted between the upstream and downstream stations of the south
drain (Table 1.3). In particular, significant declines in the concentrations of T-alkalinity
(220.0 vs. 112.0 mg CaCO3 L"), 0' (80.13 vs. 43.05 mg L"), 30.;2 (287.58 vs. 52.85 mg
L"), on“2 (173.77 vs. 48.83 mg L"), Mg+2 (44.08 vs. 17.63 mg L"), and N03’ (69.7] vs.
4.35 mg N L") were observed on 12 June (data not given in Table 1.3). Conversely,
NH; concentrations increased (0.080 vs. 0.220 mg N L'l) in the downstream direction
_ suggestive of sediment ﬂushing during periods of discharge.
Chemistry of Interstitial Water

Wetland interstitial water was very solute-rich relative to surface water; TDI
values ranged from 22.99 to 93.30 meq L'1 (Table 1.4). The range in speciﬁc

conductance along the mini-piezometer (m-p) network was 10450193 6.0 118 cm'1 with a

summer average of 1387.0 (j; 31.4) 1.18 cm". Total-alkalinity (bicarbonate only), Cl',

48

 

 

:5 mam-83 88.8 N: 8.22 64. 8 R;- ~_.-_d_8 a
so... Sod-Em 88.8 83. $3-3m 98.8 “86 Ans. .— 98 $8
2.: .— 98 may
2.: z ”.8 z-h
82 Sod-mad 38.8 82 Sod-Rod 38.8 358 r-.. z “.5 -.=z
8: 8.2-23. an: 2.2 8.8-8.8 a3: $8 p: z 95 .6:
I... $.83: :88 :2 8.88... 5.8 86 a: 98 b.
8.2 8.8-3.2 34.: 3.8 2.2-3.8 cm. 8 3.8 A...— 95 Hz
8.2 8.3-8.: 25.8 8.2 8.3-3.8 $8 :8 A .4 98 2 a:
42 : 8.88-8.3 $38 8.2 Rat-3m: 33:8 8.8. ~..-. ”.5 :5
2.2: Mada-82.8 838 8:“: 8.58-8.83 :38 3.8 n..— 88 "am
8.8. 8.3-2.2 83.: 8.2 8.8-3.8. $8.: _ .8 n: «.5 -6
c.3288: 888 38 8.8228 8m. _ 8 2:: AH. 805 98 .=<-:.
3:88.. in 8 BS c.3285.» 8.88 :2 : 9...: we .25 3%
8:80 owgm Emv :32 ownaz amv .892 3.2.3.3.—
an..8:3:: 5.8m 5828.5 5.5m

 

2.880508 88 £28: 5:888 323 3 88:38:98 .85 .Soom :onEoEom 2 5:028 .8883 05 mo 82 2:

50:98:: 883 0: 3:82:00 088 2:88 88— 2: mE3o=£ E8 Aooom 0:2. Q 28 .382 M: £982 m: £82880 8:: :o BEES
83 :86 :88 2:- .Soom :082 m: :2880 0:0 :0 829:8 3:0 83 8:8 @228— 526 2:- 68—83 358 2: 9 2:88.68
28:3 98 5:88 3:sz 03:86 03: 8 82:80:: 8338885 82:85 583 com-ta mo 098: 8:8 Ammﬁv :82 .m._ 038-.-

49

 

 

.28 2.: .28 8.8 .38 2... 2.8-22 .88 2.... ._.-. we: .0
:88 .3... 28.8 0...... .288 8.... 20.03.00 :88 0...... ._--. .. as: 2.0
._.-. ._ 0.5 .3
.228 2.2. .22.: 2.80 .22: 3.2.: 223.12.... .32.: 2.2: ._.-. z 0.5 +52
:2. 8 :22 88.8 2 a... .288 2.... 02.300 :88 w 2...: ._.-. z 0...: ..OZ
.28 2.. .28 2... .:..8 2... 3.2.8.. .2. 8 2... ._--. 95 +0.
.22: 8.2 .28 3.2 .28 .0: 8.812... .. ..8 8.2 .71. as: .2
.28: 2.2 .28 2.8 .28 2.2 82. 72.2 .28 2...: .3 0...: r as.
.22: 2.02 .28 2.2. .28 2.2. 2.22-8.2 .:..8 8.0.: ._--_ 0.5 =5
.82.: 2.28 .22.: :2 .28 2.2 2.332... .28 2.: ._-.: 95 0.00
.228 2.2. .38 2.2 .28 2.2 2.2.-2.2 .28 $2 ._.-. 0...: :0
.v. . 8 3.2 .28 .22 .38 .020 32-.....2 .02 8 2.22 ._.-. n00.6 0...: ...<-:.
. .28 :. .22 .28 :2: .88 3.2. c.5858... .v. .8 .22. .15 m... .250 02.0
am: .502 ﬁlm: .302 new: .302 025% am: 5:02 8:05.23
anon—m .302 Loan: Loan—O o—w—chm ==rﬂ I GGGN

 

 

280:8:

00m: 88:: 580808 30.0: 0: 8:030:00 49m. .2: 88.1on ”0:0:m :002: 8:2:03 0:0:m :00: 8:0 .AE cam-88. B055 8.8803
:05: A000.::0::_ 850803 0:: 80:: E 82-8 30:20: 8:2:03 :0::0 .m:0_w0: 00:8 0::_ 80828 020 0:03 850:: 5:509? .0808:
2:208: :0 888:: :0 n. é: cam-om: 8088:: 282.». 0:: w:0_0 :0802 :000 :0 80:82 2808002825.: ::0:.: ooom :0 A:0:E0:Q0m

0:3,: :0EE=m\m:::mm 0:: m::::8 80:00:00 20:08:80: 8:08:05 :0:03 8:880:85 :0: mowg: 8:0 Amm H: 8008 050:8 d.— 030..-

50

Ca”, and Mg+2 dominated the chemical composition of interstitial water in the primary
transect region (outer and inner wetland regions) of the wetland (Figure 1.2). High
concentrations of NI-L;+ were also measured and approached 4500 mg N L'l, yet NO3‘ was
generally low (<0.056 mg N L") or below detection limits (<0.030 mg N L"). The
concentration of 804'2 was also low on average, but ranged to nearly 450 mg L‘1 (Table
1.4).

Spatial trends in interstitial chemistry were evident along the study transect (Table
1.4). Mean speciﬁc conductance values were similar in the outer (1368.4 1 46.3 113 cm!)
and inner (1403.6 1 43.3 1.18 cm") regions of the wetland as was variability. In contrast,
speciﬁc conductance was much higher in the near shore region (2291.8 : 280.1 118 cm"),
whereas T-alkalinity was comparatively low relative to the outer and inner regions of the
wetland. Elevated SO4'2, Cl', and Ca+2 concentrations were the primary inﬂuence on
speciﬁc conductance measurements in the near shore region (Table 1.4). Concentrations
of Cl’, SO44, Mg”, and Na+ systematically increased shoreward from the wetland
interface. Conversely, K+ and NH4+ concentrations decreased toward shore. Dissolved
Si was on average higher in the inner and outer regions of the wetland. In contrast, NO3'
concentrations were generally below detection limits in these regions, yet much higher in
the near shore region of the wetland (Table 1.4). On 12 June, the concentration of NO3'
was highest (9.83 mg N L") at the 400-m station and much lower at the 340 and 460-m
stations (0.21 and 1.94 mg N L‘1 respectively). Near shore interstitial NO3'
concentrations remained elevated through 11 July after which concentrations were
generally below the detection limit. Similarly, N114+ concentrations at most stations

throughout the wetland declined through time.

51

Water table measurements collected from the m-p network in 2000 indicated
consistent positive vertical hydraulic gradients (V HG) within certain areas of the wetland
(Figure 1.23). In particular, the area 80-140 m and 240 m from the wetland interface
indicated localized groundwater seepage. In contrast, VHG was negative in the near
shore region of the wetland. The spatial trend was identical for piezometric head scaled
to wetland sediment elevation (Figure 1.24). Positive head differences ranged from 0.005
to 0.050 m; six stations were in equilibrium (: 0.010 m) with surface water levels
(stations 20, 120, 140, 180, 200, and 280 m from the wetland interface). Piezometric
head measurements were greatest in July and August; the most rapid decline occurred
between early June and late June in the near shore region of the wetland. Groundwater
seepage regions coincided with topographical breaks across the littoral slope. Surface
water chemical anomalies, such as high speciﬁc conductance and T-alkalinity, were
occasionally measured at these stations. The slope of the water table was approximately
10 cm per 100 m (R2=0.886, p<0.0001).

DISCUSSION

A variety of studies have demonstrated differences in surface water chemistry
between littoral and pelagic environments (Dvorak 1970; Planter 1970; Howard-Williams
and Lenton 1975; Kairesalo 1980; Klosowski 1992; Smiley and Tessier 1998). Previous
studies in exposed ﬂinging wetlands of Saginaw Bay, Lake Huron expanded upon
lake/wetland chemical differences and illustrated distinct trends in surface water
chemistry, whereby dissolved ion concentrations increased shoreward of the wetland
interface (Suzuki et al. 1995; Cardinale 1996; Cardinale et al. 1997). Similar patterns

have been noted in the more pristine coastal wetlands of northern Lake Huron (Burton

52

 

" 0.50
§% Shoreline:
I F
............ in”... . ..... . ... ... .................................. 0.00
8 ’ ’
\ D
\ O
\ --0.50 I
\ >
Wetland/Open Bay Water \ / \ .
Interface / \ L
r-l.00
”-1.50

 

 

0 40 80 120 160 200 240 280 320 360 400 440
DFWI (m)

Figure 1.23. Mean VHG (t SE) proﬁle across the study wetland in summer 2000.
Positive VHG values indicate upwelling or seepage zones and negative values indicate
downwelling or recharge zones. VHG is dimensionless; see methods for calculation.
DFWI corresponds to distance ﬁom wetland interface.

0.60
Shorelin

0.50

 

Sediment Elevation
- O- Piezometric Head

   
 
    

0.40

A : .' ,I ’ 3' I -/
» g- ' / 0.30

Wetland/Open Bay Water 0.20

Inte ace

0.10

Elevation or Piezometric Head (m)

* ' 1' Marsh Sediments 0.00

=12 . e. . . . . . . .4 - -0.10

0 40 80 120 160 200 240 280 320 360 400 440
DFWI(m)

Figure 1.24. Average (1- SE) position of the water table (piezometric head) within the
study wetland in summer 2000. Piezometric head measurements were relative to surface
water levels and re-scaled to sediment elevation (see methods). Values above sediment
surface indicate localized seepage areas; the converse is true of values below the
sediment surface. Sediment elevations were referenced to the wetland interface (0—m
station). DFWI corresponds to distance ﬁom wetland interface in meters.

 

53

unpublished). Increasing dissolved ion concentrations are thought to emanate from the
reduction in lake/wetland mixing (Suzuki et al. 1995; Cardinale 1996; Cardinale et al.
1997). Emergent vegetation provides ﬁictional resistance sufﬁcient to reduce surface
wave energy across ﬂinging wetlands. Cardinale et al. (1997) provided evidence that
prior to Scirpus stems emerging from the water surface, chemical differences across the
wetland were minor. In contrast, surface water chemical patterns were very distinct
during periods corresponding to peak plant biomass. Further, Suzuki et al. (1995)
demonstrated that spatial trends in dissolved ions did not disappear even during moderate
storm surges. However, the strength and frequency of chemical gradients noted in both
of these studies were not entirely predictable or consistent throughout the summers of
data collection.

The previous work conducted in Saginaw Bay wetlands coincided with near
average lake levels relative to the long-term (48-year) record (Figure 1.25). During this
investigation, the changes in surface water chemistry that occurred in conjunction with an
approximate 1.04 m decline in lake level suggested that lake level strongly inﬂuenced
surface water quality in fringing wetlands. For example, spatial trends in dissolved ions
across the study wetland were minor in 1997, a year that corresponded with the highest
lake stage since 1986 (Figure 1.25). Surface wave energy is related to fetch, wind
velocity, and depth (Wetzel 2001). During the high water years (1997-1998), surface
wave energy was much greater compared to low water years (1999-2000). The net effect
of higher surface wave energy resulted in a more chemically homogenous littoral
environment. In contrast, wetland surface waters became more insolated from the lake

during low water years since wave energy was not sufﬁcient to penetrate great distances

54

.8239 8x3 820 3:23.525 2 meson—8:8 n50— .8oom-3m:

33m $5 go 828 05 8:865 x388 Sam amﬁmmm 5:5 23 85: 22:3 @80st8 Am 6o3 .3 B mam—«E £05m 3:25 583
Em Ea .S 32 .3 e :35 .8 82 .3 e 2265 E 3828 36% 2 283:8 3332 .9. 36 H one; :3
ESﬁwco— 3:32ro 0:: 2.0m .32 .o=_>xommm ”SONAR: :85: 8:5 55 Bacmwmm Sm 288 omﬁm .89» “amalgam .mm._ 05m:—

ccca mag :22 mag .53 man :53 meg .53 mma—

_ P _ _ — k _ — p _

. J Wm:
4 q d u i- 4 1 u u q q u u u q a q q q a G q q q + u u u q d H d d

+ q d d u
r >54 ...—33—

 

 

 

.............................................

 

(5861 (1191 in!) was

. .3: 283.
u 3:

 

 

 

p F i? - p - bir h b h n n h p b P F L p F p r PL b h u n — p p b b P b P p p b cowb—

55

into the wetland. In fact, the variation in speciﬁc conductance measurements collected
along the study transect increased as lake stage declined. For example, the standard

deviation of speciﬁc conductance measurements increased from 11.0 uS cm’1 in 1997 to

81.8 uS cm'1 in 2000. A similar trend occurred for conservative and semi-conservative
ions; 1.6 to 3.4 mg L'1 for or between 1997 and 2000 and 0.68 to 2.04 mg L" for Mg+2
between 1998 and 2000.

Alabaster substrata experiments supported the contention that the penetration of
surface waves into the study wetland declined between 1999 and 2000 (Figure 1.16).
Annual trends in turbidity offered further support. Wetland surface water was more
turbid in 1998 (17.8:33 NTU) compared to 1999 (4511.1 NTU), suggestive of a
reduction in sediment re-suspension commonly associated with periods of lower wave
energy (Sager 1996; Hamilton and Mitchell 1997). Suzuki et al. (1995) noted turbidity
values up to 50 NTU, whereas Cardinale (1996) observed a range of 3-39 NTU,
representing additional observations coincident with lake levels higher than 1999.

Cumulative ion plots were useful for illustrating the gross changes in wetland
surface water chemistry in relation to declines in lake level (Figure 1.26). Spatial trends
were magniﬁed during the years corresponding to below average lake levels. However,
these data also illustrated the overall increase in dissolved ion content as lake levels
declined to approximately 0.56 m below average. The concentrations of all the major
anions and cations signiﬁcantly increased in 1999 and 2000 relative to the respective
previous year (Figure 1.11 and 1.13); this inter-annual trend was also evident for speciﬁc
conductance measurements. Additionally, shifts among dominant anions and cations

(proportion of total) occurred in conjunction with increased concentrations (Figure 1.27).

56

.382: E 02%an was—$3 Eot 8:8va 9 accommohg SEQ .3233 E own can 6m: .8 .o .5253 cBﬂoEBE 22>
3235:0800 ~.eOm 62me some 3 coca? ﬂ gown: 5588 < ﬁv— uza +52 com S can .3 .hng 3088 5038252 88433
macaw 22 EW Sm mm can do .mmTZ mA"...Om now EHZ can -5 use 6:9“. Sm coma .33 E menace; 225 .825 3:0 .8);
88-32 mo eoEEstEam 2.: MES“. “628:8 838mm 533 conga van—~03 .«o A74 3:: 32m :2 033—583 S32 .3; 23m:—

35 :25 :5 :23
can :3 can a: an. 3 3 a can 3." can 2: ca— ca 3 e

 

 

 

«wow I .
-.U I
57,—. D

+v— + +aZ I 6.:
«+32 mild . a...
«+5 I . . .. , . . _ : ....n

. :6—

. , c."—

 

 

 

 

 

 

 

 

 

 

 

3 we... 2 Se

. . n u n r

 

. .

 

 

9.
V
v-

57

Proportion of Major Anions

 

 

 

 

 

 

 

 

1.0
2 0.3 —
.2
5 - Ca“
.i M , 1: Mg“
§ - Nai
a [:1 K‘
¢
.5 0.4 -
I.“
G
D-
O
S: 0.2 —
ND
0.0 .
1997 1998 1999 2000

Figure 1.27. Mean annual proportions of major anions (A) and cations (B) of wetland
surface water samples collected during the period of study (1997-2000). Numbers
correspond to the proportion of total anions and cations accounted for by C1‘ and Mg+2
respectively. Proportions were determined based on ion concentrations in meq L".
Cations were not quantiﬁed in 1997 (ND represents no data). Potassium accounted for
less than 1.5%.

58

 

In particular, the proportion of total anions accounted for by C1' decreased between 1998
and 2000 while 804'2 increased. Whereas, the proportion of total cations accounted for
by Mg+2 increased consistently across years. Both trends suggested greater hydrologic
contributions by a high 804’2 and Mg+2 end member to overall wetland hydrology as lake
levels declined.

Surface water chemistry data collected from several offshore stations (1999) and
ﬂom the Quanicassee River (1999-2000) and the Saginaw River (1999) illustrated the
potential importance of water level ﬂuctuations on controlling the impact of river
discharge on wetland chemistry (Table 1.2). First, several conservative (Cl' and Na+) and
non-conservative (804’2 and N03) solutes supported the generalization that the
circulation pattern of Saginaw Bay is counter-clockwise (Budd et al. 1998). Second,
these same data illustrated that river discharge dispersed laterally and appeared to follow
the same long-shore currents that govern the circulation of the bay. This point was
important because it suggested that river discharge provides a constant ﬂux of non-point
source pollutants to ﬂinging wetlands which likely inﬂuences surface water chemistry.
This may be particularly true during years corresponding to above average lake levels
since high wave energy would increase lateral dispersion into wetlands. However,
wetland surface water chemical data and the AS experiments illustrated that low water
levels and reduced wave energy have an insolating effect on ﬂinging wetlands. Further,
river discharge is controlled by local meteorological conditions within the drainage basin.
During periods corresponding to drought (1999-2001; Beeton 2002) and below average
lake levels, river ﬂow may be greatly reduced and the contributions to coastal wetlands

much less (Figure 1.28). Therefore, it seems highly unlikely that the changes in surface

59

Avis—8% 8:

See 803 88.3% $8» 2: «EB Sega—8:8 295% «o eaten 89872; < dim—8% 20>» Saw 0: 3:365 £923.63
5 x35 BE. .8933 >229: ms 339—898 98 @6052: 0% 68523». 5:83 :22on 552:8:— Bocgwobm 3:232
mOmD 2: 88m 3:850 225 8% 26: 885 ESQ .=2 .Baﬁwmm 3 82% 265mg 2: .«o omen—Lemme 35:08 S82 .34 2:3“—

 

 

 

 

 

cccm aaa— maa— baa— waam

SfWWfNSfWWFNSfWWPNSPWWfNSfWWf

a n a n a n a n a n

whuummwﬁnmmmmnnummwﬂnuumMMauu

_—+——r——-—-—L——————|PPbLt—|Pb+bbPL—p——pn——r—b—-——L—Pb——rpb Aw

I I2:

1 lac"

. neenm
mt
s
a

1 [av 3....
(

1 12am

1 [so

.tk.r»._.__h:....:..u._____L...___L_.___..L.p:.t:.p 2:.

 

 

60

water chemistry that occurred in the study wetland over the four-year period of study
were due exclusively to riverine inputs.

The ionic composition of the agricultural drains was much more concentrated
than any of the other surface waters sampled throughout the investigation. In particular,
concentrations of S04'2 and N03' were much higher compared to the Quanicassee or
Saginaw Rivers. Lateral dispersion of drain discharge appeared to occur in the wetland,
yet a well-deﬁned outlet was evident at the wetland interface. Contrasts between
upstream and downstream ion concentrations partially supported the observation that the
drain bisected the wetland similar to a stream (Table 1.3). However, declines in solute
concentrations were not solely due to dilution or biogeochemical cycling in the wetland
because NHX concentrations increased (0.080 vs. 0.220 mg L'l) as soils were ﬂushed by
drain ﬂow. Drain discharge likely had a major impact on wetland surface water
chemistry in early spring, but this was prior to most of the sampling; ﬂow ceased
following 12 June.

Lake stage was low enough in 2000 that mean water depth within the wetland was
0.09 m, representing a 0.97 m decline relative to 1997 (Figure 1.6). As water depth
decreased consistently across years, the interaction between surface water and wetland
sediment likely increased due mainly to the reduction in water column volume. In fact,
anomalously high surface water chemical measurements were often noted within speciﬁc
areas of the study wetland. In particular, high speciﬁc conductance values and elevated
T-alkalinity, Can, and Mg+2 concentrations were measured at the 120- and 260-m
stations in 2000 (Figure 1.15, 1.17, and 1.18). These patterns supported tighter coupling

between surface water and interstitial water within these areas, since the TDI content of

61

interstitial water was high relative to surface water. Piezometric head and VHG
measurements provided additional evidence for stronger coupling in 2000 (Figure 1.22
and 1.23). Consistent positive head and VHG measurements were noted at 80-140, 180,
and 240-m from the wetland interface. However, these potential seepage areas were not
entirely consistent with the stations in which high T—alkalinity, Ca”, and Mg+2
concentrations were measured.

Interstitial water had the highest TDI content of all the water masses sampled
throughout the study. Spatial trends were less obvious for interstitial water, particularly
in the outer and inner regions of the wetland. In contrast, concentrations of Cl’, SO4'2,
NO3', Ca+2, and Mg+2 were very high in the near shore region of the wetland. This
pattern suggested that agriculture, the dominant land use adjacent to the study area, was
one potential source of these dissolved solutes (Bohlke 2002). Piezometric head
measurement in conjunction with the littoral slope (10 cm per 100 m) suggested a strong
water table gradient spanning the entire wetland (Figure 1.23). Further, elevated wetland
surface water SO4'2, Mg”, and Si concentrations in 2000 were consistent with interstitial
water as the main source since drain ﬂow ceased by late spring. Therefore, littoral
geomorphology coupled with below average lake levels increased the strength of
interaction between surface water and interstitial water (shallow groundwater) and
provided the best explanation for the increasing concentrations of major ions throughout
the four-year investigation.

Short-term water level ﬂuctuations resulting ﬂom seiches likely constituted an
important mechanism by which exchange between interstitial and surface water occurred.

For example, I observed the study wetland to completely dewater over an approximate

62

30-minute period on 12 September 2000, the direct result of a moderate seiche. The
magnitude of piezometric head and VHG should have increased instantaneously and
interstitial water would have mixed with any remaining wetland surface water (or
discharged from the wetland). Though short-term water level ﬂuctuations were not
measured, changes on the order of 10-20 cm are common to the Great Lakes (Gathman et
a1. 1999). Measurements made by Suzuki et al. (1995) indicated that the study wetland
was subject to two cycles of water level change. The ﬁrst was relatively large and had a
period of 24 hr; the second was a much smaller oscillation resulting ﬂom transverse
seiches with a period of approximately three hours. Brady (1992) observed water levels
in the study wetland decline to approximately 3-5 cm in response to a seiche during 1990,
a year corresponding with lake stage intermediate between 1998 and 1999. Additionally,
storm surges can result in short-term water level ﬂuctuations in excess of 1 m. Therefore,
it seems likely that the importance of seiche activity and storm surges on wetland surface
water chemistry increases during periods corresponding to below average lake levels.
Several biogeochemical observations also provided support for stronger
interactions between surface water and interstitial water in 2000. For example, water
table draw down generally coincides with a switch ﬂom carbon cycling processes to
mineralization (Freeman et al. 1997). In particular, drought conditions have been noted
to promote the release of inorganic nutrients (Mg+2 and Ca”) from wetland soils
(Heathwaite 1990; Freeman et al. 1993b; Freeman et al. 1997). Higher solute
concentrations in 2000 offer support for results from those studies. Additionally, sulfate
reduction may have decreased or shut down as sediments in the near shore region of the

wetland were subjected to oxidizing conditions. This would have contributed to

63

consistent measurement of high SO4'2 concentrations within this region of the wetland;
high N03“ concentrations offered further support for this contention. Alternatively, re-
oxidation of sulﬁdes is thought to occur following water table draw down (Ogden 1982;
Freeman et al. 1993b; Freeman et al. 1997). Sulﬁde oxidation in the near shore region of
the study wetland may have partially contributed to the high S04'2 concentrations. In
fact, SO4'2 concentrations measured in the near shore region increased between late
spring and summer. In contrast, consistent declines in surface water 804'2 concentrations
occurred throughout the outer and inner region of the wetland throughout spring and
summer of 2000, suggestive of increased rates of sulfate reduction associated with
surﬁcial sediments. This was not observed in previous years because concentrations
were highly consistent through time, which is often indicative of a constant turnover rate
(King and Klug 1982).

Higher SRP concentrations in 2000 supported stronger interactions between
surface water and sediments. The literature provides a variety of examples illustrating
diffusive ﬂuxes from the sediments into the water column (Carlton and Wetzel 1988;
Baldwin er al. 2000; Gardner et al. 2002) and convective circulation (James and Barko
1991 a, 1991b). In particular, drying of sediments from water table draw down often
increases the release of P (Baldwin et al. 2000; Watts 2000).

Interest in the water quality of the Great Lakes has traditionally been focused on
offshore waters. In contrast, few data on surface water chemistry of coastal wetlands
exist. The data that are available suggest that water quality in ﬂinging wetlands may
differ ﬂom offshore water (Suzuki et a1. 1985; Cardinale et al. 1997). However, speciﬁc

conductance measurements collected ﬂom various offshore stations in inner Saginaw Bay

were similar to measurements collected ﬂom the study marsh in 1997 and 1998 (Figure
1.29). In contrast, speciﬁc conductance values associated with outer Saginaw Bay were
on average 100 its cm'l lower. Temporal variability in surface water chemistry was high
at many of the inner bay stations (Nalepa et al. 1996). For example, chloride
concentrations ranged from 14.4 to 53.9 mg L'1 at station 2 and from 14.8 to 43.5 mg L'l
at station 1 between May and September of 1992 (Figure 1.29). The broad range in Cl'
concentrations noted for offshore stations in Saginaw Bay suggests that episodic
pollution events may be common features of the seasonal cycle of water quality in the
bay. The impact of water level ﬂuctuations on offshore surface water quality has not
been studied and the data collected by Nalepa et al. (1996) coincided with water levels at
or near the long-term average. In fact, most of the data available on Saginaw Bay water
quality coincides with near average water levels (Fahnenstiel et al. 1995a, Johengen et al.
1995, Nalepa et al. 1996).

Although it remains unclear if offshore water quality is related to lake level, the
shifts in wetland hydrology documented in 2000 suggest that ﬂinging wetlands may
function as nutrient and solute sources during periods of below average water levels.
Low lake levels and an exaggerated littoral slope may allow agricultural drains, which are
common throughout the Saginaw Bay watershed, to bisect ﬂinging wetlands and ﬂush
soils along the way. The cumulative effect of such a ﬂmctional shift would be highly
related to sediment storage time (i.e. number of years as a sink), which is a function of
antecedent conditions (i.e. number of years at or above average water levels). In
particular, the latter point calls attention to future water level predictions for the Great

Lakes. The period corresponding to this study (1997-2000) represented the second

65

.3209 382:5 880:8: mo 89:5: 85 5:5 8:88 29:8 8.8 x0: :28 025: 83:52 8.85:: “88:8: 88:8 :5 £23808:
Sea :8 £2 9 20:38:09 88823 688: 85 55:5 383:5 8.8 88:02: .83808: 5mm :5 5mm 2:828: moxom A92:

:2 ... 3 38:8: mooo~$ma$ .258 $5 :2 “880:8 3:888:82: :8 63— .N: 8 32:2 ”2 :8 m; «cm: $-32 :2 A885
9:: 09¢ 8m Rasmam mo 28:3 to: «883 293:: 80¢ 38:85:88 5.80 m3 8:826:00 058% .20 $2: xom .82 8:32

:3

2:.

25

can

can—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:252 $55 2855 aheaéO

gen 33 :3 32 2 «SH n 55 N <mm _ 4d”..—

Qmﬂﬁ §Q.ﬂ= WNW": §©WE “at ﬁt but
. a + l I». .
.4 .m. ., . 6 III I E .
. .. ... .. . «Jen m.~an F. . mdbn .

o .
. . . a a:
e

. n49 a can T

 

+

«dam

 

 

n.
.8: ..w
iguam .
3 58::

$5 335%....”
8:5

5:-

 

 

I

 

I

can

39
S
d
3
0

8e 3
0
u
m.
)

can we
3
m.
(u.

2::

can.

66

greatest rate of change in the last 100 years and the fastest rate of change in the last 50
years of record (Figure 1.30). The water level during 1997 was the 8th highest stage in
long-term record (31 cm below 1986), whereas 2000 corresponded to the 13th lowest
stage (31 cm higher than 1964). Water level ﬂuctuations are primarily a reﬂection of
antecedent climatic conditions within the Great Lakes basin (Larson and Schaetzl 2001).
In particular, decreased precipitation and increased evaporation resulted in a rapid drop of
lake levels between 1998 and 2000 (Beeton 2002). Additionally, several of the hottest
and driest summers on record have occurred in the last few years. Elevated summer
temperatures, in conjunction with below average precipitation, mild winters (winter of
2001-2 was the warmest on record), and reduced ice cover have substantially affected the
water balance of the upper Great Lakes (Beeton 2002). Climate change predictions for
the Great Lakes suggest that the frequency and magnitude of below average lake levels
can be expected for the future (Kling et al. 2003). Understanding how fringing wetlands
interact with large lake ecosystems is critical to anticipating future water quality issues

associated with areas of concern, such as Saginaw Bay.

67

6mm— 820Q 000—0.— :0080 _0:0¢0::0::_ 0: 00:88:00 n50: .Aooom-3$v :03: 0:0 ¢0 80:0 05 00:00:08

03880 05. .8 8.0 H 3.02 Amm H: 0m08>0 :00»-No_ 0:: 800080: 00:: 880000: 00:00 0:0 0:00 0:... .80 0 :05 0.00— 0:03
0:000: ¢0 0:00» 8030:: 0:: 80¢ 0:000:80 80008 :2 .0=_>x0mmm 0:: 80¢ 0:03 Ncoméoom :8 0:09 6:00:88 000 ..quov 298—
8:03 $058 :008 80¢ 00380 0:03 ham 3 m0:_0> _0:::< 80:03 00—04 8.: Amooméooc 0:000: 0w08 8:0:-m:0._ .omA 0:83

2.3 :03 33 2.3 :03 $3 3&— cm3 33 33 aca—

- — p - p — b _ F -

O
J q 4. I I I I I I I I I I I I I I I I I I I I I I I I I I I q I J I I I I I I I Q“ wh—

-

 

. 30. 0:000: .

.56:

3.3.—

 

 

 

 

($861 (1191 gm) aﬁms

 

8.:—

 

._u=_ 0:000:

 

 

 

D D b D IL - D D D D - D D D L - P D D P h D D D D b D D D D P D D D P l? D D D D D P I D D —l D D D P - cm.hb_

68

Chapter 2

THE INFLUENCE OF WATER LEVEL FLUCTUATIONS ON INVERTEBRATE
COMMUNITY COMPOSITION IN A GREAT LAKES COASTAL WETLAND

INTRODUCTION
Exposed ﬂinging wetlands of the Great Lakes lie at the interface between

terrestrial and aquatic ecosystems (Wetzel 1992). The interaction between adjacent
ecosystems, in‘conjunction with lakeshore geomorphology (i.e. depth, slope, sediment
type) and physical structure (i.e. macrophytes) creates high quality habitat that is quite
different from the limnetic environment (Cook and Johnson 1974; Dvorak 1987 ; Krieger
1992; Palomaki and Hellsten 1996; Diehl and Komijow 1998; Gathman et al. 1999;
Wetzel 2001). Hydrology is perhaps the most unique feature to these systems relative to
other classes of freshwater wetlands. The dissipation of surface wave energy results in
heterogeneous habitat that grades from an often turbulent outer region of the wetland to a
more quiescent near shore region. Sediment characteristics, water clarity,
hydrochemistry, and plant species shift in conjunction with reduced physical exchange,
and contribute additional habitat complexity (Suzuki et al. 1995; Cardinale 1996; Burton
er al. 2002).

A variety of studies have described the invertebrate communities of Great Lakes
coastal wetlands (see review by Gathman et a1. 1999) and several have documented
distinct invertebrate assemblages along the habitat continuum (Cardinale et al. 1997;
Cardinale et al. 1998; Stricker et al. 2001; Burton et al. 2002). Distributions are thought
to emanate from resource gradients that follow the hydrologic gradient. In particular,
epiphytic algal biomass and suspended particulates have been correlated with the

reduction in lake-wetland exchange, and these patterns are in good agreement with the

69

spatial distributions of a variety of invertebrate taxa (Suzuki et al. 1995; Cardinale 1996).
Burton et al. (2002) hypothesized that diversity was highest at intermediate levels of
exposure to surface wave energy. However, much of the empirical data that has gone
into the formulation of this hypothesis was collected during above- or near average lake
levels, periods likely corresponding to higher lake/wetland exchange. Data on
invertebrate community composition during periods of below average lake levels are
lacking.

A variety of water level ﬂuctuations are common to the Great Lakes. Over the
long-term, lake levels can ﬂuctuate on the order of 1.5 m (Burton 1984; Gathman et al.
1999). Short-term ﬂuctuations, such as seiches, are more regular in amplitude and
periodicity, typically resulting in water level oscillations between 10-20 cm (Bedford
1992; Gathman et al. 1999). The effect of ﬂuctuations of that magnitude can be
exaggerated during periods corresponding to low lake levels, resulting in periodic
dewatering of coastal wetlands (Krieger 1992). The importance of hydrologic
disturbance in lotic ecosystems has been well studied (Peckarsky 1983; Allan 1995), but
lentic systems have received far less attention (but see Wellbom et al. 1996). The
importance of long-term water level ﬂuctuations on coastal wetland invertebrate
community composition has not been studied and the impact of hydrologic disturbance
on biotic interactions remains unknown. Climate change predictions for the Great Lakes
suggest an increase in the frequency and duration of below average lake levels (Kling et
al. 2003). Understanding the effects of long-term water level ﬂuctuations on coastal

wetland invertebrate communities is critical for anticipating future changes.

70

The goal of this study was to semi-quantitatively describe the invertebrate
community of a Saginaw Bay, Lake Huron coastal wetland relative to a 0.730 m decline
in lake level over a three-year period (1998-2000). First, I describe invertebrate
community composition averaged over the spring and summer for two years to broaden
the understanding of Great Lakes coastal wetland ecology in the context of natural water
level ﬂuctuations. Speciﬁcally, I consider the study wetland as a single unit to facilitate a
concise and detailed description of the invertebrate community. Maj or shifts in
community composition within the context of lake level ﬂuctuations are illustrated by
comparing invertebrate assemblages during high (1998) and low (2000) water years.
Second, I evaluate spatial distributions of select invertebrate taxa shoreward of the
wetland/lake interface to illustrate potential dependencies between coastal wetland fauna
and wetland/lake hydrologic interactions. Lastly, spatial and temporal patterns are
evaluated in the context of hydrologic changes and several abiotic and biotic mechanisms
potentially inﬂuencing community composition are explored.

METHODS
Study Wetland

The wetland complex selected for study was located along the southeastern shore
of Saginaw Bay, Lake Huron, U.S.A. (Figure 2.1). The region of the wetland studied
encompassed only a small area of the total complex that extended approximately from the
Quanicassee River (Tuscola County) to the Sand Point/Wildfowl Bay area (Huron
County). The study area was located adjacent to Vanderbilt Park, Tuscola County,
Michigan (43° 37’N 83° 38’W). Predominant winds were out of the northwest, and the
wetland was unprotected from wind and wave exposure with a maximum fetch of 30 or

more km (Suzuki et al. 1995).

71

 

 

 

inner R

Saginaw Bay b

Saginaw R.

r“:

A

 

\ a ' l

tudy Wetland

Quanicassee R.

 

 

 

 

mammals
0 5 10 20 3O 40

Figure 2.1. Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The Quanicassee and Saginaw Rivers are referenced.

72

A transect spanning the wetland was established in spring 1997 and was used
during ensuing ﬁeld seasons (1998-2000). The transect extended 280 m perpendicular to
the shoreline, extending shoreward (S 172°) from the outer edge of the emergent plant

zone/open bay water interface hereafter referred to as the wetland interface (N 43 °36’4.4”

W 83°39’41.3”). Sampling stations were established at 20-m intervals along the study
transect (Figure 2.2). The wetland was arbitrarily sub-divided into three regions: outer
(20-140 m from the wetland interface), inner (160-280 m), and near shore (>280 m). The
emergent plant community of the outer and inner wetland regions consisted of a nearly
mono-dominant stand of three-square bulrush (Scimus pungens Vahl). Less dominant
species of bulrush, S. acutus Muhl. and S. validus Vahl, were interspersed among S.
pungens, primarily within the inner (160-280 In shoreward of the wetland interface) and
near shore (>280 m from the wetland interface) regions of the wetland. A large cattail
(Typhd angustifolia L.) complex bordered the northern edge of the study area
approximately midway between the wetland interface and shore (100-300 m). At the
onset of this study, vegetation in the near shore region of the wetland included S.
pungens, arrowhead (Sagittaria sp.), and a variety of submergent species (see Batterson
et al. 1991). As lake stage declined, the near shore region became dominated by species
typical of wet meadow/strand communities, such as sedges (Carex spp.), smartweeds
(Polygonum spp.), and cotton wood seedlings (Populus deltoides Rydb.). Several recent
studies have been conducted in this particular wetland and have provided background
information on solute chemistry, periphyton, vascular plants, and invertebrates (Brady et
al. 1995; Brady and Burton 1995; Suzuki et al. 1995; Cardinale et al. 1997; Cardinale et

al. 1998; Burton et al. 1999; Vaara 2001; Stricker et al. 2001).

73

.0030: 20:0 :00: 0:: 82.: 00:00:00 0:03 00:00:00 0:0::0::0>:: 0: ”02:0:w

0:: :: 00:0: 0:0 0:0—:03 0:: :0 0:0:w0: 0:0:mm000 0: 000: 0:0_::0>:00 w::E0Z .000::0::: 0:0003 0:: 88.: E com 330508500
02000: 003 0:20:05 0: H .000.::0::: 0:0—:03 0:: :0 :0000 0:00.30 :0 :0 00:80:00 003 w:::::00 0:0::0::0>:: .2: cam:

800:0: 0:: 0:20 0023:: E on :0 080:0 0:0 0:0 0:0:08 :_ A:0_:0:0 8-8 000.208: :0:03 .00: 5:00:01: ::0w:0::0 0:: 80¢ 00:0:00
0: 02:20: 0:0 003:0: 8020:: 30:0 0:: w:0_0 00:0:00 0:00:08: 5:05 0:203 300:0 m:_::0q0 68:00 :0 0:080:0m .~.~ 083...:

OEm—OhOn—m \\ f
:

  : : :
Pfﬁf P505002

 

2.03050

:Mug :0:=0

 

 

   

530: 0.8:... :00: a: emmA F >

 

”mu-Om 0M— maﬁa—«OB

 

ooﬂuhoz: EOIQB

 

  

>0: 5900:0303

.0._0._.H.._...0..___..0._0 :0»: :1
p

74

Habitat Measurements
Physical habitat measurements along the study transect included depth,

temperature (YSI Model 30), and stem density (no. stems m'z). Turbidity was determined
on unﬁltered surface water. samples (1998-1999) after equilibration to room temperature
(Hach model 2100A). Dissolved oxygen (DO; YSI Model 51B) was measured (between
10 am and 12 noon) at each station throughout the spring and summer of 1997-1999.
Very low water levels in 2000 prohibited accurate measurements. Concentrations of DO
(mg L") were converted to percent saturation based on theoretical temperature dependent
percent saturation values. Speciﬁc conductance (25° C; YSI Model 30) data collected to
characterize surface water chemistry in the study marsh (see chapter 1) were included to
illustrate water quality changes relative to lake level declines. Physical data available
from the spring and summer of 1997 were also included to better resolve the role of water
level ﬂuctuations on coastal wetland habitat. The elevation of each station was
determined in September 2000 using a laser level. The stage of Lake Huron at the time
of elevation measurements was used as a benchmark (N OAA real-time water level data;
http://co-ops.nos.noaa.gov/data_res.html; Essexville, MI; station #9075035).
Hydrologic Connectivity between Lake and Wetland

A relative measure of hydrologic connectivity (or exposure) between lake and
wetland was assessed during 1999 and 2000 by measuring mass loss of an alabaster
substratum (AS) at each station along the study transect (Doty 1971; Muus 1968).
Alabaster substrata were constructed of commercially available Plaster of Paris (CaSO4)
mixed to the manufacturer’s speciﬁcations, and molded into disposable plastic cups (90
mL total volume) with 1 cm diameter wooden dowels east through the center. The

substrata were allowed to air dry until masses stabilized, deployed within the marsh

75

(dowel driven into sediments), and retrieved after approximately one week. In the
laboratory, substrata were again air dried until masses stabilized. The extent of
dissolution, which was assumed to be proportional to turbulence (Muus 1968; Doty
1971), was calculated by subtracting initial mass from ﬁnal mass.
Ancillary Data

Lake Huron stage data (International Great Lakes Datum 1985) were acquired for
the nearest gauging station (Essexville, MI) from the National Oceanic and Atmospheric
Administration (NOAA; real-time water level data, http://co-ops.nos.noaa.gov/
data_res.htrnl; station #907503 5). The historical record for this gauging station extended
back to 1953, and therefore provided a 48-year chronology of lake levels within Saginaw
Bay. Longer historical records were available for the Great Lakes, but this station was
most relevant to the study area.
Invertebrates

Invertebrates were semi-quantitatively sampled along the study transect on nine
occasions from late spring through late summer (June-September) in 1998 and monthly
(J une-September) in 2000 (Table 2.1). The community was also characterized on one
occasion in mid-summer 1999 (August). Sampling intensity was greater in 1998 in an
effort to capture as much temporal variability as possible. In 2000, sampling intensity
was reduced to monthly intervals; 1999 served as a transitional year between above
average and below average water levels. A semi-quantitative approach was adopted to
facilitate collection of many samples, limit sorting time needed for sample processing,
and also make it possible to integrate three wetland habitats (epifauna, benthos, water

column). Standardized sweeps using a D-frame aquatic dip net (0.5 mm mesh net)

76

Table 2.1. List of dates invertebrates were collected ﬁ'om the study marsh throughout the
spring and summer (June-September) of 1998-2000. Sampling methods were identical
for each collection date. Sample dates were averaged by month to facilitate a concise
analysis (see methods).

 

 

1998 1999 2000
12 June
25 June 20 June
1 July
8 July 4 July
16 July
28 July
19 August 4 August
30 August 28 August
13 September 12 September

 

77

ensured consistent and similar sample sizes among stations (Vince et al. 1981; Burton et
al. 1999; Heino 2000; Burton et al. 2002). At each station, the dip net was swept across
plant stems, through the water column, and along the sediment surface over a linear
distance of approximately 2 m parallel to the shoreline. This approach resulted in the
collection of one integrated sample from each of the 14 sampling points along the study
transect (20—280 m stations; Figure 2.2) on each of the 14 dates sampled (Table 2.1). The
wetland interface (O-m station) was not sampled. Samples were transferred to containers
and ﬁlled with wetland water from each respective station and loosely capped. Samples
collected during the ﬁrst ﬁve dates of 1998 were immediately sorted upon return to the
laboratory. During ensuing trips, samples were immediately preserved in 95% ethanol
aﬁer returning from the ﬁeld.

Samples (unpreserved or preserved) were transferred to a white enamel pan and
invertebrates were sorted from the plant litter and sediments. Samples were picked in
entirety, and an effort was made to obtain at least 100 individuals. However, several
samples collected during 2000 contained a very high number of snails. In these
circumstances a sub-sample was picked in entirety and the remaining contents of the
sample were scanned to ensure a representative sample. Invertebrates were enumerated
and identiﬁed using a variety of taxonomic keys (Wiggins 1978; Burch 1980, Burch
1982; Thorp and Covich 1991; Menitt and Cummins 1996). Cladocerans accounted for
the only zooplankton observed during this study and were relatively rare.

Raw abundance data were converted to percent abundance (taxon abundance
divided by the total number of invertebrates sample'l) and analyzed in this format for

several reasons. First, samples were not collected in a quantitative fashion (i.e. m'z) and

78

abundance would therefore provide little ecological information. Second, total
abundance sample'1 was generally constrained to approximately 100 individuals; mean
number of individuals sample'1 was 87.2 (1- 2.3), 121.4 (1- 9.0), and 123.5 (_-+_- 4.8) during
1998-2000 respectively. Third, analysis of percent abundance data provided additional
information on each taxon within the context of the entire community. Fourth, this
format provided a means of standardizing samples and eliminating bias associated with
marked abundance differences between stations.

Temporal and spatial changes in invertebrate community composition were
evaluated in terms of percent abundance, taxa richness, Shannon’s diversity index, and
feeding guild structure. Invertebrates were arbitrarily classiﬁed into three groups:
abundant (>5% percent abundance), common (>2%, <5%) and uncommon (<2%). Taxa
richness (S) was calculated as the sum of operational taxonomic units sample".
Shannon’s diversity (H ’) index was calculated as follows:

H ' = i pilogropi

where, pi is the proportion of individuals found in the ith species.
A more diverse community is suggested as H’ increases. Evenness (J ') or equitability
was used in conjunction with Shannon’s index and was calculated as follows:

J’ = H’*logloS'

where, S is the total number of taxa in the sample.

Feeding guilds were based on functional groups proposed by Merritt and

Cummins (1996). Evaluation of taxa collected throughout the three-year study indicated

that the functional groups were restricted to collector-gatherers, collector-ﬁlterers,

79

shredders, scrapers, piercers, and predators. To facilitate a concise analysis of feeding
guild structure, the piercers were combined with scrapers and collector-gatherers and
collector-ﬁlterers were condensed into a generalized collector category (Table 2.2).
Temporal Patterns

Lake level ﬂuctuations were hypothesized to be an important abiotic mechanism
inﬂuencing coastal wetland invertebrate community structure. Treating the wetland as a
single unit provided a means of comparing major shifts in community composition and
dominance in relation to hydrologic changes. In this analysis, stations were treated as
replicates. The percent abundance of each taxon for each respective station was averaged
across stations for each particular sample date. Comparisons among wetland invertebrate
communities were made on a seasonal basis, such as late spring (June), early summer
(July), mid-summer (August), and late summer (September). The invertebrate
community collected when water levels were above average (1998) was quantitatively
compared to the community sampled during below average lake levels (2000). This
approach made it possible to evaluate taxonomic shiﬁs in relation to a 0.56 m decline in
lake level (0.71 m reduction in mean wetland water depth). Statistical changes in percent
abundance of select invertebrate taxa between high (1998) and low (2000) lake level
years were assessed by the Mann-Whitney U-test. This non-parametric analog to a two-
sample t test was chosen due to the prevalence of zeroes in the data and the difﬁculty in
approximating normality under such conditions. This test transforms the values to ranks
(ignoring group membership) to test that there is no shift in the centers of each group.
Only those taxa that represented at least 2% (percent abundance) of the respective

invertebrate community were statistically analyzed for temporal differences. Alpha was

80

 

 

 

 

 

 

 

 

.— 0306 002:0“:

E 00002.80

x— 058000: 0000053020: 0:03:50:

U 03009 0002000

U 0.0000509 000200: 0:0:g0:0=.0:0m

0 30:20:.

.— 0002300.

0 32:52:80

.— 00200535

U 0500::0m 30:00:00.0.—

: $20.05.:

: 30:02:20.0:

U 0:035 0000::—

U 0:203:00

.— 00500930.:

U 00.000.00.05

U 0:825:20 000:0.000:::U

.— 002032—820“:

: 0:200 300.302 230::

U 20:50:00

.— 2233 00:02:00

m 0050:0125

U 00300:.N

m 0:00:0m 002.2300:

m 00:00000k

0 05000: 82.3.0:

.— 80.80::

m 00022—0030

m 0.823% 000___0:_00.¢:U 30300.00 0:000:—

m 05:00.20»: 02:503.
9.0:”: .0:0_:0:=& 0:_::-::m\0::00 0:50: :00:O 000—9523.—

 

.Gam: 0:050:00 0:0 600:: 0580.50 0:0 00:02 :0 0000: 003 3:00:82:
30:0 .8: :0:000:: 0:0 Aom: 00:0:00 A5: 00:03 Am: :0000::0 AU: 8:00:00 0: 00:00:00: 0:03 095:0 m::000.: 0:000:00:

.2530? 3 :0::0w_:00>:: 00002:: 0:: m:_::0 00:00:00 000: 0:0::0::0>:: :0: :0::00¢_000:0 anew m::000.: 0:000:00: Nd 030:.

81

 

 

 

 

 

 

 

 

 

 

 

 

U 2.. 0:00.50.
-.. I. am 0.0000000: 00200000003

0m 0:300:03

0m 00300.3 0020.300..—

0@ 00.0: 0020»...—

0@ 500.3005

0m 000:000000000k

0m 0:00000k 0020005.?—

0m 002—0030 000003000

0 22:02.00

U 0:000:00Q 0.2020—

0 000030009 002.82 000000—

0 0000.5 22.2.00:

0 00.00.5006 002000.20”: 00002930 0003000202

0 0.0.2.8000

.— 3232: 0:00:00wzo 02.052

.— 0:00.35

m 00.00.30.002 0020003004

E 00:50:05

E 0000200:

#— 00~A00uwx 002509.20: 30:00:02.0

.— 50005030 0022:001—

m 0Eu0=00m~ 002020000000

m 000:0 0020:000< 0:0—.000

.— 00023009 002.00.00.00 00300-0002

m RN=0N000K

m 0:0:0000V 002.00».— 00030200:—

A 003002 0020..—

.: 000000002 002:00:0:0z

.— 5338: 0025332 20:90.0: 0:000:—

00000 .0003005— 0::0:-a=m\0==00 2:00,.— ..0100 000.952.»...—

 

00083 00 03.0.

82

set at 5% and all signiﬁcance tests were performed in SYSTAT (version 7.0). Data from
1999 were interpreted as representative of the invertebrate community during the
transition between above average lake levels (1998) and below average lake levels
(2000), but were not statistically evaluated to keep the analysis concise.

Enumerations of the chironomid sub-tribes Chironomini and Tanytarsini were not
performed correctly on 25 June, 1 July, 8 July, and 16 July 1998. As a result, these data
were summed for eaCh respective sampling date to facilitate complete seasonal (monthly)
descriptions of overall community composition. Had this not been done, community
composition during late spring (June) and early summer (July) would not have summed
to 100%. Fortunately, enumerations were correctly performed on 16 June and 28 July
1998 and these data provided an indication of percent abundance comprised by these taxa
during the respective months. Statistical evaluations (see below) of annual differences in
percent abundance of these taxa were conducted using the correctly enumerated data
from June and July.

Spatial Patterns

Spatial affmities for the outer and inner regions of the study wetland were
quantitatively assessed during each respective year. Only those taxa in which percent
abundance values summed (across dates within a year) to >45%, >5%, and >20% were
evaluated. These values were arbitrarily chosen since dividing by the number of dates
sampled each respective year (N =9, l, and 4) would yield a percent abundance of
approximately 5%. A total of 15, 17, and 28 taxa met this criterion each respective year.
The Mann-Whitney U test was used to compare percent abundance values of replicate

samples (stations) between the outer and inner marsh regions. Correlation analysis

83

(Pearson) was used to further evaluate spatial distributions of only those taxa in which U
tests were signiﬁcant. Replicate samples (stations) for all collection dates within each
respective year were combined to evaluate non-random spatial trends relative to distance
from the wetland interface. Alpha was set at 5% and all signiﬁcance tests were
performed in SYSTAT (version 7.0). Additionally, spatial patterns were qualitatively
compared across years. This approach provided a means of evaluating the response of
select invertebrate taxa to habitat changes that occurred in conjunction with declines in
wetland water depth.
RESULTS
Habitat Characteristics

The average stage of Lake Huron during the months of June through September
1998 was approximately 0.19 m above the long-term summer mean (Figure 2.3). This
corresponded with a 0.33 m decline from 1997, a year which represented the highest lake
stage in the 50-year record since 1986. Lake Huron water levels fell between 1997 and
2000 and corresponded to the most rapid change in the 50-year record (Figure 2.3).
Mean spring/summer lake levels declined by 0.48 and 0.71 m during 1999 and 2000
respectively relative to 1998 water levels. Changes in lake stage translated to a 0.76 m
reduction in mean water depth at the wetland interface station between 1998 and 2000.
Average water depth across the study transect decreased from 0.80 m in 1998 to 0.09 m
in 2000 (Figure 2.4). Variability was greatest in 1998 relative to later years and the
overall reduction in mean wetland water over the four—year period was nearly 1.0 m.
Daily water level ﬂuctuations were visually obvious during 2000; stations within the

inner wetland region (160-280 m) were often less than 10 cm deep (Figure 2.4). Seiche-

84

.32 83am moo—5 826 3223585 9 mucomwotoo n50— .CopEoEoméEE avowed :ozoozoo
0858.52: 3 288:8 8:: @2an .938 Ems E 33. H 362 A32: Eton new 2.8 ”mm 3 omﬁozw 3:58 a??? of
9:335 35% mo voted 05 macaw 282 093 35:08 :32 .933 #35 88.32 Enos 2: 3 wﬁncommots am 3 03.85
39¢: of was >33 mo 23» 2: maize =2 .o=_>xommm E 288 Amm NV 09% CanoEomocsc coﬁgmhﬁam =82 .m.~ oSmE

 

 

 

38 a2 :2 Ea 38:
8a E... can 5:. on: E... 8: 5:. =2. anew-05;.
comb—L.l...4411F171711f._..m...2.2.1.3.... .85.—
L _ _ _ _ _ _ M H
. _ _ _ _ _ _ l .
2.6: r. _ _ _ _ 1. ... 8.3.—
. _ _ _ _ _ . ...: . s
k _ _ _ _ _ H ...... w .m
as: L......................i_................“................................................m................¢......................H......................ﬂ................“.................................................................................................... 1.32m
. _ _ _ _ _ g Sﬁﬁﬁ. . H.
U _ _ _ _ _ H ....m H m
8.2.. i. _ _ _ _ _ _ i. .l 2:: a
. _ _ _ _ _ _ . as . m
U _ _ _ _ _ _ H H
as: .. _ _ _ _ _ _ .. was:
. _ _ _ _ _ _ . .
u _ _ _ _ _ _ H h
8.»: ....C.............C...........i...........Lt. 8.”:

 

 

 

85

 

 

 

 

 

 

 

 

 

 

 

 

 

1.06 l
0.80 .' 13°
_
i E l 00
N=60 .. . ~ A
. ., i - 0.80 5
‘ 5
a.
i 0.31 - 06° 5
N=135 ; r 040
. 0.09 r °
. + A r 0.20
N=105 :
ﬂ U l ' 0.00
1997 1998 1999

 

Figure 2.4. Box and whisker plot of water depth measurements collected within the
wetland (0-280 m) during each respective year. Box and center line represent the 25th
and 75th percentiles and the median respectively. Whiskers correspond to the 10th and
90th percentiles, circles represent outliers. Numbers above each box represent the
mean wetland depth for each respective year; number of individual measurements are
denoted below.

- 1200
- 1000
800

p
b
I-

600

No. Stems m'2

'é

 

 

 

Figure 2.5. Mean (1 SE) cumulative stem density (no. stems m'z) along the study transect
during 1997-1999. Monthly (J une-September) data were averaged. Stem densities were
not signiﬁcantly different between years within the outer wetland region (0-140 m);
signiﬁcant (p<0.05) increases were measured within the inner marsh (160-280 m) in
1999 relative to 1998. DFWI corresponds to distance from wetland interface in meters.

86

induced dewatering of the study wetland was noted on one occasion during September
2000.

Mean armual cumulative stem density was quite similar during 1997 and 1998,
but increased in the inner region of the study wetland in 1999 (Figure 2.5). Stem density
was not quantiﬁed during 2000, yet qualitative observations suggested an increase
relative to 1999. There was no signiﬁcant difference between 1998 and 1999 when the
entire study transect was considered. Contrasts between years in the outer wetland region
(20-140 m) were also not signiﬁcant. However, stem density in the inner wetland region
was signiﬁcantly greater (p<0.05) during 1999 relative to 1998. Despite the increase in
stem density in this region of the study wetland, a dramatic reduction in habitat (on a
volume basis) resulted from the sharp decline in water depth between 1999 and 2000
(Figure 2.4). Stem densities would have had to increase nearly four-fold and uniformly
across the wetland to replace habitat losses due to the decrease in water depth.

The extent of mixing between lake and wetland surface waters has been
hypothesized to inﬂuence spatial distributions of select wetland fauna (Cardinale et al.
1996; Cardinale et al. 1998). The alabaster substratum (AS) experiments were designed
to quantify hydrologic connectivity between lake and wetland. In 1999, mass loss
declined in a nearly linear fashion throughout the ﬁrst 100-120 m of the study wetland
(Figure 2.6). The dissolution of substrata was greatly reduced 120-280 m shoreward of
the wetland interface. During the 2000 experiment, a similar pattern was observed,
except that mass loss was greatly reduced approximately 60—80 m from the wetland

interface. Mean daily wind velocities during each AS experiment were not signiﬁcantly

87

 

 

 

 

 

100 . . . .
+ 1999
-V— 2000
3,3 L
3
c
.J
3
g r-
v L
0 Outer Wetland (20-140 111) Inner Wetland (160-280 m)
0 40 80 120 160 200 240 280

DFWI (m)

Figure 2.6. Mass loss (:SE) of alabaster substrata (AS) across the study marsh during
mid-summer 1999 and 2000. Arrows indicate subjective inﬂection points where mass
loss appeared to become relatively uniform thereafter. DFWI corresponds to distance

from wetland interface in meters.

88

different (p>0.05). Mean water depth in the wetland was approximately 0.22 m lower in
2000.

Average temperature differences along the study transect ranged from l.6-2.2° C
(Figure 2.7A). Temperature generally increased shoreward of the wetland interface. The
overall range in temperature was from 34.2° C (1999) to 160° C (1997). Mean
spring/ summer temperature proﬁles consistently increased from 1997 to 1999, but
declined in 2000 (Figure 2.7A). Modeled water temperatures for all of Lake Huron,
excluding Georgian Bay, suggested a similar peak during 1999 followed by a decline in
2000 (Figure 2.7B). Average air temperatures mirrored the trend in water temperatures
through 1999; data for 2000 were unavailable, but likely were lower on average relative
to the previous year (Figure 2.7C).

Percent saturation of dissolved oxygen (DO) measured in mid-morning (10 am to
12 noon) ranged from 42.6 to 149.0% during June through September of 1997-1999; high
and low measurements were made in 1999 (Figure 2.8). The outer 40 m of the study
wetland was generally at or above 100% saturation. Percent saturation typically declined
across the wetland to the 200 or 220-m stations at which point percent saturation
increased; a general decline was observed across years, particularly shoreward of the
140-m station (Figure 2.8). The sag in daytime percent saturation in the inner wetland
region was consistent across years. The trend for 2000 and the extent of diurnal
ﬂuctuations are unknown.

Speciﬁc conductance, a surrogate measure for total dissolved ion concentrations,
increased throughout the years of study, particularly between 1998 and 2000 (Figure 2.9).

Average spring/summer speciﬁc conductance values were signiﬁcantly different (p<0.05)

89

use“ 8 0 co 8 FEE: 203 8:58anan 833 :83: 8.3 33on 2:. £2qu - mOmD Eo¢ 3:350 295 Sat :93: 861—
.Amm Hv owns; :CB-w:o_ 2: o. 02:22 c5... 35% .Lo 83» at waist 9mm 239500 @8233 .853 8:3 8.“ ADV moEHSumEB

:u 83 .5>o 28 Am: 85:23an 533 3.258 35:2: :82 £82.: 5 conﬁne": 93:25 an 8536 8 mucommobg SEQ
it $3.33 mo £255» 05 2:23:85 538 has; 2: $28 @2388 8an AU ov EBEoQES SE3 $5 3 :82 SN 2:3,...

.100 {“00 .9.» we»? aééﬂ Tie/“V

 

 

  

 

 

. . . . o
. a
w
.......................... V
. N— "......
l
. 39
m
1 2d
)
O
w—D
. an
O . S
5.5—2 cchuava— + . VN
2- . 82 IOI
m- . 82 In!
nae IDI
M c 32 IOI
8
J
n
1 m .
m
d 2 .
\d
D m— i
(
an .
3
MN . n . . u r

 

 

 

2

:5 55...:
SN can 8. an. an 3. o

p u p . - . - . p . p . p o—

(3 o) diner. 1918M

:2 ID]
33 IOI - on

 

90

 

130
120
k 110 ’E
.2
‘ *0
.. g ............................................................ ‘ 100 E
" ‘ ‘ x 5
~ I .
\IIII’ mg
+ 1997
-—v— 1998 70
+ 1999
I ' I I I ﬂ I I I ' I ' I '— 60
o 40 so 120 160 200 240 280

DFWI (m)

Figure 2.8. Mean Q SE) dissolved oxygen (DO) percent saturation proﬁles across the
study wetland during June through September of 1997-2000 (N=4, 9, 6 per station for
each respective year). DF WI corresponds to distance from wetland interface in meters.

 

 

 

 

 

 

 

 

 

 

 

 

 

1200 1
0
.4000-
"-' O
E
o
3300‘
“g 0
u 600 - Q g , 3-
@4001, _.______. ,1. ‘-
'_" i o
200 ‘ 383.9 36'lﬁ.2 43']; 526.2
1997 1998 1999 2000

Figure 2.9. Box plot of speciﬁc conductance (uS cm'l) measurements collected from the
study wetland during the summers of 1997-2000 (N=60, 135, 90, and 73 respectively).
Boxes represent 25th and 75th percentiles, medians are indicated within boxes, whiskers
correspond to 10th and 90th percentiles, and circles represent outliers. Mean values are

italicized below boxes.

91

during these years relative to each respective previous year. The range in pH spanned
slightly more than two units (7.31-9.23) with a grand mean of 8.02 @003). Annual
differences between 1997-1999 were small (annual means: 7.99, 8.03, and 8.02
respectively) and, though no reliable measures of pH were obtained during 2000, a shift
in pH seemed unlikely. There was no direct evidence for constraints on invertebrate
osmo-regulatory function, despite temporal variability in pH and elevated ion
concentrations across years.

Turbidity was only quantiﬁed during June through September of 1998 and 1999;
mean values were 12.53 Q 0.69) and 3.34 (i 0.16) NTU respectively and indicated a
reduction in suspended material between these years (Figure 2.10). The range observed
during 1998 (2.7-36 NTU) was much greater than during 1999 (1 .7-9.7 NTU). Turbidity
decreased consistently across the wetland in 1998, whereas no trend was obvious during
1999. Observations made in 1997 and 2000 coupled with the 1998 and 1999 data suggest
that turbidity increases in conjunction with rising lake levels.
Invertebrate C ommunily Structure: 1998 vs. 2000

Sixty taxa were collected from the study wetland over the course of the three-year
investigation (Table 2.3). Total taxa richness (S) was lowest during 1999 and highest in
2000, reﬂective of low sampling intensity in 1999. Arachnids, annelids, and crustaceans
(amphipods) were the abundant (>5%) non-insect groups in 1998. Gastropods were also
abundant in 2000, while the percent abundance of arachnids and crustaceans decreased.
All of the major aquatic insect orders comprised a portion of the community during at
least one of the three years; Collembola and Megaloptera were uncommon (< 1%) and

only collected in 2000. Dipterans were abundant and accounted for the highest

92

 

 

 

 

 

25 -
—v— 1998
‘ + 1999
20 q -- __ T
D
[-I 154
E t ._ _-
a? 1
f2 .
f 101
3 .
h 1
5 ‘ W
0 I— I ' I V l v I v y I ' r 1
0 40 80 120 160 200 240 280

DFWI (m)

Figure 2.10. Mean (5 SE) turbidity (N TUs) proﬁles across the study wetland during June
through September of 1998 and 1999 (N=9 and 6 per station for each respective year).
DFWI corresponds to distance from wetland interface in meters.

93

 

 

 

NE. 5... E... A
NN.. 5... 8.... E
8. ..N NN ..N . m
. . . . 5.8 .... No... 5.8 N... 2.2.2::
N N N N 5.8 NNNN NN... 8N8 NN... 32.9.35
N N o N 5.8 NS. 5... ....N 5.2.:
N N N N 5. .8 SN N... . 8... 8 an... 83.2.5
N N N N 5.8 ..N...N 8.... A... .N. ..N.NN «8.2.5.
. . . . 5.8 8... N...N 5N. 8.8 8.5.2:...
N N .. N 5.8 .N.. NNN 5.8 ...... 2282.3;
N N N N .N... 8 2N 3.... 8N8 N... 5.25.5
. . c c . 5.8 N5 22.8.80:
N N N N 5.8 NN... $.N 5.8 NS. 2238......
N N N N 5.8 .3 8.. 5.8 Ni. ages»:
N N N N 5.8 NS. NN... AN... 8 SN 2283523
... N. N N 5.... NSN NEN sNN. NNNN seen
. . c o 5.8 N... 2352.5
N N N N. 8N8 EN 8.. .28 NN... 23.5.5
33-3% ..ch 33 «Na. .53 32 m3.
Amv NEE—u:— uxah .35. Aﬁcv 3:81:25. 2.3.8.—

 

 

.Noag 85855.. .582. 3:58 53:. co 3me 38.3.8 825 03 $2.55 was A. $3 .89.. @8828 9:05:25

Ara—03.02.88 v 8.8 ._ .ouZV oooméam. mo CunEoEomécsc 3883 use 9.8% on. 593385 @2338 Amv N8
.83 8 32:53:00 wig—2.. $820 .3320 ...—.33 anew 882352.. 8.8:. ..o $5 3 3.825.... .523 .355 :32 .m

5.0.. a...

. .25....
.N 0~n~wrﬁ

94

proportion of the invertebrate community during each year (Table 2.3). Coleoptera,
Ephemeroptera, Hemiptera, and Odonata comprised a higher proportion of the insect
community in 1999 and 2000 relative to 1998. Taxa richness increased for Coleoptera,
Diptera, and Hemiptera in 2000. If additional taxa had not been collected for these
orders, overall richness would have been comparable to 1998. Shannon’s diversity index
(H ’) increased across years and in conjunction with decreasing water depth in the study
wetland (Table 2.3). Evenness (J ’) also increased across years.

Nine taxa accounted for at least 2% (common or abundant) of overall community
composition in 1998, ten taxa in 2000 (Table 2.4). These taxa collectively represented
92% and 78% of the invertebrate communities each respective year. Tanytarsini and
Chironomini (sub-tribes of Diptera: Chironomidae) were not enumerated separately
during the second collection date of June 1998 and three of the four July 1998 collection
dates. Of the ﬁve remaining dates in 1998, Chironomini and Tanytarsini accounted for
approximately 18.2% (1- 4.5) and 11.3% (i 0.8) of the invertebrate community
respectively.

Abundant taxa differed across years and the only consistencies were for the sub-
tribes Chironomini and Tanytarsini, the mayﬂy Caenis (Ephemeroptera: Caenidae), and
Oligochaeta (Table 2.4). Changes in the percent abundance of Chironomini and
Tanytarsini between 1998 and 2000 were not statistically signiﬁcant. The percent
abundance of water mites (Hydracarina) and of the arnphipods Gammarus and Hyallela
declined between years; the difference was not statistically signiﬁcant for Hyallela. The
caddisﬂy Nectapsyche (Trichoptera: Leptoceridae) and the bivalve Dreissena were

common (>2%, <5%) in 1998, yet uncommon in 2000. In general, trichopteran taxa were

95

 

 

 

 

 

 

96

 

 

 

...... 8.. .....8 ...... 822.5
C Mdv vnd 2... BESHSQQ Susana—3320“ 9.02350:
.5 .5233. .8... 2% 2:8,. 82:80
.88 2.... .N...... N..... 882...... 82.8.. 53.20525...
.N . .8 NN... 82.5....
6...... N... 828......
.38 NN... 82.52.28
.28 NN... 828528
.8. .. s... 558...... 82.5.5...
-.. .N..... ..N... 822.5...
......N. NNN. 8.........5.2.....
.....8 ...... ... 5.8... 822..
.... .N...N. as 858...... .5558...
....8 NN... ...N.8 ...... 85.8285
... .... ..N... .N...... ...... 85.8.85
.5 83. 8.... 83. 8.8 5.5855 8255525
.58 ....N ...N.... ...... 82.5.8556
..- ......8 NN... 15852 822.82 555..
.88 N... 5352.5
...... ..N... .N. .8 ..N... 8558 825....
.N...... ...... 5.555%:
$5.... «.... --- 8:52....
.28 N... 5.88.. 82.5.2.2:
.88 No... 8.85....
.28 ..N... .N...... N..... 8.8.5: 82.2....
...N8 Ne... .88 N..... 823......
.88 ...... .88 No... 82:255....
3...... 8... --.. ......EeQ 2.2.—9.....Nfgu 2938...“. 83.....—
N. ..8 ...... ..NNN. 83!, 8588...... «25.8...
......N ”a... 3.525.359 £58... .256 88.0.5512...

 

.9352: com 6082...... .32 .0. 35.8.8 0.03 382805820 .0 NEEAEN. .:_m.a.>=m,_. ...... 2:805:50

......oﬁcwi .0: 358.3. N: 3...?» 283.2. 3.895.... 2.8.2. ... Newcmno ﬁchu .... .53....ch ...—3.8.88 8.. 888%.. 0.25

38. 08... .....o .mo_._:=....=o. 03.893. 2.. .... $N .98. ... 3.58.3. ...... 88. 8.8%... 8.... Eom A.» n 2. ooom 98 a M. Z. .32 ..o
A.on:.o.mom-u:=... 85.8.8 ...... main... 2.. .socwsoE. 888:3 8.... o.a.po...o>:. ..o Amm H. 3:852... .522. 32...... :32 .vd 03....

88.... 29...... 9»... 8:595. o... .o>o 83.32.
...... H. mm... ...... .3. H. .86.: .... 3.588.. :5 do... ... 8:... 5.80:8 2.... on. .... ...... .8 88.08.50 2882...... .9. 0.25 2.382.... ...... _=_Eo:o._zo u ..

 

 

 

 

 

 

 

 

 

 

 

......8 .... .88 2... 2.2.2.8..
--- .8... me... 3895...... 2.2.39.3...
.88 8... 2.8.8....
....8 8.. .88 ...... 8.55.9 822.82..
...N... m...” ......8 8... 8...... 82......
.88 .... 282.8...
$8: 8d --- $583833qu
.5... ...: .....8 ...... 2.88.. 82855...
......8 8... 82.8.2 €28.85
.538... .....8 ...... 82:82.5
.. 2.. 2... 2. .3... ...... 5.82.5 i 22.2..
.....8 ...... .....8 ...... 88285 82.8.... 2.2.8.
.8. .. ...... .... ......8 8... .....5... 82.22.:
G... C n ... ...—e... ..md 2:3..an 2.2.5..an 2.2.295. 39.88:...2
... ...... 8.2 ....n. 8.3 2828......
.....8 8... .88 ...... 822:... 58.82... 82.8....
.....8 ...... .28 ...... 2.8.5
.8... ...... ... .....8 2..” 8.8.8.82 82.82....
.88 ...... 23.8.6
.28 ..N... .88 ...... 228......
.88 8... . .....8 8... 8.8.... 82.2.2.8: 8.385.;
.818 ...... 85.8% 82.2.8...
... ....N. 8.. .....8 .... .2838... 8252.888
.88 8... .88 2.... 8...... 825.88. 58......
.88 8... 83.8.... 82.2.2.6 88.8280:
.. .8 ...... .88 ...... 5.5.8....
.88 ...... .88 8... 5.2.8... 82...... 8.282....
--- --- 8.882 26.2..
G . .... e _ ... --- 885.32 32.85.37.
.88 ...... .....8 8... 2.88.... 82:88.2
$93 £5 I. 8.3qu oughhoc «Loam—=9: 390m:—
......~ 88. 8.2-8885... .252. .8... .8682...

 

.288. ...N 2......

97

relatively unimportant during 2000, collectively accounting for less than 1.5% of the
entire community (Table 2.4). The percent abundance of Ceratopogonidae (Diptera) was
signiﬁcantly greater in 2000 relative to 1998; Enallagma (Odonata: Coenagn'onidae)
increased in 2000, but the change was not signiﬁcant. Dolichopodidae (Diptera) was
common and only collected in 2000. Three snail genera, F ossaria (Lyrnnaeidae),
Pseudosuccinea (Lyrnnaeidae), and Physa (Physidae), accounted for greater than 20% of
the entire community in 2000. The increases in percent abundance between 1998 and
2000 for F ossaria and Physa were statistically signiﬁcant; F ossaria was abundant (Table
2.4).

The percent abundance of a variety of uncommon taxa also changed as water
levels declined in the study wetland. Twenty-one new taxa were collected in 2000
relative to 1998. This was particularly evident for the insect orders Coleoptera and
Diptera (Table 2.4). For example, several genera within the family Hydrophilidae
(Coleoptera) were observed only during 2000. Seven of the new dipteran taxa collected
in 2000 accounted for approximately 6.3% of the entire community (Table 2.4).

Collectors accounted for 84.5% (i 2.5) of the invertebrate community in 1998 and
55.5% (i 5.8) in 2000. Scrapers plus piercers and predators increased roughly 24% and
6% between years. Shredders were similar among years, 3.7% (i- l.1) and 2.8% (i 0.6)
respectively.
Invertebrate Community Composition in Late Spring

Seven taxa were common or abundant in late spring 1998 and included
Hydracarina, Chironomini, Tanytarsini, Nectopsyche, Oligochaeta, Gammarus, and

Hyallela (Table 2.5). Hydracarina and Necropsyche were uncommon in June 2000 and

98

 

 

 

 

 

 

 

 

 

93.3 23 5.3 :2 05.5 2.256
$18 32¢ --- 0:003:09 002.00: anoiﬁuﬁogam In
Cmdwomd --- 002.35
ENE 36 --- 00253:.
ANNE NNd --- oae_.~=.e_.a.=m
93.8 8.0 32:...sz
93.8 36 --- SEBEE 26:00:93.—
822.50
20.3 "—... --- 002132.33:
--- --- £035 002::
280.3 8.: tease... 2.29.22.
$9 :6 Anne 3.. 0258?;
..- 8:: Ed 00:23—00:20
2.8de 3.2 GNQ 3.: EESEEU 3252.220
min—é and. A23 nmd 002:0»38200
... --.. 3003‘ 0030:05< .2035
--- --- 20:50:00
303 02 - lame :2 953:3 325.3 '
-.- --- 0.3580385
8.028 mm... --- 0.33025
€98 8.. --- 05.20% 0.02—Ease»:
..-- --- nabcetm
:53 2:0 3&0: 82.2.“:
30.8 om; --- 00200.25
--- --- 002523.50
--- --- 303:5 003.050320 20300.00 8000:—
Ammdv omd «$5.3 36 «5.33.5»: 02:50.2
8% - 2:; :2 - 2...... 2__:-._=m\2.§o has... 5.5 $225230

 

$038090“: 305—05 can 082: bmmbzc Psoccmcm 8 0:83:00

A can .: .mmocnot 823 0003380 m ”300505 000 6052—05 woo. 0:3. 08 00883 0003 A0mEEocoE5 00 0003-253 E60338.
was Ems—0:820 4:005:90 Ho: $500500 0: ”200% 50.503 00:02:50 E083 5 women”? Amodva ..L 0:005:me bﬁozmcﬂm no.“
0000300 0003 0x8 085 Eco ”0.02.5888 0:050:02: 03003.00. 05 .8 gm 300— :0 008000.500 :2: 880 00:00:05 095 0:5 .8“

u Zv coca use 3.0. n a woo. 0:3. 5 0:020? .630 05 Soc 380:8 82: 0:050:35 mo Em 3 00:00:30 0:083 .802 .m.~ 030,—.

99

 

 

 

 

 

 

 

 

 

 

G _ .8 «E 88003838:
3...». 5.2 5.88.: 328:5:
-..- --- 32.55. «62.9336
--- --- 020.2023
3.. ..-- Sawmﬁvkg 22:29
--- --- «0.030009 32.003: «62.8.
.2... 5.0 20.35 32.23::
..-- 8...... «N... «505509 02:35:30 «.02....A.E< «0.2.0.0332
25.0.8 9.3 82. «man 585%:
.88 «no 850.5: 50280.5 3:052
.38 8.0 88.5
5.8 S... 28.8 2.0 «ﬁasaz 32.83%:
--- --- 3.50.96
.28 8.0 5.323:
Q _ .8 $0 8.0:»: 32:39.2: 22:22:.
--- ..-- 53.0%.? 32.2.2.5
--.. 8a.... 2... «530.33% 32:33:30
--- --- RS? 3.0.5.0004» «3:25
--- --- «0.8.33.9 020.3330 9.030.305.
--- --- 380508..
808 ”no 0328...: 82.9.: 22:023..
--- --- «02:32 03.05
--- --- 200.332 02.200532
3.8 So 5.033.. 82.3332
--- --- 32:09 32.200
8.8 2.0 263.5
--- --- «Sanctm 03.358050: «SEE—0: «.000:—
ceen - 2:... «2... - 2:... 0......-..=m\m=:0c ....Ea: 3.0.5 320.532..

 

.33.:00. Wm 0.90...

100

.23 =ozoo=oo 2:: .3: 2: W
co Cum .3 m. : can 86 ﬂ fomém .8 38:0on :3 .32 2:... E 2% 5:00.30 9.83 2: :o Baggage b02233 S: 225 EBBSEH Ea _=_Eo:o._EU uk

 

 

 

 

umé am... n K.
MFAv NB“: " .:
mu 5 u m
-.- --- 3.3.—3...;
--- --- 3393»: 32.83.52.—
--- --- 5:53.33
3:: 3o 33:3 533......
898 86 --- 35: 3293.—
Q _ .8 cmd --- SSEMSM. 3202.5»:— 3:33.30
83 - 2:; .32 - 25.. o.....-._=m\m==oc £55— .35 32952:;

 

€283 Wm 23‘

the differences between years were statistically signiﬁcant; Gammarus and Hyallela were
not collected in June 2000. Only Chironomini, Tanytarsini, and Oligochaeta were
abundant in late spring 2000. Dolichopodidae was common in June 2000, but not
collected in 1998. Statistically, the percent abundance of Chironomini and Tanytarsini
did not differ between years; percent abundance values measured in June 1998 were
24.3% (i 3.8) and 11.3% (i 3.7) respectively. The percent abundance of
Ceratopogonidae was statistically greater in late spring 2000 relative to 1998. The
gastropod F ossaria was also abundant in June 2000, but was not collected in 1998 (Table
2.5). Nineteen new taxa were collected in spring 2000 relative to 1998 and overall taxa
richness was higher in 2000. In particular, representation increased for the insect orders
Coleoptera and Diptera, and the gastropods; most of the newly collected taxa were
uncommon. Shannon’s diversity and J’ were similar between years.
Invertebrate Community Composition in Early Summer

Thirty-one and 44 taxa were collected from the study wetland in July 1998 and
2000 respectively (Table 2.6). The increases in richness represented the addition of 15
new taxa that had not been collected during June 1998, 17 new taxa for July 2000. Most
of the new taxa collected in 1998 were caddisﬂies and snails; representation increased for
Coleoptera, Lepidoptera, Odonata, and Trichoptera in 2000. The number of abundant
taxa increased in July for both years. Hydracarina, Chironomini, Tanytarsini,
Oligochaeta, and Hyallela remained abundant in July 1998; Caenis and Gammarus were
also abundant, but the percent abundance of Nectopsyche decreased. Gammarus and
Hyallela were still not observed by early summer 2000. The percent abundance of

Dolichopodidae continued to increase during July 2000 relative to June. The dipteran

102

 

 

3Y8 and .2958 Sam 8:25 32:26

 

 

 

 

 

2.9.8 mad - 82.8.39 2.2.25 a..oE¢..oEo._._m
A28 mad --- 2.2.—E;
A38 8... 325......
.....e ...... 3252.25
2.08 86 --- oagsmgsﬁm
$38 a.» --- QESEE 32—55»...—
33. a... 82.....5
Gm.8 ha.» --- 2.6.—5:23:2—
Go.8 cc... --.. .5»qu 2.2::
2.398 9..» .5388 2.5535...
$2.. an. 02.83....
£8 8... --- 2.565855
2.6.8 5.3 «SEQ 35. _:_Eo=o._._0 SEES—9.20
1.3.836 Q. _ .8 3d oae_=owea8a.ou
as... R... 5.2.... 8223...... 22....
8N8 ovd --- 2.5.5.30
6.... 2... 3...... .2 23:3 2.25.3 .-
Am _ .8 Ed --- nasanieuh
8...... 2.». 2:53,".
3&8 $6 --- nausea 32.2.3.5»:
.8... 2... 853.3.
A28 3... A38 36 2.25:: 0:23:3—
2 ..8 cm... ..-- 2.23.3.2—
AS8 36 --- 2.2.5.3950
Gc8 cod --- 393:5 32:058.»...0 9.8.3930 83m..—
3..8 a; #2.: 3.: 2.2323: «ES—3.3.
......N - ...... .33 - ...... 2....-2Q25u 2.5... 5.5 $20.52....

 

50382.8. $05.26 23 .89.. .3336 $858.5 8 2.88:8

. x. E... .E 53:5... 8.8 8.9.3me m ”3.552.. 08 63.0288 woo. .23. 8.. BEBE 29$ ﬁaEEozozno mo 33.5-8.8 E6335...
...... 2.89.820 ..ceo..._=w.m .2. 3.83.3. m: .28» 283.2. oocavcsnm 2.8.2. ... momsﬁo Amodvm ..L .cmoamcwmm €35.86 8..
3339. 203 88. 085 Eco “8:62:58 0852.3... 03.8%.: o... 80 gm .mmo. ... 3.58.3. .2. 88. 8.36:. 2:. 20m A:
u 28 ooom can Gm n 28 .30. .23 E 9.2.25 beam 2... So... 380:8 8.3 «858.02.. .0 $5 .40 8:85.... 2.8.2. :82 6N 03m...

103

 

 

 

 

 

 

 

 

 

 

an... ...... 3583....3...
89$ mud --- «taugk 3235...?—
......... ...... 2.2.3.... 2.2.2.2...
...... ...... .....a 8... geranim
8...... ...... 3...... ...... 2:32.. «.22...
-..- --- Quahvuuab mas—034. arcs—Om—
...h... .5. ......i: 2.....21:
--- 668 and 2:32.55 323:.an «2.3.—95¢. «3:83.32
2..."... ......N 92. ma... 222.895
8...... 2... .8... ...... 35.5.... saggy... .2352
A... a. R... ...... a... $8.5
.3... ...... .32.. ...... .5388... 2.2.833.
.8... ...... 3.5.23.
9...... 2... .9... .... 3.2.23:
8.... o . ... 3...... mm... 83...: 8.........:...= 23.2.2;
5.3 N... 55.2.2.8 2.23:2...
.3... ...... .2... ...... sinszam 8...._....u..=2.u
2 n... m... 2...... ...... as... Saga}. 523.5
--- --- .33325 32.32390 3.89....qu
1:5“..43... .... 1-: a - i 3...... S... $.38...
3...... ...... 6...... 2... 5.38.... 2.2.2... 53%....2.
--- --.. 32.8.2 320..—
-u- --- 392332 25:90—5ch
...... a... .8... 2.... £325... SEES»...
u..- 3.. wrung 02.?qu
....N. S... 2.2.2....
--- --- 3.883% 323533.3— a.o..__=.o= 83....—
......~ - ...... ...... - ...... 2.5.3.25”. ...s... ......5 2535......

 

.....38. ...N 2......

104

.83 528:8 .23 as 2: .5 SA

.3 Rh: 2... Am... 3 $0.? .8 3.5.83 :5 .38.. 538:8 m3. 3.: 58 2.80 025 :o 38382.0 38280... .o: 22: 23.5.2.3. can Ego—5.30 n k

 

 

 

 

3.: mm... H k.
m: 8.: u E
3. S u m
5.8 8... 3.2.2.5.
--- A83 mod 3393»: 32.53.59..—
--- --- Sﬁfgaz
an... a; a _ .8 2... 2.333 82.22:...—
§..: 2... 93.8 a... 3a.: 2.2%.:—
Anmdv 3d --- 3833.3. 2.22.:th 32.2.25
.58 - ...... .32 - ...... 3....-._..m......=c Ea...— ..o...o ”2.6.5:...—

 

APER: 9N 033.

105

Pericoma (Psychodidae) was abundant in early summer 2000, but was not collected in
1998. Chironomini, Tanytarsini, Oligochaeta, and F ossaria remained abundant in July
2000 relative to June (Table 2.6).

Percent abundance values for Hydracarina, Chironomini, Caenis, and
Nectapsyche were higher in July 1998 relative to 2000 and these differences were
statistically signiﬁcant (Table 2.6). Tanytarsini and Oligochaeta were not signiﬁcantly
different between years, yet the percent abundance of Ceratopogonidae was signiﬁcantly
greater in 2000. Corixidae (Coleoptera) and Fossaria were not collected in July 1998; the
converse was true for the arnphipods. Other notable taxonomic differences between the
two years included an increase in the beetles, Hydrophilidae in particular, and higher
representation of the dipterans and hemipterans, particularly Corixidae and Mesovelia
(Mesoveliidae). Additionally, the odonate community shifted from primarily damselﬂies
in July 1998 (0.6%) to dragonﬂies in 2000 (2.0%), whereas the caddisﬂies and mayﬂies
declined in importance. Shannon’s diversity index increased from June to July for both
years. Diversity and J’ were higher in 2000 (Table 2.6).

Invertebrate Community Composition in Mid-summer

Total taxa richness was highest during August relative to any other month of the
summer during both 1998 (33 taxa) and 2000 (48 taxa). The number of taxa collected in
August 1999 (28) was lower than the other years, yet comparable to 1998 (Table 2.7).
Five new taxa were collected in August 1998 relative to July; nine new taxa were
collected in 2000. The number of abundant invertebrate taxa remained similar to
previous months for both years; the same was true of common taxa. Chironomini,

Tanytarsini, Caenis, and Oligochaeta were consistently abundant between years.

106

 

 

 

 

 

 

 

8.8.8 8.... .3... ...... .88 .3 2......6 82:80
.88 8... .28 E. ......8 v _ ... 2.88.5 82.8.. 8.28.8.8...
82.2.;
.88 8... 82.889
.88 S... 82.22.58
.28 ...... 828.528
.m . .8 ..N... .28.... 82.85....
.88 ...... 822......
.88 t... ................_2_....
3.83 82.2..
8.8... 2... ...... 8.... .88 ...... 2.22.8.2
.....8 ...... .88 ...... .88 8... 82.8.25
.88 2... .88 8... 822.282....
......8 S... .....8 5.8. ....88 NE. 2.22.8.5 8......2..:...u
.....n... a... .28 8.. .88 ...... 8.282.388“.
.28 8... 88...: 822.82 222..
.88 2... 8.2.22.5
.. m8 2... .88 a... 9.8 8... 85.2. 822....
.. ..8 w . ... 83.8.8.2...
.28 ...... 2.2.8....
.88 S... 2.8.... 8.2.2.25...
--- .-- --- 83.28....m
.. m8 ...... .... .8 m . ... 38...: 82.2.8:
... . .8 ... ... .....8 2.... 828:...
.88 8... 8.2.2.826
-.. -..- --- 3.82:5 035058.220 2232.5 882.—
.2.8 2.... .8... «.3 33. 8... 8288...... 2.228....
......N - 23. a... - 2.... .8. - ...... 8.2-88.8... 22...... .8... 8.62.22...

 

...—03.8%.... $05.26 ...... .89.. 2.2.3:. 9:05.95 0. 8:88.80 K. ...... .E $85."... 8.8

107

88:83.. m .88....ch .9. 8:823. m: 50>... .885 ... 39.8.... o. coax—o. 5 8%.? 5:52.88 .0 5:80.92... 2.. ... 2.. o. 32.
33 32 ...—co coon ...... was 33.. 5038.. 85.25.... .522. a. 89.83 Amodvn. ..L 38......me 2:85.35 5.. 888%.. 225 88.
80... .38 82:58:50 28.22.02: 0282.8. 2.. ..o ..\..m .23— ... 8853...... ...... 8.... 3.86... 090 20m .9: M Z. coca ...... A3

u ZV moo. Awm N Z. .33 .m=m=< ... 3.2.25 2.2.3 2.. 80¢ 888:8 88. 8.2.22.3... .0 £5 H. 8:35.... 2.3.8. :82 Sm 03....

 

 

 

 

 

 

 

 

 

 

.38. 2a... --- --- 85838338..
......8 8.... .88 8... 2.88... 8285....
.88 S... .....8 8... 8222.... 82.2.80
23.8 no... --- --- 2.2.32.5
28.8 no... --- «22m... «v.2 2.3.82Q «...-«>5
.88 ...... ......8 2.... 8.8.8.... 82...... ......8..-
.88 8... .8... 2... ......8 8.. ......s... 82.2.8:
....8 .8... .8... .... .... .... 8.... 8.8.5.6 8282...... 82.28:... 32.88....
.....8 2.... .....8 8... ......8 ....8 2828......
.88 8... .28 8... .88 ...... - 822:... 8.858.... 22.8....
.28 3.... .28 8... .88 8.... x 2.88
......8 ...... .88 a...” .. . .8 ...... 2.8.8.82 82.88....
.88 ...... 5.28.6
.38 ...... .88 3... .88 2.... 2.2.3....
.88 ...... .8... .3 .....8 t... 8.8.... 825.22.: 8.2.2.2....
.28 2... Sagas... 82.2.2... I
.88 .3... ...... 8.... 8......8 8.8 85.8.3... 8.22.8.8.
.88 8.. .28 8... 5...... 82.2.8... 32.2.0
.....8 ...... ..- .8222... 82.82.... 22.28.:
48.8 .... .88 8.. 8.38....
.....8 2.... .. ....8 ...... .....8 ...... 2.28.... 82.8.... 8.382....
...... ...... 8......2 82....
....8 8... ... 2.82.... 82.8.22
.88 8.. .88 8.. 2.88.... 82:88.:
.....8 ...... ..- 2...... 82:...
.....8 ...... ......8 ...... 82......
.88 ... .. .. _ .8 8... .528... 828238.... ......e... 38....
......m - we... 88 - ms... «28 - w=< .....-aam..=..oc 2.25.... ......o 885.522....

 

.228. Z ......

108

109

 

 

 

 

up... 5.: 3.: n L.
:2 S... 3.: u L:
3 ma mm H. m
33: 8... A38 85 :2: one 2.2.2::
..- ..-- 2 .3 :6 232:“: 32.38.52;
69.8 2:. £33.33
$9 on; 8:8 2: £323 26535:.
L8,: 3.». $2.8 25 5.8 as 2a.: 32:...—
Seé 36 --- --- Sgézusm. 333593 «Ex—9:36
833.2 32 - a? .32 - $2 2:3-..5325 :25. 5.5 2525...»:

 

.6253 2 22¢

The percent abundance of Chironomini and Oligochaeta were higher in 1998
relative to 2000 and this was statistically signiﬁcant; oligochaetes were uncommon in
1999. No signiﬁcant differences were measured for Tanytarsini or Caenis between years.
However, the percent abundance of Tanytarsini decreased across years (including 1999);
Caenis increased across years (Table 2.7). Enallagma, Gammarus, and Dreissena had
higher percent abundance values in August 1998 relative to 2000. Gammarus was also
abundant in 1999; Dreissena was not collected. The difference between years for
Enallagma was not statistically signiﬁcant, yet this taxon was particularly abundant
( 14.6%) in 1999. The dragonﬂies comprised a similar proportion of the invertebrate
community during 2000 relative to July observations.

The percent abundance of Ceratopogonidae, Fossaria, and Physa were higher in
August 2000 relative to 1998 and these differences were statistically signiﬁcant;
Ceratopogonidae increased across years (Table 2.7). Pseudosuccinea (Gastropoda) and
Stagnicola (Gastropoda) were abundant and common in August 2000 respectively, yet
these taxa were not collected in 1998 or 1999. In fact, the gastropods collectively
represented 1.8% and 0.2% respectively in August 1998 and 1999. The percent
abundances of water mites, Tanypodinae (Diptera: Chironomidae), Agraylea
(Trichoptera: Hydroptilidae), Nectapsyche, and Hyallela were highest in August 1999.

The number of uncommon taxa increased in 2000 (39) relative to 1998 (26). The
most notable changes occurred in the insect orders Coleoptera, Diptera, and Hemiptera
during 2000. The increased richness in these insect orders was abrupt; no increase prior
to 2000 had occurred. Shannon’s diversity and J’ were similar between 1998 and 1999,

but increased in 2000; values were highest in August relative to other months (Table 2.7).

110

Invertebrate Community Composition in Late Summer
Only one new taxon (Coleoptera: Curculionidae) was collected in September

1998 relative to previous months; no new taxa were collected in late summer 2000. The
abundant and common taxa collected in August 1998 remained dominant in September;
Paraponyx (LepidOptera: Pyralidae) and Hyallela were also common (Table 2.8).
Ceratopogonidae, Chironomini, Caenis, F ossaria, Pseudosuccinea, and Physa also
remained abundant or common in late summer 2000 relative to August; the percent
abundance of Chironomini and Caenis increased in September. Tanytarsini and
Stagnicola were uncommon in late summer 2000. In contrast, the percent abundance of
Berosus (Coleoptera: Hydrophilidae), Enallagma, Gammarus, Hyallela, and Gyraulus
(Gastropoda: Planorbidae) increased; Enallagma and Hyallela were common (Table 2.8).
The percent abundance of Ceratopogonidae in September 2000 remained higher
than measured in 1998, though this difference was not signiﬁcant (Table 2.8).
Statistically signiﬁcant differences between years were not measured for Chironomini,
Caenis, and Paraponyx. The percent abundances of Tanytarsini, Oligochaeta,
Gammarus, and Dreissena were signiﬁcantly higher in 1998 relative to late summer
2000; the converse was true for Enallagma, Hyallela, Physa, and Gyraulus (Table 2.8).
Taxa richness was lower in September relative to August in both years, due in
part from reduced representation by the caddisﬂies (1998) and beetles (2000). Richness
was still greater in 2000 relative to 1998; a similar trend occurred for H’ and J ’ (Table

2.8).

Functional Feeding Group Composition: 1998-2000

111

 

 

 

 

 

 

2.958 a..." 5.8 .3. e536 2.25.6
32: Rd eeisﬁ 2.2.25 2383523
~38 nod --- 26:25.
23.8 NS 82:25
3N8 X; --- 32.3.2355
2:8 36 --- 0:253:35
--- --- uﬁoutuk 32—55%.—
-.. 2.2295.
82.53.33:
--. --- .533 32::
2 we 3... .33 n2 assasﬂ
98.8 N; G. .8 3o agcoﬁcﬁ
--- 8 ..8 93 8562855
aim—.8 Eu: 3...: and _:_Eo:2EU 32529.20
$23.8 :6 85.8 $.— 325393.980
5.8 MS. 5.23. 2.23.2.2 28:5
--.. --- 22.5230
9:33 35%..» 325.3
--- --- magnaaumesh
--- --- 35323
35.8 cm." --- nausea 32.2.3.6»:
--- --. m-§§~k
& _ .8 2.6 An _ .8 m _ .o 2:53: 32.33:
--- --- 2.28:2—
--- A38 .2. 325.3950
A88 86 --- 392:5 sag—usembgu 9838.90 88m:—
--- 43.8 36 «5.33.5»: «25.3.2.
.53 - Bow :2 - 2.9m 2.3-..5228 5...“... .595 2295:}:

 

3.038%»: 32:85 28 x02: 3683c Mcoccmsm 9 2252.50 . x. was . t ammo—Eat 83 modacwaou

m .28chme 8: 35332 2 $53» 5333 852:9“ €023 E mousse—0 Amodvq ....v 685:2.“ bﬁoﬁgﬁ no.“ 338%
203 8:3 805 Eco ”82:2:an 8890:35 338%“: of «0 SN $53 3 328052 85 823 magnum—Em cab 20m A: .I. 28
ooom was A: n 5 M32 copEoEom E 95:25 355 2: 60¢ 380:8 822 3553.35 no Amm ﬂ 8:855“ E023 :82 .w.m 035.

112

 

 

 

 

 

 

 

 

 

 

 

39.: mod --- 85838333:
.88 8...: 2.28... 8222.23
p . .8 ...... 82222 2.2.2.80
.28 mm... 82.32%
......8 2.. .5... «.2 2.3228 22......
--.. .1 3383.369 uat__omm< «tomca—
.¢~.: a... .28 2a 222:. 32.222:
.38 8.... ..mm... 8.... -. 2352.6 322225 2.2.2.2... 32.8322
.38 an... .58 2.2 5.2.82.0
22.8 2... 6.8 2... 822...: 3.2.825 «2.2.3.
.38 a... $28 3.. 2.8.5
8.8 a... a . .8 m . ... stameaz 82.82....
--- --- 3.5»ka
.28 a... 2.2.23:
.28 a... 832.? 32:22.2: £2.22;
.88 2.... 52.2.5.2. 0.2.2.2....
.58 .... 5.8 a." $.52... 32:25.55
808 S... as... 2.25.23. 52.25
.1 3.. nuhcﬁzaﬁb an—aﬁhhou «5022.3on
.28 2... 2.3.8 a..." 22.38....
.88 ...... .88 2... 2.3.8.. 2.2.2... 2322......
--- --- 828.2 2.29..—
--- --- 59.8332 32.8532
2 . .8 mm... 22.8.: $2.228...
.28 .2. 2:8 32:5
308 3.. 322.50
A38 3.. --- 322.83: 323E836: 283:5: 23m:—
2...~ - ....m 23. - ....m 2...-..5..=Eu 2.5.... .320 $29522...

 

.228. ...N 22¢

113

 

 

 

 

nu... me... n K.
3; 3.: n .h
mm mm H m
$08 nmd --- 53:21.. .r
..- A23 25 3.83:»: 32.38.52.—
--- --.. £393.33
.32: S." 5.8 ”no 33:6 3.385:
23$ 36 Am _.8 m _ .c 3.3% 32%.:
A _ ”.8 Q..— --.. £33qu 323:5»; «Ex—9:30
38 - :5 :3 - 2.0m 2.27:5:25 £55. Ezo a....«_u\..__.._\£._
.9583 ad wink

 

114

Collectors were the most abundant functional feeding group throughout the spring
and summer of 1998 (Figure 2.11A). The percent abundance of collectors peaked in mid-
summer (August), coincident with maximal richness and diversity; the overall range was
77.9-88.7%. Scrapers and piercers were common during early and mid-summer 1998 yet
uncommon in late spring and late summer; percent abundance reached 2.9% (i 0.6) in
mid-summer. The piercers were mainly represented by the hydroptilid (Agraylea,
Hydroptila, and Oxyethira) caddisﬂies (Table 2.2). The percent abundance of shredders
ranged from 1.8% Q 0.4) to 6.8% (i 1.0) in 1998 and mainly reﬂected the occurrence of
Nectopsyche and Paraponyx. Shredders decreased in percent abundance throughout the
spring and summer of 1998 (Figure 2.11A). Predators comprised between 6.6% (i 0.9)
and 16.2% (i 0.9) of the invertebrate community; water mites, gyrinid beetles
(Coleoptera: Gyrinidae), Atherix (Diptera: Athericidae), Ceratopogonidae, Empididae
(Diptera), Tanypodinae (Diptera: Chironomidae), Enallagma, and 0ecetis (Trichoptera:
Leptoceridae) were the major predator taxa in 1998.

Collectors also comprised the largest pr0portion of the invertebrate community
throughout 2000; ranging ﬁ'om 43.0% (i- 6.3) to 69.2% (i 11.0). The percent abundance
of collectors declined from late spring through mid-summer and remained below 50%
during late summer (Figure 2.1 18). The change mainly reﬂected the decline in percent
abundance of chironomids and oligochaetes, despite the increase in caenid mayﬂies
throughout the spring and summer. Scrapers plus piercers were abundant in 2000 and
comprised the second largest functional feeding group; ranging from 14.2% (i 5.2) to
37.9% (i 6.6). Percent abundance of scrapers plus piercers was greatest during mid-

summer and reﬂected the peak in snails (approximately 38% of the community). Piercers

115

a...

:6”

ed?

93

93

c.3—

 

.onQSm Sn poo—08m :2“ no.“ 33% ﬂ “inwa— =oEEOQ < .AmnoﬁuE 83 8:86 on 9 383mm
203 88. 82: com 88:82 boom 8:? 3583 03>» 8385 use mcomﬁom .Amv ooom 28 A3 «92 .«o honEoEom swab:
0:3. 96:6 338% 82:58:80 3550:35 Mo 30533“ E853 coEmOQEOo 95cm msmcoom Quote—Sm owﬁo>< .:.N oSmE

..anoEom thaw—E ban. 2:...

 

9.8.695 B
969.05 +
32.2% I
293.25 B
9.38.30 I

 

 

 

 

 

yea—anom .m=w=< b5. 05:.

. céw

 

:6:—

 

mound

116

were more common in 2000 relative to 1998, percent abundance ranged from 0.2% (i
0.2) to 5.0% (i 2.1) and corixids (Hemiptera: Corixidae), coleopterans, and caddisﬂies
were important groups. Shredders comprised the smallest proportion of the invertebrate
community in 2000 (Figure 2.113). The percent abundance of predators ranged from
13.4% (i 4.7) to 19.2% (j; 2.1). Ceratopogonidae, Dolichopodidae, and Enallagma were
the major predators in 2000.

Major shifts in functional feeding group composition occurred between 1998 and
2000 (Figure 2.11). An overall reduction in the percent abundance of collectors and
greater representation by scrapers plus piercers and predators was evident throughout the
spring and summer of 2000. These differences mainly reﬂected the increase in
coleopteran, dipteran, odonate, and gastropod richness and abundance. Additionally,
declines in the percent abundance of chironornids and oligochaetes during mid- to late
summer 2000 led to an overall reduction in collector representation (Figure 2.113). In
fact, functional feeding group composition in August 1999 illustrated a transition from a
collector-dominated community (21.4% reduction relative to August 1998) to one with
much greater representation by predators (24.4% i 3.8), yet snails (<0.3%) were
uncommon and the scraper plus piercers group was best represented by the latter (Figure
2.12).
Spatial Trends of Select Invertebrate Taxa I

A total of 21 taxa exhibited statistically signiﬁcant spatial afﬁnities; only ﬁve
were signiﬁcantly associated with the outer region of the wetland. The percent
abundance of Nectopsyche was higher in the outer region during all three years (Table

2.9). The spatial afﬁnity for Gammarus was only signiﬁcant in 1998 and 1999;

117

100.0

 

  
 
 
     

80.0 ‘ ~
- Collectors
I Shredders '
- Scrapers
60.0 ‘ + Piercers -

:1 Predators

Percent

40.0 -

20.0 ‘

 

 

0.0 ‘

1998 1999 2000

Figure 2.12. Average functional feeding group composition (percent abundance) of
invertebrate communities sampled in August 1998-2000. Scrapers and

piercers were summed since these taxa were assumed to have similar food resources
(see methods).

118

 

 

 

 

 

 

.52- an; 3a.: 82%.: 83235
«8...- £3- 83. 2.2. ”325:5 82.2.55 32.322
82.- 83. . assuage
3.....- $3. Sn..- 83 :55 3.3. ueuaameaez 32.83%.— 22.355.
23.... 83. 325:3... 8222.220 :33:
can 93— :2 25m 93— 33 2:555 £55— 98.5
meets—eteu 88. D @553éaa2 3:50

 

Amodqu 2:35:me 22$ 23 E 356568 cows—280 .33“ D €85:me

.23 8:2 085 bee com 88:3 :3QO $83 85.5. 8 com: 33 ﬂax—«ca sous—2.80 .33“ ugomﬁwmméo: com cozw 8a 82.?

a o: 680: 2a 829, a ”$3508 83 38 D 3533-:52 252.3598: 05 main vommommm 82> 80:80ch Amodvav Ewe—ham?
Quocmssm .935 856508 cows—oboe cemsom mcﬁcoamoboxe 98 £0: :2on AE owNéoC :93:— 35: 2: 8 2522

came A8 9: .od 53:. 8:5 2: 53> eonMoomme 82g 8595px E359 5:9: baa—mommawmm we: 85 85 82:90:35 .mN 033.

119

Gammarus was uncommon in 2000. The oligochaetes and Physa were more commonly
associated with the outer wetland region only in 2000, whereas Tanytarsini had
statistically higher percent abundance values in 1998. All but two of the correlation
coefﬁcients were signiﬁcant (Table 2.9); near linear declines in percent abundance
relative to distance from the wetland interface were noted for Tanytarsini (1998),
Oligochaeta (2000), and Gammarus (1999).

Sixteen taxa were statistically more common in the inner region relative to the
outer region of the study wetland (Table 2.10). In 1998, irmer region taxa included
Hydracarina, Gyrinus, Hyallella, Paraponyx, and Enallagma; the latter two taxa were
also statistically signiﬁcant in 1999. Caenis was also more commonly associated with
the inner wetland in 1999. Nearly 50% of the taxa exhibiting signiﬁcant spatial afﬁnities
were associated with the year 2000 data collection: Haliplus, Berosus, Pericoma,
Stratiomyidae, Callibaetis, Hirudinea, F ossaria, Pseudosuccinea, Stagnicola, and
Gyraulus. Caenis (1999), Paraponyx (1999), Hirudinea (2000), and F ossaria (2000) had
correlation coefﬁcients (>0.500) that were indicative of near linear increases in percent
abundance relative to distance from the wetland interface (Table 2.10). Only ﬁve
correlation coefﬁcients were not signiﬁcant.

DISCUSSION
Invertebrate Community Composition: 1998-2000

Studies of littoral wetlands in small, inland lakes (Dvorak and Best 1992; Dvorak
1996) and along the shorelines of the Great Lakes (Krecker 1939; Duffy et al. 1987;
French 1988; McLaughlin and Harris 1990) have demonstrated that oligochaetes,
dipterans, crustaceans (arnphipods and isopods), and water mites commonly dominate

(>2% of entire community) invertebrate community composition. Shifts in dominant

120

 

RS ”2:. 2.32.6 3.5.2.2..

 

 

 

 

 

 

 

 

 

3...: 32. 3823.5
can: c 3d ausuunudﬁaunk
wmcé Good eusaumﬁﬂ usEosE—id aeoacbmao
in... 83 2235 2.2.235 €2._._._E<
Rm... 83. 352:5 29233
E; 3:. 2.3 33. sswssam 8223280 52.25
82. «an... :55 83. assess; 26:23 2332.3
as... 83. e55 8258
$3 :3. Essie 82.2: 2383523
an... ”5.0 26:52.25
:3 _ :3 assess 8285?. 83a:—
32. 82. ”52% 32.2.51:
32. 8: $33: 32.5.“:
5mm... mood b.3553 325:0 28.3290
52. Sod 2:822:
=ch 33 :2 .53 $3 mmﬂ 2.22.5 E55— 98.5
33.29.30 38“ D 55535.3: 2.56

 

.Amodvﬁ camocmcwﬁ 22> Eon E 356508 cots—280 558 D 38$:me

.EB 85 39: bao 89 2:033 38% 33% 8:23 9 mom: mm? mum—max corn—280 .33“ Emomimﬁéoc How Sim 8m 82?

a o: 680: 08 $23 a Amnoﬁoe 00$ 68 D 5523-582 oEoEﬁmméoz 2: maﬁa 3383 203 moocouotmc Amodvg Encamcwmm
3.853% .93: 856580 cog—0:8 seamen wamucommotoo was 805 :2on A8 9:63 538 .530 05 3 0338 :2on

AE ommécz 53.: 5:5 2: 53, 3368.0.“ 82? 3.823% :3ch 8.32 3:805:me we; 35 8:2 oﬁBotoE— .EN 2an

121

taxa among wetlands as a function of plant community type, architecture, and eco-region
are common. For example, French (1988) observed the invertebrate community
associated with a submergent wetland in Lake St. Clair to be dominated by Amphipods,
whereas McLaughlin and Harris (1990) found the chironomids to be quite abundant and
diverse within a Green Bay, Lake Michigan wetland. Studies conducted in the coastal
wetlands of Saginaw Bay, Lake Huron indicated similar dominant invertebrate groups,
but also included caenid mayﬂies (Brady 1992; Cardinale et al. 1998; Burton et al. 1999;
Stricker et al. 2001; Burton et al. 2002). Additionally, four sub-tribes of the chironomids
were generally abundant and included Chironomini, Orthocladiinae, Tanypodinae, and
Tanytarsini.

The same dominant invertebrate groups observed in previous investigations of
invertebrate community composition in Saginaw Bay coastal wetlands were collected
throughout this study (Table 2.11). Exceptions included very low representation by the
orthoclads and tanypods, and the near absence of isopods (Crustacea) throughout the
three-year period. In 1998, abundant and common taxa also included the caddisﬂy
Necropsyche (Leptoceridae), the lepidopteran Paraponyx, the damselﬂy Enallagma, and
the zebra mussel (Dreissena). Sampling intensity was greatly reduced in 1999, but these
data were very useful in interpreting community shiﬁs relative to lake level declines from
above average (1998) to approximately 0.50 m below average (2000). In fact, a variety
of taxa were similarly abundant and common in 1998 and 1999, exceptions included a
decline in Paraponyx, oligochaetes, and zebra mussels (Table 2.11). Enallagma
accounted for approximately 13% of the community in 1999 and the caddisﬂy Agraylea

(Hydroptilidae) replaced Nectapsyche as lake stage declined from levels observed in

122

 

 

 

w— a 2
x 9.5353 32:33..—
x Baku 3%.: 32m»...—
x 53335.
x ausooamenzuum
x 3.33% 323593 2.2.8.30
x 33.9.3.5 3.25;:

x x x 33.3. 2.2.2.3.
x x x 3355.5 3335530 3.5.—Ens... 33:33.32
x x 322.825

x x 3.33832 32.8895

x 3.3.3.. 32:39.6»: EoESEﬁ.
X x x 353.535 323.3330 33:25

x 3.333% 32.3..3— ...—8.3234

x 325:5 ESE—=0:
x x x «.339 32:30 «BEEoEosam
x eEeotek 32:23.3.—
.. 828.232.
x x x 2.23.59
x x x 2.3.3820 3252.850
x 325338950 ...—829
x 3.332%
x naueuun 32.2396»: ...—8:380 53»...—

x X 3.33.3: «2.53.3.

.53 23. :3 £5“... .55 :29 320.522..

 

...—30:. 28 .32 3 macaw .8888 8%..“ .3253 33 :88. comm .8333. ..o .8883 was macaw 05
3:3. 53:. 336 2.. Bo... 380:3 83. 2338.35 Aﬁcmv ..xcmAV .8888 we... AﬁemAv 33.3.53 ..o bmﬁea .355. A ..N 033...

123

1998. Chironomini, Tanytarsini, Caenis, Enallagma, and Oligochaeta remained abundant
and common taxa in 2000. However, the amphipods, Hydracarina, and the caddisﬂies
were generally uncommon and were replaced by Ceratopogonidae and the gastropods,
F ossaria, Psuedosuccinea, Stagnicola, Physa, and Gyraulus (Table 2.11). The
invertebrate community during 2000 had much greater representation in the insect orders
Coleoptera, Diptera, and Hemiptera

The shifts among abundant and common taxa from 1998 to 2000 were
inconsistent with previous studies of Saginaw Bay coastal wetlands. Some of the earliest
work indicated that damselﬂies and caddisﬂies were not numerically important
components of the invertebrate community (Brady 1992). Similar observations regarding
damselﬂies have been made in ensuing studies (Cardinale 1996; Cardinale et al. 1998;
Burton et al. 2002), yet other investigations have noted odonates to collectively
contribute up to 3.7% of the community within the inner region of a Saginaw Bay coastal
wetland (Burton et al. 1999). Additionally, samples collected during 1994 indicated that
caddisﬂies, principally members of the families Hydroptilidae and Leptoceridae
accounted for up to 6.6% of invertebrate assemblages within speciﬁc regions of the
wetland (Cardinale et al. 1998; Burton et al. 2002). Had these data been analyzed in a
collective fashion (treating the wetland as a single unit), the contribution by both groups
to overall invertebrate community composition would have been much less. However,
these data do support contentions that a variety of invertebrate taxa have distinct spatial
distributions in these types of systems (Cardinale 1996; Cardinale et al. 1998; Gathman et

al. 1999; Burton et al. 2002).

124

In previous studies of Saginaw Bay coastal marshes, the arnphipods comprised a
signiﬁcant portion of the invertebrate community (Brady 1992; Burton et al. 2002).
However, temporal shifts in dominance between Gammarus and Hyallela occurred both
in and between years. For instance, Brady (1992) only collected Gammarus. Subsequent
studies in the same wetland reported the presence of both taxa. Gammarus was reported
to be more common in early summer, while the relative abundance of Hyallela increased
by mid-summer (Burton et al. 2002). A similar pattern was observed in this study,
however the percent abundance of both arnphipods declined across years and collectively
accounted for only 3.0% of the entire community in 2000, although these taxa did
comprise >10% of the community in late summer 2000 (Table 2.4 and 2.8). The isopod,
Caecidotea, was uncommon and was not collected in 1999. In previous studies, this
taxon had only been collected from the near shore regions of the study wetland (Brady
1992; Burton et al. 2002).

In terms of gastropods, Brady (1992) observed high densities on two occasions,
but this was restricted to the genera Gyraulus and Physa during mid-summer (August)
and late summer (September). Gastropods were not a numerically important component
of the community, despite density estimates approaching 700 m'z. Similarly, subsequent
studies conﬁrmed the occurrence of planorbid and physid snails in the Saginaw Bay
coastal wetland, but contributions to the invertebrate community were generally restricted
to less than 2% (Cardinale 1996; Cardinale et al. 1998; Burton et al. 2002). Both genera
were observed in 1998; however, percent abundance was collectively less than 1% (Table
2.4). Brady (1992) also observed F ossaria, Pseudosuccinea, and Stagnicola, but

abundance was low, particularly for the latter two genera. In contrast, lymnaeid snails

125

accounted for over 18% of the community in 2000; Physidae and Planorbidae contributed
an additional 5%.

A distinct increase in taxa richness occurred in conjunction with declines in lake
level. Previous studies suggested that richness varied among coastal wetlands and was
largely dependent upon sample collection procedures, timing and intensity of sampling,
and degree of taxonomic resolution. Burton et al. (1999) sampled invertebrates in the
dominant plant communities of six coastal wetlands and found a range from 13-30 and
11-23 taxa in northern Lake Huron and Saginaw Bay wetlands respectively. Gathman et
al. (1999) synthesized much of the Saginaw Bay work and reported 69 taxa collected
from the littoral Scirpus wetland. This number was based largely on data from Brady
(1992) and Cardinale (1996) and taxonomic resolution was often at the species level.
Throughout this three-year investigation, a total of 60 taxa were collected. This was high
given the fact that taxonomic resolution was generally at the family and genus level
(Table 2.3). Richness was lowest in 1999 (28) and greatest in 2000 (54), and was within
the range reported in previous studies. Diversity (H’) followed the same pattern as
richness and was highest in 2000. The seasonal range of H ’ over the three-year study
was 0.72-1.20, which was similar to values reported by Burton et al. (1999), although
their estimate included invertebrate data from adjacent wet meadow and Typha
communities. The seasonal range in J’ (0.51-0.75) was slightly lower than that reported
by Burton et al. (1999) and probably reﬂects their inclusion of additional plant zones
(Tables 2.5-2.8).

The pronounced increase in richness observed in 2000 was largely a reﬂection of

greater representation within the class Gastropoda and insect orders Coleoptera, Diptera,

126

and Hemiptera (Table 2.3). Interestingly, many of the new taxa collected in 2000 had not
been collected in previous studies. Based on the summary by Gathman et al. (1999), a
total of 17 new families were observed in this study relative to previous descriptions by
Brady (1992) and Cardinale (1996), and this corresponded to a total of 21 taxa (Table
2.12). The four additional taxa were members of the coleopteran families Haliplidae and
Hydrophilidae. In contrast, one semi-aquatic insect order (Hymenoptera) and nine
families collected in those studies were not collected during this investigation (Table
2.12). However, taxa within all of those groups were quite rare (Gathman et al. 1999),
whereas this was not always the case with regard to the new observations made in this
study. For example, the families Hydrophilidae (Coleoptera), Dolichopodidae (Diptera),
Psychodidae (Diptera), and Aeschnidae (Odonata) accounted for 8.4% of the invertebrate
community in 2000 (Table 2.4). These data suggest that the increase in richness was not
solely a reﬂection of previously unobserved natural variation, rather changes in the
abiotic environment mediated through water level ﬂuctuations likely inﬂuenced shifts in
the invertebrate community.

Classification of invertebrate taxa relative to lake stage (i.e. year collected)
indicated that 55% of the invertebrates collected during the three-year investigation did
not exhibit a strong afﬁnity for any particular water level (Table 2.13). Four taxa
(Diptera: Atherix, Diptera: Empididae, Trichoptera: Oxyethira, Gastropoda: Pleurocera)
were strictly collected during the higher water year (1998), whereas only one taxon
(Hemiptera: Neoplea) was collected at intermediate water levels (I 999). The limpets
(Gastropoda: Ancylidae) were collected in 1998 and 1999; Tabanidae (Diptera) and

Belostoma (Hemiptera: Belostomatidae) were only collected in 1999 and 2000. A total of

127

Table 2.12. List of newly collected invertebrate taxa from the study marsh throughout the
spring and summer of 1998-2000 and taxa not collected in this study, but recorded in

previous investigations.

 

Class Order

Family

Collected in
This Study

Collected in
Previous Studies"

 

lnsecta Coleoptera

Diptera

Hemiptera

Hymenoptera
Lepidoptera
Megaloptera

Odonata

Trichoptera
Bivalvia
Gastropoda

Chrysomeliidae
Curculionidae
Elmidae
Haliplidae
Hydrophilidae
Staphylinidae
Athericidae
Dixidae
Dolichopodidae
Empididae
Ephydridae
Muscidae
Psychodidae
Sciomyzidae
Tipulidae
Belostomatidae
Hebridae
Pleidae
Veliidae

N octuidae
Corydalidae
Aeschnidae
Gomphidae
Phryganeidae
Sphaeriidae

 

Pleuroceridae

{XXXXI XXIXX

IXIXXXXI

l><><l

XXI

 

X

l><l

IXXII

IXXXIXI

l><><l

 

*Based on review by Gathman et a1. (1999); data from Brady (1992), Burton (unpublished), and Cardinale

(1996).

128

 

--- --- --- x 00......80

 

 

 

 

 

129

 

k k --- --- 255.30% 002353.20: 0.02—.50:

--- --- --- x .330 03.500

--- --.. x ......SSQD 0.2.0.... 0.0.3.0522...

x ...—.....m...

X X --- --- 03.5%..

X .-- --- --- ossiﬁcaabm

x --- --- --- 00.0.3525

x --- --- --- Seaté 002.0330..—

x 322......

x 328.223:

x --- --- .... 5.3.5 002...:

--- --- --- x LS...§..=§

--- --- --- x 00:53.33.“

--- --- x «35.3.0230

.... --- --- x {52.30.30 020.580....0

--- --- --- X 330% 03.203.33.00

--- --- X . --- 5.05! 0.3.2.2.}. ...—03.n— I-

x --- --- 22.52.00

--- --- --- x .3593 002530

x --- --- _ --- 0:50:30...

x --- .-- --- 03.53%

x --- --- 038.0% 0a.........0.....=

x --- --- ....Aueﬁm

x 3...... 8......3:

--- --- x 02:00.39

--- --- --- x 0a...:c..:0.=u

x --- ...- --- 502:5 00.....0ﬁcmrgu ...—02.00.00 2000....

--- --- x «5.30.3: 3.5.092.
I:0.mu.3< 025.0:— 0wa.3< .2050 0.....-..=m\m=:00 ....Eam .020 8595...»...—

320m .0...— 33¢.

 

.230. 00.0. .o. .03 3.82.00. 5000
main... 3.00:8 0.03 .0.0:0w mm .008: 88... 2.3.80.8. 88-32 0.03 2.. o. .0.0. 0.30. .23» $0.30 30.0. new 6.0.38.8...
69230 32.... .230. 00.0. 0. 3.8.0. A--. 00:030. .o On. 00:00.0... :0 .002... 0x0. 0.0.98.3... ..o 5.32.2320 .m..m 0...“...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.N v m mm m

--- --- --- X 2.0—.005...

--- --- X .-- 0.000.003 002.000. 00...

X --- --- 0:00.000...

--- --- --- X 00.00.00 0020.000...

--- .-- --- X 0.5... 0020»...—

X --- --- --- 0.00.5005

X --- --- --- 000.000.0000....

--- --- --- X 00.00.00... 002005.03

--- X X --- 002.0004. 000005000

-..- --- --- X 00.0...000mm

X 0008.05 0.23...

--- --- X 00.00.0000 002.003. 000.000.

x 0.00%.. 82.20.:

--- --- ..-- X 00.000000 002.05000 0000.00.00. 0005000202

--- --- --- X 0.00000N..0

--- --- --. X 000.00.... 0.0000095 02.000...

--- --- --- X 00.0000 m

--- --- --- X 0:00.033. 002.0002?»— 1

--- --- X --- 0.3.0.000

--- --- --- X 0:30.00...

--- --- --- X 00.00.00... 002.300.??— 0.0.0002...

X --- --- 50.0%....0 0022:051—

--- --- --- X 0.000000% 002002000000

--- --- --.. X 800... 00200000.. 0.00000

X --- --- --- «000.3000 002.0300 0.02—0.09.).

--- X 390500..

--- --- --- X 0......000V 002.03.. 0.02.0200.

--- X --- --- 0020.02 0020...

X --- --- --- 0.000032 002000307.

--- --- .-- X 0000000: 002:30002

X --- --- ..-- 0.....00 002..00 0.0.0.50: 0.0000.
£0.30. 0.0.000. Mug—0.5~ .0.0000 0.....-00m.00000 £50.. .020 0.00—0.50.»...—

B0.0m 00.0. 30.5.

 

.0008. m _ .0 as...

19 different taxa were collected when water depth within the wetland was lowest (2000).
Assuming that the presence or absence of a given taxon represents some degree of water
level aﬁ'mity, these data also suggest that large-scale changes in coastal wetland
hydrology had a profound effect on invertebrate community composition.

The stage of Lake Huron was the major distinction between trends in invertebrate
community composition during this study and previous investigations in Saginaw Bay
wetlands. Lake levels coinciding with the studies of Brady (1992), Cardinale (1996),
Stricker et al. (2001), and Burton et al. (2002) were intermediate between 1998 and 1999
levels, ranging from -O.31 m to +0.17 m relative to the long-term mean (Figure 2.13).
Burton et a1. (2002), however, included some data collected during 1998 and 1999, and
similar data from 1997 and 1998 are presented in Burton et al. (1999). The latter study
coincides with the ﬁfth highest lake level (0.44 m above long-term mean) observed in the
last 48 years and represents the only published record of invertebrate community
composition coinciding with lake stage higher than that observed during this study
(1998). Prior to this study, however, there were no points of reference for evaluating
changes in the invertebrate community in response to very low lake levels (below 1999
stage). Therefore, invertebrate data collected in 2000 represent new observations with
regard to coastal wetland invertebrate ecology and provide a unique perspective on
abiotic factors that influence community composition. In fact, lake stage during 2000
was 0.56 m below the 48-year average (Figure 2.13). Water levels below the year 2000
average had only been recorded on four other occasions in the 48-year record (1963-5,

2001).

Water Level Fluctuations and Changes in Habitat

131

.8:qu 8x3 320 3:39:35 9 mucoamohoo n59 Accomawma:

3:5 £5 mo 328 2: 8:365 265 ”83: .3 3 :88m ... ES .288 .3 3 ~83th .63: «3:630 .388 .3 3 cornm—

xmao 3 >35 +V 3:233 beam 2: E 382980 322%:ng womanéﬁnotgﬁ 8 3:88:00 88—32 .3. 36 H ends: :38
gargo— mEOmoEE oz: 2.0m .92 JEZBmmm :oomémm: coSE 8—3 ism Bacmmmm 8m 282 093 a?» EmmoéSm .m_.~ Sam:—

33 mag can 32 .53 mha 3.3 mean 3&— mmm—

 

 

 

(9861 ([15)! 9m) was

 

 

 

 

F . ad:

132

The relatively rapid decline in lake level observed during this study brought about
a variety of changes in the physical and chemical environment of the wetland (Figure
2.13). As lake stage declined from above average levels in 1998 to below average levels
in 2000, mean water depth in the wetland decreased by 0.71 m (Figure 2.4). Declines in
water depth during the three-year period translated to an approximate 62-74% reduction
in habitat (vertical dimension) associated with S. pungens each successive year. By year
2000, Scirpus-related habitat accounted for approximately 10% of that which was
available for colonization in 1998.

The elevation difference between wetland sediments and lake surface was
exaggerated in 2000 relative to previous years when water levels were higher. Though
seiche activity was not quantiﬁed, these ﬂuctuations are very common to the Great Lakes
and were likely prevalent throughout the study (Bedford 1992; Wetzel 2001). As water
depth in the wetland fell to levels observed in 2000 (mean = 0.09 m), the effect of seiche-
related water level ﬂuctuations increased in importance. A water level change of 10-20
cm is a conservative and often-cited estimate of seiche-related ﬂuctuations (Burton 1985;
Brady 1992; Gathman et al. 1999; Keough et al. 1999). Super-imposing this on lake
stage in 2000 suggests a very dynamic hydrologic environment in the wetland, and
personal observations supported periodic dewatering. A conservative : 0.10 m
ﬂuctuation in lake level would have resulted in no standing water within certain regions
of the wetland (Figure 2.14). Whereas, the effect of seiches on wetland water depth and
habitat would have been less important during years in which water levels were near or
above average (Figure 2.14). Seiche-mediated water level ﬂuctuations probably created a

wetting and drying cycle of short periodicity in the near shore and inner wetland regions,

133

 

 

 

 

 

 

 

 

 

1.40 . : 1 ¢ . 4- . = . r . : . r .
- + 1998 .
1.20 - —v— 1999 .
. + 2000 .
1.00 - k
A 0.80 - -
E . .
v
5 0.60 « .
a.
0 ‘ .
a 0.40 - -
0.20 - .
0.00 . ........................................................................................... _
. A .
-0020 r A I ‘ U ‘ I ‘ I L I ‘ l A 1 )
0 40 80 120 160 200 240 280
1.20 t : : . t t i t : :
-O— 1998
—v— 1999
1.00 2000
0.80 - .
J
A
g 0.60 - -
5
a.
0 - ..
Q 0.40
0.20 -
0.00 .. ........
-0.20 . . .
0 40 80 120 160 200 240 280
DFWI (m)

Figure 2.14. Mean (: SE) spring/summer (J une-September) water depth proﬁles across
the study wetland during the years of study (1998-2000) as inﬂuenced by a hypothetical
0.10m seiche-induced increase (A) and decrease (B) in water level. DFWI corresponds
to distance from wetland interface. The zero depth line (dotted) represents the sediment
surface within the marsh. Observations with a negative depth indicate the lack of
standing surface water.

134

perhaps resulting in a hydrologic regime that mimicked temporary wetlands (Williams
1997).

Declines in lake stage also resulted in reduced hydrologic connectivity between
lake and wetland (Figure 2.6) and a general decline in turbidity (Figure 2.10). The latter
was largely due to the reduction in wave-induced sediment re-suspension (Wetzel 2001).
The combined effect of reduced turbidity and lake/wetland exchange resulted in an
overall decline in seston transport into the wetland. However, a major shiﬁ from seston
and epiphytic algae (see Cardinale et a1. 1997) to benthic algae occurred as turbidity
decreased (Appendix A). The sediments were a rich source of nutrients (see chapter 1)
and likely contributed to the rapid colonization and growth of dense ﬁlamentous algal
mats (personal observation). Hydrochemical changes in the wetland also occurred in
conjunction with declines in lake stage (see chapter 1), mainly a result of increased
sediment/water interactions.

Water Level Fluctuations and Changes in Invertebrate Community Composition

Several studies in the exposed coastal wetlands of the Great Lakes have
recognized the continuum of lake-wetland exchange as an environmental gradient that
inﬂuences algal production and ﬁsh and invertebrate community composition (Cardinale
1996; Cardinale et al. 1997; Cardinale et al. 1998; Gathman et al. 1999). In fact, exposed
wetlands that are subject to high wave energy actually represent a gradation from a lotic-
like environment to true lentic or wetland habitat (Barton and Hynes 1978). Burton et al.
(2002) expanded these observations into a conceptual model that integrates aspects of
hydrology and habitat into general predictions regarding diversity and a variety of other

community attributes. The authors used empirical data to support the contention that

135

diversity is greatest at intermediate levels of disturbance (deﬁned as exposure to wave
energy). Though the foundations for these ideas stem from general ecological theory (see
for example Connell 1961 and Peckarsky 1983), the extension of these principles to
coastal wetlands provides a strong basis for understanding the ecology of these systems.
However, no studies of coastal wetlands have evaluated the disturbance effect induced
through inter-annual water level ﬂuctuations. In many lotic and lentic environments,
water level ﬂuctuations are considered a form of disturbance that provides an important
structuring force on plant (Keddy and Reznicek 1985; Keough et al. 1999) and animal
communities (Connell 1961; Peckarsky 1983).

Changes in invertebrate community composition between 1998 and 2000
supported the contention that the decline in water depth resulted in a variety of habitat-
related shifts. Most invertebrate taxa exhibited a large amount of variability with regard
to seasonal peaks in percent abundance and afﬁnity to particular regions of the wetland.
However, 21 different taxa did exhibit statistically signiﬁcant spatial afﬁnities and
provided insightful correlative evidence for shifts in habitat and habitat preferences.

Several of the taxa that were more commonly associated with the outer region of
the wetland in 1998 exhibited random distributions in 2000 (Table 2.9). For example,
Tanytarsini is a collector-ﬁlterer that is often positively correlated with seston
concentration (Dvorak 1996; Cardinale et a1. 1997). As wetland water depth and
turbidity decreased, Tanytarsini exhibited no spatial afﬁnity and in fact declined in
percent abundance. In fact, the reduction in wetland/lake exchange and colonizable
Scimus habitat likely led a variety of taxa that were more commonly associated with the

epiphytic environment, such as the chironomids and caddisﬂies, to the benthic

136

environment (Cardinale et al. 1996; Burton et al. 2002). In contrast, the caddisﬂy
Nectopsyche exhibited a habitat preference for the outer region of the wetland during all
three years, which probably reﬂected only a subtle impact on food resources due to mode
of feeding (shredder).

Taxa that were statistically more common in the inner wetland region were more
numerous (Table 2.10). Six taxa were spatially signiﬁcant in 1998 and 1999, but not in
2000. The distribution of Hydracarina was consistent with observations made by others.
For example, Davids et al. (1994) documented high densities (up to 1000 m'z) in shallow
littoral areas. Cardinale (1996) suggested that weaker swimming taxa, such as the water
mites, were excluded from the outer wave swept region of the coastal wetland. Changes
in the distributions of these six invertebrate taxa suggested that the loss of hydrologic
connectivity between lake and wetland resulted in reduced habitat preferences. Low
water levels and minimal hydrologic exchange had a homogenizing effect on overall
wetland habitat. However, 10 taxa were statistically more common within the inner
wetland region in 2000. It is interesting to note that most of these invertebrates,
particularly the coleopterans, dipterans, and gastropods, are taxa typical of temporary
wetland habitats (Williams .1997). Seiche-related water level ﬂuctuations that
periodically dewatered the near shore area of the wetland in 2000 probably inﬂuenced the
inner region as well (Figure 2.14).

Despite the compression of habitat that resulted from declines in lake level, the
plant community associated with the benthic habitat became more diverse, complexity
increased, and a new food resource (benthic algae) became available (personal

observation; Appendix A). A variety of taxa capable of exploiting ﬁlamentous benthic

137

algae increased in percent abundance. For example, the piercing caddisﬂy Agraylea
increased in percent abundance in 1999 (Table 2.7), but perhaps more dramatic was the
overall increase in scrapers in 2000 (Figure 2.11).

A larger number of taxa (19) were only collected during the low water year,
whereas comparatively fewer taxa (4) were only collected during the higher water year
(Table 2.13). This ﬁrrther indicated that the increase in new taxa relative to lake level
declines was related to changes in the abiotic environment. However, changes in the
strength of biotic interactions could not be ruled out. For instance, the increase in
damselﬂies in 1999 and 2000 suggested that predation pressure increased in response to
lake level declines. Changes in overall feeding guild structure also supported this shift;
the percent abundance of collectors decreased across years (Figure 2.15). The percent
abundance of predators may have also increased as a result of habitat compression and
the concentrating effect on prey species; this was particularly true for lie-in-wait
predators such as the odonates. Additionally, the reduction in water column likely
increased susceptibility to predation for swimming taxa, such as the water mites. Fish
exclusion has also been demonstrated to inﬂuence the abundance of invertebrate
predators as well and may have contributed to the increased percent abundance of
odonates (Crowder and Cooper 1982; Wellbom et al. 1996). Perhaps the lack of
piscivorous predators provided the necessary release for the signiﬁcant increase in snails
during 2000 (Heino 2000). Alternatively, strong competitive interactions have been
demonstrated between chironomids and snails, and the general decline of chironomids

relative to snails lent support to this possibility (Gresens 1995).

138

100.0

 

 

 

  
 
 
     

80.0 - -
- Collectors
E Shredders ’
- Scrapers

60.0 ‘ + Piercers '

C] Predators

Percent

40.0 ‘

20.0 -

 

 

0.0 ‘

1998 1999 2000

Figure 2.15. Average ﬁmctional feeding group composition (percent abundance) of
invertebrate communities during the years of study (1998-2000). Scrapers and
piercers were summed since these taxa were assumed to have similar food resources
(see methods).

139

An increase in S, H’, and J’ occurred in conjunction with declines in lake level
(Table 2.3). Many of the new taxa were restricted to the insect orders Coleoptera,
Diptera, and Hemiptera, and most of these invertebrates are commonly associated with
temporary wetlands (Williams 1997). A variety of studies have documented the
occurrence of these taxa in the inland palustrine wetlands adjacent to Saginaw Bay
(Gathman et al. 1999; Stricker et al. 2001; Burton et al. 2002). These wetlands are not in
direct connection with the lake, but water levels are likely controlled by lake stage. The
similarity among these wetland types (year 2000) suggests that the reduction in pelagic-
wetland exchange led to the transition from an exposed coastal wetland to a more
palustrine-like wetland. The adjacent inland coastal wetlands may have represented an
important source population, resulting in the immigration of the newly collected taxa to
the study wetland during 2000.

Immigration from adjacent palustrine wetlands was likely not the only reason for
increased diversity and richness in 2000. Rather, shifts in invertebrate community
composition probably resulted from the combination of changes in the wetland hydrology
(depth, permanence, etc.), the introduction of a new food resource (ﬁlamentous benthic
algae), exclusion of piscivorous predators, and changes in the intensity of biotic
interactions (competition and predation). Partitioning the effects of water level
ﬂuctuations on the above changes in wetland habitat and invertebrate community
composition was beyond the scope of this study and would require experimentation.
However, the correlative evidence presented does support the hypothesis of Gathman et
al. (1999) and Burton et al. (2002) that the habitat continuum spans a gradient from a

hydrologically harsh environment to one that is more benign (near shore region). Yet,

140

given the changes in lake level observed in this study it seems logical that hydrologic
disturbances induced by long- and short-term water level ﬂuctuations represent an
important structuring force and deserve particular attention in the study of coastal

wetland invertebrate communities.

14]

CONCLUSIONS

The results presented in the previous chapters suggest that water level ﬂuctuations
are an important structuring force that inﬂuences surface water chemistry and
invertebrate community composition in Great Lakes coastal wetlands. The 1.04 m
decline in lake level that occurred between 1997 and 2000 represented the second fastest
rate of change in the last 100 years. Climate change scenarios proposed for the Great
Lakes basin suggest an increase in the frequency and duration of below average lake
levels (Kling et al. 2003). The effects of such changes on the Great Lakes remain
unknown, but changes in the areal extent, habitat diversity, and water quality of coastal
wetlands have been proposed (Mortsch 1998). Further, loss or alteration of wetland
habitat may negatively impact ﬁsh and waterfowl production throughout the Great Lakes
ecosystem (Kling et al. 2003).

In chapter 1, I presented evidence that declines in lake level resulted in an
increase in the total dissolved ion content and a change in the overall ionic composition
of wetland surface water. The changes were signiﬁcant between years for all of the
major anions and cations. Hydrologic and chemical data supported the contention that
changes in surface water chemistry resulted primarily from increased surface
water/ sediment interactions. As mean water depth in the study wetland declined from
1.06 m in 1997 to 0.09 m in 2000, the water table became exaggerated and short-term
water level ﬂuctuations coupled surface and interstitial water. The reduction in
lake/wetland mixing resulted in shifts in overall wetland hydrology. For instance, surface
waves penetrated only short distances into the wetland in 2000 relative to 1999. This also

suggested that river discharge likely did not contribute to the increase in total dissolved

142

ion content relative to lake level declines. However, it was clear that agricultural drains
discharged into the wetland complex and, during periods of below average lake levels,
these drains effectively bisected the wetland and discharged directly into the bay. The
importance of drain discharge on Saginaw Bay water quality was clear from the ionic
content of this source water (high SO44, N03“, and NH4+ concentrations). The
implications of these observations suggest that coastal wetlands, which traditionally have
been viewed as a nutrient sinks, may act as a nutrient sources during periods
corresponding to below average lake levels.

In chapter 2, I demonstrated that hydrologic disturbance mediated through long-
and short-term water level ﬂuctuations signiﬁcantly altered habitat in the study wetland.
Changes induced by declines in lake level included a 90% reduction in colonizable
habitat, decreased lake/wetland exchange, near loss of particulate matter inputs, and
greater variability with regard to water permanence. Short- and long-term water level
ﬂuctuations acted in a synergistic fashion that led to a transition from an exposed coastal
wetland to one much more similar to a palustrine-like temporary wetland. These changes
inﬂuenced invertebrate community composition and attributes such as richness, diversity,
and evenness. A total of 60 different taxa were collected during the investigation, but 21
were only collected during low lake levels (2000). The invertebrate community
inhabiting the study wetland has been intensively sampled for the last 10 or so years, yet
a total of 17 new taxa were collected during this study. Fifteen of these taxa were
collected in 2000 when wetland water depth was lowest, and all were taxa commonly
found in temporary wetlands. Alteration of the food web was implied by way of major

shifts in feeding guild structure with increases in the percent abundance of scrapers and

143

predators. Increases in predators suggested biotic interactions increased in importance as
lake stage declined. Spatial afﬁnities of select taxa appeared to fade in relation to
reduced hydrologic connectivity between wetland and lake. Several potential causes
were discussed, but resolving abiotic and biotic interactions was beyond the scope of
these data. Experimentation on the relative importance of these interactions is warranted

if we are to expand the current knowledge of coastal wetland invertebrate ecology.

144

APPENDIX A

145

Appendix A
NUTRIENT LIMITATION PATTERNS 1N A GREAT LAKES COASTAL WETLAND

INTRODUCTION
I designed an experiment to test the hypotheses that 1) nutrient limitation

constrained periphyton growth in the study wetland, and 2) nutrient limitation was
spatially dependent and related to the extent of lake/wetland mixing. A nutrient diffusing
substratum (N DS) technique was employed to test these hypotheses. The experiment was
replicated at three locations in the study wetland to evaluate spatial dependencies. The
dissolution of alabaster substrata (AS) was used to quantify the extent of lake/wetland
mixing. Results were inconclusive because several problems arose during the
experiments. First, water depth in the study wetland was low during the 1999
experiments and also ﬂuctuated daily (seiches). This had the effect of exposing an
unknown number of NDS replicates to periodic wetting and drying. Second, the NDS
design was changed in 2000 to accommodate even lower water levels. The new design
differed in many ways from the original design (see below) and diffusion rates may have
differed; tests were not performed. Additionally, NDS were in contact with the wetland
sediments in 2000 and subject to overgrowth by ﬁlamentous benthic algae. These data
are included as an appendix to make available the experimental design and present the
interesting results that were noted.
METHODS
Study Wetland

The wetland complex selected for study was located along the southeastern shore
of Saginaw Bay, Lake Huron, U.S.A. (Figure 3.1). The portion of the wetland studied

encompassed only a small area of the total complex that extended approximately from the

146

 

 

T

‘l

‘.

 

:"qrgr
‘3r ' -

N

 

inner Q
Saginaw Bay 4. I ' é
Saginaw R. if) \m
\ ill ,
. a _.I l
in. J

m

l

1": 3 tudy Wetland

7" (5::
" ‘ I uanicassee R.

 

 

 

lGIometets

 

D 5 10 20 30

Figure '3. 1. Location of the study wetland (denoted by a star) in Saginaw Bay, Lake
Huron. The Quanicassee and Saginaw Rivers are referenced.

147

Quanicassee River (Tuscola County) to the Sand Point/Wildfowl Bay area (Huron
County). The study area was located adjacent to Vanderbilt Park, Tuscola County,
Michigan (43° 37’N 83° 38’W). Predominant winds were out of the northwest, and the
wetland was unprotected from wind and wave exposure with a maximum fetch of 30 or
more km (Suzuki et al. 1995).

The emergent plant community of the study wetland consisted of a nearly mono-
dominant stand of three-square bulrush (Scirpus pungens Vahl) that extended
approximately 300 m shoreward from the wetland interface (the interface between the
outer edge of the emergent zone and open bay water). Less dominant species of bulrush,
S. acutus Muhl. and S. validus Vahl, were interspersed among S. pungens, primarily
within the inner (160-280 m shoreward of the wetland interface) and near shore (300-500
In shoreward of the wetland interface) regions of the wetland. A large cattail (T ypha
angustifolia L.) complex bordered the northern edge of the study area adjacent to the
inner wetland region. At the onset of this study, vegetation within the near shore region
of the wetland (3 00—500 m from open water) included S. pungens, arrowhead (Sagittaria
sp.), and a variety of submergent species (see Batterson et al. 1991). As lake stage
declined, the near shore region became dominated by species typical of wet
meadow/strand communities, such as sedges (Carex spp. ), smartweeds (Polygonum spp.),
and cotton wood (Populus deltoides Rydb.) seedlings. Several recent studies have been
conducted in this particular wetland and have provided critical background information
on solute chemistry, periphyton, vascular plants, and invertebrate communities (Brady

1992; Brady et al. 1995; Brady and Burton 1995; Suzuki et al. 1995; Cardinale 1996;

148

Cardinale et a1. 1997; Cardinale et al. 1998; Burton et al. 1999; Stricker et al. 2001;
Vaara 2001; Burton et al. 2002).

A transect bisecting the S. pungens wetland was established in J une 1997 and
used during ensuing ﬁeld seasons (1998-2000) for sampling wetland surface water and
invertebrates (Figure 2.2). The transect extended 280 m perpendicular to the shoreline,
extending shoreward (S 172°) from the outer edge of the emergent plant zone/open bay

water interface hereafter referred to as the wetland interface (N 43°36’44” W

83°39’41 .3”). In 1999, three perpendicular transects were established along the primary
transect to experimentally evaluate nutrient limitation patterns in the study wetland. The

perpendicular transects were located at 60, 120, and 180 m from the wetland interface
(Figure 2.2).
Nutrient Limitation

Nutrient limitation was assessed experimentally in 1999 and 2000 using a nutrient
diffusing substratum (NDS) technique (F airchild et al. 1985; Pillsbury et al. 2000). The
NDS used in the 1999 experiment was a 10-cm clay pot (201.4 : 5.08 mL total volume;
652.9 cm2 surface area). Pots were pre-leached and the opening ﬁtted with a 63 -cm
diameter acrylic bezel. Each bezel had two l-cm diameter holes drilled, one to accept a
wooden dowel rod for positioning the NDS into the sediments and a second to ﬁll the pot
with the appropriate agar-based nutrient treatment. The bezel was sealed to the base of
each pot using a silicone adhesive. Vinyl tape was used to form a watertight seal around
the wooden dowel in the region that contacted the bezel. A number 000 rubber stopper

sealed the hole used for ﬁlling each pot. The NDS design was changed in 2000 to a 3-cm

149

.2 SEED 093 36323 333 335385 3:3 83:5 .3 3:3 2: 8m 388:8 833 8.93% 533 oust—am :02?

E 3:236 3 3383380 8.83% 3:3 mac—85 430338 3 58:0 203 $308.35 Eon 303.83 3:3 $0853 333233388 2: m5?
383% >335 22> Eon .35 68.3335 3:283 2: 80¢ E o3 3:3 .5 cm. .8 cc 3828.: ﬁns—ma 23 9 33335an 3232388
308:3: b33383 8:3 mac—m 3803333 80>» 3325898 396 533533 $53.23 333532 .3233 8-8 338535 633

.83 53333303 2: .8 333385 E owm 333385 68:35 bwﬁta BE. A332: .326 wcsoummn 88:3: .3 umaaeuaom .N.m 25mm..—

......

2:225 ..\\. W f
\5... > P < > > r > r P
......im > f: > > > r

8 ea... ....._ _ P 3553.»an

> > > ; \tlmaz > >

, ... \ .5350»: «532:0
\ ... \ .3 $30 .528
3. \

\

O

> 3...... .

P > . \ . an.
> y a > » > . ammunﬁhﬁ.

150

clay pot (20.6 i 0.5 mL total volume; 66.4 cm2 surface area) due to a 0.22 m decline in
mean wetland water depth relative to 1999.

A 2% agar (Bacto-agar) solution was used as the media in the NDS experiments.
Seven treatments were used in each experiment following the design of Fairchild et al.
(1985): control (no nutrient amendment), 0.05-M N, 0.50-M N, 0.05-M P, 0.50-M P,
0.05-M N+P, and 0.50-M N+P. Nitrogen amendments were in the form of nitrate
(N aNO3) and P amendments were in the form of phosphate (Na2HP04). Each treatment
was replicated three times and three separate experiments were conducted within the
wetland to evaluate spatial patterns. The clay pots were evenly spaced along the
perpendicular transects and adjacent pots (treatments) were chosen at random. After a
four week incubation period, the clay pots were removed from the ﬁeld, transported to
the laboratory on ice, and immediately frozen prior to chlorophyll a and ash ﬁee dry mass
(AFDM) determinations. The experiments were replicated three times, August 1999,
July 2000, and August 2000.

Laboratory analyses of the pot experiments differed between years because of the
change in NDS design. In 1999, periphyton was removed from the entire surface of each
pot in a known volume of de-ionized water under low light conditions. A 2-mL sub-
sample of the resultant slurry was ﬁltered through a membrane ﬁlter (0.45 pm). The
ﬁlter was carefully removed, folded, and placed in a 15-mL centrifuge tube containing 5
mL of MgC03 buffered 90% aqueous acetone. Chlorophyll a was extracted for 24 h and
quantiﬁed ﬂuorometrically (Turner Designs Model IO-OOOR) following acidiﬁcation
(0.1N HCl) to correct for pheopigments (W etzel and Likens 2000). An additional ﬁve 2-

mL sub-samples were ﬁltered through a single pre-weighed glass ﬁber ﬁlter (GP/F) for

151

determination of AF DM. Samples were dried at 85°C for 24 h, re-weighed, and
combusted at 550°C for 4 hours. After cooling, ﬁlters were again re-weighed to
determine loss on ignition.

In 2000, periphyton was removed from the entire surface of each pot in a known
volume of de—ionized water under low light conditions. However, the entire volume of
water and algal material was ﬁltered through a membrane ﬁlter (0.45 pm). Chlorophyll a
concentrations were determined ﬂuorometrically, but via the non-acidiﬁcation technique
(Welschmeyer 1994).

Hydrologic Connectivity between Lake and Wetland

The potential for boundary layer constraints on periphyton production was
evaluated by measuring mass loss of an alabaster substratum (AS). This approach
provided a relative measure of surface wave energy (or exposure) at each station along
the study transect and was assessed in conjunction with NDS experiments (Muus 1968;
Doty 1971). Alabaster substrata were constructed of commercially available Plaster of
Paris (CaSO4) mixed to the manufacturer’s speciﬁcations, and molded into disposable
plastic cups (90 mL total volume) with 1 cm diameter wooden dowels cast through the
center. The substrata were allowed to air dry until masses stabilized, deployed into the
wetland (dowel driven into sediments), and retrieved after approximately one week. In
the laboratory, substrata were again air dried until masses stabilized. The extent of
dissolution, which was assumed to be proportional to turbulence (Muus 1968; Doty
1971), was calculated by subtracting ﬁnal mass from initial mass. Mean daily wind
speeds were retrieved from the National Weather Service (N WS; www.nws.com) and

correspond to the Essexville station.

152

Data Analysis and Statistics
Nutrient limitation experiments were analyzed by AN OVA using chlorophyll a

(1999 and 2000) and AF DM (1999 only) as dependent variables. Separate ANOVAs
were performed for each of the three perpendicular transects (60, 120, and 180 m ﬁ'om
the wetland interface) and for each of the three experiments (August 1999, July 2000, and
August 2000). Concentrations of chlorophyll a (mg cm'z) were log-transformed to better
approximate normality. Dunnett’s test (one-tailed) was used for all control-treatment
pair-wise comparisons. Alpha was set at 5% and all signiﬁcance tests were performed in
SYSTAT (version 7.0).
RESULTS
Nutrient Limitation Patterns

In 1999, chlorophyll a responses ranged from 0.037-0.461 mg cm'2 on nutrient
amended pots compared to 0.048-O.304 mg cm"2 on control pots (Figure 3.3A). At the
60-m transect, both low and high levels of N and the high N+P treatments resulted in a
statistically signiﬁcant (p<0.05) increase in chlorophyll a. Unlike the pattern between the
two N+P treatments, no dose response was observed for the N treatments. Phosphorous
addition did not result in increased periphyton biomass relative to the control. The
pattern was similar at the 120-m transect, but within treatment variability was high and
control-treatment pair wise comparisons were not signiﬁcant (Figure 3.3A). The low and
high phosphorus and low N+P treatments resulted in no signiﬁcant (p>0.05) difference in
chlorophyll a concentration compared to the control. There was again no dose response
among any of the treatments. At the ISO-m transect, the low N and low N+P treatments
were signiﬁcantly different (p<0.05) relative to the control (Figure 3.3A). The responses

measured for the low N and low N+P treatments increased on average as distance from

153

 

0.200
G Control A)

’/////1 Low N
\\\\\\‘ High N 0

”"5“ E [-1th p<o,os

J p < 0.05
0.100

Chl a (mg cm'z)

 
 

 

 

 

 

 

 

 

c
.1:
.A
P
”‘=.+ a
‘W.

  

t\\\\\\\\\\\\\\\\\\§

 

 

0.000 -

‘-

 

1b

0.75 %
[:1 Control B)

_LowN
-HighN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0-60 i m LowP '
a High P
0?" [rm]: Low N+P
E 0.45 1 m High N+P ..
"3 ...
5 p>0.05
E
. p > 0- .
a 030 p < 0.05 "ST
< I I
F
1 _
o.oo~—- . ﬂ , w
60 120 180
DFWI (m)

Figure 3.3. Mean (: SE) periphyton chlorophyll a concentrations from the August 1999
NDS experiment (A). Mean (j; SE) AF DM concentrations from the August 1999 NDS
experiment (B). DF WI corresponds to distance from wetland interface in meters. N and
P correspond to nitrogen and phosphorous treatments. Probabilities from AN OVA are
indicated above each experiment; signiﬁcant pair-wise treatment comparisons relative to
the control (Dunnett's) are denoted by circles.

154

the wetland/lake interface increased. Ash ﬁ'ee dry mass associated with the high N+P
treatment was signiﬁcantly greater relative to the control at the 60-m transect (Figure
3.38).

The concentrations of chlorophyll a measured in the July 2000 experiment were
much lower compared to the August 1999 experiment; variability was also low (Figure
3.4A). Several signiﬁcant differences were noted at the 60-m transect and included the
low N, low and high P, and low N+P treatments. At the 120-m transect, the low N+P
treatment was signiﬁcantly different from the control. No differences were noted at the
180-m transect; concentrations of chlorophyll a were similar among transects (Figure
3.4A).

Concentrations of chlorophyll a measured on control and treatment pots were
higher in August relative to the July 2000 experiment; variability was high (Figure 3.48).
However, concentrations were still lower compared to measurements made in August
1999. Analysis of variance indicated no signiﬁcant treatment effects at any of the

transects.

Spatial Extent of Surface Wave Energy and Lake/Wetland Mixing
Mass loss observed for the AS declined in a near linear fashion 100-120 In (outer

region of the wetland) shoreward of the wetland interface in 1999 (Figure 3.5). Mass loss
was greatly reduced beyond this region and suggested equilibrium dissolution was more
important than mass loss associated with physical abrasion (surface waves). A similar
pattern occurred during the 2000 experiment, except a substantial reduction in

lake/wetland mixing was suggested approximately 60-80 m from the wetland interface

155

 

0.05

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[:1 Control A)
Low N
High N
0,04 . m Low? ,.
a High P
[[111]] Low N+P
é" 35% High N+P
E 0.03 4 -
u
90
E
z 0
E 0.02 a -
U
0 or . ' < 0 05
. p < 0.05 P ' p > 0.05 b
' " EM
0.05 i i i
B) [:1 Control
Low N
x\\\\~ High N
0.04 r m LowP -
E High P
A [HID Low N+P
"3E __ E High N+P
o 0.03 - -
90
E a
v
Q r!“
5 0.02 “ F b
0.01 a p > 0.05 -
E p > 0.05
0,, mg a Fm
120 ' 180

 

DFWI (m)

Figure 3.4. Mean (1 SE) periphyton chlorophyll a concentrations from the July 2000 NDS
experiment (A). Mean Q SE) periphyton chlorophyll a concentrations from the August
2000 NDS experiment (B). DFWI corresponds to distance from wetland interface in
meters. N and P correspond to nitrogen and phosphorous treatments. Probabilities from
ANOVA are indicated above each experiment; signiﬁcant pair-wise treatment
comparisons relative to the control (Dunnett's) are denoted by circles.

156

 

 

 

 

10000 T ' l ‘ l ' T ' l ' I ' l ' I
+ 1999
+ 2000
80.0 - .
i
3.9 60.0 - ~
2
Q
..i
a l
g 40.0 - -
L.
20.0 1 -
i
0.0 l ' I T I Y I V l V I ' T ' l
0 40 80 120 160 200 240 280

DFWI (m)

Figure 3.5. Mean mass loss (1- SE) of alabaster substrata (AS) across the study wetland
during mid-summer (August) 1999 and 2000. Arrows indicate subjective inﬂection
points where mass loss appears to stabilize throughout the remaining portion of the study
transect. DF WI corresponds to distance from wetland interface (wetland/open bay water

interface) in meters.

157

(Figure 3.5). Mean daily wind velocities dming each AS experiment were not
signiﬁcantly different (p>0.05; two-sample t test).
DISCUSSION

Surface water chemistry proﬁles along the study transect illustrated distinct
spatial trends for a variety of important solutes (see chapter 1). It was originally
hypothesized that the establishment of chemical gradients resulted principally from the
reduction in lake/wetland exchange. Extension of this idea to constraints on wetland
algal productivity suggested that patterns in nutrient limitation might arise along the same
chemical continuum. In fact, the impetus for the NDS experiments was based on spatial
differences in algal biomass documented in Saginaw Bay coastal wetlands (Cardinale
1996; Cardinale et al. 1997; Suzuki et al. 1995). Cardinale (1996) proposed that
diffusion constraints related to the reduction in physical exchange (i.e. lake/wetland
exchange) limited epiphyton production in the inner region of the wetland relative to the
more wave-swept outer region. The NDS experiments were designed to evaluate
potential nutrient limitation in conjunction with boundary layer constraints resulting from
the reduction in wetland/lake exchange.

The 1999 experiment illustrated nutrient limitation in the study wetland.
Phosphorous limitation, which is commonly limiting in freshwater systems (Fairchild et
al. 1985; Wetzel 2001), was not demonstrated. In contrast, the experiment suggested N
limitation on periphyton grth and the strongest evidence occurred at the 60-m transect.
This pattern was contrary to expected results. These data suggested that boundary layer
constraints on nutrient diffusion might not limit periphyton production. In 1999, lake-

wetland exchange occurred up to 100-120 m from the lake/wetland interface (Figure 3.5).

158

Internal recycling of nutrients within the algal/bioﬁlm complex may be more efﬁcient
than sequestering nutrients from the water column, which is contingent on nutrient
supply. Surface water chemical data indicate that N and P are in low supply by July and
August (see chapter 1).

The lack of a dose response for any of the nutrient amended treatments was
curious. It was hoped that repeating the experiment in 2000 would provide stronger
evidence for nutrient limitation and perhaps illustrate potential reasons for discrepancies
between this study and similar research (F airchild et a1. 1985; Pillsbury et al. 2000).
However, the 2000 experiments were even more inconclusive. Water depth in the
wetland declined from 0.31 m to 0.09 in between 1999 and 2000. The NDS design was
changed to accommodate low water levels, but no efforts were made to demonstrate that
diffusion rates were similar for the two designs. Low concentrations of chlorophyll a
measured in the 2000 experiments suggested the second design was likely ﬂawed.
However, benthic algal mats were dense in 2000 and overgrowth of the clay pots was
evident (personal observation). Light limitation may have hampered periphyton accrual
on the NDS in 2000.

The experiments indicated that nutrient limitation may occur in Great Lakes
coastal wetlands and may be related to the extent of lake/wetland exchange. However,
the disparity between experiments conducted in 1999 and 2000 illustrated the importance

of lake stage on algal productivity and community type (epiphytic versus benthic).

159

LITERATURE CITED

160

LITERATURE CITED

Allan, J .D. 1995. Stream Ecology: structure and function of running waters. Chapman
& Hall, New York.

Baldwin, D.S., A.M. Mitchell, and G.N. Rees. 2000. The effects of in situ drying on
sediment-phosphate interactions in sediments from an old wetland.
Hydrobiologia 431 :3-12.

Barton, DR. and H.B.N Hynes. 1978. Wave-zone macrobenthos of the exposed
Canadian shores of the St. Lawrence Great Lakes. Journal of Great Lakes
Research 4:27-45.

Batterson, T.R., C.D. McNabb, and F .C. Payne. 1991. Inﬂuence of water level changes
on distribution of primary producers in emergent wetlands of Saginaw Bay.
Michigan Academy of Science, Arts and Letters 23:149-160.

Baxter, C. and RR. Hauer. 2003. Measuring groundwater-stream water exchange: new
techniques for installing minipiezometers and estimating hydraulic conductivity.
Transactions of the American Fisheries Society 132:493-502.

Bedford, K.W. 1992. The physical effects of the Great Lakes on tributaries and marshes.
Journal of Great Lakes Research 18:571-589.

Beeton, A.M. 2002. Large freshwater lakes: present state, trends, and future.
Environmental Conservation 29:2 1 -3 8.

Bierrnan, V.J., D.M. Dolan, and R. Kasprzyk, and J .L. Clark. 1984. Retrospective
analysis of the response of Saginaw Bay, Lake Huron, to reductions in
phosphorous loadings. Environmental Science and Technology 18:23-31.

Bishop, CT. 1990. Historical variation of water levels in Lakes Erie and Michigan-
Huron. Journal of Great Lakes Research 16:406-425.

Bohlke, J .K. 2002. Groundwater recharge and agricultural contamination.
Hydrogeology Journal 10:153-179.

Brady, VJ. 1992. The invertebrates of a Great Lakes coastal marsh. MS. Thesis.
Michigan State University, East Lansing, Michigan.

Brady, VJ. and TM. Burton. 1995. The microcrustacean community of a Saginaw Bay
coastal emergent marsh. Pages 325-341. In: M. Munawar, T. Edsall, and J.
Leach (eds), The Lake Huron ecosystem: ecology, ﬁsheries and management.
SPB Academic Publishing, Amsterdam, The Netherlands.

161

Brady, V.J., B.J. Cardinale, and TM. Burton. 1995. Zebra mussels in a coastal marsh:
the seasonal and spatial limits of colonization. Journal of Great Lakes Research
21(4):587-593.

Brazner, J .C. 1997. Regional, habitat, and human development inﬂuences on coastal
wetland and beach ﬁsh assemblages in Green Bay, Lake Michigan. Journal of
Great Lakes Research 23:36—5 l.

Budd, J .W., W.C. Kerfoot, and AL. Maclean. 1998. Documenting complex surface
temperature patterns from advanced very high resolution radiometer (AVHRR)
imagery of Saginaw Bay, Lake Huron. Journal of Great Lakes Research 24:582-
594.

Burch, J .B. and J .L. Tottenham. 1980. North American Freshwater Snails: Species list,
ranges, and illustrations.

Burch, J .B. 1982. North American Freshwater Snails: Identiﬁcation keys, generic
synonymy, supplemental notes, glossary, references, and index.

Burton, TM. 1984. The effects of water level ﬂuctuations on Great Lakes coastal
marshes. Pages 3-13. In: H.H. Prince and RM. D’Itri (eds.), Coastal Wetlands.
Lewis Publishers, Chelsea, MI.

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

Burton, T.M., C.A. Stricker, D.G. Uzarski. 2002. Effects of plant community
composition and exposure to wave action on invertebrate habitat use of Lake
Huron coastal wetlands. Lakes & Reservoirs: Research and Management 7:255-
269.

Cardinale, B.J. 1996. The effect of a pelagic-littoral mixing gradient on an epiphytic
invertebrate community. MS. Thesis. Michigan State University, East Lansing,
MI.

Cardinale, B.J., T.M. Burton, and V.J. Brady. 1997. The community dynamics of
epiphytic rnidge larvae across the pelagic-littoral interface: do animals respond to
changes in the abiotic environment? Canadian Journal of Fisheries and Aquatic
Sciences 54:2314-2322.

Cardinale, B.J., V.J. Brady, and TM. Burton. 1998. Changes in the abundance and

diversity of coastal wetland fauna from open water/macrophyte edge towards
shore. Wetlands Ecology and Management 6:59-68.

162

Carlton, R.G. and R.G. Wetzel. 1988. Phosphorous ﬂux from lake sediments: effect of
epipelic algal oxygen production. Limnology and Oceanography 33:562-5 70.

Connell, J .H. 1961. The inﬂuence of interspeciﬁc competition and other factors on the
distribution of the bamacle Chthamalus stellatus. Ecology 42:710-723.

Cook, D.G. and MG. Johnson. 1974. Benthic macroinvertebrates of the St. Lawrence
Great Lakes. Journal of the Fisheries Research Board of Canada 31:763-782.

Crowder, LB. and WE Cooper. 1982. Habitat structural complexity and the interaction
between bluegills and their prey. Ecology 63:1802-1813.

Davids, C., E.H.T. Winkel, and 01 de Groot. 1994. Temporal and spatial patterns of
water mites in Lake Maarsseveen I. Netherlands Journal of Aquatic Ecology
28:11-17.

Devito, KI. and AR. Hill. 1999. Sulphate mobilization and pore water chemistry in
relation to groundwater hydrology and summer drought in two conifer swamps on
the Canadian Shield. Water, Air, and Soil Pollution 113:97-114.

- Diehl, S. and R. Komijow. 1998. Inﬂuence of submerged macrophytes on trophic
interactions among ﬁsh and macroinvertebrates. Pages 24-45. In: E. Jeppesen,
M. Sondergaard, M. Sondergaard, and K. Christoffersen (eds.), The Structuring
Role of Submerged Macrophytes in Lakes. Springer Verlag, New York.

Doty, MS. 1971. Measurement of water movement in reference to benthic algal growth.
Botanica Marina 14:32-35.

Duffy, WC. and TR. Batterson. 1987. The St. Mary's River, Michigan: an ecological
proﬁle. US. Fish and Wildlife Service Biological Report 85 (7.10). Washington,
DC.

Dvorak, J. 1970. Horizontal zonation of macrovegetation, water properties and
macrofauna in a littoral stand of Glyceria aquatica (L.) Wahlb. in a pond in South
Bohemia. Hydrobiologia 35: 17-30.

Dvorak, J. 1987. Production-ecological relationships between aquatic vascular plants
and invertebrates in shallow waters and wetlands — a review. Arch. Hydrobiol.
Beih. Ergebn. Limnol. 27: 181-184.

Dvorak, J. and E.P.H. Best. 1982. Macro-invertebrate communities associated with the
macrophytes of Lake Vechten: structural and functional relationships.
Hydrobiologia 95:1 15-126.

Dvorak, J. 1996. An example of relationships between macrophytes, macroinvertebrates
and their food resources in a shallow eutrophic lake. Hydrobiologia 339:27-36.

163

Fahnenestiel, G.L., G.A. Lang, T.F. Nalepa, and TH. Johengen. 1995. Effects of zebra
mussel (Dreissena polymorpha) colonization on water quality parameters in
Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21 :43 5-448.

F airchild, G.W., R.L. Lowe, and W.B. Richardson. 1985. Algal periphyton grth on
nutrient diffusing substrates: an in situ bioassay. Ecology 66:465-472.

Freeman, C., M.A. Lock, and B. Reynolds. 1993. Impacts of climatic change on
peatland hydrochemistry: a laboratory based experiment. Chemical Ecology 8:
49-59.

Freeman, C., G. Liska, N.J. Ostle, M.A. Lock, S. Hughes, B. Reynolds, and J. Hudson.
1997. Enzymes and biogeochemical cycling in wetlands during a simulated
drought. Biogeochemistry 39: 177-187.

French, J RP. 1988. Effect of submersed aquatic macrophytes on resource partitioning
in yearling rock bass (Ambloplites rupestris) and pumpkinseeds (Lepomis
gibbosus) in Lake St. Clair. Journal of Great Lakes Research 14:291-300.

Gardner, W.S., L. Yang, J .B. Cotner, T.H. Johengen, and PL Lavrentyev. 2002.
Nitrogen dynamics in sandy freshwater sediments (Saginaw Bay, Lake Huron).
Journal of Great Lakes Research 27(1):84-97.

Gafny, S. and A. Gasith. 1993. Effect of low water level on the water quality of the
littoral zone in Lake Kinneret. Water Science and Technology 27:363-371.

Gathman, J .P., TM. Burton, and B.J. Armitage. 1999. Coastal marshes of the upper
Great Lakes: distribution of invertebrate communities in response to
environmental variation. Pages 949-994. In: Batzer, D.P., R.B. Rader, and SA.
Wissinger (eds.), Invertebrates in Freshwater Marshes of North America. John
Wiley & Sons, New York.

Gathman, J .P. 2000. Abiotic and biotic inﬂuences on aquatic invertebrate community
structure in a Lake Huron coastal wetland. Ph.D. Dissertation, Michigan State
University, East Lansing, Michigan.

Gresens, SE 1995. Grazer diversity, competition and the response of the periphyton
community. OIKOS 73:336-346.

Hamilton, DP. and SF Mitchell. 1997. Wave-induced shear stresses, plant nutrients
and chlorophyll in seven shallow lakes. Freshwater Biology 38:159-168.

Heathwaite, AL. 1990. The effect of drainage on nutrient release from fen peat and its

implications for water quality; a laboratory simulation. Water, Air, and Soil
Pollution 49: 159-173.

164

Heino, J. 2000. Lentic macroinvertebrate assemblage structure along gradients in spatial
heterogeneity, habitat size and water chemistry. Hydrobiologia 418:229-242.

Hoerling, M. and A. Kumar. 2003. The perfect ocean for drought. Science 299:691-
694.

Horsch, GM. and HG. Stefan. 1988. Convective circulation in littoral water due to
surface cooling. Limnology and Oceanography 33: 1068-1083.

, Howard-Williams, C. and GM. Lenton. 1975. The role of the littoral zone in the
functioning of a shallow tropical lake ecosystem. Freshwater Biology 5:445-459.

Huddart, P.A., F.J. Longstaffe, and AS. Crowe. 1999. 8D and 6180 evidence for inputs
to groundwater at a wetland coastal boundary in the southern Great Lakes region
of Canada. Journal of Hydrology 214:18-31.

James, W.F. and J .W. Barko. 1991a. Estimation of phosphorous exchange between
littoral and pelagic zones during nighttime convective circulation. Limnology and
Oceanography 36:179-187.

James, W.F. and J .W. Barko. 1991b. Littoral-pelagic phosphorous dynamics during
nighttime convective circulation. Limnology and Oceanography 36:949-960.

J ansson, M., R. Andersson, H. Berggren, and L. Leonardson. 1994. Wetlands and lakes
as nitrogen traps. Ambio 23:320-325.

Johengen, T.H., T.F. Nalepa, G.L. Fahnenstiel, and G. Goudy. 1995. Nutrient changes
in Saginaw Bay, Lake Huron, aﬁer establishment of the zebra mussel (Dreissena
polymorpha). Journal of Great Lakes Research 21:449-464.

Jude, DJ. and J. Pappas. 1992. Fish utilization of Great Lakes coastal wetlands. Journal
of Great Lakes Research 18:651-672.

Kairesalo, T. 1980. Diurnal ﬂuctuations within a littoral plankton community in
oligotrophic Lake Paajarvi, southern Finland. Freshwater Biology 10:533-53 7.

Kashian, DR. and TM. 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.

Keddy, PA. and A.A. Reznicek. 1986. Great Lakes vegetation dynamics: the role of

ﬂuctuating water-levels and buried seeds. Journal of Great Lakes Research
12225-3 6.

165

Keough, J .R., T.A. Thompson, G.R. Guntenspergen, and DA. Wilcox. 1999.
Hydrogeomorphic factors and ecosystem responses in coastal marshes of the
Great Lakes. Wetlands 19:821-834.

King, GM. and M.J. Klug. 1982. Comparative aspects of sulfur mineralization in

sediments of a eutrophic lake basin. Applied Environmental Microbiology
43:1406-1412.

Kolak, J .J ., D.T. Long, J .M. Matty, G.J. Larson, D.F. Sibley, and TB. Councell. 1999.
Ground-water, large-lake interactions in Saginaw Bay, Lake Huron: A
geochemical and isotopic approach. GSA Bulletin 111(2): 177-188.

Krecker, RH. 1939. A comparative study of the animal population of certain submerged
aquatic plants. Ecology 20:553-562.

Kling, G.W., K. Hayhoe, L.B. Johnson, J .J. Magnuson, S. Polasky, S.K. Robinson, B.J.
Shuter, M.M. Wander, D.J. Wuebbles, D.R. Zak, R.L. Lindroth, S.C. Moser, and
ML. Wilson. 2003. Confronting Climate Change in the Great lakes Region:
Impacts on our Communities and Ecosystems. Union of Concerned Scientists,
Cambridge, Massachusetts, and Ecological Society of America, Washington, DC.

Klosowski, S. 1992. Temporal and spatial variation of habitat conditions in the zonation
of littoral plant communities. Aquatic Botany 43:199-208.

Krieger, K.A. 1992. The ecology of invertebrates in Great Lakes coastal wetlands:
current knowledge and research needs. Journal of Great Lakes Research 18:634-
650.

Larson, G. and R. Schaetzl. 2001. Origin and evolution of the Great Lakes. Journal of
Great Lakes Research 27:518-546.

Liston, CR. and S. Chubb. 1985. Relationships of water level ﬂuctuations and ﬁsh.
Pages 121-134 in H.H. Prince and F .M. D’Itri, editors. Coastal wetlands:
proceedings of the ﬁrst Great Lakes coastal wetlands colloquium. Lewis
Publishers, Inc., Chelsea, Michigan.

Long, D.T., T.P. Wilson, M.J. Takacs, and DH. Rezabek. 1988. Stable-isotope
geochemistry of saline near-surface ground water: East-central Michigan basin.
Geological Society of America Bulletin 100: 1568-1577.

McLaughlin, DB. and H.J. Harris. 1990. Aquatic insect emergence in two Great Lakes
marshes. Wetlands Ecology and Management 1:111-121.

Merritt, R.W. and K.W. Cummins. 1996. An Introduction to the Aquatic Insects of
North America. 3rd edition. Kendall/Hunt, Dubuque, Iowa. Pp. 862.

166

Mortsch, L.D. 1998. Assessing the impact of climate change on the Great Lakes
shoreline wetlands. Climatic Change 40:391-416.

Muus, B.J. 1968. A ﬁeld method for measuring “exposure” by means of plaster balls: a
preliminary account. Sarsia 34:61-68.

Nalepa, T.F., G.F. Fahnenstiel, M.J. McCormick, T.F. Johengen, G.A. Lang, J .F.
Cavaletto, and G. Goudy. 1996. Physical and chemical variables of Saginaw
Bay, Lake Huron in 1991-93. NOAA Data Report ERL-GLERL 91. Great Lakes
Environmental Research Laboratory, Ann Arbor, Michigan.

Ogden, J .G. 1982. Seasonal mass balance of major ions in 3 small watersheds in a
maritime environment. Water, Air, and Soil Pollution 17: 119-130.

Palomaki, R. and S. Hellsten. 1996. Littoral macrozoobenthos biomass in a continuous
habitat series. Hydrobiologia 339185-92.

Peckarsky, BL. 1983. Biotic interactions or abiotic limitations: A model of lotic
community structure. Pages 196-254. In: Fontaine IH, TD. and SM. Bartel]
(eds.), Dynamics of Lotic Ecosystems. Ann Arbor Science, Ann Arbor, MI.

Pillsbury, R.W., R.L. Lowe, Y.D. Pan, and J.L. Greenwood. 2002. Changes in the
benthic algal community and nutrient limitation in Saginaw Bay, Lake Huron,
during the invasion of the zebra mussel (Dreissena polymorpha). Journal of the
North American Benthological Society 21:238-252.

Planter, M. 1970. Physico-chemical properties of the water of reed-belts in Mikolaj skie,
Taltowisko, and Sniardwy Lakes. Pol. Arch. Hydrobiol. 17:337-356.

Prince, H.H., P.I. Padding, and R.W. Knapton. 1992. Waterfowl use of the Laurentian
Great Lakes. Journal of Great Lakes Research 18:673-699.

Qiu, S. and A.J. McComb. 1996. Drying-induced stimulation of ammonium release and
nitriﬁcation in reﬂooded lake sediment. Marine and Freshwater Research 47:531-
536.

Sager, E.P.S. 1996. Factors limiting the light environment in Cootes Paradise, Hamilton
Harbour and other coastal marshes of Lake Ontario. Water Quality Research
Journal of Canada 31:553-575.

Smiley, EA. and A.J. Tessier. 1998. Environmental gradients and the horizontal
distribution of microcrustaceans in lakes. Freshwater Biology 39:397-409.

Stricker, C.A., T.M. Burton, V.J. Brady, B.J. Cardinale, and JP. Gathman. 2001.

Invertebrate communities of Great Lakes’ coastal wetlands, Saginaw Bay, Lake
Huron. Verh. Intemat. Verein. Limnol. 27:3440-3443.

167

Suzuki, N., S. Endoh, M. Kawashima, M. Itakura, C.D. McNabb, F .M. D'Itri, and T.R
Batterson. 1995. Discontinuity bar in a wetland of Lake Huron's Saginaw Bay.

Journal of Freshwater Ecology 10:111-123.

Thorp, J .H. and AP. Covich. 1991. Ecology and Classiﬁcation of North American
Freshwater Invertebrates. Academic Press, New York. Pp. 911.

Vaara, A.M. 2001. The effects of water level ﬂuctuations on two coastal emergent
wetland plant communities of Saginaw Bay, Lake Huron. M.S. Thesis, Michigan
State University, East Lansing, Michigan.

Vanni, M.J. W.H. Renwick, J .L. Headworth, J .D. Auch, and M.H. Schaus. 2001.
Dissolved and particulate nutrient ﬂux from three adjacent agricultural

watersheds: A ﬁve-year study. Biogeochemistry 54:85-114.

Vince, S.W., I. Valiela, and J .M. Teal. 1981. An experimental study of the structure of
herbivorous insect communities in a salt marsh. Ecology 62: 1662-1678.

Watts, C]. 2000. Seasonal phosphorous release from exposed, re-inudated littoral
sediments of two Australian reservoirs. Hydrobiologia 431 :27-39.

Wellbom, G.A., D.K. Skelly, and BE. Werner. 1996. Mechanisms creating community
structure across a freshwater habitat gradient. Annual Review of Ecology and
Systematics 27:337-63.

Welschmeyer, NA. 1994. Fluorometric analysis of chlorophyll a in the presence of
chlorophyll b and pheopigments. Limnology and Oceanography 39:1985-1992.

Wetzel, R.G. 1990. Land-water interfaces: Metabolic and limnological regulators.
Verh. Intemat. Verein. Limnol. 2426-24.

Wetzel, R.G. 1992. Wetlands as metabolic gates. Journal of Great Lakes Research
18(4):529-532.

Wetzel, R.G. and GE. Likens. 2000. Limnological Analyses. 3rd edition. Springer-
Verlag, New York. 429 pp.

Wetzel, R.G. 2001. Limnology: Lake and River Ecosystems. 3rd edition. Academic
Press, New York. 1006 pp.

Wiggins, GB. 1978. Larvae of the North American Caddisﬂy Genera (Trichoptera).
University of Toronto Press, Toronto. Pp. 401.

Williams, DD. 1997. Temporary ponds and their invertebrate communities. Aquatic
Conservation: Marine and Freshwater Ecosystems 72105-117.

168

Winter, TC. 1981. Effects of water-table conﬁguration on seepage through lakebeds.
Limnology and Oceanography 26:925-934.

169

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

11111111111111