......»

.
Q.

.. «Madrigal.
P... £135.... ..

.
1%.... Sam: .. ..
: , ....Ie.) 536mb
.

I .br-
:v- 1‘ 1
) .Cfr.»
......{1
J .. “(Mantaflfnhfl
Hkfiw;¢ ......u
33)”...
. 3. "luv”:
3%.... .m
. ; ,
. . 5411.114.“
. .11.“. .
a? _ . . .
. .. .
..
.

 

 

.5: .r 9 .r}
A 3.7-.)E..¥’.S
i 17. 5‘55) . v}.
.‘ 5'4 ‘3‘: .I...
‘16! E) A rt. ....r'
L... . . .

3.9.3... 3.... ...
q . .
3:25:48?! )2! lflh’r Liv?
illiti‘ltog . null!-
51-1...‘-lslv A5 I...
X ......I.\

.... 9...}
5.33... r 1....
5...... ta .5. «

Er}

219?»... ii .
1“

5:“.

£7 .5...
13....
1‘. I

 

v v.
...-{1.1. 4‘

 

...:kusalxnfi 1..
. . . . {Ilium Atlnv:
. ‘11}..-15 ail...
. . 1.2.23.1... 3‘9. . 2:.
. I 6“. In)" ifl..’n
i. x. .7}... . L ..o..... {fauna
P. . 1 .bliflakflrn 25.1.... . .
-... ...X. ... ...iu... 34.3, ..

. ...... His-2.1). ........I.s.......ué

 

(

174:3:

’) , / v
.x w

Illlllllllllllllllllll

This is to certify that the

thesis entitled

Factors Affecting Bird use of Wetlands Created
for Mitigation

presented by
Michael Robert Wasilco

has been accepted towards fulfillment
of the requirements for

M.S. degreein Fish. & Wildl.

SNCLQM

Major professor

Date December 15, 1999

 

0-7639 MS U i: an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
Unlverslty

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 chlRC/DataDmpBS-p.“

FACTORS AFFECTING BIRD USE OF WETLANDS CREATED FOR
MITIGATION

By

Michael Robert Wasilco

A THESIS

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

MASTER OF SCIENCE
Department of Fisheries and Wildlife

1 999

FACTORS AFFECTING BIRD USE OF WETLANDS CREATED FOR MITIGATION
By

Michael Robert Wasilco

In order to determine the response of birds to wetland habitat created as
mitigation for wetlands lost to development, avian species use of 28 constructed wetlands
in south-central Michigan was assessed through point count surveys during 3 May-1
October, 1998. Factors hypothesized to affect avian use and diversity were compared on
the basis of species numbers recorded at wetlands with different levels of each factor.

A total of 129 species was recorded. The factors that were linked to significant
differences in total species numbers between wetlands were wetland size and distance to
cropland. Larger wetlands tended to have more species present. The number of species
present increased as distance from the wetland to the nearest cropland decreased.
Wetland bottom contour was responsible for the greatest difference between wetlands not
only in the number of species/visit, but also in percentages of habitat types available.
Contour of the wetland basin was also linked to differences between wetlands on the
basis of the number of species present within different species groups. The average
number of species seen/visit was lowest at wetlands with steep slopes and smooth
bottoms. Wading birds species were most diverse at wetlands with steep slopes and
diverse bottoms. Waterfowl species diversity was highest with gentle slopes and diverse
bottoms. Shorebird species varied the most at sites with diverse bottoms regardless of

slope.

ACKNOWLEDGEMENTS

This study was funded by a grant from the Michigan Department of Natural
Resources Nongame Fund Small Grants Program. Additional equipment and supplies
were provided by Michigan State University. Special thanks to Scott Winterstein, my
major professor, for all his guidance and assistance during this project. I also thank my
committee members Donald Beaver and William Taylor for their advice and
understanding. Special thanks are owed to the Michigan Department of Transportation,
Environmental Section, especially Erica Staton and Dave Schuen for their assistance in
getting access to the study wetlands and for all their help in providing the background
information on these wetlands. Thanks also go to Gregory Soulliere for his advice and

reviews of the rough drafis.

iii

TABLE OF CONTENTS

Page
LIST OF TABLES ....................................................................... v
LIST OF FIGURES ..................................................................... vi
INTRODUCTION ....................................................................... 1
OBJECTIVES ............................................................................ 3
STUDY AREA ........................................................................... 3
METHODS ................................................................................ 5
RESULTS ................................................................................. 13
DISCUSSION ............................................................................. 47
CONCLUSIONS .......................................................................... 54
RECOMMENDATIONS ................................................................ 55
APPENDIX A .............................................................................. 56
APPENDIX B .............................................................................. 65
LITERATURE CITED ................................................................... 72

Table

A1

Bl

BZ

B3

LIST OF TABLES

Page
The ten most common species seen at study wetlands between 3 May - 1
October, 1998 ............................................................................ 16
Results of ANOVA tests for influence of wetland size, % emergent cover,
% open, and % semipermanently flooded on number of species groups ......... 35

Results of AN OVA (P-values) tests for influence of wetland size, % emergent
cover, % open, and % semipermanently flooded on number of species observed
within each species groups .............................................................. 36

P-values associated with tests for differences in species use among 4 classes of
wetland basin contour configurations ................................................. 38

P-values associated with tests for differences in habitat variables among 4
classes of wetland basin contour configurations ..................................... 39

Results of comparisons of avian species use as a function of maximum depth. 46
The common and scientific names of all species recorded during the study. The

number of observations is the number of different points the species was
recorded at during the study period, and has no reflection on the number of

individuals recorded each time ......................................................... 57
The species that constituted each of the 7 wetland species groups ............... 63
Levels of each habitat factor within each of the study wetlands ................... 66
Distances (m) from each wetland to the nearest area of each surrounding

habitat ...................................................................................... 69
Summary of species use data for each wetland ....................................... 70

Figure

10

ll

12

13

LIST OF FIGURES

Page
Location of study wetlands ............................................................ 4

Number of the 129 avian species observed at each of the study wetlands.
These totals include all species seen from survey points, including birds that
were seen outside the wetland boundaries ............................................ 14

Number of the 110 avian species observed using each of the study wetlands... 15

The number of species seen in each of the study wetlands in each of 4 size
classes. The difference among wetlands is significant (F=5.32; df=3;
P=0.0059) .................................................................................. 17

Number of bird species observed as a function of the percent of wetland area
supporting emergent vegetation (n=28; R2=0.056; P=0.2259) ..................... 19

Number of species of birds observed as a function of the amount of wetland
area(ha) supporting emergent vegetation (n=28; R2=O.295; P=0.0028) .......... 20

Number of species observed as a function of the percent of wetland area that
was open water/non-emergent habitat (n= 28; R2=O.O45; P=0.2800) ............ 21

Number of species of birds observed as a function of the size of the open water
component of a wetland (n=27; R2=0.131; P=0.0641) .............................. 22

Number of bird species observed as a function of the percent of wetland area
that is continuously flooded in a normal year (n=28; R2=O.057; P=0.2230). 23

Number of species of birds observed as a function of the amount of area (ha)
that is continuously flooded during a normal year (n=27; R2=O.139;
P=0.0555) ................................................................................. 24

Percent of wetland area that is continuously flooded in a normal year among
different size classes of wetlands (F=1.65; df=3; P=0.2044) ....................... 26

Percent of wetland area covered by open water or non-emergent vegetation
among different size classes of wetlands (F=1.94; df=3; P=0. 1493) .............. 27

The number of avian species present in a wetland as a function of the distance
to the nearest road (n=27; R2=O.000; P=O.9663) ................................... 28

vi

Figure
14

15

16

17

18

19

20

21

22

LIST OF FIGURES continued

Page
The number of species of birds observed in wetlands as a function of the
distance to the nearest cropland habitat (n=28; R2=O.166; P=0.0311) ............ 29
The number of avian species observed using wetlands as a function of the
distance to the nearest adjacent wetland (n=28; R2=0.O26; P=O.4l41) ............ 30
The number of species of birds observed in a wetland as a function of the
distance from that wetland to the nearest forested habitat (n=28; R2=0.Ol6;
P=O.5271) .................................................................................. 32

The number of avian species present at the study wetlands as a function of the
distance from the nearest source of regular human disturbance (n=28;
R2=O.046; P=0.2750) ..................................................................... 33

The density of avian species observed within planned (n=21; mean = 9.23
species/ha) and unplanned (n=7; mean = 6.56 species/ha) wetlands (F =O.99;

df=1; P=O.3300). Note that within categories wetlands increase in size from

left to right ................................................................................ 34

Average number of avian species observed/visit in each wetland in each of 4
basin contour configurations (F=4.23; df=3; P=0.0161) ............................ 40

Percent of wetland area semipermanently flooded in each wetland in each of
four basin contour configurations (F=15.51; df=3; P=0.0001) ..................... 42

Percent of wetland area supporting emergent vegetation in each wetland in
each of 4 basin contour configurations (F=15.41; df=3; P=0.0001) .............. 43

Percent of wetland area containing open water/non-emergent vegetation in

each wetland in each of 4 basin contour configurations (F=19. l 3; df=3;
P=0.0001) ................................................................................. 44

vii

INTRODUCTION

Wetlands perform many important functions in a landscape. Wetlands provide
floodwater storage, water filtration, nutrient trapping, and ground-water recharge in
addition to many other functions. Wetlands providing these functions can vary widely in

their ability to provide plant and wildlife habitat.

As more and more wetlands have been lost to development, through draining and
filling, this important wildlife habitat has also been lost. Michigan has lost at least 50
percent of its wetland areas since the late 17003 (Mitsch and Gosselink 1993). Since the
early 19703 there has been increasing protection of the remaining wetland resources
(Beck 1994, Young 1996). In recent years it has become mandatory to mitigate for any
wetland losses that cannot be prevented (Brinson and Rheinhardt 1996). This mitigation
is often in the form of creating new wetlands, or restoring and enhancing existing
wetlands. However, creating wetlands to replace drained or filled wet areas is a relatively
new science (Young 1996). A limited amount of work has been done to assess the
effectiveness of this mitigation (Mitsch and Wilson 1996). The wetlands created as
mitigation for losses to development are often created with specific objectives of
providing a wetland of some type, but not necessarily the type lost or a type that is useful

to wildlife (Zedler 1996).

Wetland bird species likely to use created wetlands include: Pied-billed Grebe
(Podilymbus podiceps), Great Blue Heron (Ardea herodias) and other wading birds;

Canada Goose (Branta canadensis), Mallard (Anas platyrhynchos) and other ducks;

Virginia Rail (Rallus limicola), Sora (Porzana carolina), Solitary Sandpiper (Tringa
solitaria) and other shorebirds; and blackbirds (Icteridae), Marsh Wren (Cistothorus
palustris), Swamp Sparrow (Melospiza geogiana) and other songbirds. Some of these
species nest in wetlands while others (e.g., herons) only feed there. Each of these
species requires a different set of wetland characteristics to meet their habitat needs
throughout the year. For example, pied-billed grebes require wetlands with a
combination of open water and abundant emergent vegetative cover (Storer 1991).
Canada Geese only require a pond with very little emergent cover surrounded by
vegetation for grazing. The shorebirds require shallow open water or mudflat areas to
forage in, and the songbirds require heavy emergent vegetation for nesting. Ideally, a
constructed wetland would provide suitable habitat for many of these wetland bird
species, while still providing the other required wetland functions. Relatively few studies

have looked at bird use of created wetlands for species other than waterfowl.

A major problem with wetland mitigation is the time lapse between the loss of the
original wetland and the creation of a replacement wetland (Brinson and Rheinhardt
1996). The wetland loss usually occurs before the new wetland basin is capable of
adequately supporting wildlife. Many studies of created wetlands don’t begin to assess
the success of the creation project until at least two years after the wetland was created
because it often takes the vegetation this long to become fairly well established, and then
many studies only monitor the wetland for a few years (Mitsch and Wilson 1996). In
many cases wetland evaluators consider bird use as only a minor part of the wetland

monitoring (Dave Schuen, Michigan Dept. of Transportation (MDOT), pers. comm).

In the past, wetlands have often been created to replace lost habitat for many
species of wildlife with little assessment of whether wildlife actually use these created
wetlands. With the increasing amount of development impacting wetlands and
corresponding mitigation, it is necessary to evaluate the wetland characteristics that can
be created to encourage use of these wetlands by many species of wildlife. This study
examined some of the factors thought to influence avian use of created wetlands. The
results of this study will make it easier for wetland designers to create wetlands with

characteristics that will attract and be used by a greater variety of wetland birds.

OBJECTIVES

To explore the relationship of avian use of wetlands created for mitigation, this

study had the following objectives.
1. Determine which physical and spatial factors present in created wetlands are
correlated with avian diversity.

2. Determine which wetland design characteristics can be used to attract avian
use of a wetland.

STUDY AREA

This study examined the avian use at 28 wetland sites located throughout south-
central Michigan. Wetlands were located in the following counties: Midland (l
wetland), Tuscola (1), Clinton (10), Eaton (2), Oakland (8), Macomb (5), and Livingston
(1) (Figure 1.). These wetlands range in size from 0.6 ha to 17.6 ha and are located in a

variety of land use areas ranging from primarily agriculture and forest to mostly

3

 

 

 

 

 

 

 

 

 

 

 

 

25

 

 

 

 

l,23,2 I

 

 

 

11,12,13
3,4

 

 

 

27
28

 

 

22,26

 

 

15,16,17,18,l9

 

l

 

 

 

Figure 1. Location of study wetlands.

 

suburban/urban. Michigan Department of Transportation (MDOT) created all 28
wetlands. Most were planned and created as mitigation for wetlands lost to road
construction. Several were not planned and were created when fill material was removed
for use in road construction. Due to this link with road construction, all of the wetlands
studied were located near major roadways, with several actually within the Cloverleaf of

entrance/exit ramps.

METHODS

Avian use data were collected by conducting point surveys from 3 May to 1
October, 1998. Wetlands contained enough survey points to thoroughly sample the
wetland habitats located at that site. Survey points were established by examining a
diagram of the wetland and visiting the wetland to determine the points that would allow
for the entire wetland to be completely visually surveyed from the fewest possible points.
A small, round wetland with only a narrow fringe of emergent vegetation would only
need a single point. A large wetland with large areas of emergent cover and many points
and fingers extending into the surrounding uplands would require as many as 6 survey
points. Survey points were a minimum of 50 meters apart on a small densely vegetated
wetland and a maximum of 200 meters apart at a large open lake-like wetland with no

emergent vegetation along the edges.

Surveys were conducted at points by counting and recording all species seen or
heard from that survey point during a 20-minute period that began with my arrival at that
point. The length of time at each point was suggested by Gibbs and Melvin (1993) to

increase the chances of detecting some of the more secretive species such as rails and

5

grebes. Whenever possible the numbers of each species seen were also recorded. All
point surveys were conducted between 15 minutes prior to sunrise and 4 hours after
sunrise. It was possible to survey up to five nearby wetlands in a single morning. In
wetland groups where individual wetlands were always surveyed together on the same
date, the first wetland surveyed each time was varied to reduce the possible chances of
the same birds being pushed from one wetland to the other if they flushed due to the
presence of the observer. All wetlands were surveyed nine times with at least 10 days

allowed to pass between consecutive surveys at individual wetlands.

An attempt was made to use call/response survey methods for the final five
minutes of each twenty-minute survey period in hopes of detecting any non-detected
species. The location of wetlands so close to busy roads and the accompanying noise of
traffic made this impractical. In most cases the sparseness of cover in the wetlands made
the call/response survey unnecessary as all species present would be detected visually.
The only wetlands heavily vegetated enough to require this method had individuals of the

targeted species detected without the use of calls.

Bird species seen outside a wetland during a survey were also recorded as having
been seen but not using the wetland habitat. This was done to maximize the chances of
recording all the possible species attracted to a wetland. Any species recorded outside a
wetland that was later seen using the wetland would be counted as a species attracted to
the wetland. However, some species were recorded outside wetlands, but never in

wetlands and are most likely not attracted by the wetland habitat.

Avian species within wetlands were also recorded any time the site was visited.
These instances included visits to do vegetation sampling and visits to measure water
depth as well as any time that I happened to be at the site. These species were recorded
separately from species seen on point surveys, and were used in the analysis only for total
species numbers, since wetlands were not all visited the same number of times for

vegetation sampling.

Habitat data were collected by recording water depth (cm), flooding regime
(upland, temporary, seasonal, semipermanent), vegetation type (upland, emergent,
aquatic bed, submerged, none-mud, none-open water), vegetation height (cm), and
relative vegetation density (none, very sparse, sparse, medium, dense, very dense) every
3 meters along transects through the wetlands. Each wetland contained a minimum of
four transects. Large wetlands had transects located every 100 m and each transect was
at least 20 meters from the nearest shore. Wetlands too small to have transects 100 m
apart had four evenly spaced transects at least 10 m from the nearest parallel edge. Each
transect ran fiom one edge of the wetland to the opposite edge. Vegetation data were
collected during August, September and October, 1998. Wetland boundaries were
delineated following standard procedures (Environmental Laboratory 1987), although in
most cases it was obvious where the wetland ended and the upland began due to the fact
that these were constructed wetlands. Wetland vegetation type and flooding regime

classes follow those outlined in Cowardin et. al. (1979).

Water level changes were monitored in most of the wetlands by establishing a
depth gauge during May and recording water levels (cm) each time the wetland was
surveyed. Several wetlands not receiving a gauge had no easily accessible standing
water. One wetland had too much human activity to establish a water depth gauge due to
the very high risk of tampering. Several of the wetlands that did receive gauges needed

to have the gauge moved due to complete drying at the site of the gauge.

There are many factors that may affect the avian diversity of a created wetland.
The following is a list of the factors that I examined:

Age of the wetland

Average and Maximum depth of the wetland

Size of wetland

Distance from other wetlands

Distance from surrounding habitat and land use types.

Distance from disturbances (human activity)

Composition of vegetation (natural and planted) in wetland

Bottom contour configuration (smooth slope vs. diverse depths across
wetland)

wsswewwr

The study wetlands were all at least 2 years old at the start of the study with a
maximum age of 10 years according to MDOT records. It is unknown if this is measured
from the completion of the excavation of the wetland or from the time when the wetland
filled with water for the first time. The newest wetlands already had at least two growing
seasons worth of vegetation present and in some cases had more vegetation than older

wetlands.

The average and maximum depths (cm) of wetlands were determined from the

transect data and standardized to a common date using the depth gauge readings to adjust

the transect data. Maximum depth for several wetlands had to be estimated since there
were areas too deep to measure along some transects. Average depth for the wetlands
that completely dried during the summer are treated as zero once the wetland was dry.
Due to the amount of variation of wetland depth depending on recent rainfall only
maximum depths could be standardized to a common date (June 20, average date of

gauge placement).

Bottom contour configuration is closely linked to maximum and average depth.
Transect data were used to determine the level of smoothness of the bottom of each
wetland. This factor is quite subjective in its measurement, since there is no easy way to
standardize it. I decided to use four categories: gentle smooth, steep smooth, gentle
diverse, steep diverse. Gentle slopes varied <10 cm between transect points and steep
slopes varied 2 10 cm between transect points. Smooth bottoms had a continuous depth
gradient from shallow to deep to shallow across a wetland. Diverse bottoms had a
bottom gradient from shallow to deep too shallow to deep to shallow across a wetland.
The difference between shallow and deep had to be at least 30 cm and constitute more

than 1 point out of character to be considered as diverse.

Wetland sizes (ha) were estimated from the transect data. Lengths of transects
were used along with a measurement of the length of the wetland perpendicular to the
transects to estimate the number of m2 contained in the wetland. This estimate was then
converted to hectares. I compared the sizes obtained by this method to the sizes of the

few wetlands that had been measured separately on the construction diagrams and found

it to be a relatively accurate estimate considering that I was comparing it to projected and

not actual sizes.

The distance (111) to the nearest wetland and other surrounding habitats was
measured if less than 100m and estimated for distances >100m. The distance recorded
was the shortest distance between the nearest edge of the wetland to the closest edge of
the other wetland or habitat. Distance to nearest human disturbance was recorded as the
shortest distance (m) from an edge of the wetland to the nearest occupied building.

Roads were not considered as a source of disturbance since the birds ignored the traffic
and almost all of the wetlands offered limited access to pedestrians, with the exception of

two wetlands open to public fishing.

Data analysis was done using only the numbers of species seen within the wetland
boundaries, and excludes observations of species outside wetlands or merely seen flying
over. Wetland factors were tested on the basis of total species numbers observed within a
wetland, and on the basis of the number of different wetland species groups that were
recorded. Species observed were placed into 8 groups; 7 groups of wetland species
(Appendix B) and the remaining group contained any species not included in the first
seven. The wetland groups were: 1) diving birds (loons, grebes, coots), 2) waders
(herons, bittems, egrets, cranes), 3) waterfowl (ducks and geese), 4) raptors, 5) rails, 6)
shorebirds, 7) gulls/terns, 8) other species. The different wetland factors examined were

tested on the basis of how they influenced the total number of species recorded as well as

10

the number of species groups present. The wetland factors were also examined for

influence within each species group.

Comparisons of wetland factors (size, habitat type, and bottom contour) that could
be separated into classes were tested using One-Way Analysis of Variance (AN OVA).
Wetlands were grouped into four size classes: small (<1 ha), medium (1-3 ha), large (3-6
ha), and very large (>6 ha). Wetland size categorizations were based on the groupings
that became evident when the sizes were graphed. Wetlands were also grouped into four
classes (525%, 26-50%, 51-75%, and >75%) to examine the effect of the different habitat
types (vegetation types and flooding regimes). Percentage class for a wetland was
determined by the number of transect points within that wetland containing the habitat
type being examined compared to the total number of points. The class boundaries were
set at the quartile boundaries to maintain an easily distinguished and commonly used
breakpoint for percentage classes. Graphing the data led me to consider combining the
middle two classes into a single class (26% - 74%), but to maintain equal class sizes I
retained the original classes. Wetland 28 was excluded form the analysis of bottom
contour and maximum depth since I did not have that data from transects. Wetland 28
was excluded from the analysis of area of open water and area of semiperrnanent flooding
since this wetland contained about three times as much area of these habitats as any other

wetland.

Wetland factors of a more continuous nature were examined using Linear

Regression. Most of the factors examined in this study required the use of regression

ll

since any classes determined for testing with an AN OVA would have been highly
arbitrary. Factors tested with regressions were the distances to other habitats, as well as
the amount of each habitat type available within a wetland. Sample size for all
regressions was 28 unless noted otherwise. An alpha level of 0.05 was used to determine

significance for all comparisons except as noted below.

The analysis of large-scale wetland characteristics (size, percent coverages) effect
on the species group level resulted in 28 pairwise comparisons. A Bonferroni adjustment
was made to control the experimentwise Type I error rate at 0.100 (Sokal and Rohlf

19952240). Therefore, the significance of statistical tests involved in this analysis was

based on or = 0.004.

RESULTS

A total of 129 species (Appendix A.) were recorded during the point surveys.
This includes all the species seen and heard from points both inside and outside the
wetlands. The number of species recorded at each wetland ranged from 21 to 62 (Figure
2.). When only species seen within the wetlands were included the total number of
species seen drops to 110 with a range of 12 to 48 different species for individual
wetlands (Figure 3.). The ten most commonly observed species are listed in Table l with

the number of times the species was observed.

Wetlands were grouped into four size classes: small (n=8), medium (n=10), large
(n=5), and very large (n=5). I examined the effect of wetland size on the total number of
species observed using an One-Way Analysis of Variance (ANOVA) and found there is a
significant effect (F =5.32; df=3; P=0.0059). The smallest wetlands averaged 24.1 total
species with the average total increasing with size; medium (25.9), large (33.8) and very
large wetlands (41.8 total species) (Figure 4.). I also found that size significantly affected
the average number of species recorded per survey (F =10.44; d%3; P=0.0001). Very
large wetlands averaged 24.8 species/survey followed in order by large (19.2), medium

(9.8) and small (8.33).

This difference in vegetation characteristics between wetlands led me to examine

for vegetation effects on the number of species observed. These factors were examined

13

 

dogs—on wed—83 05 023.5 58 203 35 «WEB menus—05 .358
>023 82m :08 360% =e “6305 283 32:. dean—8.5 33m 05 mo :08 an antenna 860% 53a as 05 mo c3852 .N onE

.enEaz ouou cease;
emzemmuewmwwmRoweetefliflfi:2 a e a o m e e. «F

or

ON

on

,0?

 

 

 

on

uaas segoads Io reqwnN

l4

acne—83 33m 05 mo :08 m5? 83030 860% 53a o: 05 .«o 89:52 .m ecu—mi

eeeoeeese;
wNNNmNvaNMNNNFNommwwrnrmFmrvrmFNF :9. m w n m m v m N_.

 

 

_ _ _ o
— _ _ 2
l . N
om n
w
a.
a
I.
- on m.
S
d
a
m.
- on w
t l ll- 1 l I ll- om

 

om

 

15

Table 1. The ten most common bird species seen at study wetlands from 3 May — 1

October, 1998.

 

 

Number of
Common Name Scientific Name ObservationsIll
Song Sparrow Melospiza melodia 321
Red-winged Blackbird Agelaius phoeniceus 303
Mallard Anas platyrhynchos 251
Savannah Sparrow Passerculus sandwichensis 243
Killdeer Charadrius vociferus 208
Barn Swallow Hirundo rustica 198
Spotted Sandpiper Actitis macularia 146
American Goldfinch Carduelis tristis 141
Great Blue Heron Ardea herodias 123
Tree Swallow T achycineta bicolor l 14

 

“ Each species was seen at nearly all sites on at least one visit. The number of times that
the species was recorded at a point is listed to the right of the scientific name. This total
is not the number of individuals recorded, but the number of times the species was

recorded during point counts.

16

.8831“ .Rnee fife Eocene
E 35:03 macaw oonaobmo BE. .8320 ofim v mo :08 5 $5303 beam 05 mo :03 E :03 86on mo 538:: 25. .v cam...“

aqueous-=0;
meNmPoFNNVNvnNFvvoFmepnNthoNNPNowwmemhmwhww

noes moods jo requmu

 

Am: ob omcmn >cm>l

An; 0-8 e93!

 

lA An; m.: 523.20

7 :2 3:95!

 

 

l7

using linear regression since the vegetation characteristics could not be easily categorized
for analysis. The first factor that I examined was the percent of wetland area supporting
emergent vegetation (Figure 5.) and found that it was not significant (R2=0.056;
P=0.2259) although there is a tendency for more species being observed at higher
percentages of emergent vegetation. The amount of area supporting emergent vegetation
was significant in determining how many species were detected (R2=0.295; P=0.0028)
with more species observed as hectares of emergent cover increased regardless of what
percent of total area was emergent (Figure 6.). Percent open water/non-emergent
vegetation (Figure 7.) was also found to be non-significant (R2=0.045; P=0.2800) with a
trend toward lower species numbers as open water increases. However, the size of the
open water habitat was significant (n=27; R2=0.131; P=0.0641) with the number of
species increasing as area of open water increased (Figure 8.). Percent of wetland area
that was continuously flooded throughout a normal year was tested (Figure 9.) and found
to be non-significant (R2=0.057; P= 0.2230) with a trend toward lower numbers of
species at higher percentages. The relationship between size of the habitat that was
continuously flooded and number of species was significant (n=27; R2=0.139; P=0.0555)
with the number of species increasing as size increased (Figure 10). These are the same
trends as seen for open water and should be expected since the percent open water is
likely very closely linked to the percent area continuously flooded as can be seen in

Figures 11 and 12.

After finding that all the vegetation characteristics were not significant, I tested

for differences in vegetation characteristics between size classes. There was no

18

.ammmfa ”Sesame fins

32890 538%? “cognac wfitonasm 38 98—83 .«o amoeba 05 mo note—5m a ma cotomno 86on 23 mo 5:822 .m 03E

no; cease; be e\..

 

 

 

 

 

 

 

oo_. om ow on om om ow om cm or o
b H F IF » h u h o
l or
o o
o 0
t e 9 ON
o e #
ll 0
I,
0 «t on “v
o o c a
l o m.
o a
0 S
o o...
o
O O O
on

 

 

 

 

om

l9

.9831. was"...

mwmué newfiowo.» ...omBEo wfitcmmsm 3: no... 28:03 mo .52.... on. ..o .8305... a mm 320on 3.3 ..o 8.3% ..o 89:52 .o 02%....

ooooér

eczauouc> 2.09.2.5 .o :5 no.5

 

 

 

 

 

 

 

oooo.N_. 00092. 00006 00006 00006 oooo.~ 00006
E b b p . b o
or
O
«a
O 0 0 ON
0
0 O m
w
\\ m.
t on m.
. s
e o. m.
a
O ov
O
O
O
O
\ om

 

om

 

20

.8821. was"...
”mm "5 .839. .cowuofifcosbfia come we? 35 «Pa 93:95 mo Beacon 2: .«o 55:3 5 we wotomno 860% Mo .3852 H 03E
3.... .0 2.3.8..

oo_. cm on on cm on o... om cm 0.. o
p _ P L h _ _ H L o

 

 

o_.

 

 

O
N

O
t")
segoeds jo .quulnN

 

 

 

 

 

 

 

 

 

cm

21

.235an

:m _ .oHNM ”RUE wage? a mo 30:383 833 mono 05 .«o ofim 05 mo 86an a mu 3203c 3.3 we 86on mo 898:2 .w 0.5me

ooomé

.335 :onO “—0 35 no.2

 

 

 

 

 

 

 

 

 

 

 

coco;~ ooomd oooo.n ooom.~ oooo.~ ooom... ooooé ooomd ooood

_ .r p h _ p h b 0

or

O
O O
O 0 how N
n
w
1:1 m
o 6 u.“ on m.
0 S
O

. m
. '0
11 o a

cc

0 O O
on

 

 

 

om

 

22

.883"; KBan fins
.80» 38.8: a 5 Bacon Emacsfiwqoo mm 35 38 @5303 mo :8ch 08 mo .8on a ma 32030 860% BB mo 33:52 .a 2&5

«a; “Euro? no .\.
oow om on on om cm. 0? on em or o

 

 

or

 

 

 

 

0

I
:01:

 

 

 

O
O
senaeds

 

ow

 

om

 

 

 

ow

23

.888"; gawk fins so»
388 a mama—u wowooa Ema—02:88 mm 35 Ag 88 no “5585 05.3 530:3 a ma Btomno $85 mo 86on mo “38:2 .2 2:me

umuooE >=cccuE.on-_Emm $5 ~2<

 

 

 

 

 

 

88.8 88.x. 888 888 88.... 88.8 888 88. F 88.8
_ r p _ p p _ o
S
O O
O O
O
8 N
n
w
a.
a
J
8 m...
S
d
a
m.
a
I s

 

O
V

 

 

 

 

 

 

 

cm

24

significant difference in percent of area semi-permanently flooded between size classes
(Figure 11. F=1.65; df=3; P=O.5099) or percent area covered by open water/non-
emergent vegetation (Figure 12. F =1 .94; df=3; P=O.5314). This lack of difference is due
to the large amount of variance in wetlands. There were semi-permanently flooded and
temporarily flooded wetlands in all size categories with percent open water ranging from
O-95%. The percent area supporting emergent vegetation was also not significantly
affected by wetland size 03:1.53; df=3; P=0.2330). It should be noted that percent

emergent cover is the reciprocal of percent open water/non-emergent.

I examined the relationships between the munber of species recorded at a
wetland and the distance from that wetland to other habitats or land use types. I found
that the distance to the nearest road had no effect on species numbers (n=27; R2=0.OOO;
P=O.9663, Figure 13.). Wetland 20 was excluded from the analysis of distance from
roads since it was nearly three times as far from a road as the next most distant wetland.
The distance to the nearest upland/grassland was also not significant (R2=0.005;
P=O.7148) due to the fact that almost all my wetlands had upland bordering at least one

edge.

Distance to the nearest cropland was significant (R2=0.166; P=0.0311) with
increasing numbers of species as cropland distance lessened (Figure 14.). Distance to
nearby wetlands was not significantly correlated to the number of species (R2=0.026;
P=0.4141) although species numbers tended to decrease as distance increased (Figure

15.). Distance to forest habitat was also not significant (R2=O.Ol6; P=0.5271) with no

25

Agnoufimumu
”we. TB $85303 mo momma—o ofim Beanbag macaw an» 388 a E Becca b38538 mm 85 m2“ USN—$3 mo 800qu .: oBmE
230 3.23;
wumwmrorumvm v m NF: w 5 m mwmmnw m om N Remove m n 3:?

papooH Alsnonunuog 231v %

Am; mA 283 >ao>l

a; 98 85.8
7 :2 8-: 6282.“.

:2 2 =95.

 

26

.33;de Hub". ”23an
mug—83 mo momma? onm 805%? waoaa magnum? Eombfioéoa 8 H895 none 3 @2960 8.3 US$03 mo :5ch .fi onE

888335
832888,. .0. 31¢ :mmrnwaomwm 8892..“ 52:.

W‘ \[4
Am: at 09“.. Bo>l V

uonuoBoA zuaBJoura-uomrozam uado luv 94

:2 8-8 823-7

a; mi 5382.“;

 

V
:2 E :95! l

 

27

our

.888";

68an mnmuav 38 823: 05 8 3:8me 2: mo 5on a 8 98:03 a 5 Samoa 86on 53m «0 898:: 2:. .2 oSwE

uuom Eat 25 3.335

 

 

 

 

 

 

 

 

 

 

03 cm om ov om o
_ r r k r o
o_.
O
O
0 cm
0 an
. m
S
lllllllllr 0” w
0 O O O m.
. a
o ov
O
O
O
on

 

 

cm

28

.2 88am 88AM
mwmufi «35a: cam—no.8 823g 05 8 858% 05 mo c383 a ma 35:95 5 3.2.530 v.33 mo 86on mo .5959: BE. .3 035

tea—no.5 8215 3:520

 

 

 

 

 

 

 

 

 

 

 

 

 

 

89 8.: 8m. 88 8» 80 8v 8m 8
. b p L! p p _ o
o.
O O
O
o 4 cm
0’
o #
. w
i 0 . .... a
a
o m.
o ..l a
s
O
9.
O O Q
Q
I .- 8
8

 

 

 

29

com

.338"; smirk

mwmncv Esau? 308.28 8088 o5 8 35$va 2: MO 3383 a mu avg—83 mam: 33030 86on :38 .«o 838:: BE. .2 onE

33:03 39.32 8 :5 00:30.9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

com 00? Don CON cow 0
a a F F r o
w or
.
O .9 .
«0, ON
.0 O O
I, O .
a"
m.
1| III”: on s
o m
0 6C. 9 _I
o a
O
.
l. O?
.
.9
. 9
O
l, l - ll 8

 

 

om

 

30

real trend evident (Figure 16.). The distance to the nearest occupied building and
therefore potential regular human disturbance was also not significant (R2=O.O46;
P=0.2750) and interestingly showed a slight trend toward higher species totals closer to

buildings (Figure 17).

I also compared the planning and construction methods used to create the
wetlands and found that the unplanned wetlands (6.56 species/ha; n= 7) did not differ
significantly (F =O.99; df=1; P=0.3 3) in species density from the planned wetlands (9.23
species/ha; n=21) (Figure 18). I could only compare these wetlands on a species/ha basis
due to the fact that planned wetlands tended to be larger than unplanned wetlands with

very little overlap in size.

The different factors tested for influence on total species numbers were also tested
for affects on the species guild groups. The results of these tests are shown in Table 2.
None of the factors tested were found to have a significant effect on the number of

species groups (Table A2.) present in the study wetlands.

The same factors were tested for influence on the number of species present from
within each species group (Table 3.). Only wetland size and percent of wetland area
semi-permanently flooded were found to affect the number of species present within a
group. Raptors (F =3.56; df=3; P=0.0291) and shorebirds (F =4.04; df=3; P=0.0184) both
showed differences between wetland sizes. Raptor species were more common on very

large wetlands than any other size with an average of 2.0 species/wetland versus

31

.Chmmdum 6_o.on~m mwmncv «850:
0080.8.“ 80.80: 05 8 38:03 35 8on 005086 05 mo cognac a an 28:03 a E 32030 «.85 no 860% m0 038:: 05. .2 03E

“330... .020”. E9: :5 3:305

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oom omN CON omw oow on o
p P P b P o
or
O O
. a
o
l 1» . Ill- ON
6 o #
m.
III! S
9 IE- on d
O a
. . 4 m.
o s
O
1 r o 8
Av
o
. o a"
l I- 8

 

 

 

 

 

 

 

32

.8880": 63.0"”: gang 8500320 525:
Rama .00 00.508 8080: 0% 80¢ 00586 05 m0 250:3 0 mm mun—0:03 $058 05 «0 E0805 820058 530 m0 0028:: SF .2 0.39m

0053.305 50:5: 59.1.5 005305

 

 

 

 

 

 

 

 

can com com cat com com 00— o
_ k p h p _ o
[I or
O A.
O O
O
0 O 0 cm
0
o, o #
o w.
o 0 I!“ ll, 0 on .m.
9 ‘ 01D], 0
O O m.
S
O
O
L- 3.
O
n 0 Av
\ L - x - i I- on

 

 

 

 

oo

33

 

.Emt 9 :2 80¢ 0N8 5 08.0205 8050003 80:03:00 :02? 85 082 .8026": 3&0 amen": 80:000.,» AS32009.
000 n :02: KHE 00:53:: 0:0 $082009. mmd n :02: JNME 00:53 5505 00?:030 8.060% 530 00 38:00 2F .M: 055

280 05:25
mNmNmForNNNr—«vwwrmmvkmmNvNomovmrkanFvwvaNonw

00.0

S

d

a

8.9 m.

a

s

o

a.

s

a

. w

82 P

W

e
00:52:- 1

‘ : .1 i 1 \\ 1 ‘\; l x W 8.8

7
00ccm_ac3D 7

2‘ 1 w ‘ w 1: ‘ 1:1 1 \ ‘ 00.3

34

Table 2. Results of AN OVA tests for influence of wetland size, % emergent cover, %
open, and % semipermanently flooded on number of species groups.

 

 

Factor P-value Significant
Wetland Size 0.1589 No
% Emergent 0.7034 No
% Open 0.3428 No
% Semipermanently Flooded 0.6882 No

 

35

95:83.00 «00:0m:0m :0.“ 80080:: 003 voodvm :0 00800 8:02.038 ......
3:02.038 £00850 _..

 

Sofie .. vowed mound wow _.o oammd Ew—d ammo 00000E
m_0:0:08:0:::0m a\..
002.0 Good 223 ... wmmod owned ... 386 good :0:O ..\o
0053 wmomd Emmd $3.0 Smwd wmmod gmmd 0:0w:08m X.
336 ... £56 wmcvd ... Emod flamed Nmovd 336 0N5 0:00.03
080E855 80:50:25 £00m 8:00:0M 050.880? 8.0000? 800m wfizm 800$

 

000:» 806008 5000 5505 0030890 82008 .00 00:5: :0 0008a
30:05:80.:808 .x. 0:0 .:000 ..\o .:0>00 E09080 o\o .0n8 05:03 .00 00:0:c:_ 00m 883 ......Amofigiv <>OZ< m0 80:80.»: .m 030,—.

36

averages of 0.6, 0.6, and 0.8 species/wetland for small, medium, and large wetlands
respectively. Shorebird species were more varied on large wetlands with 7.6
species/wetland than small (2.6 spp./wetland), medium (4.5 spp./wetland), or very large
(5.2 spp./wetland) wetlands. Shorebird species were most common (F=3.94; df=3;
P=0.0204) at wetlands with 26-50% of area normally flooded with an average of 9.0
species/wetland. This compares averages of 4.0, 4.9, and 3.1 species/wetland at wetlands
with 0-25%, 51-75%, and >75% of area normally flooded. Wetlands with 26-50% of
their area normally flooded may produce the desired mudflat habitat more readily than

the other categories of wetlands examined.

The contour of the wetland bottom turned out to be the most important factor
examined in this study. Wetlands were grouped into 4 bottom contour classes: 1) gentle
slope/smooth bottom (n=8); 2) gentle slope/diverse bottom (n=3); 3) steep slope/smooth
bottom (n=9); and 4) steep slope/diverse bottom (n=7). Bottom contour was found to be
significant in 10 of 18 comparisons (Table 4, Table 5). The average number of species
seen per visit was significantly lower (F=4.23; df=3; P=0.0161) at wetlands with a steep
slope and smooth bottom (7.1 species/wetland/visit) than at wetlands with the other
contour types (13.7 gentle smooth, 16.0 gentle/diverse, 19.9 steep/diverse)(Figure 19).
However, total species numbers were not affected by bottom contour (F=2.32; df=3;
P=0.1023) although steep/smooth bottoms had the lowest average total (22.3 species)
compared to 31.6, 30.7, and 34.0 species for gentle/smooth, gentle/diverse, and

steep/diverse respectively.

37

Table 4. P-values associated with tests for differences in species use among 4 classes of

wetland basin contour configurations.

 

 

P-valuea Best contour combinationb
Total # species 0.1023 All but steep/smooth
Avg. # species/visit 0.0161 All but steep/smooth
Species/ha 0.4949 Steep/smooth
# of species groups (8)c 0.1245 No best combination
Diving birds (5)d 0.0523 Steep/diverse
Waders (5)d 0.0440 Steep/diverse
Waterfowl (12)d 0.01 78 Gentle/diverse
Raptors (10)d 0.1718 Gentle/smooth
Rails (2)d 0.1943 Diverse bottom
Shorebirds (13)d 0.0006 Diverse bottom
Gulls/tems (4)d 0.1511 Steep/diverse

 

3‘ Results of one-way AN OVA with n=27.

b Wetlands were classed as having either gentle or steep slopes and smooth or diverse
bottoms. The best combination had the highest average level for the variable examined.
° The total number of groups that were possible (7 wetland and 1 other).

d The total number of species possible within the groups.

38

Table 5. P-values associated with tests for differences in habitat variables among 4
classes of wetland basin contour configurations.

 

 

P-valueal Best contour combinationb

Wetland size 0.1448 None
Wetland Age 0.0111 Gentle=young, Steep=old
% semiperrnanetly

flooded 0.0001 Lowest in gentle/smooth
°/o emergent vegetation 0.0001 Highest in gentle/smooth
% open water/nonemergent

vegetation 0.0001 Highest in steep/smooth
Maximum depth 0.0002 Lowest in gentle/smooth
Average Depthc 0.0008 Highest in steep/smooth

 

“ Results of one-way ANOVA with n=27.

b Wetlands were classed as having either gentle or steep slopes and smooth or diverse
bottoms. The best combination had the strongest influence on the variable examined.
° The average depth from the transect data. The average depths were not standardized
and cannot be compared directly.

39

.9080!“ 8&0
”3.0an 80003500 0:00:00 :800 v 00 0000 E 0:000? 0000 E “80> :0: 0020800 800008 :030 00 00:5: 0w000>< .3 0:35

2300:0003
m_. V m 0 0 mm 0 mwhmom N rwomwr m N VNNF v mNOFNva rpm—.mwkw

 

00.0

00.0

000—.

00.0w

 

00.0w

 

 

 

000m

3 4° # afieJaAv

  
  

 

00.020300wa

 

00.00

salaad

585030208

 

00.00
020>5$=000U

 

1r 00.00

 

SooEw\0_EwOI

 

 

00.0»~

40

Bottom contour had a significant effect (F =3. 1 6; df=3; P=0.0440) on the number
of species of wading birds detected at a wetland with the highest average for
steep/diverse wetlands (2.0 species/wetland) followed by steep/smooth and gentle/diverse
at 1.3 species/wetland and gentle/smooth with 0.9 species of wading birds per wetland.
Waterfowl species numbers were also significantly influenced (F=4. 12; df=3; P=0.0178)
by wetland bottom contour. The number of different species of waterfowl were highest
at wetlands with diverse bottoms (5.0 and 4.3 species/wetland for gentle and steep slopes
respectively) compared to wetlands with smooth bottoms (2.0 and 3.0 species/wetland for
gentle and steep slopes respectively). Shorebird species were also significantly (F =8.45;
df=3; P=0.0006) more varied at wetlands with diverse bottoms (7.3 and 7.42
species/wetland for gentle and steep slopes respectively) compared to smooth bottoms

(3.5 and 2.7 species/wetland for gentle and steep slopes respectively).

Wetland bottom contour was also significantly related to many of the other
wetland factors that were examined. Wetlands with gentle slopes were significantly
(F=4.64; df=3; P=0.0111) younger (3.8 and 3.0 years vs. 6.8 and 5.3 years for smooth
and diverse respectively) than steep sloped wetlands. Bottom contour significantly
affected percentages of wetland area supporting emergent vegetation (F=15.51; df=3;
P=0.0001, Figure 20), open water (F=15.41; df=3; P=0.0001, Figure 21), and
semiperrnanent flooding (F=19. l 3; df=3; P=0.0001, Figure 22). Bottom contour was also
significant in determining average (F =7.98; df=3; P=0.0008) and maximum depths
(F=10.43; df=3; P=0.0002) of wetlands. Wetlands with gentle slopes and smooth

bottoms were shallowest on average (2.0 cm avg., 42.2 cm max.) compared to

41

.Qooodum MmH00
:m.m Mun: 300050050 0:00:00 :80: 500 00 0000 E 000003 0000 E 000000 380.008.8058 00.0 0:000? .00 0:00:00 .0N 050E

3000:0003
Mr V m 0 0 MN 0 mwhmom N Fwomww m \. VNNV F mNOFNva wrormwhv

 

 

 

 

 

 

 

 

0202900000.

291v puenaM moored

 

 

 

 

-1- S
Ecosmaooawl
0202900000D m 00

g 1 om
,

EooEQEEool

 

 

 

00F

42

.Cooodnm mm&0 20.2an
30005500 58:00 £000 0 .00 0000 E 080003 0000 3 5000002, E09080 0580008 000 0:000? 00 0:00:00 Am 0530

2800:0003
mp v m 0 0 MN 0 mwhuou N vwommr m N VNNr v mNOFNer Frmvmwnw

 

 

 

43

 

 

 

 

1 1111M1111J
020303005'
11- 1-11 1 585030200 1 1

11 11 110 0000>5\00000D '11 11

 

231v puenaM moored

 

 

 

 

 

 

 

00v

 

208.on anus a; We aoueswfis
59:00 303 v m0 :000 E 05:03 :08 E :0w80w0> 309080.:050803 :oqo mEES:00 02: 05:03 90 0:020.” .NN 2&5

£30 2.2.35

mpv m m ammomwhwmww PNomme NVNNFFmNoFNNvFFFoFaFNF

 

 

 

 

 

 

 

 

080203005-

58503020!

0902 90:00 O D

 

 

 

11” :1“ £8 E 22:5 0 l

 

 

ON

on

om

oow

231v puenaM auamad

44

gentle/diverse (35.5 cm avg., 144.7 cm max.), steep/smooth (68.5 cm avg., 139.7 cm

max.), and steep/diverse (33.6 cm avg., 150.0 cm max.).

Maximum depth of wetlands was found to be not significant for all comparisons
(Table 6). Wetland age and average depth were not examined as factors that may
influence bird use of a wetland due to a number of confounding factors. Analysis of
wetland age is confounded by changes in wetland designs as the science of wetland
creation has grown and continues to learn from the past. Average depth was not used
because I could not standardize the average depths so wetlands could be compared on an
even basis. This was due to several factors. The limits of time kept me from collecting
all habitat data at the same time. This was important since the wetland water levels
constantly changed in response to rainfall or lack thereof. This meant that the exposed
areas of mudflats was constantly changing and therefore would affect the calculation of

average depth, but not maximum depth.

45

Table 6. Results of comparisons of avian species use as a function of maximum depth of
study wetlands.

 

 

n R2 P-value
Total number of species 27 0.004 0.7409
Average number of species/visit 27 0.015 0.5437
Total Species/ha 27 0.037 0.3356

 

46

DISCUSSION

It should be noted that the 129 species recorded during this study is nearly 32% of
the 409 species accepted as having ever been recorded in the state of Michigan (Byrne
1997). Even the 110 species recorded within the wetlands is an impressive 27% of the
state list. These percentages are even more impressive when you consider that they were
recorded during a single season at a limited habitat. The study period missed much of the
migration season and only looked at wetland habitats. The study wetlands were
obviously attracting some species from the surrounding habitats as I recorded many

species that are not necessarily considered wetland species.

It is interesting to note that although a total of 129 species were recorded during
this study, a maximum of 62 species was recorded at a single wetland. This makes it
clear that the mix of species present at individual wetlands could vary greatly. These
different mixes likely result from the varying mixes of habitats present in and around
each wetland and likely cannot be explained by any one factor. The factor with the

strongest influence on species numbers is wetland size.

It is not surprising that wetland size was significantly linked to the number of
species recorded within a wetland. Other studies of created and natural wetlands in
Michigan and elsewhere have found that large wetlands tend to have higher species
diversity and numbers than do small wetlands (Soulliere and Monfils 1996, Wasilco and
Soulliere 1995, Brown and Dinsmore 1986) as can be expected by Island Biogeography

Theory. These wetlands in many cases are probably acting as habitat islands in an

47

otherwise dry landscape (Cody 1985). The fact that species numbers observed in
medium and small size wetlands were so similar is most likely due to a combination of
many factors including the fact that wetlands ranged from small, forested wetlands to
large open water wetlands and large emergent wetlands. Most of the medium sized
wetlands were open water/emergent wetlands while several of the small wetlands were
forested or heavily vegetated. The average number of species seen per wetland visit was
also significantly linked to wetland size, likely for the same reasons as total species
numbers were. If there are more total species there then the higher the average number

should be.

It should also come as no surprise that the amount of area supporting different
habitat types was significantly related to species diversity at wetlands. The area of a
habitat type present in a wetland is limited by the size of the wetland. The larger the
wetland the larger each habitat can be. The more important factor to consider when
comparing two wetlands is the percent of wetland area each habitat type constitutes.
These percentages had no significant effect on total species present, but they are

important in determining which species are likely to be present.

It was somewhat unexpected to find that the distance to the nearest cropland
habitat was significant in determining the number of species present at a wetland. The
fact that the distance to cropland was the only significant factor other than wetland size
might be due to the fact that that was the only habitat type that was not found within 1000

m of every wetland site. Therefore, the reduced species totals may have more to due with

48

the fact that most wetlands without nearby cropland were located in a suburban/urban

setting.

It was also unexpected to find that the distance to the nearest wetland was not a
significant factor in determining species numbers. The more isolated a wetland the fewer
species one would expect (Brown and Dinsmore 1986). This may have not been a
significant factor in this study due to the relative closeness of other wetlands to each

study site. The farthest a bird would have to travel to reach another wetland was 500 m.

The distance from the nearest road was also a factor that I thought might be
significant at the start of the study. The close proximity of all but one wetland (most less
than 50 m) to major roadways likely lessened the impact seen at the study wetlands. This
factor may have been more important if some of the wetlands had been isolated from
roadways. The proximity to roads may have kept some species away, but it seemed like

the passing cars did not disturb the species that were present.

It should be noted that bottom contour affected the average number of species
present, but not the total number. This could be taken to mean that the contour of a
wetland basin is important in determining how many species will be present on a regular
basis, but not in determining how many different species might stop in during a season.
This may be due to the fact that wetland bottom contour was important in determining the
percentages of each habitat type present as well as the average and maximum depths of

each wetland.

49

This influence of wetland basin contour on habitat availability should be fairly
obvious. A wetland with a smooth sloping bottom is going to look like a bowl filled with
water. If the slope is gentle the wetland may be fairly shallow over a large percentage of
its area and be well vegetated. However, if the slope is steep, the wetland may only be
shallow at the very edge, and support only a narrow ring of vegetation. This is what
many older created wetlands look like (Mitsch and Wilson 1996). This is often due to
the fact that these wetlands were created in the most economical way possible since it is
easier to make a nice smooth basin than a basin with a diverse structure (Hollands 1990)
and it is difficult to predict the final result of a planned wetland (McKinstry and
Anderson 1994, Erwin 1990, Niering 1990). Wetlands with diverse bottom contours
provide a better mix of habitats throughout the wetland. Again, a gentle slope may lead
to a shallower wetland that is well vegetated, but in a more random pattern as water
depths and flooding regimes vary across the wetland. The same is true for steep slopes.
An added benefit of steep sloped diverse bottomed wetland is the fact that it tends to have
deeper holes that will remain flooded when the rest of the wetland has dried due to

drought.

Another factor influencing and strongly linked to the percentages of wetland
supporting emergent vegetation or open water is the percent of the wetland that is
semipermanently flooded. This is especially true for the open water/non-emergent
vegetation habitat. Areas that are mudflats when dry are by definition open water when
flooded. Many of the study wetlands had a fringe of emergent vegetation around the

edges with open areas toward the center. Many if not most of the open areas were open

50

due to flooding during much of the growing season. Areas that are continuously flooded
past a certain depth are unlikely to support emergent vegetation due to the inability of the
vegetation to become established in standing water (Fredrickson and Taylor 1982,
Bishop et al. 1979). The changing area of flooded habitat in a wetland is key to
determining the patterns of vegetation types that will become established (van der Valk

1981).

The significant differences seen within the species groups that were tested were
all likely due to the habitat preferences of the species within those groups. Wading bird
species were most abundant in wetlands that had lower amounts (<75%) of emergent
vegetation and high amounts (>25%) of open water as well as steep slopes and diverse
bottoms. The wading bird species were dominated by great blue heron and used the
study wetlands mainly as foraging sites. Great blue herons prefer to forage in open areas
often from exposed mud flats and sandbars (Short 1985). Wetlands with steep slopes
and diverse bottoms tend to have more mudflats and sandbars adjacent to deeper water
that would contain the small fish and other invertebrates that the herons feed on. It
should be noted that these results are based upon the observations of a maximum of three

species of wader at any one wetland.

Raptors were most varied at wetlands with higher amounts (>5 0%) of emergent
vegetation and low amounts (<50%) of open water. This is likely not a true preference as
these results are based on a maximum of 3 species of raptors recorded at any one wetland

and no raptors being recorded at 12 of the 28 wetlands examined. Raptors were also

51

recorded as using the wetland if they were seen flying within sight of the wetland since

they hunt over such large areas.

Waterfowl species were most numerous in larger (>1 ha) wetlands with diverse
bottoms with >25% open water. The diverse bottoms caused this open water to be
interspersed with emergent vegetation. This is the expected habitat preferred by
waterfowl. There were four main species of waterfowl present throughout the study,
mallards, canada geese, wood ducks, and blue-winged teal. These are also the four most
commonly breeding waterfowl species in southern Michigan (Brewer et al. 1991). These
species were all observed raising broods in more than one wetland as well as using the
wetlands for feeding and resting. Canada geese and blue-winged teal were found to be
nesting in the upland cover within 20 m of the wetlands. The other species of waterfowl

recorded were likely using the wetlands as migratory stopover areas.

Shorebird species were most numerous on large (3-6 ha) wetlands with 25-50% of
the wetland area semipermanently flooded and 25-75% of the area open water/non-
emergent vegetation in a diversely contoured basin. This combination of habitat factors
tends to produce large areas of mudflats, which are the preferred foraging habitats of
most species of shorebirds (Arnold 1994). Two species of shorebirds used the wetlands
for reproduction and raising young (killdeer and spotted sandpiper). Most of the

Shorebird species were using the wetlands as migration stopovers and foraging sites.

52

The study wetlands provided nesting and brood rearing habitat for many species.
This was especially true for the diving bird, waterfowl and rail species, which are
dependent on wetlands to raise their young. There nearby forested areas provided nesting
sites for wood ducks and the adjacent upland grass areas provided the nesting sites for the
mallards, tea] and geese. In the case of the diving birds and rails the wetland vegetation

provided the necessary habitat for nesting as well as brooding.

The most important fiinction the study wetlands played from a wildlife standpoint
was providing foraging habitat for both resident and migrant birds. This was probably
especially true during the late summer and fall migration. During the spring migration,
there is often water in all of the temporary and seasonal wetlands, but these are dry by the
time the fall migration commences. The study wetlands for the most part still contained
at least some water during the fall migration. These wetlands were heavily used by
shorebirds as well as other migrant and resident birds as foraging locations. During the
dry summer of the study season, many upland species were observed visiting the study

wetlands for a drink of water, demonstrating the importance of wetlands for all wildlife.

53

CONCLUSIONS

The main conclusion to be drawn from this study is that the bigger and more
diversely contoured the wetland the better. There are likely many other factors that affect
the number of species of birds that will use a wetland. These include the vegetation
within the wetland as well as the habitats that surround the wetland. The distance to
another wetland is also likely to have an effect if the wetland is very isolated. The most
important conclusion of all is that some wetland habitat is better than none at all since
even the smallest, most poorly vegetated wetland that I looked at still was used by at least

12 species of birds.

Further studies should concentrate on determining which factors are important
within wetland size categories. Hopefully, this will allow managers to create
recommendations for maximizing species diversity within different size classes of
wetlands, since the creation of a wetland is expensive and no one will want to create a
larger wetland than they have to. It is important to note that while each size of wetland
may attract a different number of species, they may all attract different species depending

on the vegetation in the wetland.

54

RECOMMENDATIONS

The best recommendation that can be provided at this point is to create wetlands
that replace the same type of wetland habitat as was lost. If this is not possible then it
seems best to create the largest, most diverse wetland possible. However, it is likely
better to create a small, forested wetland than a large open pond. The best solution is
probably to create a large wetland with a diversely contoured bottom and areas of both
steep and gentle slopes to provide the best mix of habitat types to attract the most diverse

collection of avian species.

55

APPENDIX A

56

Table A1. The common and scientific names of all species recorded during the study.
The number of observations is the number of different points the species was recorded at
during the study period, and has no reflection on the number of individuals recorded each

time.

 

 

Number of

Common Name Scientific Name Observations
Song Sparrow Melospiza melodia 321
Red-winged Blackbird Agelaius phoeniceus 303
Mallard Anas platyrhynchos 251
Savannah Sparrow Passerculus sandwichensis 243
Killdeer Charadrius vociferus 208
Barn Swallow Hirundo rustica 198
Spotted Sandpiper Actitis macularia 146
American Goldfinch C arduelis tristis 141
Great Blue Heron Ardea herodias 123
Tree Swallow Tachycineta bicolor 114
Bank Swallow Riparia riparia 96
Eastern Kingbird Tyrannus tyrannus 95
Canada Goose Branta canadensis 91
Common Grackle Quiscalus quiscula 90
Blue-winged Teal Anas discors 79
Common Yellowthroat Geothlypis trichas 78
Solitary Sandpiper Tringa solitaria 69
American Robin T urdus migratorius 61
Mourning Dove Zenaida macroura 58

57

Table A1 (continued).

 

 

Number of

Common Name Scientific Name Observations
Pied-billed Grebe Podilymbus podiceps 56
Lesser Yellowlegs Tringaflavipes 49
Least Sandpiper Calidris minutilla 45
Cedar Waxwing Bonbycilla cedrorum 45
Eastern Meadowlark Sturnella magna 43
Yellow Warbler Dendroica petechia 42
Chimney Swift C haetura pelagica 41
Northern Rough-winged Stelgidopteryx serripennis 40
Swallow

Wood Duck Aix sponsa 36
European Starling Sturnus vulgaris 34
Greater Yellowlegs T ringa melanoleuca 33
American Coot F ulica americana 31
American Crow Corvus brachyrhynchos 31
Ring-billed Gull Larus delawarensis 3O
Belted Kingfisher Ceryle alcyon 28
Green Heron Butorides virescens 27
Gray Catbird Dumetella carolinensis 24
Green-winged Teal A nas crecca 23
Indigo Bunting Passerina cyanea 22
Red-tailed Hawk Buteojamaicensis 21

58

Table A1 (continued).

 

 

Number of

Common Name Scientific Name Observations
Black-capped Chickadee Parus atricapillus 21
Double-crested Phalacrocorax carbo 20
Willow Flycatcher Empidonax traillii 20
Blue Jay C yanocitta cristata 18
Common Snipe Gallinago gallinago 18
Northern Cardinal Cardinalis cardinalis 16
F orster's Tern Sternaforsteri 15
Downy Woodpecker Picoides pubescens 14
Baltimore Oriole Icterus galbula 13
Rock Dove Columba livia 12
Sora Porzana carolina 11
Pectoral Sandpiper Calidris melanotos 9
Eastern Bluebird Sialia sialis 9
Yellow-rumped Warbler Dendroica coronata 9
Great Egret Ardea alba 8
Bobolink Dolichonyx oryzivorus 8
House Sparrow Passer domesticus 8
Hooded Merganser Lophodytes cucullatus 6
Swamp Sparrow Melospiza georgiana 6
Brown-headed Cowbird Molothrus ater 6
Chipping Sparrow Spizella passerina 6

59

Table A1 (continued).

 

 

Number of
Common Name Scientific Name Observations
Eastern Phoebe Sayornis phoebe 6
Ruddy Duck Oxyurajamaicensis 5
Warbling Vireo Vireo gilvus 5
American Kestrel F alco sparverius 4
Vesper Sparrow Pooecetes gramineus 4
Semipalmated Sandpiper Calidris pusilla 4
Semipalmated Plover C haradrius semipalmatus 4
Sandhill Crane Grus canadensis 4
Rusty Blackbird Euphagus carolinus 4
Red-bellied Melanerpes aurifi'ons 4
Turkey Vulture C athartes aura 3
Eastern Wood-pewee Contopus virens 3
Tufied Titmouse Parus bicolor 3
American Black Duck Anas rubripes 3
American Pipit Anthus rubescens 3
Dowitcher species Limndromus spp. 3
Ruby-throated Hummingbird Archilochus colubris 3
Cooper's Hawk Accipiter cooperii 3
Grasshopper Sparrow Ammodramus savannarum 3
Northern Flicker Colaptes auratus 3
Great Crested Flycatcher Myiarchus crinitus 3

60

Table A1 (continued).

 

 

Number of

Common Name Scientific Name Observations
House Wren Troglodytes aedon 2
Western Sandpiper Calidris mauri 2
American Wigeon Anas americana 2
White-breasted Nuthatch Sitta carolinensis 2
Black Tern Chlidonias niger 2
Northern Harrier Circus cyaneus 2
Lincoln’s Sparrow Melospiza lincolnii 2
White-throated Sparrow Zonotrichia albicollis 2
Dunlin Calidris alpina 2
Gadwall Anas strepera 2
Palm Warbler Dendroica palmarum 2
Purple Martin Progne subis 2

American Bittem
Yellow-billed Cuckoo
House Finch

Virginia Rail
White-crowned Sparrow
Hairy Woodpecker
LeConte's Sparrow
Solitary Vireo

Broad-winged Hawk

Botaurus lentiginosus
C occyzus americanus
C arpodacus mexicanus
Rallus limicola
Zonotrichia leucophrys
Picoides villosus
Ammodramus Ieconteii
Vireo solitarius

Buteo platypterus

61

Table A1 (continued).

 

Common Name

Scientific Name

Number of
Observations

 

Common Loon
Caspian Tern

Field Sparrow

Merlin

Northern Pintail
Ring-necked Pheasant
Ring-necked Duck
Red-breasted Nuthatch

Sharp-shinned Hawk

Gavia immer

Sterna caspia
Spizella pusilla

F alco columbarius
Anas acuta
Phasianus colchicus
A ythya collaris

S itta canadensis

Accipiter striatus

1

l

 

62

Table A2. The species that constituted each of the 7 wetland species groups.

 

Species Group

Common Namea

 

Diving Birds

Waders

Waterfowl

Raptors

Common Loon
Pied-billed Grebe

Horned Grebe
Double-crested Cormorant
American Coot

American Bittem
Great Blue Heron
Great Egret
Green Heron
Sandhill Crane

Canada Goose
Wood Duck
Green-winged Teal
American Black Duck
Mallard

Northern Pintail
Blue-winged Teal
Gadwall

American Wigeon
Ring-necked Duck
Hooded Merganser
Ruddy Duck

Turkey Vulture
Osprey

Northern Harrier
Sharp-shinned Hawk
Cooper’s Hawk
Red-shouldered Hawk
Broad-winged Hawk
Red-tailed Hawk
American Kestrel
Merlin

63

Table A2 (continued).

 

 

Species Group Common Nameal
Rails
Virginia Rail
Sora
Shorebirds

Semipalmated Plover
Killdeer

Greater Yellowlegs
Lesser Yellowlegs
Solitary Sandpiper
Spotted Sandpiper
Semipalmated Sandpiper
Western Sandpiper
Least Sandpiper
Pectoral Sandpiper
Dunlin

Dowitcher Species
Common Snipe

Gulls/T ems
Ring-billed Gull
Caspian Tern
Forster’s Tern
Black Tern

 

a Any species listed in Table A1 and not listed here was placed into the eighth species
group for comparisons on all species groups.

APPENDIX B

65

02200050 m c: 9.30 a a» a $3. 2

 

500890—800 m 2 mod 0 3 E m _ mm.“ _ _
”0008900000 0 mm on; Q 2 m0 vad 3
00039080 a 8 8% mm an cm «3.5 a
0E0>5E08m w m2 omde 3 cm H fl mmmhd w
£008mE000m w #2 0 _ . _ m 00 no mm mm m wd N.
0E0>mea000m x mm _ m _ . 8 mm R g g 030; 0
58800800 0 3 ~02 on S ”N $33 0
0000>5E08m m _ w fl an . _ _ 2 M3 0m womwd v
0839330 m S 0 8.: 3. mm 2 025. M
585080 0 ow E. _ m K : mm 53.: N
0003500000 N 0m 00.0 mm C ox on 8.0 _
0.500030 00?. A.Aaov500a «9:35:09 w5000E 00003 :0030w0> 05006 000003
£00m 0:000? 8:80:02 0w000>< 0:0:0::aw:0m .x. 0090 .x. 0:0w008m °\°

 

0000003 >030 05 00 £000 £505 00000.0 00030: 0000 .«0 £0.64 .5 030,—.

66

0000330350 N 3N NY: 3” m0 om Nammd 0N

 

80200800 m a: :00 om 00 s. can? 8
08800050 m 2 80 o N 00 39.0 mm
5880085 0 03 8.2 K S om 083 8
5889085 0 5 8.8 i a a 88.0 on
58800080 0 3 $0 2 o E SE 2
58808on m we 8.2 _ mm a a 800.0 2
58800050 m 8 2:. mm a S Echo :
58800080 m R of m o 8 80$ 2
5860880 0 m2 :1 : 8 a w 8%.: 2
58500080 m 8 00m 2 a E 083 E
geezeaoem m x: 03m 00 Q 3 com; 2
00:00:00 003.. ..Aaovfiga «Afiovfiga w5000E 00003 :0030w0> 053% 0:000?
500m 000003 808082 0w80>< 000008000303 .x. 0000 .x. 0:0w008m o\c

 

€85.80 5 03.3

67

0:000? 0:0—00:0— 0w:0_ 0:0 :0: 0808000 0:00:00, 0:0? 00:»:

00:30: 80000 00 :0 0m0:0>00 800:0: 00:. 0:00 00% 0:0 080 0w:0_ b0> 0:: 0: 0:0 mm 0:000? :0 080:0:00 :0: 0:0? 0088:. ...
80:00:00 082:: 0: ?0=0:m 0: :000

0: ?0=0:m 0: 82:: 00080 0 800 8000: 00 :0 200:: 00 00w:0:0 :00 8080.: 0 w:0_0 :80: 0:0 :00 0:08 :0: 803A :0 0:00 8
080:0 0 00: 08000: 0m:0>_Q .?0: 0 8 0:80: 8080:: 00:00:00 :00?:0: 00:00.00 803A 00: 00:20 :80 .::0:00:w 8000m\0:0_m 0
$080 00 :0 0:0 00 0: 8008800 0080 800000 w8?0:w :0 008:2 0

0808000 :00 0:000 0: 00800 :08 0:0 00008000 0:0? 800» :A 00:0Q 00080::

:0 00:0:::00:0 0:00 8:80:08 0:0 00w:0:0 0w:0w 0:00 :0 0000: £080 00 w:_::0 0:00 8:80:08 00 0: 0: 0080:0080 a
.00000 000:800 0: :0: :00 00:000? 00 .300 :08800 0 0: 0080:0080: :0: 0:0 0:00 8080.: 80.0 00030.00 8

 

 

-- n -- -- v: m: N 000.: ”mm
5880080 : mm: 8.? a a n 8:: R
0008m\:00:m 0: :mm omdm: mm mm : ommm: mm

00089050 m 0w vmd m _ an 08.3 mm
080800 U0?: n2830:0Q aA80v0:0a w::000_: 0:03 :0000w0> @0000 0:000 3
800m 0:000 >7 8:80:02 0w0:0>< 8808:0880 c\o :0:O c\o ::0w:08m .x.

 

€8,880 :0 2%:

68

Table BZ. Distances (m) from each wetland to the nearest area of each surrounding
habitat.

 

 

Upland
Wetland Road Forest Wetland Cropland House“ Grass
1 20 130 20 110 550 0
2 70 1 50 1000 155 o
3 25 200 120 130 400 o
4 15 225 120 110 370 0
5 110 16 26 850 310 0
6 40 26 60 350 75 0
7 55 1 90 600 200 0
8 51 18 60 670 55 0
9 20 17 60 450 60 0
10 55 0 5 10 320 0
11 20 150 12 35 320 o
12 45 170 170 45 200 o
13 33 150 12 5 320 0
14 4o 0 250 30 70 35
15 14 57 500 700 225 0
16 21 0b 40 70 160 0
17 30 1 500 1000+ 190 0
18 45 15 10 1000+ 500 o
19 3 0b 50 1000+ 300 0
20 300 10 40 1000 300 0
21 75 15 20 1000» 30 0
22 27 1 5 1000 170 0
23 20 130 20 110 600 o
24 25 180 110 170 500 0
25 20 o 250 1000+ 120 35
26 15 260 170 1000+ 300 0
27 37 50 30 170 0 0
28 30 15 30 170 0 0

 

’ The distance to the nearest house represents the distance to the nearest source of regular
human disturbance. This distance was set to zero for wetlands 27 and 28 since they were
open to the public for fishing and were used by people daily.

b These wetlands were forested and the distance to forested habitat was less than zero
since >50% of the wetland supported mature trees.

69

o 0 o o v _ 0 0 end 0N MN
o m o m N _ o 0 50.3 _N NN
o m o N m N o 0 :05 Nm :N
o m o o v N o m 0%: cm cN
o _ _ _ o o o m 50.: NN 2
o 0 o o _ _ _ 0 :0 S w:
o N o _ 0 _ o :0 mmé : S
o N o N 0 0 o 0 cod: N0 3
0 0 o N m _ o m mm.w :N m:
o m o _ 0 0 o v 00. : Nm 3
_ m N N m N 0 N. mmdv >0 2
0 w H o b N N 0 mmd: mm N:
_ 0 o o v 0 o 0 cos 0N S
o w o m N _ o 0 mNAN w: 2
N w c o 0 m 0 m 50.3 mm a
N m o o m N N m 3N: om w
0 m o o m N 0 m 86 RN 5
N w o _ 0 m _ o 2.2 :m 0
o m o o m 0 o m 00.0 m: m
o 0 o 0 m N o v 09:: 3 0
o 2 o o 0 0 o m 0m.NN 0m m
o m c 0 m N _ m 3.: Nm N
o 0 0 0 m o o 0 09m ON —

 

8:68:00 00220 20: 83%: 36.060? 828? 0000 .385 80228030 .8060 0:006?
w8>5 000025 0 00:03: 00:.

 

0:000? 0000 :0: 800 00: 880:0 .00 @0880 .mm 030%

70

0:000: 0::0:w 000:0 0:000? h :0 :80: 00 :0 0:0 6
003:0 8:0: 000 w::.::0 000:0 :0 008:: 00:03:. a 1
0000800 00:000? :0 :0: 000:0 c: : :0 0:0 0:000? 00 w::m: 000:000: 00:00:m a 7

 

m 0 o : 0 N : 0 0:..VN 00 wN
o o o o m : o N mud N: :N
: o o o N o : m mfm N: 0N
: m o m m : o m ooNN D: mN
: e: o N m N o m : :.0N hm 0N

 

080.0330 00:50:25 005: 0:00:00 :?0.:0:03 00003 000m: o0::05 A100200003 a00:00:m 0:000?
w::>:Q 00:00:m 0 00:02: 000:.

 

06:50:80 8 03S

LITERATURE CITED

Arnold, K. A. 1994. Common Snipe. Pages 117-125 in T. C. Tacha and C. B. Braun,
eds. Migratory shore and upland game bird management in North America. Allen
Press, Lawrence, Kansas. 223 pp.

Beck, R. E. 1994. The movement in the United States to restoration and creation of
wetlands. Nat. Resour. J ., 34(4):781-822.

Bishop, R. A., R. D. Andrews, and R. J. Bridges. 1979. Marsh management and its
relationship to vegetation, waterfowl, and muskrats. Proc. Iowa Acad. Sci.
86:50-56.

Brewer, R., G. A. McPeek, and R. J. Adams, Jr. 1991. The atlas of breeding birds of
Michigan. Michigan State University Press, East Lansing, Michigan 594 pp.

Brinson, M. M., and R. Rheinhardt. 1996. The role of reference wetlands in firnctional
assessment and mitigation. Ecol. Appl., 6(1):69-76.

Brown, M. and J. J. Dinsmore. 1986. Implications of marsh size and isolation for marsh
bird management. J. Wildl. Manage. 509:392-397.

Byrne, A. M. 1997. The Michigan bird list. Michigan Birds and Natural History 4:339-
346.

Cody, M. L. 1985. An introduction to habitat selection in birds. Pages 3-56 in M. L.
Cody, ed. Habitat selection in birds. Academic Press, Orlando, Florida. 558 pp.

Cowardin, L. M., V. Carter, F. C. Golet, and E. T. LaRoe. 1979. Classification of
wetlands and deepwater habitats of the United States. US. Fish Wildl. Serv. Pub.
FWS/OBS-79/31, Washington, DC. 103 pp.

Environmental Laboratory. 1987. Corps of Engineers Wetland Delineation Manual.
Technical Report Y-87-1. US. Army Engineer Waterways Experiment Station,
Vicksburg, Miss., USA.

Erwin, K. L. 1990. Wetland evaluation for restoration and creation. Pages 429-449 in J.
A. Kusler and M. E. Kentula, eds. Wetland creation and restoration: the status of
the science. Island Press, Washington, DC. 594 pp.

Fredrickson, L. H. and T. S. Taylor. 1982. Management of seasonally flooded
impoundments for wildlife. US. Fish and Wildl. Serv. Res. Pub. 148. 30pp.

72

Gibbs, J. P. and S. M. Melvin. 1993. Call-response surveys for monitoring breeding
waterbirds. J. Wildl. Manage. 57(1):27-34.

Hollands, G. G. 1990. Regional analysis of the creation and restoration of kettle and
pothole wetlands. Pages 281-298 in J. A. Kusler and M. E. Kentula, eds.
Wetland creation and restoration: the status of the science. Island Press,
Washington, DC. 594 pp.

McKinstry, M. C., and S. H. Anderson. 1994. Evaluation of wetland creation and

waterfowl use in conjunction with abandoned mine lands in northeast Wyoming.
Wetlands, 14(4):284-292.

Mitsch, W. J ., and J. G. Gosselink. 1993. Wetlands. Second edition. Van Nostrand
Reinhold, New York, New York, USA.

Mitsch, W. J ., and R. F. Wilson. 1996. Improving the success of wetland creation and
restoration with know-how, time, and self-design. Ecol. Appl., 6(1):77-83.

Niering, W. M. 1990. Vegetation dynamics in relation to wetland creation. Pages 479-
486 in J. A. Kusler and M. E. Kentula, eds. Wetland creation and restoration: the
status of the science. Island Press, Washington, DC. 594 pp.

Short, H. L. 1985. Habitat suitability index models: great blue heron. US. Fish Wildl.
Serv. Bio. Rep. 82(10.99). 23 pp.

Soulliere, G. J. and M. J. Monfils. 1996. Waterbird use of a constructed wetland
complex in eastern upper Michigan (year 3). Mich. Dept. Nat. Resour., Wildl.
Div. Rep. No. 3262. 12 pp.

Sokal, R. R. and F. J. Rohlf. 1995. Biometry: the principles and practice of statistics in
biological research. Third ed. W. H. Freeman and Company, New York. 887 pp.

Storer, R. W. 1991. Pied-billed grebe, Podilymbus podiceps. Pages 100-101 in R.
Brewer, G. A. McPeek, and R. J. Adams, editors. The atlas of breeding birds of
Michigan. Michigan State University Press, East Lansing, Michigan, USA.

Wasilco, M. R. and G. J. Soulliere. 1995. Initial waterbird use of created wetlands in
eastern Upper Michigan. Mich. Dept. Nat. Resour., Wildl. Div. Rep. No. 3223.

10pp.

van der Valk, A. G. 1981. Succession in wetlands: a Gleasonian approach. Ecology
62:688-696.

Young, P. 1996. The “new science” of wetland restoration. Environ Sci. Technol.,
30(7):292A-296A.

73

Zedler, J. B. 1996. Ecological issues in wetland mitigation: An introduction to the
forum. Ecol. Appl., 6(1):33-37.

74