EVALUATING THE INFLUENCE WHITE-TAILED DEER HAVE ON WETLAND
VEGETATION TYPES WITH RESPECT TO WETLAND BIRDS AT SHIAWASSEE
NATIONAL WILDLIFE REFUGE, SAGINAW, MICHIGAN

By

Stephanie E. Longstaff

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
Fisheries and Wildlife - MASTER OF SCIENCE
2013

ABSTRACT
EVALUATING IMPACTS WHITE-TAILED DEER HAVE ON WETLAND VEGETATION
TYPES WITH RESPECT TO WETLAND BIRDS AT SHIAWASSEE NATIONAL WILDLIFE
REFUGE, SAGINAW, MICHIGAN
By
Stephanie E. Longstaff
White-tailed deer (Odocoileus virginianus), can be keystone herbivores in forest
ecosystems and greatly impact forest conditions. Results of previous deer herbivory research
could cause biologists concern when managing wetlands as suitable habitat for migratory
waterfowl since high levels of herbivory can negatively impact plant communities. There is a
knowledge gap on how deer use of wetland vegetation types may affect bird use of those same
wetlands. Our objectives were to 1) evaluate white-tailed deer habitat suitability within a
landscape dominated by wetlands, 2) quantify and compare use of wetland vegetation types by
white-tailed deer to use by wetland bird communities, and 3) quantify white-tailed deer
herbivory within various wetland types. We classified 1 x 1 m aerial imagery data in ArcGIS
v10.0 and developed a white-tailed deer habitat suitability index (HSI) model. We conducted
driving surveys during crepuscular hours alternating morning and evening sampling times per
week from May – August 2011 and 2012 in 3 wetland types. We constructed exclosures and
paired them with open areas in moist soil, perennial marsh, lakeplain prairie, and bottomland
hardwood forest vegetation types where we measured horizontal cover, vertical cover, species
richness, total above ground biomass, and seed biomass. Wetlands provide highly suitable deer
habitat, deer are using wetlands but not influencing wetland bird communities, and current deer
herbivory is not negatively impacting the composition and structure of wetland plant
communities. These results are being used to help guide management decision making at
Shiawassee National Wildlife Refuge.

ACKNOWLEDGEMENTS

I would like to thank all of the organizations responsible for helping fund this project and
make it possible. Primary funding was provided by the U.S. Fish and Wildlife Service,
Shiawassee National Wildlife Refuge, Michigan State University, and Ducks Unlimited.
Without their support this project would have never gotten off the ground.
A big thanks goes to my major advisor Dr. Rique Campa for helping guide me through
graduate school and shape me into a better scientist and researcher. I really appreciate all of his
encouragement, enthusiasm, advice, and wisdom throughout this project. It was a wonderful
opportunity to get to work with him on such a fantastic project and has opened many doors for
the future. I would also like to thank all of my committee members Dr. Scott Winterstein, Dr.
Shawn Riley (Department of Fisheries and Wildlife), Dr. Catherine Lindell (Zoology
Department), and Eric Dunton (Shiawassee National Wildlife Refuge) for their advice and
support. I would also like to thank Dr. Alexandra Locher at Grand Valley State University for
her considerable knowledge and help with all of the GIS and habitat suitability work.
Thank you to all the staff at Shiawassee National Wildlife Refuge, you really helped
make my field seasons possible. A special thanks to Steve Kahl, refuge manager, and Eric
Dunton, refuge wildlife biologist, for answering my endless questions and helping shape this
project into applicable results for the management of Shiawassee National Wildlife Refuge. I
would also like to thank Chris Haggard for his many long hours and hard work in the field
constructing countless exclosures, harvesting plants in the hot summer sun while being eaten by
mosquitoes, and even providing us with a fun field trip to the hospital along with a great story. I
iii

would also like to thank Don Poppe and Jim Bush for joining our field team on occasion, you
kept me laughing both summers and your hard work was greatly appreciated.
I would also like to thank all of my friends and family for supporting me throughout
this process. Thanks to all my friends for listening and supporting me along the way. Thanks to
my parents Tom and Debbie Longstaff and Vicki Longstaff, and my brothers Greg and Chris
Longstaff for always encouraging and supporting me regardless if they knew exactly what I was
doing, or why I was working outside instead of in an office. Thanks to my roommates Alexa
Wilson and Pratap Sankur. You guys were a great support system and always knew how to keep
me going. A special thank you goes to Ray Hummel for your endless patience, words of
encouragement, listening to countless presentations, editing papers, being the best friend anyone
could ask for, and always knowing how to make me smile and brighten my day. You really are
the volunteer of the year. I couldn’t have done this without your support.

iv

TABLE OF CONTENTS
LIST OF TABLES………………………………………………………………………….……vii
LIST OF FIGURES………………………………………………………………………….…...xi
PREFACE……………………..……………………………………………………………..….xiii
INTRODUCTION………………………………………………………………………………...1
OBJECTIVES…………………………………………………………………………..…3
LITERATURE CITED……………………………………………………………………4
CHAPTER 1: Quantifying white-tailed deer habitat suitability in wetland vegetation types with
potential impacts to waterfowl
INTROUDUCTION……………………………………………………………………………....8
STUDY AREA…………………………………………………………………………………..10
METHODS………………………………………………………………………………………11
RESULTS………………………………………………………………………………………..15
DISCUSSION……………………………………………………………………………………17
MANAGEMENT IMPLICATIONS…………………………………………………………….21
TABLES AND FIGURES………...……………………………………………………………..23
LITERATURE CITED….…………………………………………………………………….…26
CHAPTER 2:Habitat interaction patterns of white-tailed deer and wetland bird communities
within wetland vegetation types
INTRODUCTION……………………………………………………………………………….31
STUDY AREA…………………………………………………………………………………..32
METHODS………………………………………………………………………………………34
RESULTS…………………………………………………………………………………..……35
DISCUSSION……………………………………………………………………………………38
MANAGEMENT IMPLICATIONS……………………………………...……………………..41
TABLES AND FIGURES...……………………………………………………………………..42
LITERATURE CITED…………………………………………………………………………..52
CHAPTER 3:Effects of white-tailed deer herbivory within wetland vegetation types on a
landscape dominated by wetlands
INTRODUCTION……………………………………………………………………………….56
STUDY AREA…………………………………………………………………………………..58
METHODS………………………………………………………………………………………59
RESULTS………………………………………………………………………………………..62
Moist Soil Vegetation Type………………………………………………...……………62
Perennial Marsh Vegetation Type……………………………………………………….66
Lakeplain Prairie Vegetation Type ……………………. ……………………………….68
Bottomland Hardwood Forest Vegetation Type…………………………………………69
DISCUSSION……………………………………………………………………………………70
MANAGEGMENT IMPLICATIONS……………………………………………………….…..73
v

TABLES AND FIGURES..…………………………………………………………………..…74
LITERATURE CITED………………………………………………………………………….90
APPENDICES
APPENDIX A: GPS locations and data sheets for long term bottomland hardwood forest
monitoring herbivory study.…………………………………...………………………..95
APPENDIX B: Research and outreach activities………………………………………………100

vi

LIST OF TABLES

Table 1.1.

Average habitat suitability index (HSI) values (and standard errors) for
Shiawassee National Wildlife Refuge, the surrounding landscape, and specific
vegetation types within the refuge (2011-2012 growing season, Michigan)…….23

Table 1.2.

Percent of each habitat suitability index (HSI) value classifications for outside
Shiawassee National Wildlife Refuge, within Shiawassee National Wildlife
Refuge, and 5 vegetation types within Shiawassee National Wildlife Refuge
(2011-2012 growing season, Michigan). Low = 0-33.0, medium-low =33.1-53.0,
medium-high= 53.1-68.0, and high=68.1-86.0 HSI values……………………...24

Table 2.1.

Age class of each management unit for 2011 and 2012 field seasons at
Shiawassee National Wildlife Refuge. Age classes are defined as, early (<1yr
since treatment), mid (1-2 years since treatment), and late (>2yrs since treatment).
An X indicates the management unit was part of the age class………………….42

Table 2.2.

Correlation coefficients between average number of white-tailed deer and average
bird species richness for moist soil (MSU), perennial marsh (PM) and lakeplain
prairie (LPP) vegetation types at Shiawassee National Wildlife Refuge, 2001 and
2012, for monthly sampling and 15 day sampling
periods……………………………………………………………………………42

Table 2.3.

Averages (standard errors) for moist soil (MSU), perennial marsh (PM), and
lakeplain prairie (LPP) wetland vegetation types for number of deer and bird
species richness for monthly sampling periods at Shiawassee National Wildlife
Refuge, 2011 and
2012………………………………………………………………………………43

Table 2.4.

Species composition of moist soil, perennial marsh, and lakeplain prairie wetland
vegetation types at Shiawassee National Wildlife Refuge. E, M, and L indicates
the species was observed at least one time in the early, mid or late successional
age class of the moist soil units, PM indicates it was observed in the perennial
marsh units, and LPP indicates it was seen in the lakeplain prairie units. An X
indicates it was observed within all of the vegetation types and
units……………………………………..………………………………………..44

Table 2.5.

Average number of deer and average bird species richness (standard errors) for
early (<1year since treatment), mid (1-2 years since treatment), and late (>2years
since treatment) age classes of the moist soil vegetation type at Shiawassee
National Wildlife Refuge, 2011 and
2012………………………………………………………………………………45

Table 2.6.

Correlation coefficients for average number of deer and average bird species
richness for early (<1year since treatment), mid (1-2 years since treatment), and
late (>2years since treatment) age classes of the moist soil vegetation type on

vii

monthly and 15day sampling periods at Shiawassee National Wildlife Refuge,
2011 and 2012………………………………………………………………….45
Table 3.1.

Average horizontal cover (standard errors) for moist soil (MSU) vegetation types
at Shiawassee National Wildlife Refuge 2011 and 2012. A two tailed t-test was
used to compare average horizontal cover between all exclosures, all open areas,
core exclosures vs. core open areas, and edge exclosures vs. edge open areas
(P>0.1, α=0.10)…………………………………………………………………..74

Table 3.2.

Mean percent vertical cover (standard errors) for the moist soil (MSU) vegetation
type and each management unit individually within the moist soil vegetation type
at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was
used to compare means between all exclosures, all open areas, core exclosures vs.
core open areas, and edge exclosures vs. edge open areas (P>0.1, α=0.10)……..75

Table 3.3.

Mean species richness (standard errors) for the moist soil (MSU) vegetation type
and each management unit individually within the moist soil vegetation type at
Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was
used to compare means between all exclosures, all open areas, core exclosures vs.
core open areas, and edge exclosures vs. edge open areas (P>0.1, α=0.10)……..76

Table 3.4.

Mean above ground biomass production (standard errors) for the moist soil
(MSU) vegetation type and each management unit individually within the moist
soil vegetation type at Shiawassee National Wildlife Refuge, 2011 and 2012. A
two tailed t-test was used to compare means between all exclosures, all open
areas, core exclosures vs. core open areas, and edge exclosures vs. edge open
areas (P>0.1,α=0.10)……………………………………………………………77

Table 3.5.

Mean seed biomass production (standard errors) for the moist soil (MSU)
vegetation type and each management unit individually within the moist soil
vegetation type at Shiawassee National Wildlife Refuge, 2011 and 2012. A two
tailed t-test was used to compare means between all exclosures, all open areas,
core exclosures vs. core open areas, and edge exclosures vs. edge open areas
(α=0.10)…………………………………………………………………………78

Table 3.6.

Mean horizontal cover (standard errors) for the perennial marsh (PM) vegetation
type and each management unit individually at Shiawassee National Wildlife
Refuge, 2011 and 2012. A two tailed t-test was used to compare means between
all exclosures, all open areas, core exclosures vs. core open areas, and edge
exclosures vs. edge open areas (P>0.1, α=0.10)…………………………………79

Table 3.7.

Mean percent vertical cover (standard errors) for the perennial marsh (PM)
vegetation type and each management unit individually at Shiawassee National
Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means
between all exclosures, all open areas, core exclosures vs. core open areas, and
edge exclosures vs. edge open areas (P>0.1, α=0.10)…………………………80

Table 3.8.

Mean species richness (standard errors) for the perennial marsh (PM) vegetation
type and each management unit individually at Shiawassee National Wildlife
viii

Refuge, 2011 and 2012. A two tailed t-test was used to compare means between
all exclosures, all open areas, core exclosures vs. core open areas, and edge
exclosures vs. edge open areas (P>0.1, α=0.10)…………………………………81
Table 3.9.

Mean total above ground biomass (standard errors) for the perennial marsh (PM)
vegetation type and each management unit individually at Shiawassee National
Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means
between all exclosures, all open areas, core exclosures vs. core open areas, and
edge exclosures vs. edge open areas (P>0.1, α=0.10)…………………………..82

Table 3.10

Mean total seed mass (standard errors) for the perennial marsh vegetation type
and each management unit individually at Shiawassee National Wildlife Refuge,
2011 and 2012. Pool 1A was the only perennial marsh management unit where
seeds were present. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge
exclosures vs. edge open areas (P>0.1, α=0.10)…………………………………82

Table 3.11.

Mean horizontal cover (standard errors) for the lakeplain prairie (LPP) vegetation
type and each management unit individually at Shiawassee National Wildlife
Refuge, 2011 and 2012.A two tailed t-test was used to compare means between
all exclosures and all open areas (P>0.1, α=0.10)………………………………83

Table 3.12.

Mean vertical cover (standard errors) for the lakeplain prairie (LPP) vegetation
type and each management unit individually at Shiawassee National Wildlife
Refuge, 2011 and 2012. A two tailed t-test was used to compare means between
all exclosures and all open areas (P>0.1, α=0.10)……………………………….83

Table 3.13

Mean species richness (standard errors) for the lakeplain prairie (LPP) vegetation
type and each management unit individually at Shiawassee National Wildlife
Refuge, 2011 and 2012. A two tailed t-test was used to compare means between
all exclosures and all open areas (P>0.1, α=0.10)……………………………….84

Table 3.14.

Mean above ground biomass (standard errors) for the lakeplain prairie (LPP)
vegetation type and each management unit individually, at Shiawassee National
Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means
between all exclosures and all open areas (P>0.1, α=0.10)……………………...84

Table 3.15.

Mean seed biomass (standard errors) for the lakeplain prairie (LPP) vegetation
type and each management unit individually, at Shiawassee National Wildlife
Refuge, 2011 and 2012. A two tailed t-test was used to compare means between
all exclosures, all open areas (P>0.1, α=0.10)…………………………………...85

Table 3.16.

Mean horizontal cover (standard errors) for the bottomland hardwood forest
vegetation type and each stand individually, at Shiawassee National Wildlife
Refuge, 2011 and 2012. 3 height strata were used to make measurements <0.5m,
0.5-1.0m, and >1.0m. A two tailed t-test was used to compare means between all
exclosures and all open areas (P>0.1, α=0.10)…………………………………..86

ix

Table 3.17.

Mean vertical cover (standard errors) for the bottomland hardwood forest
vegetation type and each stand individually at Shiawassee National Wildlife
Refuge, 2011 and 2012. 3 height strata were used to make measurements <1.0m,
1.0-2.0m, and >2.0m. A two tailed t-test was used to compare means between all
exclosures and all open areas (P>0.1, α=0.10)…………………………………..87

Table 3.18.

Mean overstory species richness (standard errors) for the bottomland hardwood
forest vegetation type and each stand individually, at Shiawassee National
Wildlife Refuge, 2011 and 2012. A two-tailed t-test for used to compare means
between exclosures and open areas (P>0.1) (α=0.10)…………………………88

Table A.1.

Location of exclosures and paired open areas within forested stands at Shiawassee
National Wildlife Refuge. Open area locations are referenced in cardinal
direction and distance from exclosure…………………………………………...96

Table A.2.

Example data sheet for continued collection of horizontal cover, vertical cover,
and overstory species richness at each exclosure and open area within the forest
exclosure sites……………………………………………………………………97

Table A.3.

List of species observed within exclosures and open areas at bottomland
hardwood forest sites. An X indicates when a species was
present……………………………………………………………………………98

x

LIST OF FIGURES

Figure 1.1

Habitat suitability index (HSI) model results of Shiawassee National Wildlife
Refuge and the surrounding landscape with SNWR boundary and impoundments
delineated. HSI scale ranges from 0-86…………………………………………25

Figure 2.1.

Driving survey routes at Shiawassee National Wildlife Refuge for sampling bird
species richness and deer use of wetland vegetation types during 2011 and 2012.
Arrows indicate direction driven and green stars indicate starting/ending point of
the survey………………………………………………………………………...46

Figure 2.2.

Average white-tailed number of deer present in moist soil (MSU), perennial
marsh (PM), and lakeplain prairie (LPP) wetland vegetation types over time on a
monthly basis at Shiawassee National Wildlife Refuge, 2011 and
2012………………………………………………………………………………47

Figure 2.3.

Average bird species richness present in moist soil (MSU), perennial marsh (PM),
and lakeplain prairie (LPP) wetland vegetation types over time on a monthly
average at Shiawassee National Wildlife Refuge, 2011 and 2012………………47

Figure 2.4.

Average number of white-tailed deer present in moist soil (MSU), perennial
marsh (PM), and lakeplain prairie (LPP) wetland vegetation types over time on a
15 day average at Shiawassee National Wildlife Refuge, 2011 and
2012………………………………………………………………………………48

Figure 2.5.

Average bird species richness present in moist soil (MSU), perennial marsh (PM),
and lakeplain prairie (LPP) wetland vegetation types over time on a 15 day
average at Shiawassee National Wildlife Refuge, 2011 and 2012………………48

Figure 2.6.

Average number of white-tailed deer present in the moist soil vegetation type per
age class; early (<1year since treatment), mid (1-2 years since treatment), and late
(>2years since treatment) over monthly sampling periods at Shiawassee National
Wildlife Refuge, 2011 and 2012…………………………………………………49

Figure 2.7.

Average bird species richness in the moist soil vegetation type per age class; early
(<1year since treatment), mid (1-2 years since treatment), and late (>2years since
treatment) over monthly sampling periods at Shiawassee National Wildlife
Refuge, 2011 and 2012…………………………………………………………49

Figure 2.8.

Average number of white-tailed deer present in the moist soil vegetation type per
age class; early (<1year since treatment), mid (1-2 years since treatment), and late
(>2years since treatment) over monthly15 day sampling periods at Shiawassee
National Wildlife Refuge, 2011 and 2012……………………………………….50

Figure 2.9.

Average bird species richness in the moist soil vegetation type per age class; early
(<1year since treatment), mid (1-2 years since treatment), and late (>2years since

xi

treatment) over 15 day sampling periods at Shiawassee National Wildlife Refuge,
2011 and 2012……………………………………………………………………50
Figure 2.10.

Average bird abundance counts for perennial marsh (PM) and moist soil
vegetation type successional stages; early (>1year since treatment), mid (1-2 years
since treatment), and late (<1 year since treatment)…………………………..…51

Figure 3.1.

Deer estimate numbers and deer hunter harvest numbers for Shiawassee National
Wildlife Refuge…………………………………………………...……………...89

Figure A.1.

Aerial image of Shiawassee National Wildlife Refuge, refuge boundary outlined
in red. Each numbered circle indicates approximately where an exclosure and
paired open area are located within the forested stands………………………….99

xii

PREFACE

This thesis is organized into 3 chapters and 2 appendices. The chapters generally follow
the guidelines for manuscripts submitted to The Wildlife Society Bulletin. Chapters 1-3 were
formatted as complete manuscripts so some redundancy may occur (e.g. study area description
and some content within the introductions). Chapter 1 describes results on habitat suitability of
wetland vegetation types for white-tailed deer within Shiawassee National Wildlife Refuge and
the surrounding area. Chapter 2 describes results on deer use and bird species richness within
different wetland vegetation types at Shiawassee National Wildlife Refuge. Chapter 3 describes
results on white-tailed deer herbivory within 4 wetland vegetation types. Appendix A provides
data sheets and locations of forest exclosures to help the Shiawassee National Wildlife Refuge
staff to continue monitoring these sites and obtain data for a long-term study. Appendix B
highlights how the results of this research project have been disseminated within the scientific
community and also with outreach projects. Photo documentation of exclosure sites has been
archived at Shiawassee National Wildlife Refuge and with Dr. Henry Campa, III at Michigan
State University. An animal use exemption form dated December 7, 2010 addressed to Dr.
Henry Campa, III, was received from the Institutional Animal Care and Use Committee because
we were not handling wildlife for this study.

xiii

INTRODUCTION
White-tailed deer (Odocoileus virginianus) are adaptable organisms, which inhabit and
use many different vegetation types and environments from forests and agricultural fields to
urban settings (Rooney 2001). Deer are an important part of the Midwest landscape because
they can be keystone herbivores in forest ecosystems and, thereby, potentially impact wildlife
communities (Vercauteren and Hygnstrom 1998, Rooney 2001, Rooney and Waller 2003). Deer
have also been shown to negatively impact forest vegetation types by altering the composition,
structure, and biomass of understory vegetation and reducing recruitment and regeneration of
woody plant species (Marquis and Grisez 1978, Horsely and Marquis 1982, Rooney 2001,
Rooney and Waller 2003). Historic removal of native predators, such as the grey wolf (Canis
lupus) and cougar (Felis concolor), changes in hunting regulations, and habitat modification
have all also contributed to the expansion of deer populations throughout their range (Rooney
2001).
Degradation of vegetation due to browsing can be a concern for natural resource
managers because a correlation between the loss of understory and reductions in abundance and
diversity of insects, mammals and migratory birds has been documented for forest vegetation
types (Rooney 2001, Rooney and Waller 2003). Deer herbivory can also cause a shift from one
vegetation type to another. In severe cases, where deer herbivory has negatively affected plant
communities by reducing the structure, composition or productivity, deer have been able to set
back succession by consuming the entire understory leaving only ferns and less palatable plants
(Rooney 2001, Urbanek et al. 2012). Deer habitat, however, is not composed of just forest
vegetation types but also encompasses many different wetland vegetation types (Pusateri 2003,
Hiller 2007).
1

Deer use of wetlands can cause concern for natural resource managers because wetland
vegetation types can provide important life requisites for many bird species including waterfowl,
wading birds and shorebirds (Burger et al. 1996, Hafner 1997, Colwell and Taft 2000, Steven et
al. 2003, Stafford et al. 2010, O’Neal et al. 2012). The National Wildlife Refuge System, run by
the U.S. Fish and Wildlife Service, is one way the federal government acquires land to create
habitat and conserve these important wetlands for migratory birds and endangered species, and is
the only federal agency responsible for creating and maintaining habitat for migratory birds
(USFWS 2012). Nationally, there is a minimum of one National Wildlife Refuge per state, and
many of the National Wildlife Refuges are located along major flyways in an effort to conserve
important land for migratory birds (USFWS 2012).
Wetland vegetation types provide important habitat for feeding stopover sites during
migration for waterfowl as well as important over-wintering sites (Stafford et al. 2010).
Shorebirds also use wetlands for feeding and foraging sites along migration routes (Bookhout et
al. 1989, Burger et al. 1996). Large numbers of shorebirds and waterfowl can be seen using
different wetland vegetation types during fall and spring migration times for foraging and resting
sites, most heavily from May through June and September through October (Burger et al 1996,
Chaulk and Turner 2007). While shorebirds and waterfowl can be seen using wetlands heavily
during spring and fall migration, wading birds are also dependent on wetland vegetation types
for feeding sites (Colwell and Taft 2000). Wetland vegetation types are also important resting,
breeding and brood rearing sites for waterfowl and wading birds (Hafner 1997, Stevens et al.
2003). Wetland vegetation types are often optimal sites for nesting and brood rearing because of
the amount of structural diversity they can provide for different species (Hafner 1997, Stevens et
al. 2003).
2

The challenges presented by white-tailed deer herbivory have not necessarily been
limited to forest vegetation types. If deer can affect the distribution and abundance of plant and
wildlife species while changing the community structure at more than one trophic level in forest
vegetation types (Hanley 1996, Waller and Alverson 1997), then it can be projected they may
change other vegetation types in a similar manner. Coniferous and deciduous forest vegetation
types can be important vegetation types for white-tailed deer habitat; however mixed wetland,
lowland shrub and lowland deciduous are also important vegetation types (Hiller 2007). Whitetailed deer have also been documented to use wetland vegetation types extensively to fulfill life
requisites (Larson et al. 1978, Hiller 2007, Gubanyi et al. 2008, Clements et al. 2011), but a
knowledge gap exists on how a landscape dominated by wetlands provides suitable habitat for
white-tailed deer.
OBJECTIVES
One goal of this study was to investigate how white-tailed deer were potentially
impacting a landscape dominated by wetlands. Another goal was to describe the results for the
manager and biologist at Shiawassee National Wildlife Refuge to help them make more effective
management decisions that will help them maintain their goals and objectives for Shiawassee
National Wildlife Refuge. The objectives of this study were to
1. Evaluate white-tailed deer habitat suitability within a landscape dominated by wetland
vegetation types to help inform deer and waterfowl management decision making,
2. Quantify and compare use of wetland vegetation types by white-tailed deer to use by wetland
bird communities, and
3. Quantify white-tailed deer herbivory within a landscape dominated by wetlands
3

LITERATURE CITED

4

LITERATURE CITED
Bookhout, T.A., K.E. Bednarik, and R.W. Kroll. The Great Lakes. Habitat Management for
Migrating and Wintering Waterfowl in North America.Lubbokck, Texas: Texas Tech.
University Press, 1989. 131-156. Print.
Burger, J., L. Niles, and K.E. Clark. 1996. Importance of beach, mudflat and marsh habitats to
migrant shorebirds on Delaware Bay. Biological Conservation 79:283-292.
Chaulk, K.G., and B. Turner. 2007. The timing of waterfowl arrival and dispersion during spring
migration in Labrador. Northeastern Naturalist 14:375-386.
Clements, G.M., S.E. Hygnstrom, J.M. Gilsdorf, D.M. Baasch, M.J. Clements, and K.C.
VerCauteren. 2011. Movements of white-tailed deer in riparian habitat: implications for
infectious diseases. Journal of Wildlife Management 74:1436-1442.
Colwell, M.A., and O.W. Taft. 2000. Waterbird communities in managed wetlands of varying
water depth. Waterbirds 23:45-55.
Gubanyi, J.A., J.A. Savidge, S.E. Hygnstrom, K.C. VerCauteren, G.W. Garabrandt, and S.P.
Korte. 2008. Deer impact on vegetation in natural areas on Southeastern Nebraska.
Natural Areas Journal 28:121-129.
Hafner, H. 1997. Ecology of wading birds. Colonial Waterbirds 20:115-120.
Hanley, T.A. 1996. Potential role of deer (Cervidae) as ecological indicators of forest
management. Forest Ecology and Management 88:199-204.
Hiller, T.L. 2007. Land-use patterns and population characteristics of white-tailed deer in an
agro-forest ecosystem in south central Michigan. Dissertation, Michigan State University,
East Lansing, USA .
Horsely, S.B. and D.A. Marquis. 1982. Interference by weeds and deer of Allegheny hardwood
reproduction. Canadian Journal of Forest Research 13:61-69.
Larson, T.J., O.J. Rongstad, and F.W. Terbilcox. 1978. Movement and habitat use of white-tailed
deer in southcentral Wisconsin. Journal of Wildlife Management 42:113-117.
Marquis, D.A. and T.J. Grisez. 1978. The effect of deer exclosures on the recovery of vegetation
in failed clear cuts in the Allegheny plateau. U.S. Department of Ag. Forest Service. Res.
Note NE-270.
O’Neal, B.J., J.D. Stafford, R.P. Larkin. 2012. Stopover duration of fall-migrating dabbling
ducks. Journal of Wildlife Management 76:285-293.
Pusateri, J.S. 2003. White-tailed deer population characteristics and landscape use patterns in
southwestern lower Michigan. M.S. Thesis, Michigan State University, East Lansing,
USA.
Rooney, T.P. 2001. Deer impacts on forest ecosystems: a North American perspective. Forestry
74:201-208.
Rooney, T.P. and D.M. Waller. 2003. Direct and indirect effects of white-tailed deer in forest
ecosystems. Forest Ecology and Management 181:165-176.
5

Stafford, J. D., M.M. Horath, A.P. Yetter, R.V. Smith, and C.S. Hine. 2010. Historical and
contemporary characteristics and waterfowl use of Illinois river valley wetlands.
Wetlands 30:565-576.
Stevens, C.E., T.S. Gabor, and A.W. Diamond. 2003. Use of restored small wetlands by breeding
waterfowl in Prince Edward Island, Canada. Restoration Ecology 11:3-12.
U.S. Fish and Wildlife Service. 2012. National Wildlife Refuge System. NWRS-Land.
Department of the Interior.<www.fws.gov/refuges/land/>Accessed 25 February 2012.
Urbanek, R.E., C.K. Nielsen, G.A. Glowacki, and T.S. Pruess. 2012. Effects of white-tailed deer
(Odocoileus virginianus Zimm.) herbivory in restored forest and savanna plant
communities. American Midland Naturalist 167:240-255.
Vercauteren, K.C. and S.E. Hygnstrom. 1998. Effects of agriculture activities and hunting on
home ranges of female white-tailed deer. Journal of Wildlife Management 62:280-285.
Waller, D.M. and W.S. Alverson. 1997. The white-tailed deer: a keystone herbivore. Wildlife
Society Bulletin 25:217-226.

6

CHAPTER 1

Quantifying white-tailed deer habitat suitability in wetland vegetation types with potential
impacts to wetland bird communities

7

INTRODUCTION
White-tailed deer (Odocoileus virginianus) are adaptable organisms, which inhabit and
use many different vegetation types and environments from forests and agricultural fields to
urban settings (Rooney 2001). Deer densities throughout the Midwest have greatly increased
2

from pre-settlement historic estimates of ~3.2-4.2 deer/ km to estimates of 10-25deer/km

2

commonly seen today (Rooseberry et al. 1998, Rooney 2001, Rooney and Waller 2003, Kraft et
al. 2004, Gubanyi et al. 2008). Deer are an important part of the Midwest landscape because
they can be keystone herbivores in forest vegetation types which ultimately can impact wildlife
communities (Vercauteren and Hygnstrom 1998, Rooney 2001, Rooney and Waller 2003). Deer
have also been shown to negatively impact forest vegetation types by altering the composition,
structure, and biomass of understory vegetation and reducing recruitment and regeneration of
woody plant species (Marquis and Grisez 1978, Horsely and Marquis 1982, Rooney 2001,
Rooney and Waller 2003).
The challenges of white-tailed deer herbivory has not necessarily been limited to forest
vegetation types. If deer can affect the distribution and abundance of species while changing the
community structure at more than one trophic level in forest vegetation types (Hanley 1996,
Waller and Alverson 1997), then it can be projected they may change other vegetation types in a
similar manner. Coniferous and deciduous forest vegetation types can be important vegetation
types for white-tailed deer habitat; however mixed wetland, lowland shrub and lowland
deciduous are also important vegetation types (Hiller 2007). White-tailed deer have also been
documented to use wetland vegetation types extensively to fulfill life requisites (Larson et al.

8

1978, Hiller 2007, Gubanyi et al. 2008, Clements et al. 2011), but a knowledge gap exists on
how a landscape dominated by wetlands provides suitable habitat for white-tailed deer.
Understanding how much wetland vegetation types contribute to deer habitat is important
because wetlands are often managed for species other than white-tailed deer (Gray et al. 1999,
USFWS 2010). For example, wetland vegetation types within the National Wildlife Refuge
System are managed for waterfowl because waterfowl use wetlands for feeding, resting, and
nesting areas (USFWS 2010). Areas with restricted hunting opportunities and riparian areas
often have high deer densities, which can be a concern for National Wildlife Refuge mangers and
biologists since refuges often have these same characteristics (USFWS 2010, Clements et al.
2011).
Food availability within wetland vegetation types is also important because migratory
waterfowl use wetlands as stopover sites along migration routes for resting and replenishing
energy reserves (Bookhout et al. 1989). Waterfowl feed largely on seeds, roots and, herbaceous
growth of aquatic and wetland plants (Low and Bellrose 1944). Just 2 hours of flight can result
in 0.5-1.5 days of feeding and resting before some waterfowl are able to continue migration,
while 14 hours of flight can result in a range of 5-12 days of feeding and rest (Bookhout et al.
1989). Because aquatic and wetland plants also provide foods for white-tailed deer,
understanding how wetland vegetation types contribute to white-tailed deer habitat suitability is
therefore, essential in understanding their potential impact to waterfowl in landscapes dominated
by wetlands. Classifying a landscape into different vegetation types can help managers predict
successional trajectory or site potential allowing them to make more informed wildlife habitat
management decisions based on ecological characteristics that lead to conservation and
sustainability of natural resources on a landscape scale (Felix et al. 2004, Felix et al. 2007, Felix
9

and Campa 2010). Our objective was to evaluate white-tailed deer habitat suitability within a
landscape dominated by wetland vegetation types to help inform deer and waterfowl
management decision making.
STUDY AREA
Our study area was located at Shiawassee National Wildlife Refuge, SNWR, in the
eastern lower peninsula of Michigan in Saginaw County (USFWS 2010, USDA 1997). SNWR
sits on many different commonly flooded soils such as sandy-loams, Misteuay silty clays,
fluvaquents and mucks along with some soils that are infrequently flooded; sloan silt loam and
sloan-ceresco complexes (USDA 2009). The 3,845 ha refuge is composed primarily of 75%
bottom-land hardwood forests while the remaining 25% is rivers, perennial marshes, moist soil
units, lakeplain prairies, and crop lands (USFWS 2010). The adjacent landscape was composed
of wetland vegetation types, agriculture fields, and deciduous forest vegetation types. SNWR
experiences temperate climate conditions with an average yearly temperature of 8.3°C,
temperatures ranging from 8.0°C to 32.2°C during the growing season with highs and lows
throughout the year observed in August and January respectively (National Weather Service
2007).
SNWR is an important stopover site for over 270 species of migratory birds annually and
is located on some of Michigan’s most productive wetlands. Thousands of birds have been
observed daily using the refuge during fall and spring migrations (USFWS 2010). The only
current management strategies being conducted are water manipulations of the perennial marshes
and moist soil units accompanied by spraying, mowing, and disking moist soil units as needed to

10

set back succession (Gray et al. 1999, Steven Kahl, SNWR manager, Eric Dunton, SNWR
wildlife biologist personal correspondence).
METHODS
We used a white-tailed deer habitat suitability index, HSI, model developed by Felix
(2003) in combination with previous research on deer habitat use to evaluate the habitat
suitability of SNWR and the surrounding landscape. This model allows mangers to assess the
landscape’s potential to provide spring and summer habitat, fall and winter food and thermal
cover sources for white-tailed deer. We used the portion of the model for assessing spring and
summer habitat because we wanted to understand how the areas managed for waterfowl and
surrounding area corresponded to suitable deer habitat during the growing season (Felix 2003,
Hiller 2007). These results would potentially be useful for managers to understand if deer would
be competing with waterfowl for food sources and how deer might potentially be impacting the
plant communities within those managed impoundments.
We used high resolution (1 x 1 m) 4-band imagery (2010) from the National Agriculture
Inventory Program (NAIP), distributed by the Farm Service Agency with Department of
Agriculture to classify land cover within the SNWR and surrounding area. Images are captured
during the agricultural growing season and within each quarter quad tile ≤ 10% cloud cover is
present when the image is captured which allows for identifying and classifying different
vegetation types (USDA 2011). The 2010 NAIP imagery for Saginaw County, Michigan was
reclassified, using ERDAS Imagine, into 15 different vegetation types, and validated based on
field observations. Vegetation types included: agriculture, cattails, (Typha spp.) deciduous, elmash-cottonwood (Ulmus-Fraxinus-Populus deltoids), aquatic emergent, forest, mudflat, open,

11

pine (Pinus), silver maple-sugar maple-oak (Acer saccharinum-Acer saccharum – Quercus),
upland hardwood, upland hardwood – oak, urban, water, wet-open.
To analyze and quantify HSI, we used the same 3 cover classification variables
developed in the Felix (2003) white-tailed deer HSI model (coniferous, deciduous, and wetland)
and added 2 more cover classification variables (agriculture edge and urban) based on research
indicating deer frequently use these cover types to fulfill life requisites (Williamson and Hirth
1985, Miranda and Porter 2003, Stewart et al. 2006, Hiller 2007). The coniferous variable
received the highest suitability index score (100%) when a minimum of ≥10% cover was present,
using the pine classification of the cover type variables. Hiller (2007) suggested white-tailed
deer, in southern Michigan, select areas with some portion of their home range containing
coniferous forest vegetation type regardless of season because of the need for thermal cover.
Adult female does and young fawns have been seen selecting this vegetation type which leads to
the conclusion it is an important life requisite (Pusateri 2003, Hiller 2007). In southern
Michigan, deer are not seen selecting for coniferous forest types as often as in more northern
climates, because milder climatic conditions reduces the need for thermal cover and coniferous
forest vegetation types are less abundant (Van Deelen 1995, Felix 2003, Hiller 2007).
The deciduous forest variable received suitability scores of 100% when cover ranged
from 20-70% on the landscape. The score decreased when <20% and >70% of the landscape
was deciduous forest. In areas where no deciduous forest was present, the habitat suitability
score was 0.0% because in Michigan white-tailed deer habitat always includes deciduous forests
(Pusateri 2003, Hiller 2007). Deciduous forests are an important vegetation type for white-tailed
deer because they are a significant source of food such as browse and mast (Johnson et al. 1995,
Kraft et al. 2004, Collins and Battaglia 2008).
12

The agriculture edge variable had a 100% suitable value when the agricultural field edge
was ≤90m from a forested cover type edge with habitat suitability decreasing as the agriculture
and forest edge become further apart than 90m. Deer prefer to use portions of agriculture fields
no wider than 180m across or 90m away from forest edges (Williamson and Hirth 1985, Braun
1996). Agriculture vegetation types are not considered the most important vegetation type.
However, adult female deer in the Midwest have been documented using them during all seasons
of the year (Vercauteren and Hygnstrom 1998, Pusateri 2003, Stewart et al. 2006, Hiller 2007).
Numerous researchers have documented that white-tailed deer will frequently use urban
areas for food and cover if they contained vegetation types that provided those life requisites
(Hiller 2007, Storm et al. 2007, Rhoads et al. 2010). For example Hiller (2007) documented
adult female white-tailed deer used urban areas in an exurban landscape but <8% of their home
range was composed of this vegetation type. These findings indicated that deer will use urban
areas when available but they are not essential for fulfilling life requisites (Hiller 2007, Storm et
al. 2007, Rhoads et al. 2010). The urban variable received 100% white-tailed deer habitat
suitability when the percentage of urban development was ≤20%, was 0 when >20% urban
development was present.
The wetland variable had highest white-tailed deer habitat suitability when landscapes are
composed of 10-25% wetlands. Suitability decreases if landscapes had <10% wetlands and
>25% wetlands (Miranda and Porter 2003, Pusateri 2003, Hiller 2007, Gubanyi et al. 2008,
Clements et al. 2011). Previous research has documented white-tailed deer using numerous
wetland types to fulfill life requisites of hiding, feeding and thermal cover during all times of the
year (Rongstad and Tester 1969, Larson et al. 1978, Pusateri 2003, Hiller 2007). Hiller (2007)
and Pusateri (2003) observed adult female deer and fawns using numerous wetland vegetation
13

types throughout the growing and non-growing seasons including lowland shrub, lowland
deciduous forests, and bottomland hardwood forests.
All 5 suitability index variables were combined in the following equation to obtain the
overall habitat suitability for SNWR and the surrounding area.
HSI = (conifer suitability + 2* deciduous suitability + 2* agriculture edge suitability + urban
suitability + 2*wetland suitability)/8
The ecological contributions of conifer cover was not weighted in this model because of the lack
of coniferous forest vegetation types and relatively mild winter weather conditions when
compared to more northern regions of Michigan, which led to the belief white-tailed deer would
select other vegetation types to fulfill their life requisite needs. Wetland and agriculture edge
suitability was weighted as twice as much as other vegetation types because of the availability on
the landscape, the availability of numerous habitat components along edges, and personal visual
observations of white-tailed deer frequently using these vegetation types.
We combined data from the vegetation types to assess these 5 cover classifications
variables as follows: 1) Coniferous forest = pine; 2) Deciduous forest = deciduous, elm – ash cottonwood, silver maple - sugar maple-oak, upland hardwood, upland hardwood-oak; 3)
Agriculture edge = agriculture; 4) Urban = open, urban; 5) Wetlands = cattails, emergent,
mudflat, wet-open. Areas classified as water received no value and were not incorporated into
the model because waterways were not considered vegetation types, and therefore, did not
contribute to the overall HSI. Using the classified imagery, we used a circular roving window
with a radius of 900 m to quantify habitat suitability of each pixel centered within the average

14

home range size of a female white-tailed deer in southern Michigan (i.e., 100ha; Pusateri 2003;
Hiller 2007). Suitability indices were created for each variable in the model and then combined
using Map Algebra in the Spatial Analyst Toolset (ArcMap v 10) to create an overall HSI model
for the landscape. HSI values were obtained on a scale of 0-100% with 0% corresponding to no
suitable habitat and 100% corresponding to the habitat fulfilling 100% of the life requisite needs
for white-tailed deer.
We first assessed habitat suitability within the SNWR boundary and compared mean
pixel HSI values to the area surrounding SNWR within a 100ha buffer zone, using a two-tailed ttest. We then compared mean pixel HSI values on a finer scale among 5 different wetland
vegetation types within SNWR using a Tukey’s honestly significant difference (HSD) test. We
were able to conduct the fine scale analysis within SNWR’s boundary because personal field
observations and refuge vegetation classifications allowed us to adjust general land cover classes
(e.g., emergent wetland) to more specific vegetation classifications (e.g., cattail, mudflat).
Personal field observations could not be done outside the refuge boundary, which impeded finer
scale date analysis among different wetland vegetation types for outside the refuge boundary.
We used the natural breaks classification scheme in ArcMap to evaluate habitat suitability on a
finer scale. We chose 4 classes, low 0.0-33 HSI, medium-low 33.1-53, medium-high 53.1-68,
and high 68.1-86.
RESULTS
We ran the HSI model for SNWR and the surrounding area (Fig. 1.1). There was no
significant difference between mean pixel HSI values of the SNWR (63.8%) and the surrounding
area (64.6%) (P>0.10) (Table 1.1). Within SNWR, the south and east portions had the highest

15

habitat suitability values (Fig. 1.1). There is also a pocket of medium-low habitat suitability
bordering the south side of the largest water body, the Shiawassee River, running through the
northern half of SNWR. Another pocket of medium-low habitat exists in the northeast most
portion of the refuge. This portion of the refuge is surrounding by urban development to the
north, east, and west, with deciduous forest vegetation types located to the south west (Fig. 1.1).
The remainder of the refuge (36%) is classified as medium-high habitat. Proportionally within
the refuge boundary there are 0%, 24%, 36%, and 40% suitable habitat in the low, medium-low,
medium-high and high categories, respectively (Table 1.2).
The area surrounding SNWR (within 100ha from SNWR boundary) was classified as
high habitat suitability on the south and east sides of the refuge, medium-high with small pockets
of medium-low habitat on the west side, medium-high and high suitability on the north west side,
and medium-low with low suitability on the north east side of SNWR’s boundary (Fig. 1.1).
Proportionally we saw 8% low habitat suitability, 12% medium-low habitat suitability, 24%
medium-high habitat suitability, and 56% high habitat suitability (Table 1.2).
When a finer scale analysis was taken to assess white-tailed deer habitat suitability within
SNWR’s boundary, we evaluated 5 of the primary vegetation types; moist soil, perennial marsh,
lakeplain prairie, bottomland hardwood forest, and agriculture. We found moist soil vegetation
types to have the highest average HSI of 73.8% followed by bottomland hardwood forest,
lakeplain prairie, agriculture and perennial marsh with 70.7%, 69.7%, 63.6% and 53.3%
respectively (Table 1.1). Moist soil vegetation types showed 0.0% low and medium-low habitat
suitability, 32.0% medium-high habitat suitability, and 68% high habitat suitability. The
perennial marsh vegetation type had the lowest average HSI with 0.0% low habitat suitability,
48.0% medium-low habitat suitability, 52.0% medium-high habitat suitability and 0.0% high
16

habitat suitability on the landscape (Table 1.2). The lakeplain prairie vegetation type had 0.0%
low and medium-low habitat suitability, 28.0% medium-high habitat suitability, and 72.0% high
habitat suitability on the landscape (Table 1.2). Bottomland hardwood forests vegetation had
0.0% low and medium-low habitat suitability, 36.0% and 64.0% habitat suitability was found in
the medium-high and high classifications respectively (Table 1.2). The agriculture vegetation
type had 0.0% low habitat suitability, 28.0% habitat suitability for medium-low and mediumhigh classifications, and 44% high habitat suitability (Table 1.2).
We compared the average HSI values among the different vegetation types within SNWR
boundary to one another and found, moist soil vegetation types offered 20.5% and 11.2%
significantly greater habitat suitability than perennial marsh and agriculture vegetation
respectively but did not offer significantly different habitat suitability than lakeplain prairie and
bottomland hardwood forest vegetation types. Lakeplain prairie vegetation types offered 16.4%
and 7.1% significantly greater suitable habitat than perennial marshes and agriculture fields
respectively. Bottomland hardwood forests offered 17.4% and 8.1% significantly greater
suitable habitat than perennial marshes and agriculture fields respectively. Agriculture fields
offered 9.3% significantly greater suitable habitat than perennial marshes (Table 1.1).
DISCUSSION
Managing white-tailed deer in landscaped dominated by wetlands may be just as
challenging for natural resource managers as it is for those managing deer in landscapes
dominated by forest vegetation types. White-tailed deer have the potential to be keystone
herbivores and can dramatically change the structure, composition, and productivity of plant
communities in forest vegetation types (Marquis and Grisez 1978, Horsely and Marquis 1982,

17

Rooney 2001, Rooney and Waller 2003). Pusateri (2003) and Hiller (2007) documented that
deciduous forest vegetation types are important components of white-tailed deer habitat but
wetland vegetation types can also encompass a large portion of deer home ranges. Landscapes
dominated by some wetland types can provide areas of high habitat suitability for white-tailed
deer. On SNWR’s landscape 76% of SNWR was classified as medium-high to high habitat
suitability while the area surrounding SNWR was 80% medium-high to high habitat suitability.
Landscapes, such as many National Wildlife Refuges, dominated by wetlands that provide high
HSI values and large amounts of the adjacent landscape proportionally as medium-high and high
habitat suitability for white-tailed deer could see negative impacts to the plant communities from
high deer denisties similar to those affects seen by previous research from Marquis and Grisez
(1978), Rooney (2001), and Rooney and Waller (2003).
SNWR and the surrounding area are very similar in terms of vegetation types they offer
with wetlands, agriculture, and forest vegetation types present in both areas and SNWR does not
offer significantly more suitable habitat than the surrounding area. Bordering the west side of
SNWR is Shiawassee River State Game area owned by the Michigan Department of Natural
Resources which is also managed specifically for waterfowl including the American bittern
(Botaurus lentiginosus), mallard (Anas platyrhynchos), and wood duck (Aix sponsa) (DNR
2012). The northeast area bordering SNWR is the city of Saginaw, MI which accounts for the
proportion of low deer habitat suitability (Fig.1.1, Table 1.2). In our study area, deer may not be
seeking out SNWR for habitat because the surrounding area provides just as suitable habitat as
SNWR, so deer would potentially be able to fulfill all of their life requisites outside SNWR’s
boundary. Other National Wildlife Refuges or landscapes dominated by wetlands may not have
similar vegetation types on their boarders. This landscape composition could be a problem for
18

managers in those areas since deer in lower habitat suitability areas adjacent to those boundaries
may potentially move opportunistically to higher habitat suitability within refuge boundaries
when seeking food sources or cover.
VerCauteren and Hygnstrom (1998), and Stewart et al. (2006), and Hiller (2007) found
that deer will frequently use agricultural vegetation types for summer food sources. Wetlands
that are adjacent to agriculture areas may receive higher HSI values than those that are not. Our
results showed the moist soil and lakeplain prairie units had the highest and third highest average
habitat suitability values respectively (Table 1.2). These units were adjacent to agricultural fields
and relatively small in size (<60.3 ha) contributing to the amount of edge that white-tailed deer
usually seek to meet their life requisites (Williams and Hirth 1985, Stewart et al. 2006). The size
and juxtaposition to agricultural fields probably explains why these units received significantly
higher HSI scores than the perennial marsh units and the agriculture fields. Stewart et al. (2006)
found that deer prefer to feed in the edge of fields surrounded by forests. All of the moist soil
and lakeplain prairie units were bordered on at least one side by forest vegetation types which
may also have contributed to their high habitat suitability scores. Managers and biologists of
National Wildlife Refuges may expect to see higher habitat suitability for white-tailed deer in
wetlands with similar characteristics.
The perennial marsh vegetation type had the lowest average habitat suitability primarily
because some units were relatively large (>130.3ha). This configuration provided less edge
when compared to the core of the units, decreasing habitat suitability because it left more area
>90m from the edge of the impoundment. Williams and Hirth (1985), Braun (1996), and Stewart
et al.( 2006) all documented that deer prefer to feed and use areas within the edges of units,
perhaps because it provides easier access to cover from predators. The larger size of these units
19

and that they were not adjacent to agriculture fields, favorite summer feeding sources for whitetailed deer, probably attributed to the low HSI values. The perennial marsh units also held water
longer than moist soil units throughout the year. Drawdowns of water typically started in late
May or early June and water was replaced in late August to early September. Some years, some
units would not be drawn down at all to leave pools of water and more niches for resident and
migratory waterfowl, which decreased the availability of habitat for white-tailed deer and
decreased habitat suitability scores (Gray et al. 1999, Eric Dunton and Steve Kahl, Wildlife
Biologist and Refuge Manager, personal correspondence). The use of water management or
water level manipulations could potentially be used as a management technique to deter deer
herbivory that may impact waterfowl food sources and habitat. National Wildlife Refuge
managers may expect to find wetland units within their landscape that are not adjacent to
agriculture fields and are larger with more of the units >90m from the edge to provide lower
habitat suitability values for white-tailed deer.
The bottomland hardwood forests had the second highest average HSI value (Table 1.1)
and were also the largest (237.9 ha) managed units within the refuge. The edge component is not
important for these units because, being forested units they provide cover throughout the entire
unit. These units shared west borders with agriculture lands which contributed to their high
suitability (Williamson and Hirth 1985, Stewart et al. 2006). Another factor contributing to their
habitat suitability to deer was they contained mast producing tree species such as white oak and
swamp white oak (Quercus spp.). Having mast producing trees in a forest increases deer habitat
suitability. Johnson et al. (1995) documented mast is an important food source for white-tailed
deer and can compose a large portion of their fall and winter diets. On a landscape with

20

bottomland hardwood forests that contain mast producing tree species, natural resource managers
can expect these areas to be highly suitable habitat for white-tailed deer.
Wetland habitat suitability for white-tailed deer seems to depend greatly on the proximity
to agricultural fields, the size of the wetland units, and the amount of water that may be on
management units. The closer the wetland vegetation types are to preferred feeding sites, like
agriculture fields and mast producing trees, the greater the habitat suitability (Johnson et al.
1995, Miranda and Porter 2003, Stewart et al 2006). The larger units with less total area of the
wetland units <90m from edge of the unit the lower habitat suitability since white-tailed deer
seem to prefer to feed and use areas of fields within edges (Williamson and Hirth 1985, Braun
1996, Stewart et al. 2006). Overall the majority of landscapes dominated by wetlands provide
medium-high to high habitat suitability for white-tailed deer.
MANAGEMENT IMPLICATIONS
Understanding how landscape components contribute to white-tailed deer habitat
suitability, and ultimately to white-tailed deer abundance, is important for managers since locally
abundant deer numbers in some areas can make natural resources management a challenge.
Attempting to manage white-tailed deer on National Wildlife Refuges, such as SNWR, or
landscapes dominated by agricultural lands and wetlands in the face of other management
priorities can be challenging (Xie et al. 2001, Gubanyi et al. 2008). Providing National Wildlife
Refuge managers and wildlife biologists, in landscapes dominated by wetlands, with a whitetailed deer habitat suitability model can be useful to demonstrate areas of relatively high to low
habitat suitability and where they might expect the greatest impacts from deer, since landscapes
dominated by wetlands are often managed for species other than white-tailed deer. If white-

21

tailed deer use is substantial within wetland vegetation types, white-tailed deer may potentially
change the vegetation composition and structure and, therefore, the successional trajectories of
those wetland vegetation types (Horsely and Marquis 1982, Hanley 1996, Roosenberry and Wolf
1998, Rooney 2001, Rooney and Waller 2003).
With a better understanding of how landscape components contribute to deer habitat
suitability, managers and biologists will be able to more effectively plan field/impoundment size
with wetland restoration or mitigation in mind (Hanley 1996). Managers and biologist can
anticipate higher levels of deer herbivory in wetland fields/impoundments that provide highest
habitat suitability values. This will help managers when making decisions and providing habitat
for migratory and resident waterfowl within wetland vegetation types.
National Wildlife Refuges face challenges when managing their landscapes because one
goal of the U.S. Fish and Wildlife Service is to provide adequate habitat for resident and
migratory waterfowl (USFWS 2010). Meeting this goal may be compromised in the face of
locally abundant deer using the same vegetation types as bird communities. Using this model
will help them understand how their landscape can provide habitat for white-tailed deer and will
help tailor management decision making at a landscape level.

22

Table 1.1. Average habitat suitability index (HSI) values ( standard errors) for Shiawassee
National Wildlife Refuge, the surrounding landscape, and specific vegetation types within the
refuge (2011-2012 growing season, Michigan).

Vegetation Type

Average HSI value (%)

Outside SNWR boundary

64.6 (3.1)

Within SNWR boundary

63.8 (2.5)
a

Moist Soil

73.8 A (1.5)

Perennial Marsh

53.3 B (1.7)

Lakeplain Prairie

69.7 A (0.6)

Bottomland Hardwood Forest

70.7 A (0.9)

Agriculture

62.6 C (2.6)

a

Means with the same letters were not statistically different from one another (Tukey’s honestly
significant difference (HSD) test, HSD =5.01).

23

Table 1.2. Percent of each habitat suitability index (HSI) value classifications for outside
Shiawassee National Wildlife Refuge, within Shiawassee National Wildlife Refuge, and 5
vegetation types within Shiawassee National Wildlife Refuge (2011-2012 growing season,
Michigan). Low = 0-33.0, medium-low =33.1-53.0, medium-high= 53.1-68.0, and high=68.186.0 HSI values.
Vegetation Types

HSI Classifications
% Low

% MediumLow

% MediumHigh

% High

Outside SNWR Boundary

8

12

24

56

Within SNWR Boundary

0

24

36

40

Moist Soil

0

0

32

68

Perennial Marsh

0

48

52

0

Lakeplain Prairie

0

0

28

72

Bottomland Hardwood Forest

0

0

36

64

Agriculture

0

28

28

44

24

City of Saginaw, MI

Legend
SNWR
Legend Boundary
Impoundment Boundary
SNWR Boundary
Water
Impoundment Boundary
Habitat Suitability Index
Water
Value Habitat Suitability Index Value

High :
High (86%)86

LowLow : 0
(0%)
Figure 1.1. Habitat suitability index (HSI) model results of Shiawassee National Wildlife Refuge
and the surrounding landscape with SNWR boundary and impoundments delineated. HSI scale
ranges from 0-86. For interpretation of the references to color in this and all other figures, the
reader is referred to the electronic version of this thesis.
25

LITERATURE CITED

26

LITERATURE CITED
Bookhout, T.A., K.E. Bednarik, and R.W. Kroll. The Great Lakes. Habitat Management for
Migrating and Wintering Waterfowl in North America. Lubbokck, Texas: Texas Tech.
University Press, 1989. 131-156. Print.
Braun, K.F. 1996. Ecological factors influencing white-tailed deer damage to agricultural crops
in northern Michigan. Thesis, Michigan State University, East Lansing, USA.
Clements, G.M., S.E. Hygnstrom, J. M. Gilsdorf, D. M. Baasch, M. J. Clements, and K. C.
VerCauteren. 2011. Movements of white-tailed deer in riparian habitat: implications for
infectious diseases. Journal of Wildlife Management 74:1436-1442.
Collins, B. and L.L. Battaglia. 2008. Oak regeneration in southestern bottomland hardwood
forests. Forest Ecology and Management 255:3026-3034.
DNR Wildlife Division. 2012. Crow Island State Game Area Habitat Division. Michigan
Department of Natural Resources.
Felix, A.B. 2003. Development of landscape-scale models to describe habitat potential of whitetailed deer (Odocoileus virginianus) in Michigan. M.S. Thesis, Michigan State
University, East Lansing,USA.
Felix, A.B. and H. Campa, III. 2010. Relating ecological properties of habitat types to
differences in aspen stand structure and succession for managing timber and wildlife
resources. Northern Journal of Applied Forestry 27:13-20.
Felix, A.B., H. Campa, III, K.F. Millenbah, S.R. Winterstein and W.E. Moritz. 2004.
Development of landscape-scale habitat-potential models for forest wildlife planning and
management. Wildlife Society Bulletin 32:795-806.
Felix, A.B., D.P. Walsh, B.D. Hughey, H. Campa, III, and S.R. Winterstein. 2007. Applying
landscape-scale habitat-potential models to understand deer spatial structure and
movement patterns. The Journal of Wildlife Management 71:804-810.
Gray, M.J., R.M. Kaminski, G. Weerakkody, B.D. Leopold, and K.C. Jensen. 1999. Aquatic
invertebrate and plant responses following mechanical manipulations of moist-soil
habitat. Wildlife Society Bulletin 27:770-779.
Gubanyi, J.A., J.A. Savidge, S.E. Hygnstrom, K.C. VerCauteren, G.W. Garabrandt, and S.P.
Korte. 2008. Deer impact on vegetation in natural areas on Southeastern Nebraska.
Natural Areas Journal 28:121-129.
Hanley, T.A. 1996. Potential role of deer (Cervidae) as ecological indicators of forest
management. Forest Ecology and Management 88:199-204.
Hiller, T.L. 2007. Land-use patterns and population characteristics of white-tailed deer in an
agro-forest ecosystem in south central Michigan. Dissertation, Michigan State University,
East Lansing, USA .
27

Horsely, S.B. and D.A. Marquis. 1982. Interference by weeds and deer of Allegheny hardwood
reproduction. Canadian Journal of Forest Research 13:61-69.
Johnson, A.S., P.E. Hale, W.M. Ford, J.M. Wentworth, J.R. French, O.F. Anderson, and G.B.
Pullen. 1995. White-tailed deer foraging in relation to successional stage, overstory type,
and management of southern Appalachian forests. American Midland Naturalist 133:1835.
Kraft, L.S, T.R. Crow, D.S. Buckley, E.A. Nauertz and J.C. Zasada. 2004. Effects of harvesting
and deer browsing on attributes of understory plants in northern hardwood forests, Upper
Michigan, USA. Forest Ecology and Management 199:219-230.
Larson, T.J., O.J. Rongstad, and F.W. Terbilcox. 1978. Movement and habitat use of white-tailed
deer in southcentral Wisconsin. Journal of Wildlife Management 42:113-117.
Low, J.B. and F.C. Bellrose, Jr. 1944. The seed and vegetative yield of waterfowl food plants in
the Illinois river valley. Journal of Wildlife Management 8: 7-22.
Marquis, D.A. and T.J. Grisez. 1978. The effect of deer exclosures on the recovery of vegetation
in failed clear cuts in the Allegheny plateau. U.S. Department of Ag. Forest Service. Res.
Note NE-270.
Miranda, B.R. and W.F. Porter. 2003. Statewide habitat assessment for white-tailed deer in
Arkansas using satellite imagery. Wildlife Society Bulletin 31:715-726.
National Weather Service (IWIN) and the National Oceanic Atmospheric Administration
(NOAA). 2007. Climate for Saginaw, MI.
<http://www.rssweather.com/climate/Michigan/Saginaw> Accessed 12 Dec. 2010
Pusateri, J.S. 2003. White-tailed deer population characteristics and landscape use patterns in
southwestern lower Michigan. Thesis, Michigan State University, East Lansing, USA.
Rhoads, C.L., J.L. Bowman, and B. Eyler. 2010. Home range and movement rates of female
exurban white-tailed deer. Journal of Wildlife Management 74:987-994.
Rongstad, O. J. and J.R. Tester. 1969. Movement and habitat use of white-tailed deer in
Minnesota. Journal of Wildlife Management 33:366-379.
Rooney, T.P. 2001. Deer impacts on forest ecosystems: a North American perspective. Forestry
74:201-208.
Rooney, T.P. and D. M. Waller. 2003. Direct and indirect effects of white-tailed deer in forest
ecosystems. Forest Ecology and Management 181:165-176.
Rooseberry, J.L. and A. Woolf. 1998. Habitat-population density relationships for white-tailed
deer in Illinois. Wildlife Society Bulletin 26:252-258.
Stewart, C. M., W.J. Mcshea, and B.P. Piccolo. 2006. The impact of white-tailed deer on
agricultural landscapes in 3 national historic parks in Maryland. Journal of Wildlife
Management 71:1525-1530.
28

Storm, D.J., C.K. Nielsen, E.M. Schauber, and Alan Woolf. 2007. Space use and survival of
white-tailed deer in an exurban landscape. Journal of Wildlife Management 71:11701176.
United States Department of Agriculture. 1997. National Agricultural Stats Service. State and
County Profiles.
<http://www.agcensus.usda.gov/Publications/1997/County_Profiles/Michigan/index.asp>
Accessed 8 March 2010.
United States Department of Agriculture. 2009. Natural Resource Conservation Service. Web
Soil Survey. <http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx>. Accessed
12 December 2010
United States Department of Agriculture. 2011. Aerial Photography Field Office; Imagery
Programs.<http://www.fsa.usda.gov/FSA/apfoapp?area=home&subject=prog&topic=nai.
> Accessed July 18, 2012.
U.S Fish and Wildlife Service. 2010. National Wildlife Refuge System. Shiawassee National
Wildlife Refuge. Department of the Interior. <http://www.fws.gov/midwest/shiawassee/>.
Accessed 12 December 2010.
Van Deelen, T. 1995. Seasonal migrations and mortality of white-tailed deer in Michigan’s
Upper Peninsula. Dissertation, Michigan State University, East Lansing, USA.
Vercauteren, K.C. and Scott E. Hygnstrom. 1998. Effects of agriculture activities and hunting on
home ranges of female white-tailed deer. Journal of Wildlife Management 62:280-285.
Waller, D.M. and W.S. Alverson. 1997. The white-tailed deer: a keystone herbivore. Wildlife
Society Bulletin 25:217-226.
Williamson, S. J. and D. H. Hirth. 1985. An evaluation of edge use by white-tailed deer. Wildlife
Society Bulletin 13:252-257.
Xie, J., J. Liu, and R.t Doepker. 2001. DeerKBS: a knowledge-based system for white-tailed deer
management. Ecological Modeling 140:177-192.

29

CHAPTER 2

Habitat interaction patterns of white-tailed deer and wetland bird communities within wetland
vegetation types

30

INTRODUCTION
Wetland vegetation types provide important life requisites for many different species of
birds including waterfowl, wading birds and shorebirds (Burger et al. 1996, Hafner 1997,
Colwell and Taft 2000, Steven et al. 2003, Stafford et al. 2010, O’Neal et al. 2012). The
National Wildlife Refuge System, run by the U.S. Fish and Wildlife Service, is how the federal
government acquires land to create habitat and conserve these important wetlands for migratory
birds and endangered species, and is the only federal agency responsible for creating and
maintaining habitat for migratory birds (USFWS 2012). Nationally there is a minimum of one
National Wildlife Refuge per state, and many of the National Wildlife Refuges are located along
major flyways in an effort to conserve important land for migratory birds (USFWS 2012).
Wetland vegetation types provide important habitat for feeding stopover sites during
migration for waterfowl as well as important over-wintering sites (Stafford et al. 2010).
Shorebirds also use many wetland types for feeding and foraging sites along migration routes
(Bookhout et al. 1989, Burger et al. 1996). Large numbers of shorebirds and waterfowl can be
seen using different wetland vegetation types during fall and spring migration times for foraging
and resting sites, most heavily from May through June and September through October (Burger
et al 1996, Chaulk and Turner 2007). While shorebirds and waterfowl depend on wetlands
heavily during the migration seasons, wading birds are also dependent on wetland vegetation
types for feeding sites (Colwell and Taft 2000). Wetland vegetation types are also important
resting, breeding and brood rearing sites for waterfowl and wading birds (Hafner 1997, Stevens
et al. 2003). These vegetation types are often optimal sites for nesting and brood rearing because
of the amount of structural diversity they can provide and array of niches available for nesting
(Hafner 1997, Stevens et al. 2003).

31

Waterfowl, wading birds and shorebirds are not the only species that use wetlands to
fulfill life requisites. White-tailed deer (Odocoileus virginianus) have been documented using
wetland vegetation types for hiding cover and foraging (Miranda and Porter 2003, Pusateri 2003,
Hiller 2007, Gubanyi et al. 2008, Clements et al. 2011). Wetland vegetation types are among the
most important vegetation types for white-tailed deer in south-central Michigan because they
provided dense hiding and thermal cover and food, and in Minnesota white-tailed deer use
wetlands for thermal cover and food resources throughout winter home ranges (Rongstad and
Tester 1969, Hiller 2007).
Natural resource managers faced with maintaining wetlands for waterfowl, shorebirds,
and wading birds may be challenged because, a knowledge gap exists on how white-tailed deer
use of wetland vegetation types may affect the composition of wetland bird species communities
using those same wetlands. White-tailed deer could potentially be competing with wetland bird
species for space requirements to meet their resting, reproductive, and food requirements (Hafner
1997, Russel et al. 2001, Sovada 2001, Pusateri 2003, Hiller 2007). Our objective was to
quantify and compare use of wetland vegetation types by white-tailed deer to use by wetland bird
communities.
STUDY AREA
The study area was located at Shiawassee National Wildlife Refuge (SNWR), in the
eastern lower peninsula of Michigan in Saginaw County (USFWS 2010, USDA 1997). SNWR
sits on many different commonly flooded soils such as sandy-loams, Misteuaysilty clays,
fluvaquents and mucks along with some soils that are infrequently flooded; sloan silt loam and
sloan-ceresco complexes (USDA 2009). The refuge is approximately 3,845 ha and is 75%

32

bottom-land hardwood forests while the remaining 25% is rivers, perennial marshes, moist soil
units, lakeplain prairies, and crop lands (USFWS 2010). The adjacent landscape was composed
of wetland vegetation types, agriculture fields (corn, soybeans, winter wheat), and deciduous
forest vegetation types. SNWR experiences temperate climate conditions with an average yearly
temperature of 8.3°C, temperatures ranging from 8.0°C to 32.2°C during the growing season
with highs and lows throughout the year observed in August and January respectively (National
Weather Service 2007).
SNWR is an important stopover site for over 270 species of migratory birds annually and
is located on some of Michigan’s most productive wetlands. Thousands of birds have been
observed at the refuge during fall and spring migrations as well as resident bird communities
(USFWS 2010). Current management strategies being conducted are invasive species
management, water level manipulations in the various habitat types (i.e., perennial marshes,
floodplain forest, and moist soil units), and active moist soil management (e.g., spraying,
mowing, and disking) as needed to set back succession (Gray et al. 1999, Steve Kahl, SNWR
manager, Eric Dunton, SNWR wildlife biologist personal correspondence).
SNWR, perhaps like other National Wildlife Refuges (e.g. De Soto National Wildlife
Refuge), has experienced issues related to high deer densities and deer herbivory including,
complaints from nearby farmers about agricultural damage from white-tailed deer, high doe:buck
ratios, low fawn weights and extensive damage to forests. In an effort to combat these issues,
SNWR has reduced antlerless deer densities by using a regulated hunting season on the refuge
and issuing permits on a yearly basis since early 1980’s (Steven Kahl, SNWR manager, Eric
Dunton, SNWR wildlife biologist, personal correspondence).

33

METHODS
To accomplish our objective we conducted driving surveys around SNWR (Fig.2.1). We
planned to conduct a minimum of 5 surveys per impoundment per month. Surveys were
conducted from sunrise to 3 hours after, and from 3 hours before to sunset, alternating morning
and evening sampling times each week (Sitar 1996). Surveys were conducted May through
August of 2011 and 2012 after water had been drawn off impoundments. This approach ensured
deer would have the potential to use sites and making the comparison of deer use and wetland
bird species richness possible. Survey routes alternated starting and ending points every time
they were conducted, to reduce bias so the same impoundments were not reached at the same
time during every survey (Sitar 1996).
During the driving survey, Monarch ATB 8x powered lenses were used to scan fields for
wildlife. Date, time, impoundment name and vegetation type, species, and number per species of
wildlife were recorded for each designated impoundment. Moist soil, perennial marsh, and
lakeplain prairie vegetation types were the wetland vegetation types surveyed. Although the
bottomland hardwood forest wetland vegetation type occurred on the refuge it was not included
in this analysis because of the inability to observe bird species in this vegetation type from the
survey routes. These impoundments contained mature trees dominated by silver maple (Acer
saccharinum) and cottonwood (Populus deltoides) and were large (>125ha) making it hard to see
into the impoundments making the driving survey very biased to only wildlife using the edge of
impoundments. Due to the amount of disturbance caused by walking through these sites and the
long settling period needed before observations could take place, other surveys such as point
count, or transect surveys were not used to analyze bird species richness and deer use of the
bottomland hardwood forests.
34

All observational data were averaged for bird species richness and white-tailed deer use
for moist soil, perennial marsh and lakeplain prairie wetland vegetation types. A coarse scale
analysis was first used by averaging data on a monthly basis to see if correlations existed
between average number of deer and average wetland bird species richness. A finer scale
analysis was used to see if any distinct temporal patterns of deer and bird use emerged when data
were averaged for 15 day sampling periods. Data within the moist soil vegetation type were
averaged on successional age to see if any correlations existed among the different age classes of
the moist soil management units and deer and wetland bird use. We used 3 different age classes;
early (<1year since treatment), mid (1-2years since treatment), and old (>2years since treatment)
(Table 2.1). Treatment is defined as when the management unit was set back to the earliest
successional stage by first spraying the units with a general herbicide to kill all plants, mowing
the unit, and/or then disking the unit. We were not able to analyze the other vegetation types
based on age class because, within the perennial marsh vegetation type, we were only able to use
one management unit because of sightability issues (Table 2.1). Both the lakeplain prairie
management units were only used in the 2012 field season and had undergone a burn
immediately prior to the start of sampling in 2012 so they were considered the same age class,
early (Table 2.1). Correlation analysis was run on the data per wetland vegetation type to
compare average wetland bird species richness to average deer use. This was conducted to
investigate if deer use impacted the number of wetland bird species using the same wetland
vegetation types.
RESULTS
The month with the highest average number of deer observed per survey route was
August, with an average of 9.1 deer observed within the perennial marsh vegetation type (Table
35

2.3). The lakeplain prairie vegetation type had the lowest observed average deer use with
averages of 0.0 seen in July and August, while the moist soil vegetation type had more deer use
than the lakeplain prairie units but less than the perennial marsh impoundments with averages
ranging from 2.1 to 3.4 from May through August (Table 2.3).
The lakeplain prairie vegetation type also had the lowest average bird species richness
observed of 0.0 throughout the entire sampling time. The perennial marsh vegetation type had
the highest averages of bird species richness observed throughout the entire sampling period
ranging from 2.7 to 5.0 and had the highest bird species richness average overall of 5.0 in May
(Table 2.3). The moist soil vegetation types had higher bird species richness averages than the
lakeplain prairie vegetation type but less than the perennial marsh vegetation type with averages
ranging from 0.0 in August to 0.8 in May (Table 2.3). The most common bird species seen
throughout the sampling period were Canada goose (Branta canadensis) and Mallard (Anas
platyrhynchos) but many other species were observed (Table 2.4).
When looking first at the coarse scale analysis we found no definite correlations between
the average number of deer and average bird species richness using a wetland vegetation type.
The correlation coefficient for the moist soil vegetation type between average number of deer
and bird species richness was the largest at 0.45, while the perennial marsh vegetation type had a
correlation coefficient of -0.10, and the lakeplain prairie vegetation type had a correlation
coefficient of no value because no birds were observed within those impoundments (Table 2.2).
In addition, there were no trends between deer use of wetland vegetation types and bird species
richness observed (Fig. 2.2, Fig. 2.3).

36

When analyzing data on a finer scale, we found no correlations between the number of
deer using a wetland vegetation type and bird species richness observed for the same vegetation
type. The moist soil vegetation type again had the largest positive correlation coefficient (0.02)
while the perennial marsh vegetation type had a negative correlation coefficient (-0.05). We
were not able to calculate a correlation coefficient value for the lakeplain prairie vegetation type
because there were no recorded observations for bird species richness. No trends over time
were observed within the finer scale analysis between bird species richness and deer use of
wetland vegetation types (Fig. 2.4, Fig. 2.5). There did not seem to be a positive correlation
when average deer numbers increased along with average species richness or a negative
correlation when average deer numbers increased, average bird species richness decreased (or
vice versa).
When analyzing the data based on age class of the management units within the moist
soil vegetation type we found the average number of deer observed to be highest in the late age
class in the June 1st start date of the 15 day sampling period (8.3) and the lowest in the late age
class in the month of May (0.0) (Table 2.5). We found bird species richness to be the highest in
the early age class in May (2.2) and lowest in the early, mid and, late age classes in July and
August (0.0) (Table 2.5). Based on a correlation analysis only one strong correlation occurred
between average deer and bird species richness, within the early age class for the monthly
sampling period (0.94) (Table 2.6). The mid and late age classes had correlation coefficients
between average deer and bird species richness of 0.49 and -0.29, respectively (Table 2.6).
Based on the analysis for the 15 day sampling periods we found no strong correlations between
average deer and species richness. Within the early age class we found a correlation coefficient
of 0.50, 0.03 within the mid age class and -0.21 within the late age class (Table 2.6). No trends
37

over time were observed between average deer and bird species richness observed on a monthly
basis (Fig. 2.6, Fig. 2.7). No trends were observed over time for the finer scale analysis within
the 15 day sampling periods for deer use and bird species richness within any of the age classes
for the moist soil vegetation type (Fig. 2.8, Fig. 2.9). We also looked at total abundance of birds
within moist soil age classes, perennial marshes and lakeplain prairie vegetation types. We
found no trends with total abundance observed from May to August (Fig. 2.10).
DISCUSSION
We observed the highest bird species richness in May, probably because that time of the
year is when birds are completing their spring migrations (Burger et al. 1996, O’Neal et al.
2012). The low species richness observed throughout June, July and August at SNWR was
probably due to the fact that migration of waterfowl, wading birds or shorebirds does not begin
to occur typically until September (Burger et al. 1996, Chaulk and Turner 2007, O’Neal et al.
2012). The perennial marsh units had the highest average number of bird species present. This
is probably due to the fact that these units held water longer throughout the year than the moist
soil and lakeplain prairie units. Draw downs of water started in June for the perennial marsh
units, as opposed to May for the moist soil units and the lakeplain prairie units did not hold
standing water (personal correspondence, Steve Kahl, SNWR manager and Eric Dunton, SNWR
biologist). The lack of water may have reduced the availability of feeding and adequate nesting
habitat for shorebirds, waterfowl, and wading birds for the moist soil and lakeplain prairie
vegetation types which is why we saw lower species richness than in the perennial marshes
(Hafner 1997, Stevens et al. 2003, Colwell and Taft 2007, Stafford et al. 2010).

38

We found only one strong correlation between deer and bird species richness in the early
age class of the monthly sampling period within the moist soil vegetation type. This correlation
was positive and throughout our entire sampling period (May-August) the number of deer and
species richness decreased which is why this correlation coefficient was so high. One reason for
finding low correlation coefficients within the remainder of the analysis might have been the
limitations with the methods used to conduct the study. As the vegetation grew during the
growing season it became harder to visually see wildlife within those units after mid-June. The
driving survey route was conducted on roads and dikes that overlooked the managed
impoundments but this did not always offer a suitable view or an advantage for sighting wildlife
in impoundments with dense vegetation. This reduced the ability to see wildlife and count them,
so some wildlife may have been using impoundments when a zero value was recorded. We
might recommend doing occupancy modeling or developing a detection probability for each
wetland vegetation type for future studies similar to this one. This might allow for a better
understanding of what species of wetland birds and how per species are using wetland vegetation
types.
Another reason for low correlation values might have been because deer are herbivores
and, therefore, do not threaten waterfowl, wading birds or shorebirds (Butler and Vennesland
2001, Russel et al. 2001, Sovada et al. 2001, West and Messmer 2004). Deer and birds were
seen within 15m of one another feeding and using the wetlands. White-tailed deer were probably
not perceived as a threat to the different bird species using those same wetland vegetation types
so their behavior and use of those wetlands probably was not altered when white-tailed deer were
present. The presence of feeding white-tailed deer within wetland vegetation types was probably
also not perceived as a threat to waterfowl, wading birds and shorebirds because they all feed on
39

different food sources. Waterfowl feed on seeds, roots, and shoots of aquatic and wetland plants
(Low and Bellrose 1944, O’Neal et al. 2012), wading birds feed on small fish and aquatic
organisms (Butler and Vennesland 2000), and shorebirds feed on small invertebrates within
mudflats (Burger et al. 1996). Daigel et al (2004) and Stewart et al. (2006) documented deer
feeding heavily in crop fields during the summer months when they were available, indicating
croplands are some preferred food sources for white-tailed deer. Johnson et al. (1995) also found
woody plant species materials (i.e.. leaves and shoots) to compose a large portion of white-tailed
deer diets. Crops and woody plants were not within the boundary of any of the wetland
vegetation types within our study area but those food sources were within SNWR and in close
proximity to the wetlands. We can infer the deer we saw within this study were probably
traveling through the wetland sites, maybe foraging on some plants but were most likely seeking
out those preferred or favorite food sources nearby. Because wading birds and shorebirds do not
potentially feed on the same food resources as white-tailed deer, deer probably were not
perceived as a threat or competition for food resources to the shore and wading bird species.
This information is helpful to wildlife managers and biologists who work in landscapes
dominated by wetland vegetation types because white-tailed deer use does not seem to be
impacting bird species use within wetland vegetation types. We did not see increased bird
species richness when fewer deer were present. The only strong correlation we found was a
positive trend and as deer numbers decreased so did bird species richness. However the results
presented from this study can only be considered baseline data and very minimal counts.
Because observing the wildlife was difficult after mid-June because of vegetation height, more
wildlife could have been present and not accounted for. More in depth analysis on how different
species are affected by the presence or absence of white-tailed deer would be needed to make
40

more definite conclusions if white-tailed deer influence wetland bird behavior when using
wetland vegetation types.
MANAGEMENT IMPLICATIONS
The presence of white-tailed deer within wetland vegetation types does not seem to have
a negative effect on the bird species composition using wetland vegetation types at SNWR. Thes
results can be helpful for wetland managers and biologists. The implications of these results is
that although white-tailed deer may be using the same wetland vegetation types as waterfowl,
wading bird or shorebird species that are being managed for, white-tailed deer are not negatively
impacting the use of the wetlands for those birds. Deer were mostly likely using wetlands as
travel corridors to reach areas where more preferred food sources were available. Pusateri
(2003) and Hiller (2007) also documented deer using wetland vegetation types for habitat within
the summer and winter months. Deer may also be using some wetland vegetation types during
the spring and summer for thermal cover, explaining why we saw them so frequently within
different wetland types. Managers can then concentrate their management efforts into habitat
requirements for the desired species. Managers and biologists do not have to take management
efforts to keep white-tailed deer out of wetland vegetation types because deer use of those
wetlands does not appear to be impacting bird use of the same wetlands.

41

Table 2.1. Age class of each management unit for 2011 and 2012 field seasons at Shiawassee
National Wildlife Refuge. Age classes are defined as, early (<1yr since treatment), mid (1-2
years since treatment), and late (>2yrs since treatment). An X indicates the management unit
was part of the age class.
2011
Vegetation Type

Unit Name

Early

Mid

2012
Late

Early

MSU 1

X

X*

MSU 2E

Mid

X

Late

Moist Soil

MSU 2W

X*
X

X

X

MSU 6

X

X

Perennial Marsh
Pool 1A

X

X

SU 1

N/A

N/A

N/A

X

SU 3

N/A

N/A

N/A

X

Lakeplain Prairie

* MSU 1 underwent treatment mid-summer 2012, so for the first half of 2012 sampling period it
was considered late age class and during the second half of 2012 sampling period it was
considered early age class.

Table 2.2. Correlation coefficients between average number of white-tailed deer and average
species richness for moist soil (MSU), perennial marsh (PM) and lakeplain prairie (LPP)
vegetation types at Shiawassee National Wildlife Refuge, 2001 and 2012, for monthly (n=8)
sampling and 15 day (n=4) sampling periods.
Wetland Vegetation Types
MSU
PM
LPP

Correlation Coefficient
Monthly Sampling Period 15 day Sampling Period
0.45
0.02
-0.10
-0.05
No value
No value

42

Table 2.3. Averages (standard errors) for moist soil (MSU), perennial marsh (PM), and lakeplain
prairie (LPP) wetland vegetation types for number of deer and bird species richness for monthly
sampling periods at Shiawassee National Wildlife Refuge, 2011 and 2012.
Start Date of
Sampling
Period

Average Deer per Vegetation
Type

Average Species Richness per
Vegetation Type

May 1
June 1
July 1
August 1

MSU
3.4 (1.1)
2.2 (1.2)
3.4 (2.4)
2.1 (2.0)

PM
5.2 (1.4)
3.5 (1.7)
5.3 (2.1)
9.1 (3.1)

LPP
2.2 (0.7)
0.7 (0.1)
0.0 (0.0)
0.0 (0.0)

MSU
0.8 (0.9)
0.5 (0.4)
0.3 (0.3)
0.0 (0.0)

PM
5.0 (1.0)
2.7 (0.6)
3.3 (0.6)
2.9 (0.5)

LPP
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)

May 15
June 1
June 15
July 1
July 15
August 1

3.4 (1.1)
1.8 (0.4)
2.6 (0.5)
4.4 (1.3)
2.5 (0.6)
2.1 (0.6)

5.2 (1.4)
3.0 (1.1)
4.0 (1.4)
5.0 (1.9)
5.5 (1.2)
9.1 (3.1)

2.2 (0.7)
1.0 (0.1)
0.3 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)

0.8 (0.2)
0.8 (0.2)
0.2 (0.1)
0.3 (0.2)
0.2 (0.1)
0.0 (0.0)

5.0 (1.0)
3.5 (1.0)
1.8 (0.4)
4.3 (0.8)
2.3 (0.8)
2.9 (0.5)

0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)

43

Table 2.4. Species composition of moist soil, perennial marsh, and lakeplain prairie wetland
vegetation types at Shiawassee National Wildlife Refuge. E, M, and L indicates the species was
observed at least one time in the early, mid or late successional age class of the moist soil units,
PM indicates it was observed in the perennial marsh units, and LPP indicates it was seen in the
lakeplain prairie units. An X indicates it was observed within all of the vegetation types and
units.
Species

Monthly Sampling Period

Common Name
White-tailed Deer

Scientific Name
Odocoileus virginianus

May
X

June
X

July
E,M,
L,PM

August
E,M,L,P
M

Blue Winged Teal

Anas discors

*PM

PM

PM

PM

Canada Goose
Dunlin
Great Blue Heron

Branta canadensis
Calidris alpina
Ardea herodias

L,PM

PM

PM

Ardea alba

E,M,L
,PM
M,PM

E,M,
PM
E,PM

L,PM

Great Egret
Green Heron
Herring Gull
Killdeer

Butorides vicescens
Larus argentatus
Charadrius vociferus

L, PM
E, PM
E,M,L,
PM
E,M,
PM
PM

PM

PM
PM

PM
PM

PM

E,M,
E,M,
PM
PM
Lesser Yellowlegs Tringa flavipes
PM
Mallard
Anas platyrhynchos
E,M,L, E,PM L,PM L,PM
PM
Northern Shoveler Anas clypeata
E
Pied-billed Grebe
Podilymbus podiceps
PM
PM
PM
PM
Sandhill Crane
Grus canadensis
E,M
E,M,L M
*No wetland bird species were observed during any sampling periods within the lakeplain prairie
vegetation types

44

Table 2.5. Average number of white-tailed deer and average bird species richness (standard
errors) for early (<1year since treatment), mid (1-2 years since treatment), and late (>2years
since treatment) age classes of the moist soil vegetation type at Shiawassee National Wildlife
Refuge, 2011 and 2012.
Start Date of
Sampling
Period

Average Deer per Age Class

Average Species Richness per Age
Class

May 1
June 1
July 1
August 1

Early
4.0 (1.6)
2.1 (1.0)
2.1 (0.9)
1.2 (1.0)

Mid
0.6 (0.3)
1.2 (0.6)
1.2 (0.8)
0.7 (0.7)

Late
0.0 (0.0)
5.1 (1.6)
4.3 (1.6)
5.5 (2.8)

Early
2.2 (0.6)
0.8 (0.4)
0.0 (0.3)
0.0 (0.0)

Mid
0.3 (0.1)
1.2 (0.6)
0.0 (0.0)
0.0 (0.0)

Late
0.6 (0.3)
0.6 (0.2)
0.3 (0.1)
0.0 (0.0)

May 15
June 1
June 15
July 1
July 15
August 1

4.0 (1.6)
1.1 (1.0)
3.1 (1.1)
2.4 (1.6)
1.8 (0.6)
1.2 (1.0)

0.6 (0.3)
1.0 (0.4)
1.4 (0.8)
1.7 (0.9)
0.7 (0.7)
0.7 (0.7)

0.0 (0.0)
8.3 (2.9)
1.8 (0.5)
3.5 (1.0)
5.0 (1.9)
5.5 (2.8)

2.2 (0.6)
1.3 (0.3)
0.3 (0.2)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)

0.3 (0.1)
2.0 (1.2)
0.4 (0.2)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)

0.6 (0.3)
0.9 (0.3)
0.3 (0.1)
0.5 (0.1)
0.2 (0.1)
0.0 (0.0)

Table 2.6. Correlation coefficients for average number of white-tailed deer and bird species
richness for early (<1year since treatment), mid (1-2 years since treatment), and late (>2years
since treatment) age classes of the moist soil vegetation type on monthly (n=8) and 15day (n=4)
sampling periods at Shiawassee National Wildlife Refuge, 2011 and 2012.
Age Class
Early
Mid
Late

Correlation Coefficient
Monthly Sampling Period
15 day Sampling Period
0.94
0.50
0.49
0.03
-0.29
-0.21

45

Legend
SNWR Boundary
Impoundment Boundary
Alternate Starting Points of Survey
Survey Route
Figure 2.1. Driving survey routes at Shiawassee National Wildlife Refuge (SNWR) for sampling
bird species richness and deer use of wetland vegetation types during 2011 and 2012. Arrows
indicate driving route and green stars indicate starting/ending point of the survey.
46

Average Number of Deer

Average Number of Deer per Wetland
Vegetation Type
10.0
8.0
6.0

LPP

4.0

MSU

2.0

PM

0.0
May

June
July
Sampling Period (Month)

August

Average bird species richness

Figure 2.2. Average number of white-tailed deer present in moist soil (MSU), perennial marsh
(PM), and lakeplain prairie (LPP) wetland vegetation types over time on a monthly basis at
Shiawassee National Wildlife Refuge, 2011 and 2012.

Average Bird Species Richness per Wetland
Vegetation Type
6.0
5.0
4.0
3.0

LPP

2.0

MSU

1.0

PM

0.0
May

June
July
Sampling Period (Month)

August

Figure 2.3. Average bird species richness present in moist soil (MSU), perennial marsh (PM),
and lakeplain prairie (LPP) wetland vegetation types over time on a monthly average at
Shiawassee National Wildlife Refuge, 2011 and 2012.

47

Average Number of Deer

Average Number of Deer per Wetland
Vegetation Type
10.0
8.0
6.0

LPP

4.0

MSU

2.0

PM

0.0
May 15 June 1 June 15 July 1 July 15 August 1
Start Date of Sampling Period (15 days long)

Average Species Richness

Figure 2.4. Average number of white-tailed deer present in moist soil (MSU), perennial marsh
(PM), and lakeplain prairie (LPP) wetland vegetation types over time on a 15 day average at
Shiawassee National Wildlife Refuge, 2011 and 2012.

Average Bird Species Richness per Wetland
Vegetation Type
6.0
5.0
4.0
3.0
2.0
1.0
0.0

LPP
MSU
PM
May 15 June 1 June 15 July 1 July 15 August 1
Start Date of Sampling Period (15 days long)

Figure 2.5. Average bird species richness present in moist soil (MSU), perennial marsh (PM),
and lakeplain prairie (LPP) wetland vegetation types over time on a 15 day average at
Shiawassee National Wildlife Refuge, 2011 and 2012.
48

Average Number of Deer

Average Number of Deer per Age Class of Moist
Soil Vegetation Type
6.0
5.0
4.0
3.0
2.0
1.0
0.0

Early
Mid
Late
May

June

July

August

Sampling Period (Month)

Average Bird Species Richness

Figure 2.6. Average number of white-tailed deer present in the moist soil vegetation type per age
class; early (<1year since treatment), mid (1-2 years since treatment), and late (>2years since
treatment) over monthly sampling periods at Shiawassee National Wildlife Refuge, 2011 and
2012.

Average Bird Species Richness per Age Class of
Moist Soil Vegetation Type
2.5
2.0
1.5

Early

1.0

Mid

0.5

Late

0.0
May

June
July
Sampling Period (Month)

August

Figure 2.7. Average bird species richness in the moist soil vegetation type per age class; early
(<1year since treatment), mid (1-2 years since treatment), and late (>2years since treatment) over
monthly sampling periods at Shiawassee National Wildlife Refuge, 2011 and 2012.

49

Average Number of Deer

Average Number of Deer per Age Class of
Moist Soil Vegetation Type
10.0
8.0
6.0

Early

4.0

Mid

2.0

Late

0.0
May 15 June 1 June 15 July 1

July 15 August 1

Start Date of Sampling Period (15 days long)

Average Bird Species Richness

Figure 2.8. Average number of white-tailed deer present in the moist soil vegetation type per age
class; early (<1year since treatment), mid (1-2 years since treatment), and late (>2years since
treatment) over monthly15 day sampling periods at Shiawassee National Wildlife Refuge, 2011
and 2012.

Average Bird Species Richness per Age Class of
Moist Soil Vegetation Type
2.5
2.0
1.5
1.0
0.5
0.0

Early
Mid

Late
May 15 June 1 June 15 July 1 July 15 August
1
Start Date of Sampling Period (15 days long)

Figure 2.9. Average bird species richness in the moist soil vegetation type per age class; early
(<1year since treatment), mid (1-2 years since treatment), and late (>2years since treatment) over
15 day sampling periods at Shiawassee National Wildlife Refuge, 2011 and 2012.
50

Average Bird Abundance (count)

Average Bird Abundance Per Monthly Sampling
Period for Moist Soil Successional Stages and
Perennial Marshes
35
30
25
20
15
10
5
0

Early
Mid
Late
PM
May

June

July

August

Sampling Period (Month)

Figure 2.10. Average bird abundance counts for perennial marsh (PM) and moist soil vegetation
type successional stages; early (>1year since treatment), mid (1-2 years since treatment), and late
(<1 year since treatment).

51

LITERATURE CITED

52

LITERATTURE CITED
Bookhout, T.A., K.E. Bednarik, and R.W. Kroll. The Great Lakes. Habitat Management for
Migrating and Wintering Waterfowl in North America. Lubbokck, Texas: Texas Tech.
University Press, 1989. 131-156. Print.
Butler, R.W., and R.G. Vennesland. 2000. Integrating climate change and predation risk with
wading bird conservation research in North America. Waterbirds 23:535-540.
Burger, J., L. Niles, and K.E. Clark. 1996. Importance of beach, mudflat and marsh habitats to
migrant shorebirds on Delaware Bay. Biological Conservation 79:283-292.
Chaulk, K.G., and B. Turner. 2007. The timing of waterfowl arrival and dispersion during spring
migration in Labrador. Northeastern Naturalist 14:375-386.
Clements, G.M., S.E. Hygnstrom, J.M. Gilsdorf, D.M. Baasch, M.J. Clements, and K.C.
VerCauteren. 2011. Movements of white-tailed deer in riparian habitat: implications for
infectious diseases. Journal of Wildlife Management 74:1436-1442.
Colwell, M. A., and O.W. Taft. 2000. Waterbird communities in managed wetlands of varying
water depth. Waterbirds 23:45-55.
Daigle, C., M. Crete, L. Lesage, J. Ouellet, and J. Hout. 2004. Summer diet of two white-tailed
deer, Odocoileus virginianus, populations living in low and high density in Southern
Quebec. Canadian Field-Naturalist 118:360-367.
Gubanyi, J.A., J.A. Savidge, S.E. Hygnstrom, K.C. VerCauteren, G.W. Garabrandt, and S.P.
Korte. 2008. Deer impact on vegetation in natural areas on Southeastern Nebraska.
Natural Areas Journal 28:121-129.
Hafner, H. 1997. Ecology of wading birds. Colonial Waterbirds 20:115-120.
Hiller, T.L. 2007.Land-use patterns and population characteristics of white-tailed deer in an
agro-forest ecosystem in south central Michigan. Dissertation, Michigan State University,
East Lansing, USA .
Miranda, B.R. and W.F. Porter. 2003. Statewide habitat assessment for white-tailed deer in
Arkansas using satellite imagery. Wildlife Society Bulletin 31:715-726.
O’Neal, B. J., J.D. Stafford, R.P. Larkin. 2012. Stopover duration of fall-migrating dabbling
ducks. Journal of Wildlife Management 76:285-293.
Pusateri, J.S. 2003. White-tailed deer population characteristics and landscape use patterns in
southwestern lower Michigan. Thesis, Michigan State University, East Lansing, USA.
Rongstad, O. J., and J.R. Tester. 1969. Movement and habitat use of white-tailed deer in
Minnesota. Journal of Wildlife Management 33:366-379.
Rooney, T.P. 2001. Deer impacts on forest ecosystems: a North American perspective. Forestry
74:201-208.
53

Russel, F.L., D.B. Zippin, and N.L. Fowler. 2001. Effects of white-tailed deer (Odocoileus
virginianus) on plants, plant populations, and communities: a review. The American
Midland Naturalist 146:1-26.
Sitar, K.L. 1996. Seasonal movements, habitat use patterns, and population dynamics of whitetailed deer (Odocoileus virginianus) in an agricultural region of northern lower Michigan.
M.S. Thesis. Department of Fisheries and Wildlife, Michigan State University. East
Lansing, Michigan.
Socada, M.A., R.M. Anthony, and B.D.J. Batt. 2001. Predation on waterfowl in arctic tundra and
prairie breeding areas: a review. Wildlife Society Bulletin 29:6-15.
Stafford, J.D., M.M. Horath, A.P. Yetter, R.V. Smith, and C.S. Hine. 2010. Historical and
contemporary characteristics and waterfowl use of Illinois rive valley wetlands. Wetlands
30:565-576.
Stevens, C.E., T.S. Gabor, and A.W. Diamond. 2003. Use of restored small wetlands by breeding
waterfowl in Prince Edward Island, Canada. Restoration Ecology 11:3-12.
U.S. Fish and Wildlife Service. 2012. National Wildlife Refuge System. NWRS-Land.
Department of the Interior.<www.fws.gov/refuges/land/>Accessed 25 February 2012.
West, B.C. and T.A. Messmer. 2004. Impacts and management of duck-nest predation: the
manager’s view. Wildlife Society Bulletin 32:772-781.

54

CHAPTER 3

Effects of white-tailed deer herbivory within wetland vegetation types on a landscape dominated
by wetlands

55

INTRODUCTION
Over the past few decades white-tailed deer densities (Odocoileus virginianus) have
increased to levels well above the historic range of variability (Rooney 2001). Pre-settlement
-2

deer density estimates in North America were found about 3.1 to 4.2 deer per km (Rooney
2

2001) while common densities of 10 deer/ km frequently occur in the Midwest (Rooney and
Waller 2003, Kraft et al 2004). Historical removal of native predators, such as the grey wolf
(Canis lupus) and cougar (Felis concolor), changes in hunting regulations, and habitat
modification have all also contributed to the expansion of deer populations throughout their
range (Rooney 2001). In recent decades, logging of late successional forests allowed for
substantial regeneration of new growth and pioneer species which are preferred food for whitetailed deer (Russell et al. 2004, Rooney 2001, Waller and Alverson 1997).
High densities of deer can be a challenge for natural resource managers because deer can
act as keystone herbivores (Waller and Alverson 1997, Rooney 2001). The structural integrity of
forest vegetation types can be modified or compromised in cases of high deer densities and
intense herbivory. Vertical structure is an important deer habitat component because it directly
influences the quantity and quality of the habitat and can influence how wildlife selects habitat
(Haukos et al. 1998, Felix et al. 2007). Important understory components such as saplings,
seedlings, herbs and shrubs are prone to heavy browsing pressure and in some instances can be
completely eliminated (Rooney 2001).
Deer herbivory can also cause a shift from one vegetation type to another. In severe
cases, instances where deer herbivory has negatively affected plant communities by reducing the
structure, composition or productivity, deer have set back succession by consuming the entire
understory leaving only ferns and less palatable plants: American beech (Fagus grandifolia) and
56

hay scented fern (Dennstaedtia punctilobula) (Rooney 2001). Recruitment and regeneration
failure of preferred tree species of white-tailed deer is also commonly seen. Eastern hemlock
(Tsuga Canadensis) and oak (Quercus spp.) forests have undergone compositional shifts due to
browsing pressure from white-tailed deer with seedling regeneration showing a negative linear
correlation with browsing pressure (Rooney and Waller 2003). Deer herbivory can also limit
seedling growth, and negative correlations have been observed between the number of seedlings
present in a landscape and deer abundance (Waller and Alversons 1997). These trends have also
been seen in northern white-cedar (Thuja occidentalis) and yellow birch (Betula alleghaniensis),
which are also nutritionally important and favored food species for white-tailed deer (Rooney
and Waller 2003, Barnes and Wagner 2008).
When deer herbivory changes the structure and composition of ecosystems, productivity
is also jeopardized. The absolute and relative abundance of woody and herbaceous species has
been negatively correlated with deer herbivory (Waller and Alverson 1997). In areas protected
from deer herbivory it is not uncommon to see increases in the number of leaves per plant, plant
height, numbers and sizes of fruits, flowers, and seeds, and the proportion of flowering and
fruiting individuals (Kraft et al. 2004, Urbanek et al. 2012). Total forest biomass production can
also be reduced by deer herbivory because seedling growth at ground cover level can be stunted
(Kraft et al 2004). Impacts of deer and elk herbivory may have a time limit and do not
necessarily affect forest structure over the long term in the same manner as deer herbivory does
over the short term (Raymer 2000).
White-tailed deer may use other vegetation types besides forests to fulfill their life
requisites. White-tailed deer to use a variety of wetland vegetation types for feeding, and hiding

57

cover (Pusateri 2003, Hiller 2007). If deer herbivory can negatively impact forest vegetation
types then it can be surmised they can have similar effects on other vegetation types.
A knowledge gap exists on how deer herbivory impacts wetland vegetation types. This is
a concern to wetland managers and National Wildlife Refuge managers because wetlands are
often not managed for white-tailed deer, but are managed for resident and migratory waterfowl
(USFWS 2010). Understanding how deer herbivory impacts wetland vegetation types will help
managers plan and implement more effective management strategies in landscapes dominated by
wetlands. The objective of this study was to quantify white-tailed deer herbivory within a
landscape dominated by wetlands.
STUDY AREA
Our study area was located at SNWR in the eastern lower peninsula of Michigan in
Saginaw County (USFWS 2010, USDA 1997). SNWR sits on many different commonly
flooded soils such as sandy-loams, Misteuaysilty clays, fluvaquents and mucks along with some
soils that are infrequently flooded; sloan silt loam and sloan-ceresco complexes (USDA 2009).
The refuge is 3,845 ha and 75% of it is composed bottom-land hardwood forests while the
remaining 25% is rivers, perennial marshes, moist soil units, lakeplain prairies, and crop lands
(USFWS 2010). The adjacent landscape was composed of wetland vegetation types, agriculture
fields, and deciduous forest vegetation types. SNWR experiences temperate climate conditions
with an average yearly temperature of 8.3°C, temperatures ranging from 8.0°C to 32.2°C during
the growing season and highs and lows throughout the year observed in August and January
respectively (National Weather Service 2007).
SNWR is an important stopover site for over 270 species of migratory birds annually and
is located on some of Michigan’s most productive wetlands. Thousands of birds have been

58

observed at the refuge during fall and spring migrations as well as many resident bird species
(USFWS 2010). The only current management strategies being conducted are water
manipulations of the perennial marshes and moist soil units accompanied by spraying, mowing
and disking moist soil units as needed to set back succession (Gray et al. 1999, Steven Kahl,
SNWR manager, Eric Dunton, SNWR wildlife biologist, personal correspondence).
SNWR, perhaps like other national wildlife refuges, has experienced issues related to
high deer densities and deer herbivory including, complaints from nearby farmers about
agricultural damage from white-tailed deer, high doe:buck ratios, low fawn weights and
extensive damage to forests. In an effort to combat these issues, SNWR has reduced antlerless
deer densities by implementing a structured hunting season on the refuge and issuing harvesting
permits on a yearly basis since the early 1980’s (Steven Kahl, SNWR manager, Eric Dunton,
SNWR wildlife biologist, personal correspondence).
METHODS
To quantify white-tailed deer herbivory within moist soil, perennial marsh, lakeplain
prairie, and bottomland hardwood forest vegetation types we selected random sites to construct
exclosures in each vegetation type and paired them with areas open to herbivory to assess
browsing effects from May-August 2011 and 2012. We used 4 moist soil units, 2 perennial
marsh units, 2 lakeplain prairie units, and 3 bottomland hardwood forest units. Before exclosure
construction each management unit within moist soil and perennial marsh vegetation types were
stratified into core (>90m from the edge of the unit) and edge (<90m from the edge of the unit).
We used this stratification because research by Williamson and Hirth (1985), Braun (1996),
Vercauteren and Hygnstrom (1998), and Stewart et al. (2006) suggest deer use the edge of fields
for feeding more heavily than the core area. The size of the lakeplain prairie vegetation type
management units did not allow stratification into core and edge. Bottomland hardwood forests
59

were not stratified into core and edge because previous research did not indicate deer would use
the edge of these vegetation types more than the core. A minimum of 6 exclosures and paired
open areas were placed into each unit, equal within the core and edge stratification. In units
larger than 13ha we used 1exclosure per 2.15ha, again with equal numbers within the core and
edge stratification. Exclosures within the bottomland hardwood forest vegetation type were 20m
X 20m X 2.4m according to other researchers who investigated the influence of large herbivores
on forest communities (Campa et al. 1992) and 3 X 3m in the moist soil, perennial marsh and
lakeplain prairie vegetation types (Rodríguez –Pérez and Green 2006). Paired open area sites
were placed randomly 10m away from each exclosure in an effort to reduce bias if deer are
attracted to or deterred from exclosures (Braun 1996).
Bottomland hardwood forest exclosures were constructed out of wire mesh farm fencing
with 5.0 X 7.5cm openings. These exclosures were constructed to start a long term monitoring
study on how deer herbivory affects bottomland hardwood forests. Exact locations of each
exclosure and paired open area and vegetation sampling data sheets for continued data collection
are in Appendix A. Moist soil, perennial marsh, and lakeplain prairie exclosures were
constructed using 4 corner t-post polymesh fencing with 1.87 X 1.87cm openings. These
exclosures were only 1.52m high because we assumed the size of o the exclosures (3m X 3m)
would not encourage deer jumping into them. During the study deer were never observed within
these exclosures. There were also no tops placed on these exclosures to allow bird use and
fencing was approximately 0.22m above the ground to allow for small mammals to enter the
exclosures to ensure we were only isolating deer herbivory.
We evaluated horizontal cover in every exclosure and paired open area in every
management unit using a profile board in late July of 2011 and 2012. To quantify horizontal
60

cover we used 5 categories of cover when making our ocular estimates; 0-20%, 20-40%, 4060%, 60-80%, and 80-100%. We quantified horizontal cover with the moist soil, perennial
marsh, and lakeplain prairie vegetation types for <1.0m height strata. Within the bottomland
hardwood forest vegetation type we quantified horizontal cover within 0-0.5m, 0.5-1.0m and
>1.0m height strata. A minimum of 3 measurements were taken at each exclosure and open area.
Measurements were averaged per management unit and per vegetation type.
Vertical cover was quantified using the line intercept method during late July of 2011 and
2012 (Canfield 1941). Within moist soil, perennial marsh, and lakeplain prairie vegetation types
line intercepts were 3m in length and vertical cover was quantified for all vegetation <1.0m in
height. Within the bottomland hardwood forest vegetation type, we assessed vertical cover
within 4 different height strata with a 10 m transect; <1.0m, 1.0-2.0, and>2.0m. We took 3
vertical cover measurements in each exclosure and paired open area for all vegetation types and
then averaged these measurements. Measurements were then averaged on a per management
unit and vegetation type basis.
We evaluated the plant species richness of moist soil, perennial marsh, lakeplain prairie,
and bottomland hardwood forest vegetation types during late July of 2011 and 2012. Within the
moist soil, perennial marsh, and lakeplain prairie vegetation types the number of species present
in a 0.5m2 randomly placed plots within exclosures and open areas was recorded (Atkinson et al.
2010). Within the bottomland hardwood forests we took overstory species richness
measurements within 20 X 20m plots.
We measured total above ground and seed biomass of moist soil, perennial marsh, and
lakeplain prairie vegetation types within all exclosures and open areas during mid-August of
2011 and 2012. Vegetation was clipped at ground level within a 0.5m2 area placed randomly

61

within the exclosure and paired open area. Vegetation was separated by species and then dried at
100οC to a constant dry weight before being massed (Atkinson et al. 2010). Total biomass was
massed with seed heads still attached to plants and seed heads were removed and massed
separately to obtain seed biomass measurements.
We also visually documented the structure, composition, and productivity of each
exclosure and paired open area of moist soil, perennial marsh and lakeplain prairie vegetation
types by photographing each exclosure on a monthly basis with the first photo being taken when
the exclosures were constructed and the last photo taken when the exclosures were taken down.
Photos were taken from the same direction and at times during the day when there was the same
amount of light to reduce potential mirages from light or distance. These photos are archived at
Michigan State University with Dr. Henry Campa, III and at SNWR.
We analyzed plant composition, structure, and productivity data by first looking at the
averages of each vegetation type and making 3 comparisons within each vegetation type. We
compared all exclosures to all open areas, core exclosures to core open areas, and edge
exclosures to edge open areas. Then we analyzed each management unit or impoundment
individually within the vegetation types doing the same 3 comparisons for all variables. We used
two-tailed t-tests to make all comparisons (α=0.10) between exclosures and paired areas open to
foraging.
RESULTS
Moist Soil Vegetation Type
Within the moist soil vegetation type there were no significant differences for the average
horizontal cover between all exclosures (80-100%) and all open areas (80-100%), between all

62

core exclosures (80-100%) and all core open areas (80-100%), nor between all edge exclosures
(80-100%) and all edge open areas (80-100%) (P>0.10, α=0.10) (Table 3.1). We found no
significant differences between any of the management units when comparing all exclosures to
all open areas, core exclosures to core open areas, and edge exclosures to edge open areas
because the mean value for horizontal cover was 80-100% for every moist soil management unit
and every category.
There were no significant differences in vertical cover for moist soil units between all
exclosures (96.6%) compared to all open areas (96.9%), core exclosures (97.2%) compared to
core open areas (97.2%), or when edge exclosures (96.6%) were compared to edge open areas
(96.1%) (P> 0.10, α=0.10) (Table 3.2). We found no significant differences among the means
within each moist soil management unit; MSU1, MSU 2E, MSU 2W, MSU 6. There are no
significant differences within MSU 1 between all exclosures (95.8%) and all open areas (95.8%),
between core exclosures (93.3%) and core open areas (93.3%), and between edge exclosures
(98.3%) and edge open areas (98.3%) (P>0.1, α=0.10) (Table 3.2). Within MSU 2E there were
no significant differences in vertical cover among all exclosures (95.0%) compared to all open
areas (94.2%), core exclosures (98.3%) to core open areas (98.3%), and edge exclosures (91.7%)
to edge open areas (90.0%) (P>0.1, α=0.10) (Table 3.2). Within MSU 2W, we found no
significant differences among the means when comparing all exclosures to all open areas, core
exclosures to core open areas, and edge exclosures to edge open areas because all means were
exactly the same at 100% (P>0.1, α=0.10) (Table 3.2). Looking at MSU 6 we found no
significant differences between all exclosures (98.7%) and all open areas (98.3%), core
exclosures (100.0%) to core open areas (100.0%), and edge exclosures (97.4%) to edge open
areas (96.6%) (P>0.1, α=0.10) (Table 3.2).
63

Species richness was analyzed making the same 3 comparisons between core and edge
exclosures, among each moist soil management unit and as a whole for the moist soil vegetation
type. Within the moist soil vegetation type there was no significant difference between species
richness when comparing all exclosures (5.7) to all open areas (5.3), core exclosures (5.9) to core
open areas (5.9), and edge exclosures (5.0), to edge open areas (5.3) (P>0.1, α=0.10) (Table 3.3).
There were no significant differences between average species richness within MSU 1 when
comparing between all exclosures (6.6) to all open areas (6.0), core exclosures (6.0) to core open
areas (6.3), and edge exclosures (6.0) to edge open areas (7.0) (P>0.1, α=0.10) (Table 3.3). In
addition, there were no significant differences between average species richness with MSU 2E,
MSU 2W and MSU 6, respectively when comparing between all exclosures (5.5,5.3, and 5.5) to
all open areas (5.1, 5.3, and 5.1), core exclosures (5.3, 7.2, and 5.3) to core open areas (5.3, 6.2,
and 6.0), and edge exclosures (5.6, 3.5, and 5.0) to edge open areas (5.0, 4.3, and 4.8) (P>0.1,
α=0.10) (Table 3.3).
There were no significant differences in total above ground biomass among the moist soil
vegetation type when comparing all exclosures (160.0kg/ha) to all open areas (178.7kg/ha), all
core exclosures (157.1kg/ha) to all core open areas (183.3kg/ha), and all edge exclosures
(163.3kg/ha) to all edge open areas (174.1kg/ha) (P>0.1, α=0.10) (Table 3.4). When analyzing
MSU 1 individually, we found no significant differences between above ground biomass
between all exclosures (99.1kg/ha) to all open area (141.8kg/ha), core exclosures (138.4kg/ha) to
core open areas (102.0kg/ha), and edge exclosures (145.0kg/ha) to edge open areas (96.4kg/ha)
(P>0.1, α=0.10) (Table 3.4). Within MSU 2E we found no significant differences when
comparing all exclosures (143.3kg/ha) to all open areas (152.2kg/ha), core exclosures
(140.7kg/ha) to core open areas (164.2kg/ha), and edge exclosures (146.2kg/ha) to edge open
64

areas (38.5kg/ha) (P>0.1, α=0.10) (Table 3.4). Within MSU 2W we found no significant
difference when comparing all exclosures (159.8kg/ha) to all open areas (169.0kg/ha), core
exclosures (148.4kg/ha) to core open areas (164.0kg/ha), and edge exclosures (173.1kg/ha) to
edge open areas (173.1kg/ha) (P>0.1, α=0.10) (Table 3.4). Within MSU 6 we found no
significant differences in biomass between all exclosures (201.0kg/ha) to all open areas (227.
6kg/ha), core exclosures (199.7kg/ha) to core open areas (236.1kg/ha), and edge exclosures
(202.6kg/ha) to edge open areas (218.1kg/ha) (P>0.1, α=0.10) (Table 3.4).
No significant differences (P>0.10, α=0.10) in seed biomass occurred on moist soil
vegetation types between all exclosures (39.4kg/ha) and all open areas (35.9kg/ha), all core
exclosures (33.8kg/ah) and all core open areas (39.5kg/ha), and all edge exclosures (49.2kg/ha)
and all edge open areas (31.8kg/ha) (Table 3.5). Similarly, no significant differences in seed
biomass production occurred on MSU 1 between all exclosures (12.7kg/ha) to all open areas
(24.2kg/ha), core exclosures (8.6kg/ha) to core open areas (14.8kg/ha), and edge exclosures (33.5
kg/ha) to edge open areas (15.0kg/ha) (Table 3.5). Within MSU 2E we found no significant
differences between the mean seed biomass production of all exclosures (6.8kg/ha) and all open
areas (11.2kg/ha), core exclosures (5.3kg/ha) and core open areas (11.4kg/ha), and edge
exclosures (9.7kg/ha) and edge open areas (11.3kg/ha) (Table 3.5). Within MSU 2W we found
no significant differences between the means of seed biomass when comparing all exclosures
(33.3kg/ha) to all open areas (31.5kg/ha), core exclosures (21.5kg/ha) to core open areas
(35.1kg/ha), and edge exclosures (49.9kg/ha) to edge open areas (28.6kg/ha) (Table 3.5). Within
MSU 6 we found no significant differences when comparing mean seed biomass of all
exclosures (75.6kg/ha) to all open areas (58.0kg/ha) and core exclosures (63.3kg/ha) to core open
areas (65.55kg/ha). Moist soil unit 6 did have significantly less (P<0.10, α=0.1) seed biomass
65

production in the open areas (46.8kg/ha) in the edge region than within the exclosures
(133.4kg/ha) in the edges (Table 3.5).
Perennial Marsh Vegetation Type
There were no significant differences in the percent horizontal cover of the perennial
marsh vegetation type when comparing all exclosures (60-80%) to all open areas (60-80%), all
core exclosures (20-40%) to all core open areas (20-40%), and all edge exclosures (80-100%) to
al edge open areas (80-100%) (P>0.1, α=0.10) (Table 3.6). Similarly, there were no significant
differences in horizontal cover within each individual management unit, Butch’s Marsh or Pool
1A (P>0.1, α=0.10) (Table 3.6).
There were no significant differences (P>0.10) in the mean percent vertical cover within
the perennial marsh vegetation type between all exclosures (54.5%) to all open areas (55.8%), all
core exclosures (30.8%) to all core open areas (34.1%), and all edge exclosures (78.3%) to all
edge open areas (77.4%) (P>0.1, α=0.10) (Table 3.7).Within Butch’s Marsh, we found no
significant differences in the mean percent vertical cover within all exclosures (64.1%) compared
to all open areas (64.1%), core exclosures (36.6%) to core open areas (36.6%), and edge
exclosures (91.6%) to edge open areas (91.6%) (P>0.1, α=0.10) (Table 3.7). Similarly, within
Pool 1A, there were no significant differences (P>0.10) in mean percent vertical cover between
all exclosures (45.0%) to all open areas (47.6%), core exclosures (25.0%) to core open areas
(31.8%), and edge exclosures (65.0%) to edge open areas (63.3%) (Table 3.7).
When comparing mean species richness within the perennial marsh vegetation type we
found no significant differences between all exclosures (3.6) and all open areas (3.6), all core
exclosures (4.0) and all core open areas (4.0), or all edge exclosures (3.3) and all edge open areas
66

(3.3) (P>0.1, α=0.10) (Table 3.8). Looking at Butch’s Marsh individually we found no
significant differences in mean species richness between all exclosures (2.0) and all open areas
(1.6), core exclosures (2.5) and core open areas (1.5), and edge exclosures (1.6) and edge open
areas (1.6) (P>0.1, α=0.10) (Table 3.8). Looking at just Pool 1A we found no significant
differences between all exclosures (4.5) and all open areas (4.7), core exclosures (4.6) and core
open areas (5.0), and edge exclosures (4.4) and edge open areas (4.4) (P>0.1, α=0.10) (Table
3.8).
When comparing mean total above ground biomass production we found no significant
differences with the perennial marsh vegetation type when comparing all exclosures
(296.0kg.ha) to all open areas (313.2kg/ha), all core exclosures (249.7kg/ha) to all core open
areas (238.9kg/ha), and all edge exclosures (346.0kg/ha) to all edge open areas (399.3kg/ha)
(P>0.1, α=0.10) (Table 3.9). Looking at Butch’s Marsh individually we found no significant
differences when comparing the means of total above ground biomass of all exclosures
(535.1kg/ha) to all open areas (663.6kg/ha), core exclosures (387.0kg/ha) to core open areas
(571.5kg/ha), and edge exclosures (683.6kg/ha) to edge open areas (737.3kg/ha) (P>0.1, α=0.10)
(Table 3.9). Within Pool 1A we found no significant differences between the means of the total
above ground biomass when comparing all exclosures (247.1kg/ha) to all open areas
(243.1kg/ha), core exclosures (219.8kg/ha) to core open areas (185.7kg/ha), and edge exclosures
(265.3kg/ha) to edge open areas (314.8kg/ha) (P>0.1, α=0.10) (Table 3.9).
When comparing the means of seed biomass production within the perennial marsh
vegetation type, Pool 1A was the only management unit which produced any seed mass. We
found no significant differences between mean seed mass production when comparing all
exclosures (25.0kg/ha) to all open areas (29.3kg/ha), core exclosures (12.7kg/ha) to core
67

exclosures (25.2kg/ha), and edge exclosures (36.3kg/ha) to edge open areas (32.7kg/ha) within
Pool 1A (P>0.1, α=0.10) (Table 3.10).
Lakeplain Prairie Vegetation Type
The lakeplain prairie vegetation type had 2 management units; SU1 and SU3. These
units were small and yielded no edge making only the comparisons between exclosures and open
areas possible. We first looked at horizontal cover and found no significant differences with the
lakeplain prairie vegetation type, SU1 or SU3 between all exclosures and all open areas because
the mean horizontal cover values were 80-100% for all measurements (P>0.1, α=0.10) (Table
3.11).
Comparing the mean vertical cover values for the lakeplain prairie vegetation type, SU1,
and SU3 found no significant differences between all exclosures and open areas. All mean
values were 100% for vertical cover (P>0.1, α=0.10) (Table 3.12).
Within the lakeplain prairie vegetation type we found no significant differences between
the exclosures (3.8) and open areas (4.8) for the mean species richness values (P>0.1, α=0.10)
(Table 3.13). We found no significant differences of the mean species richness values within
SU1 and SU 3 when comparing mean exclosure values of 3.6 and 4.1 respectively to mean open
area values of 4.8 and 4.5 respectively (P>0.1, α=0.10) (Table 3.13).
When comparing the total above ground biomass of the lakeplain prairie vegetation type
we found no significant differences when comparing the means of exclosures (206.4kg/ha) to
open areas (179.6kg/ha) (P>0.1, α=0.10) (Table 3.14). Within SU1 and SU3 we found no
significant differences when comparing the exclosure means of 193.9kg/ha and 221.0kg.ha
respectively to the open area means of 173.1kg/ha and 189.7kg/ha respectively (P>0.1, α=0.10)
68

(Table 3.14). We also found no significant difference between the mean seed biomass of the
exclosures (29.4kg/ha) and open areas (23.8) within the lakeplain prairie vegetation type (P>0.1,
α=0.10) (Table 3.15). When comparing the mean seed biomass of SU1 and SU3 we found no
significant difference when comparing the exclosure means of 25.7kg/ha and 31.7kg.ha to the
open area means of 26.5kg/ha and 18.1kg.ha respectively (P>0.1, α=0.10) (Table 3.15).
Bottomland Hardwood Forest Vegetation Type
Within the bottomland hardwood forest vegetation type we found no significant
differences between all exclosure and open areas or within any individual stand for mena
horizontal cover within the <0.5m height strata, 0.5-1.0m height strata, and >1.0m height strata
because all values were 0-20% (P>0.1, α=0.10) (Table 3.16). When looking at mean vertical
cover we found no significant differences for the bottomland hardwood forest vegetation type
between exclosures (37. 9%) values and open areas (36.8%) within the >1.0m height strata
(P>0.1, α=0.10) (Table 3.17). We found no significant differences for mean vertical cover
within any of the individual stands for the >1.0m height strata between exclosures and open areas
and report exclosure means 80.6%, 14.0%, 16.1%, 29.6%, 46.6%, and 39.35within stands 1-6
respectively and open area means of 77.0%, 17.5%, 15.7%, 31.6%, 46.0%, and 37.6%within
stands 1-6 respectively (P>0.1, α=0.10) (Table 3.17). Within the 1.0-2.0m height strata we
found no significant differences in mean vertical cover between the exclosures and open areas
with the bottomland hardwood forest vegetation type or within any of the individual stands
because all values were 0.0% (P>0.1, α=0.10) (Table 3.17). Within the <2.0m height strata we
found no significant differences of the mean vertical cover between exclosures (85.7%) and open
areas (84.2%) for bottomland hardwood forest vegetation type (P>0.1, α=0.10) (Table 3.17). We
found no significant differences for mean vertical cover within any of the individual stands for
69

the <2.0m height strata between exclosures and open areas and report exclosure means of 83.3%,
83.3%, 83.3%, 88.3%, 86.6%, and 88.3% for stands 1-6 respectively and open area means of
90.0%, 85.0%, 80.0%, 81.6%, 91.6%, and 88.3%within stands 1-6 respectively (P>0.1, α=0.10)
(Table 3.17).
Within the bottomland hardwood forest vegetation type we found no significant
differences for species richness between exclosures (3.7) and open areas (3.1) (P>0.1, α=0.10)
(Table 3.18). We also found no significant differences between any of the individual stands
exclosures and open areas and found exclosure means for species richness to be 3.0, 3.0, 4.0, 3.0,
4.0, and 4.0 for exclosures in stands 1-6 respectively and means within the open areas to be 3.0,
2.0, 3.0, 4.0, 2.0, and 4.0 within stands 1-6 respectively (P>0.1, α=0.10) (Table 3.18).
In the future we would also recommend sampling stem densities of woody species
at <1m, 1-3m and >3m height strata. We would also recommend recording average diameter at
breast height in trees >3m, because deer prefer to feed on the leaves and shoots of new growth of
woody plant species and measuring this variable within exclosures and open areas might show if
browsing pressure is too high (Johnson et al. 1995).
DISCUSSION
White-tailed deer at high densities can be a keystone herbivore and negatively impact
forest community composition, structure, and biomass, however, little is known on the effects
deer herbivory can have on the wetland vegetation type they use to fulfill life requisites (Kraft et
al. 2004, Rooney and Waller 2003, Rooney 2001). Stewart et al. (2006) found that deer feed
heavily in the edge of croplands that were surrounded by forests. This may be why in MSU 6 we
found the exclosures in the edge region had significantly more seed biomass than the edge open
70

areas. Williamson and Hirth (1985) and Braun (1996) also found deer foraging to negatively
impact the edges of crop fields more than the core areas of those fields. The reason we may have
not seen significant difference between above ground biomass but only in the seed biomass in
MSU 6 between the edge exclosures and opens areas is that deer sometimes feed on the flowers
of plants (Urbanek 2012, Rooney 2001). If the flower of a plant is eaten late enough in the
growing season the plants may not be able to produce another flower and therefore seed
productivity (Kettering et al. 2009). We may have been seeing this describe occurrence in MSU
6 which is why seed biomass was significantly different within edge exclosures and open areas
but total above ground biomass was not.
Our results quantifying minimal to no effects of deer herbivory on wetland plant
vegetation types may be attributed to the lack of preferred spring and summer foods within these
vegetation types. Hiller (2007), Miranda and Porter (2003), and Pusateri (2003) all found whitetailed deer selecting for agriculture vegetation types when foraging for spring and summer foods.
Hiller (2007) and Pusateri (2003) did document deer using wetland vegetation types during the
spring and summer months for feeding and hiding cover but they were not the most important
vegetation types but Johnson et al. (1995) found white-tailed deer do prefer to feed on new
growth of woody plant species. Dostaler et al. (2010) and Daigel et al. (2004) found that when
agriculture was present in a landscape and available to white-tailed deer for feeding, agricultural
crops made up a larger portion of the diet than woody plant species, forbs and grasses than in
areas where agricultural crops were limited. These results suggest agricultural crops may be
acting as preferred foods, over wetland plant species, for white-tailed deer within this region of
Michigan because of their abundance immediately adjacent to the SNWR.

71

Russel et al. (2001) commented that the effects of deer herbivory within forest vegetation
types are not immediate effects but compounding and being to show after years of heavy
browsing. In future years we may anticipate different results than from our initial 2 years within
the bottomland hardwood forest vegetation type due to the exclusion of deer from plants within
the exclosures, since negative impacts to regeneration and recruitment of forest vegetation sites
is often seen after a longer time period (Raymer 2000, Russell et al. 2001).
Additionally, we also may not have seen any negative effects to plant species richness,
horizontal and vertical cover, total above ground biomass, or seed biomass because deer
densities at SNWR have been are highly managed over the last 15 years. Biologists at SNWR
try to maintain estimated deer density of ~1.5deer/ha within the refuge boundary. To meet this
goal, they issue specific hunting permit regulations every fall. The estimated number of deer has
been greatly reduced through hunting efforts from ~4.2 deer/ha in the early 1990’s to ~1.5
deer/ha today (Fig. 3.1). If we had conducted this study in the early 1990’s we may have seen
drastically different results. Urbanek et al.’s (2012) exclosure study documented high deer
densities (3.0 deer/ha) within savanna vegetation types negatively impacted savanna plant
communities by reducing the number of flowering individuals and species diversity. Anderson
et al. (2004) also found the community composition within a tallgrass prairie to decrease in times
of intense deer herbivory (6.0 deer/ha). Although this study did not show how deer herbivory
could be a challenge to manage within wetland vegetation types previous research has showed
that deer densities at high levels are what cause changes to the plant communities.

72

MANAGEMENT IMPLICATIONS
High densities of white-tailed deer can pose challenges to natural resource managers in
landscapes dominated by wetlands where deer densities are not highly managed. Those
managers may want to avoid planting costly and highly preferred plant species for waterfowl
within the edge of fields or management units. Natural resource managers may expect to find
fewer flowering and fruiting individuals in wetland landscapes where deer densities are high but
when deer densities are lower we may not see negative impacts of deer herbivory to a diversity
of wetland plant communities.
Overall low densities of deer (~1.5 deer/ha) on our study area did not negatively impact
wetland plant communities. In certain instances we may see more effects of herbivory around
the edge of management units or fields when they are surrounded by preferred vegetation types
like deciduous forest vegetation types. Managing for lower deer densities in landscapes
dominated by wetlands may help reduce potential negative impacts to wetland plant communities
and help avoid seeing plant community changes therefore allowing biologists to achieve other
goals and objectives such as maintaining habitat for migratory and resident waterfowl.

73

Table 3.1. Average horizontal cover <1.0m (standard errors) for moist soil (MSU) vegetation types at Shiawassee National Wildlife
Refuge 2011 and 2012. A two tailed t-test was used to compare average horizontal cover between all exclosures, all open areas, core
exclosures vs. core open areas, and edge exclosures vs. edge open areas (α=0.10).
Unit

Mean Horizontal Cover (%)

All MSU

All
Exclosures
80-100 (0.3)

All Open
Areas
80-100 (0.3)

Core
Exclosures
80-100 (0.0)

Core Open
Areas
80-100 (0.4)

Edge
Exclosures
80-100 (0.4)

Edge Open
Areas
80-100 (0.0)

MSU 1

80-100 (0.3)

80-100 (0.0)

80-100 (0.4)

80-1000 (0.0)

80-100 (0.0)

80-100 (0.0)

MSU 2E

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

MSU 2W

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

MSU 6

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

80-100 (0.0)

74

Table 3.2. Mean percent vertical cover <1.0m (standard errors) for the moist soil (MSU) vegetation type and each management unit
individually within the moist soil vegetation type at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was
used to compare means between all exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open
areas (α=0.10).
Unit

Mean Vertical Cover (%)

All MSU

All
Exclosures
96.6 (0.4)

All Open
Areas
96.9 (0.5)

Core
Exclosures
97.2 (0.4)

Core Open
Areas
97.2 (0.4)

Edge
Exclosures
96.6 (0.5)

Edge Open
Areas
96.1 (0.6)

MSU 1

95.8 (0.5)

95.8 (0.5)

93.3 (0.5)

93.3 (0.6)

98.3 (0.3)

98.3 (0.3)

MSU 2E

95.0 (0.3)

94.2(0.7)

98.3 (0.3)

98.3 (0.3)

91.7 (0.5)

90.0 (0.3)

MSU 2W

100 (0.0)

100 (0.0)

100 (0.0)

100 (0.0)

100 (0.0)

100 (0.0)

MSU 6

98.7 (0.2)

98.3 (0.2)

100 (0.0)

100 (0.0)

97.4 (0.3)

96.6 (0.4)

75

Table 3.3. Mean species richness (standard errors) for the moist soil (MSU) vegetation type and
each management unit individually within the moist soil vegetation type at Shiawassee National
Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge
open areas (α=0.10).
Unit

Mean Species Richness

All MSU

All
Exclosures
5.7 (0.1)

All Open
Areas
5.3 (0.1)

Core
Core Open
Exclosures Areas
5.9 (0.2)
5.9 (0.1)

Edge
Exclosures
5.0 (0.2)

Edge Open
Areas
5.3 (0.3)

MSU 1

6.6 (0.8)

6.0 (0.5)

6.0 (0.6)

6.3(0.5)

6.0 (1.5)

7.0 (1.0)

MSU 2E

5.5 (0.6)

5.1 (0.5)

5. 3 (0.8)

5.3 (0.8)

5.6 (0.9)

5.0 (0.7)

MSU 2W

5.3 (0.7)

5.3 (0.5)

7.2 (0.7)

6.2 (0.8)

3.5 (0.6)

4.3 (0.5)

MSU 6

5.5 (0.4)

5.1 (0.4)

5.3 (0.6)

6.0 (0.6)

5.0 (0.4)

4.8 (0.6)

76

Table 3.4. Mean above ground biomass production (standard errors) for the moist soil (MSU) vegetation type and each management
unit individually within the moist soil vegetation type at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was
used to compare means between all exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open
areas (α=0.10).
Unit

Mean Above Ground Biomass (kg/ha)

All MSU

All
Exclosures
160.0 (3.0)

All Open
Areas
178.7 (3.8)

Core
Exclosures
157.1 (4.0)

Core Open
Areas
183.3 (5.3)

Edge
Exclosures
163.3 (4.5)

Edge Open
Areas
174.1 (5.5)

MSU 1

99.1 (51.1)

141.8 (83.7)

138.4 (67.8)

102.0 (121.1)

145.0 (77.7)

96.4 (119.3)

MSU 2E

143.3 (54.4)

152.2 (56.8)

140.7 (72.9)

164.2 (79.0)

146.2 (83.7)

38.5 (83.1)

MSU 2W

159.8 (56.8)

169.0 (74.9)

148.4 (63.9)

164.0 (95.4)

173.1 (96.6)

173.1 (115.0)

MSU 6

201.0 (62.2)

227.6 (79.5)

199.7 (92.1)

236.1 (188.5)

202.6 (84.0)

218.1 (106.8)

77

Table 3.5 Mean seed biomass production (standard errors) for the moist soil (MSU) vegetation type and each management unit
individually within the moist soil vegetation type at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was used
to compare means between all exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open areas
(α=0.10).
Unit

Mean Seed Biomass (kg/ha)

All MSU

All
Exclosures
39.4 (7.0)

All Open
Areas
35.9 (5.9)

Core
Exclosures
33.8 (5.2)

Core Open
Areas
39.5 (6.2)

Edge
Exclosures
49.2 (9.5)

Edge Open
Areas
31.8 (5.7)

MSU 1

12.7 (6.5)

24.2 (12.9)

8.6 (7.2)

14.8 (7.6)

33.5 (10.0)

15.0 (24.8)

MSU 2E

6.8 (3.8)

11.2 (7.9)

5.3 (3.3)

11.4 (9.5)

9.7 (8.5)

11.3 (13.8)

MSU 2W

33.3 (22.0)

31.5 (18.2)

21.5 (11.2)

35.1 (26.9)

49.9 (46.5)

28.6 (25.3)

MSU 6

75.6 (20.6)

58.0 (18.4)

63.3 (26.7)

65.55 (27.5)

133.4 (33.3)*

49.8 (24.2)*

*Averages followed by * are significantly different from one another.

78

Table 3.6. Mean horizontal cover <1.0m (standard errors) for the perennial marsh (PM) vegetation type and each management unit
individually at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open areas (α=0.10).
Unit

Mean Horizontal Cover (%)

All PM

All
All Open
Exclosures Areas
60-80 (0.8) 60-80 (0.8)

Core
Exclosures
20-40 (0.9)

Core Open
Areas
20-40 (1.2)

Edge
Exclosures
80-100 (0.6)

Edge Open
Areas
80-100 (0.6)

Butch’s Marsh

60-80 (0.8) 60-80 (0.8)

0-20 (1.3)

20-40 (1.3)

80-100 (0.0)

80-100 (0.0)

Pool 1A

60-80 (0.8) 60-80 (0.8)

20-40 (0.6)

20-40 (1.0)

80-100 (1.3)

80-100 (1.3)

79

Table 3.7. Mean percent vertical cover <1.0m (standard errors) for the perennial marsh (PM) vegetation type and each management
unit individually at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open areas (α=0.10).
Unit

Mean Vertical Cover (%)

All PM

All
Exclosures
54.5 (17.7)

All Open
Areas
55.8 (17.7)

Core
Exclosures
30.8 (20.8)

Core Open
Areas
34.1 (23.3)

Edge
Exclosures
78.3 (17.7)

Edge Open
Areas
77.4 (17.5)

Butch’s Marsh

64.1 (17.3)

64.1 (17.3)

36.6 (27.2)

36.6 (27.2)

91.6 (3.3)

91.6 (3.3)

Pool 1A

45.0 (18.2)

47.6 (18.1)

25.0 (14.4)

31.8 (19.5)

65.0 (32.5)

63.3 (31.8)

80

Table 3.8. Mean species richness (standard errors) for the perennial marsh (PM) vegetation type and each management unit
individually at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open areas (α=0.10).
Unit

Mean Species Richness

All PM

All
All Open
Exclosures Areas
3.6 (0.4)
3.6 (0.5)

Core
Exclosures
4.0 (0.6)

Core Open
Areas
4.0 (0.7)

Edge
Exclosures
3.3 (0.6)

Edge Open
Areas
3.3 (0.9)

Butch’s Marsh

2.0 (0.3)

1.6 (0.2)

2.5 (0.5)

1.5 (0.5)

1.6 (0.3)

1.6 (0.3)

Pool 1A

4.5 (0.4)

4.7 (0.5)

4.6 (0.6)

5 (0.4)

4.4 (0.6)

4.4 (1.1)

81

Table 3.9. Mean total above ground biomass (standard errors) for the perennial marsh (PM) vegetation type and each management unit
individually at Shiawassee National Wildlife Refuge, 2011 and 2012. A two tailed t-test was used to compare means between all
exclosures, all open areas, core exclosures vs. core open areas, and edge exclosures vs. edge open areas (α=0.10).
Unit

Mean Total Above Ground Biomass (kg/ha)

All PM

All
Exclosures
296.0 (39.3)

All Open
Areas
313.2 (38.1)

Core
Exclosures
249.7 (38.6)

Core Open
Areas
238.9 (37.8)

Edge
Exclosures
346.0 (40.2)

Edge Open
Areas
399.3 (37.5)

Butch’s Marsh

535.1 (283.1)

663.6 (251.4)

387.0 (424.8)

571.5 (586.6)

683.6 (402.2)

737.3 (183.0)

Pool 1A

247.1(90.7)

243.1 (89.5)

219.8 (149.9)

185.7 (109.0)

265.3 (118.1)

314.8 (154.1)

Table 3.10. Mean total seed mass (standard errors) for the perennial marsh vegetation type and each management unit individually at
Shiawassee National Wildlife Refuge, 2011 and 2012. Pool 1A was the only perennial marsh management unit where seeds were
present. A two tailed t-test was used to compare means between all exclosures, all open areas, core exclosures vs. core open areas, and
edge exclosures vs. edge open areas (α=0.10).
Unit
Pool 1A

All
Exclosures
25.0 (13.6)

All Open
Areas
29.3 (9.0)

Mean Seed Biomass (kg/ha)
Core
Core Open Edge
Exclosures Areas
Exclosures
12.7 (6.1)
25.2 (15.1) 36.3 (26.8)

82

Edge Open
Areas
32.7 (12.4)

Table 3.11. Mean horizontal cover <1.0m (standard errors) for the lakeplain prairie (LPP)
vegetation type and each management unit individually at Shiawassee National Wildlife Refuge,
2011 and 2012.A two tailed t-test was used to compare means between all exclosures and all
open areas (α=0.10).
Unit

Mean Horizontal Cover (%)
Exclosure

Open Areas

All LPP

80-100 (0.0)

80-100 (0.0)

SU1

80-100 (0.0)

80-100 (0.0)

SU3

80-100 (0.0)

80-100 (0.0)

Table 3.12. Mean vertical cover <1.0m (standard errors) for the lakeplain prairie (LPP)
vegetation type and each management unit individually at Shiawassee National Wildlife Refuge,
2011 and 2012. A two tailed t-test was used to compare means between all exclosures and all
open areas (α=0.10).
Unit

Mean Vertical Cover (%)
Exclosure

Open Areas

All LPP

100 (0.0)

100 (0.0)

SU1

100 (0.0)

100 (0.0)

SU3

100 (0.0)

100 (0.0)

83

Table 3.13. Mean species richness (standard errors) for the lakeplain prairie (LPP) vegetation
type and each management unit individually at Shiawassee National Wildlife Refuge, 2011 and
2012. A two tailed t-test was used to compare means between all exclosures and all open areas
(α=0.10).
Unit

Mean Species Richness
Exclosure

Open Areas

All LPP

3.8 (0.4)

4.6 (0.5)

SU1

3.6 (0.3)

4.8 (0.5)

SU3

4.1 (0.5)

4.5 (0.6)

Table 3.14. Mean above ground biomass (standard errors) for the lakeplain prairie (LPP)
vegetation type and each management unit individually, at Shiawassee National Wildlife Refuge,
2011 and 2012. A two tailed t-test was used to compare means between all exclosures and all
open areas (α=0.10).
Unit

Mean Above Ground Biomass (kg/ha)
Exclosure

Open Areas

All LPP

206.4 (2.9)

179.6 (2.0)

SU1

193.9 (67.3)

173.1 (56.0)

SU3

221.0 (100.7)

189.7 (71.1)

84

Table 3.15. Mean seed biomass (standard errors) for the lakeplain prairie (LPP) vegetation type
and each management unit individually, at Shiawassee National Wildlife Refuge, 2011 and 2012.
A two tailed t-test was used to compare means between all exclosures, all open areas (α=0.10).
Unit

Mean Seed Biomass (kg/ha)
Exclosure

Open Areas

All LPP

29.4 (3.1)

23.8 (2.5)

SU1

25.7 (9.9)

26.5 (10.1)

SU3

31.7 (16.6)

18.1 (6.9)

85

Table 3.16. Mean horizontal cover (standard errors) for the bottomland hardwood forest vegetation type and each stand individually,
at Shiawassee National Wildlife Refuge, 2011 and 2012. 3 height strata were used to make measurements <0.5m, 0.5-1.0m, and
>1.0m. A two tailed t-test was used to compare means between all exclosures and all open areas (α=0.10).
Unit

Mean Horizontal Cover (%)
<0.5m

0.5-1.0m

>1.0m

Exclosures

Open Areas

Exclosures

Open Areas

Exclosures

Open Areas

All Stands

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

1

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

2

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

3

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

4

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

5

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

6

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

0-20 (0.0)

86

Table 3.17. Mean vertical cover (standard errors) for the bottomland hardwood forest vegetation type and each stand individually at
Shiawassee National Wildlife Refuge, 2011 and 2012. 3 height strata were used to make measurements <1.0m, 1.0-2.0m, and >2.0m.
A two tailed t-test was used to compare means between all exclosures and all open areas (α=0.10).
Stand #

Mean Vertical Cover (%)
<1.0m

1.0-2.0m

>2.0m

Exclosures

Open Areas

Exclosures

Open Areas

Exclosures

Open Areas

All Stands

37.9 (0.9)

36.8 (0.6)

0.0 (0.0)

0.0 (0.0)

85.7 (0.8)

84.2 (1.6)

1

80.6 (0.3)

77.0 (0.4)

0.0 (0.0)

0.0 (0.0)

83.3 (0.4)

90.0 (0.3)

2

14.0 (0.2)

17.5 (0.5)

0.0 (0.0)

0.0 (0.0)

83.3 (0.3)

85.0 (0.2)

3

16.1 (0.2)

15.7 (0.4)

0.0 (0.0)

0.0 (0.0)

83.3 (0.3)

80.0 (0.3)

4

29.6 (0.2)

31.6 (0.4)

0.0 (0.0)

0.0 (0.0)

88.3 (0.3)

81.6 (0.2)

5

46.6 (0.4)

46.0 (0.5)

0.0 (0.0)

0.0 (0.0)

86.6 (0.6)

91.6 (0.4)

6

39.3 (0.2)

37.6 (0.3)

0.0 (0.0)

0.0 (0.0)

88.3 (0.3)

88.3 (0.5)

87

Table 3.18. Mean overstory species richness (standard errors) for the bottomland hardwood
forest vegetation type and each stand individually, at Shiawassee National Wildlife Refuge, 2011
and 2012. A two-tailed t-test for used to compare means between exclosures and open areas
(α=0.10).
Stand #

Mean Overstory Species Richness
Exclosure

Open Area

All Stands

3.5 (0.1)

3.1 (0.2)

1

3 (0.0)

3 (0.0)

2

3 (0.0)

2 (0.0)

3

4 (0.0)

3 (0.0)

4

3 (0.0)

4 (0.0)

5

4 (0.0)

3 (0.0)

6

4 (0.0)

4 (0.0)

88

Number of Deer

Number of Observed and Harvested Deer at
SNWR
1600
1400
1200
1000
800
600
400
200
0

Year
Aerial Survey Results (SGA & SNWR)
Shiawassee NWR Registered Harvest

Figure 3.1 Deer estimates and hunter harvest numbers of Shiawassee National Wildlife Refuge

89

LITERATURE CITED

90

LITERATURE CITED
Anderson, R.C., D. Nelson, M.R. Anderson, and M.A. Rickey. 2004. White-tailed deer
(Odocoileus virginianus) browsing effects on quality of trallgrass prairie community
forbs. Proceedings of the 19th North American Prairie Conference.
Atkinson, R.B., J.E. Perry, G.B. Noe, W.L. Daniels, and J. Cairns Jr. 2010. Primary productivity
in 20-year old created wetlands in southwestern Virginia. Wetlands. 30:200-210.
Barnes, V.B. and W.H. Wagner, Jr. 2008. Michigan trees: A guide to the trees of Michigan and
Great Lakes region. University of Michigan Press, Ann Arbor, Michigan, USA. 448pp.
Braun, K.F. 1996. Ecological factors influencing white-tailed deer damage to agricultural crops
in northern Michigan. M.S. Thesis. Department of Fisheries and Wildlife, Michigan State
University. East Lansing, Michigan.
Campa, H., III, J.B. Haufler, and E. Beyer, Jr. 1992. Effects of simulated ungulate browse on
aspen characteristics and nutritional qualities. The Journal of Wildlife Management
56:158-164.
Canfield, R.H. 1941. Application of the line intercept method in sampling range vegetation.
Journal of Forestry 39:388-394.
Daigle, C., M.Crete, L. Lesage, J. Ouellet, and J. Hout. 2004. Summer diet of two white-tailed
deer, Odocoileus virginianus, populations living at low and high density in Southern
Quebec. Canadian Field Naturalist 118:360-367.
Dostaler, S., J. Ouellet, J. Therrien, and S.D. Cote. 2010. Are feeding preferences of white-tailed
deer related to plant constituents? Journal of Wildlife Management 75:913-918.
Felix, A.B., D.P. Walsh, B.D. Hughey, H. Campa, III, and S.R. Winterstein. 2007. Applying
landscape-scale habitat-potential models to understand deer spatial structure and
movement patterns. The Journal of Wildlife Management 71:804-810.
Gray, M.J., R.M. Kaminski, G. Weerakkody, B.D. Leopold, and K.C. Jensen. 1999. Aquatic
invertebrate and plant responses following mechanical manipulations of moist-soil
habitat. Wildlife Society Bulletin 27:770-779.
Hiller, T.L. 2007.Land-use patterns and population characteristics of white-tailed deer in an
agro-forest ecosystem in south central Michigan. Dissertation, Michigan State University,
East Lansing, USA .
Haukos, D.A., H.Z. Sun, D.B. Wester, and L.M. Smith. 1998. Sample size, power, and analytical
considerations for vertical structure data from profile boards in wetland vegetation.
Wetlands 18:203-215.
Johnson A.S., P.E. Hale, W.M. Ford, J.M. Wentworth, J.R. French, O.F. Anderson, and G.B.
Pullen. 1995. White-tailed deer foraging in relation to successional stage, overstory type,
and management of southern Appalachian forests. American Midland Naturalist 133:1835.

91

Kettering, K.M., C.W. Weekley, and E.S. Menges. 2009. Herbivory delays flowering and
reduces fecundity of Liatrisohlingerae (Asteraceae), an endangered, endemic plant of the
Florida scrub. Journal of the Torrey Botanical Society 136:350-362.
Kraft, L.S, T.R. Crow, D.S. Buckley, E.A. Nauertz and J.C. Zasada. 2004. Effects of harvesting
and deer browsing on attributes of understory plants in northern hardwood forests, Upper
Michigan, USA. Forest Ecology and Management 199:219-230.
Miranda, B.R. and W.F. Porter. 2003. Statewide habitat assessment for white-tailed deer in
Arkansas using satellite imagery. Wildlife Society Bulletin 31:715-726.
National Weather Service (IWIN) and the National Oceanic Atmospheric Administration
(NOAA). 2007. Climate for Saginaw, MI.
<http://www.rssweather.com/climate/Michigan/Saginaw> Accessed 12 Dec. 2010
Pusateri, J.S. 2003. White-tailed deer population characteristics and landscape use patterns in
southwestern lower Michigan. Thesis, Michigan State University, East Lansing, USA.
Raymer, D.F.N. 2000. Effects of elk and white-tailed deer browsing on aspen communities and
wildlife habitat quality in northern lower Michigan: an 18 year evaluation. Ph.D.
Dissertation. Department of Fisheries and Wildlife, Michigan State University. East
Lansing, Michigan.
Rodrígues-Pérez, H. and A.J. Green. 2006. Waterbird impacts on widgeongrass Ruppia
maritime in a Mediterranean wetland: comparing bird groups and seasonal effects. Okios
112: 525-5334.
Rooney, T.P. 2001. Deer impacts on forest ecosystems: a North American perspective. Forestry
74:201-208.
Rooney, T.P. and D.M. Waller. 2003. Direct and indirect effects of white-tailed deer in forest
ecosystems. Forest Ecology and Management 181:165-176.
Russell, F.L., D.B. Zippin, and N.L. Fowler. 2001. Effects of white-tailed deer (Odocoileus
virginianus) on plants, plant populations and communities: A review. The American
Midland Naturalist 146:1-26.
Stewart, C.M., W.J. Mcshea, and B.P. Piccolo. 2006. The impact of white-tailed deer on
agricultural landscapes in 3 national historic parks in Maryland. Journal of Wildlife
Management 71:1525-1530.
Urbanek, R.E., C.K. Nielsen, G.A. Glowacki, and T.S. Pruess. 2012. Effects of white-tailed deer
(Odocoileus virginanus Zimm.) herbivory in restored forest and savanna plant
communities. American Midland Naturalist 167:240-255.
United States Department of Agriculture. 1997. National Agricultural Stats Service. State and
CountyProfiles.<http://www.agcensus.usda.gov/Publications/1997/County_Profiles/Mich
igan/index.asp> Accessed 8 March 2010.
92

U.S Fish and Wildlife Service. National Wildlife Refuge System. Welcome to the National
Wildlife Refuge System. Department of the Interior, 23 Feb. 2010.Web. 12 Dec. 2010. <
http://www.fws.gov/refuges/about/welcome.html >.
Vercauteren, K.C. and S.E. Hygnstrom. 1998. Effects of agriculture activities and hunting on
home ranges of female white-tailed deer. Journal of Wildlife Management 62:280-285.
Waller, D.M. and W.S. Alverson. 1997. The white-tailed deer: a keystone herbivore. Wildlife
Society Bulletin 25:217-226.
Williamson, S.J. and D.H. Hirth. 1985. An evaluation of edge use by white-tailed deer. Wildlife
Society Bulletin 13:252-257.

93

APPENDICES

94

APPENDIX A

Locations of forest exclosures and sample data sheet for continuation of herbivory monitoring
within bottomland hardwood forest vegetation type

95

Table A.1. Location of exclosures and paired open areas within forested stands at Shiawassee
National Wildlife Refuge. Open area locations are referenced in cardinal direction and distance
from exclosure.
Exclosure
Number

Location (Degrees Minutes Seconds)

Location (from exclosure)

Exclosure

Open Area

ο

ο

1

N 43 21’ 2.6” W 84 00’ 30.9”

2

N 43 20’ 12.3” W 84 00’ 30.3”

3

N 43 20’ 20.6” W 84 01’ 0.3”

4

N 43 21’ 33.9” W 83 59’ 16.7”

5

N 43 21’ 23.9” W 83 59’ 29.3”

6

N 43 22’ 3.9” W 83 57’ 28.2”

10m from west side

ο

ο

10m from south side

ο

ο

10m from south side

ο

ο

10m from north side

ο

ο

10m from south side

ο

ο

96

10m from west side

Table A.2. Example data sheet for continued collection of horizontal cover, vertical cover, and
overstory species richness at each exclosure and open area within the forest exclosure sites. We
recommend following the methods described in Chapter 3.
Exclosure #________
Variable
Horizontal Cover
(<0.5m)
Horizontal Cover
(0.5-1m)
Horizontal Cover
(>1.0m)
Vertical Cover
(<1.0m)
Vertical Cover
(1-2m)
Vertical Cover
(>2.0m)
Species Richness

Location ___________________________
Measurement #
Exclosure
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1

List of Overstory Species Present

97

Open Area

Table A.3. List of species observed within exclosures and open areas at bottomland hardwood
forest sites. An X indicates when a species was present.
Species
Common Name

Scientific Name

American Beech

Fagus grandifolia

American Elm

Ulmus americana

Basswood

Tilia Americana

Cottonwood

Site Number

Populus deltoides

1

2

3

4

5

6

X

X

X

X

X

X

X
X

X

X
X

X

Shagbark Hickory Carya ovate

X

Silver Maple

Acer saccharinum

X

Sugar Maple

Acer saccharum

White Oak

Quercus alba

X
X

X

X

X

X
X

98

X

Legend
SNWR Boundary
Exclosure number and approximate location
Figure A.1. Aerial image of Shiawassee National Wildlife Refuge, refuge boundary outlined in

red. Each numbered circle indicates approximately where an exclosure and paired open area are
located within the forested stand.
99

APPENDIX B

Research and outreach activities

100

RESEARCH ACTIVITES
PROFESSIONAL PRESENTATIONS (name in bold indicates presenter)
Graduate Student Organization Research Symposium (Department of Fisheries and Wildlife, MSU) January 2013 – East Lansing MI
- Evaluating deer impacts to wetland vegetation types
S. Longstaff, H. Campa, III, A. Locher, S.R. Winterstein, E. Dunton, and S. Kahl
Stewardship Network Conference – January 2013 – East Lansing, MI
- Hunting not just recreation but also conservation: evaluating how deer hunting opportunities
help conserve wetland vegetation types and waterfowl food resources at Shiawassee National
Wildlife Refuge
S. Longstaff, H. Campa, III, A. Locher, S.R. Winterstein, E. Dunton, and S. Kahl
Midwest Fish and Wildlife Conferences – December 2012 – Wichita, Kansas
- Evaluating wetland use by white-tailed deer and the habitat suitability wetland vegetation types
provide for them
S. Longstaff, H. Campa, III, A. Locher, S.R. Winterstein, E. Dunton, and S. Kahl
Midwest Fish and Wildlife Conference – December 2011 – Des Moines, Iowa
- Behind the Fence: Evaluating how white-tailed deer impact wetland vegetation types
S. Longstaff, H. Campa, III, S.R. Winterstein, E. Dunton and S. Kahl

OUTREACH ACTIVITIES
PRESENTATIONS (name in bold indicates presenter)
Ducks Unlimited – March 2013 – Annarbor, MI
-Bucks and Ducks: Evaluating how wetlands are impacted by white-tailed deer
S. Lonstaff, H. Campa, III, S.R. Winterstein, E. Dunton, S. Kahl, and A. Locher
Shiawassee National Wildlife Refuge – January 2013- Saginaw, MI
-Yearly research update
S. Longstaff, H. Campa, III, S.R. Winterstein, E. Dunton, S. Kahl, and A. Locher
Friends of Shiawassee National Wildlife Refuge Seminar Series – June 2012 – Saginaw, MI
- Within the Levees: Taking a look at some of the research going on at Shiawassee National
Wildlife Refuge
S. Lonstaff, H. Campa, III, S.R. Winterstein, E. Dunton, S. Kahl, and A. Locher
-A presentation given to the general public highlighting current research project being conducted
at Shiawassee National Wildlife Refuge. One of the goals of the presentation was to help bridge the gap
between the scientific community and the non-science community. The presentation helped show why
research was being conducted and what the results would potentially be used for.

101

Shiawassee National Wildlife Refuge – January 2012- Saginaw, MI
-Yearly research update
S. Longstaff, H. Campa, III, S.R. Winterstein, E. Dunton, S. Kahl

SUBMITTED ARTICLES
S. Longstaff. 2014. A Buck in the Muck?: Evaluating how wetlands provide habitat for
white-tailed deer and how deer use affects birds within those wetlands. Spotlight.
- Article will be published in the upcoming 2014 issue of Spotlight Magazine.
PROFESSIONAL MEMBERSHIPS
Member of The Wildlife Society (2011- present)
Member of the North Central Section
Member of the Michigan Chapter

Member of the Graduate Student Organization (GSO) (2010-present)
- An organization for the graduate students in the Department of Fisheries and Wildlife at
Michigan State University. The organization offers opportunities for students to develop themselves
professionally, get involved in graduate student life, and maintain their health and wellness while
attending Michigan State University. The organization also provides students with travel funding and
fundraising opportunities along with weekly seminars and other educational opportunities to help gain
exposure to the entire field of Fisheries and Wildlife.

Spotlight Committee Member (2010 – present)
- Spotlight is a non-profit magazine run by volunteer graduate students in the Department of Fisheries and
Wildlife at Michigan State University. The magazine is written, designed, edited, published, and
distributed by Fisheries and Wildlife graduate students. It highlights the research and community
outreach projects the department is currently working on.
- Layout Designer: 2010 – present
- Editor: 2011 – present
- Distribution Manager: 2011 – present

102