\s

.531).

A“. I... .1- h-
L‘.‘ “Wu: K.

 

Q .. _ ' , 13.. T V . . ,

.wwz.‘ . 2+2, , . ‘ . . 4|..né...n:..!‘:.l.. L- “a“ a . . a

‘ , - .. , . . : .. .‘ z. .
. . , 4 .rnanwww £35.59:

 

"WP‘ZY‘Q

,-

m }
pgk...‘ .) «.2

U}

This is to certify that the
thesis entitled

Development of landscape-scale models to describe habitat
potential of white-tailed deer (Odocoileus virginianus) in
Michigan

presented by

Alexandra B. Felix

has been accepted towards fulfillment
of the requirements for the

MS. degree in Fisheries and Wildlife

 

 

\T(\§’b\’\\/\ KIRK {Avfi ; l i l
)Major‘fi’rofessor’s Signature
Oil i4] '0 3

l T

Date

MSU is an Affinnative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

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

DATE DUE DATE DUE DATE DUE
MAY 2005 {g g| ‘ '0 7 0 3 1 3
AUG 3 0

m1 3

IAY03
26

APR” 3 {00%

 

am cJCIHC/DaloDuo.p65-p.15

DEVELOPMENT OF LANDSCAPE-SCALE MODELS To DESCRIBE HABITAT
POTENTIAL OF WHITE-TAILED DEER (ODOCOILEUS VIRGINIANUS) IN
MICHIGAN.

By

Alexandra B. Felix

A THESIS
Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Fisheries and Wildlife

2003

ABSTRACT
DEVELOPMENT OF LANDSCAPE-SCALE MODELS TO DESCRIBE HABITAT
POTENTIAL OF WHITE-TAILED DEER (ODOCOILEUS VIRGINIANUS) IN
MICHIGAN
By

Alexandra B. Felix

The Michigan Department of Natural Resources has addressed a need to re-
evaluate Michigan’s deer population goals with a consideration of how the range of biotic
and abiotic factors within landscapes influence deer habitat selection and subsequently
affects populations. This study was initiated to contribute to the understanding of the
relationship between white-tailed deer populations and their habitat based on the
ecological potential of habitat types and how they can be managed in Michigan. Habitat
types were defined in 3 areas of Michigan using digital vegetation and soil databases.
empirical vegetation attribute data, and ecological classification systems. Compositional.
structural, and geological information were identified for each habitat type and stored in a
database. Three deer habitat requirements were identified from literature and a habitat
classification key was developed, which identifies habitat suitability values associated
with a range of habitat conditions within landscapes. Landscape-scale models were
constructed that quantify habitat suitability for each seral stage within habitat types. The
highest suitability that a habitat type could attain is indicative of habitat potential (i.e.. the
capability of an area being or becoming habitat based on biological and geological
characteristics). The spatial distribution of habitat potential can be projected in a GIS and
combined with deer demographic data to establish realistic deer management goals based

on the ecological potential of landscapes and population patterns.

ACKNOWLEDGEMENTS

I would like to thank the Michigan Department of Natural Resources, Wildlife
Division for providing funding for this project. I would also like to thank my committee
members Dr. Rique Campa, Dr. Scott Winterstein, and Dr. Kelly Millenbah (Fisheries
and Wildlife), Dr. Pete Murphy (Plant Biology), and Dr. Bill Moritz (Michigan
Department of Natural Resources, Wildlife Division) for advice, support and
encouragement throughout the long duration of this project. Thanks Rique, my major
advisor, and Kelly, my other advisor, for providing me with the opportunity to take this
project and run with it. I’ve learned a lot from working on this project and I’ve been
fortunate to have had tremendous opportunities to present this research at national and
international conferences. Thanks, also, Rique, for giving me advice and opportunities.
aside from presenting my research, that have helped shaped me professionally.

Thanks to Sarah Panken and Kristy Wallmo for “Eco-Deer support”! Thanks to
Sarah Panken (again), Mark Garner, and Brandi Hughey for working with me on
numerous “satellite projects” that stemmed from our research and was able to be
presented at professional conferences. Thanks to Michelle Gerwitz and Sara Kolesar for
assisting me with grunt work. Thanks to fellow grad students for listening to all my
grudges. Thanks to my family (including the in-laws) for all their encouragement and
support despite not understanding exactly what it is that I do and why I am doing it.
Thanks, especially, to my sister, for her support in her own unique way. Finally, a big 12-
point thanks to my best pal and husband, Craig, for being my audience when I practiced

presentations, standing by me all the way. and always making me laugh.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vi
LIST OF FIGURES ............................................................................... viii
INTRODUCTION ................................................................................... l
OBJECTIVES ........................................................................................ 6
STUDY AREA DESCRIPTION .................................................................. 7
METHODS
Obtaining necessary data ................................................................. .l 1
Land Cover ........................................................................ 11
Soils ................................................................................. 13
Land Type Associations ......................................................... 13
Ecoregions ......................................................................... 14
Building a habitat type coverage and quantifying deer habitat potential ........... l4
Habitat types ....................................................................... 14
Delineation of habitat type boundaries ......................................... 16
Identifying successional pathways of habitat types .................................... 17
Validation of habitat types ................................................................ 18
Process for quantifying habitat potential ................................................ 20
Habitat requirement: fall and winter food ..................................... 21
Browse—-All study areas ................................................ 23
Mast—southern Lower Peninsula and northeastern Lower
Peninsula ......................................................... 25
Mast—Upper Peninsula ................................................. 29
Site quality ................................................................ 33
Fall and winter food potential .......................................... 36
Habitat requirement: thermal cover ............................................. 38
Forest composition ...................................................... 38
Forest structure ........................................................... 40
Thermal cover potential ................................................ 43
Habitat requirement: spring and summer habitat ............................. 44
Spring and summer habitat potential .................................. 49
RESULTS
Southern Lower Peninsula ............................................................... 50
Successional pathways ........................................................... 50
Habitat suitability curves for each successional trajectory .................. 52
Habitat potential .................................................................. 52

(RESULTS Cont’d)

Northern Lower Peninsula ............................................................... 54
Successional pathways ........................................................... 54
Habitat suitability curves for each successional trajectory .................. 57
Habitat potential .................................................................. 58
Upper Peninsula ........................................................................... 62
Successional pathways ........................................................... 62
Habitat suitability curves for each successional trajectory .................. 64
Habitat potential .................................................................. 64
DISCUSSION
Consideration of agriculture in habitat potential models ............................ 68
Value of habitat typing ................................................................... 69
Applications of habitat potential models ............................................... 72
Understanding the relationship between habitat potential and deer
demographics ............................................................ 73
Understanding the relationship between habitat potential and deer
spatial structure .......................................................... 75
Recommendations for applying the habitat potential model to managing
wildlife populations and habitat ................................................ 76
Landscape-level .................................................................. 77
Ecosystem-level .................................................................. 78
Species-level ...................................................................... 81
Validation .................................................................................. 86
Habitat type themes .............................................................. 86
White-tailed deer habitat potential models .................................... 87
Future directions ........................................................................... 89
APPENDIX .......................................................................................... 91
LITERATURE CITED ........................................................................... 151

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

LIST OF TABLES

Climate in Michigan by study area (Sommers 1977). Ranges are the
minimum and maximum of the average values for each specified region.
SLP = southern Lower Peninsula, NLP = northeastern Lower Peninsula, UP
= Upper Peninsula .................................................................... 9

Guide to equations constructed quantify white-tailed deer habitat potential in
3 study areas of Michigan. SLP = southern Lower Peninsula, NLP =
northeastern Lower Peninsula, UP = Upper Peninsula ......................... 22

Possible successional pathways and proportion of the southern Lower
Peninsula study area within each successional trajectory (Traj.). Dominant
overstory vegetation is identified for early-, mid-, and late-successional
stages. A = aspen, AB = American elm, BA = black ash, Bas = basswood,
BC = black cherry. Bee = beech. Bir = white birch. CO = black oak, Cw =
cottonwood. For = forbs. G = grass species. Hck = hickory species, LBr =
lowland brush, RM = red maple. R0 = red oak, Sdg = seedges, Shr = shrub,
SM = sugar maple, SiM = silver maple, SWO = swamp white oak, WA =
white ash, Wil = willow species. WO = white oak. WP = white pine. Open
Water includes all lakes and rivers ................................................ 50

Possible successional pathways and proportion of the northeastern Lower

Peninsula study area within each successional trajectory (Traj.). Dominant
overstory vegetation is identified for early-. mid-, and late-successional
stages. A = aspen, AB = American elm. BA = black ash, Bas = basswood,
Bee = beech, BF = balsam fir. Bir = white birch, BP = balsam poplar, BS =
black spruce, C = cedar, G = grass species, H = hemlock, JP = jack pine,
LBr = lowland brush, 0 = oak species. RM = red maple. R0 = red oak, RP
= red pine, Shr = shrub, SM = sugar maple, T = tamarack WP = white
pine. Open Water includes all lakes and rivers ................................. 55

Possible successional pathways and proportion of the Upper Peninsula
study area within each successional trajectory (Traj.). Dominant

overstory vegetation is identified for early-, mid-, and late-successional
stages. A = aspen. AB = American elm, BA = black ash, Bas = basswood,
Bee = beech, BF = balsam fir, Bir = white birch, BP = balsam poplar, BS =
black spruce, C = cedar, G = grass species. H = hemlock, [W = ironwood, JP
= jack pine. LBr = lowland brush. RM = red maple, R0 = red oak, RP = red
pine, Shr = shrub, Spg = Sphagnum, SM = sugar maple, T = tamarack, WA
= white ash, WP = white pine. YB = yellow birch. Open Water includes all
lakes and rivers ...................................................................... 62

vi

Appendix 1. Scientific names of overstory vegetation that was mentioned in the text of
this thesis ............................................................................. 92

Appendix 2. List of available databases and descriptions investigated for the development
of habitat types ...................................................................... 93

Appendix 4a. Habitat classification key describing fall and winter food habitat
requirements (browse component), suitability indices (SI) on a scale of 0-1
(1 = high and 0 = poor), and sources ............................................ 96

Appendix 4b. Habitat classification key describing fall and winter food habitat
requirements (mast component). suitability indices (SI) on a scale of 0-1
(1 = high and 0 = poor), and sources ............................................ 97

Appendix 5a. Habitat classification key describing thermal cover habitat requirements
(forest composition component), suitability indices (SI) on a scale of 0-1 (1
= high and O = poor), and sources ................................................ 99

Appendix 5b. Habitat classification key describing thermal cover habitat requirements
(forest structure component), suitability indices (SI) on a scale of 0-1 (1
= high and O = poor), and sources ................................................ 99

Appendix 6. Habitat classification key describing spring and summer habitat

requirements. suitability indices (SI) on a scale of 0-1 (1 = high and O =
poor), and sources .................................................................. lOl

vii

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

LIST OF FIGURES

Location of three study areas in Michigan used for modeling white-tailed
deer habitat potential. SLP = southern Lower Peninsula, NLP = northeastern
Lower Peninsula, UP = Upper Peninsula ........................................... 8

Location of areas with the same successional pathway in southern Michigan
(Barry, Calhoun. and Eaton Counties). Vegetation codes for early, middle,
and late successional stages are as follows: A = aspen, AB = American elm,
BA = black ash, Bas = basswood. BC = black cherry, Bee = beech, Bir =
white birch, CO = black oak, Cw = cottonwood, For = forbs. G = grass
species. Hck = hickory species, LBr = lowland brush, RM = red maple, R0
= red oak, Sdg = seedges, Shr = shrub, SM = sugar maple, SiM = silver
maple, SWO = swamp white oak. WA = white ash, Wil = willow species.
WO = white oak, WP = white pine ............................................... 51

Distribution of white-tailed deer fall and winter food potential throughout
a 1,872 mi2 area in southern Lower Peninsula, Michigan (Barry, Calhoun.
and Eaton Counties). Fall and winter food potential categories are classified
based on natural breaks ............................................................ 53

Distribution of white-tailed deer spring and summer habitat potential
throughout a 1,872 mi2 area in southern Lower Peninsula. Michigan (Barry,
Calhoun. and Eaton Counties). Spring and summer habitat potential
categories are classified based on natural breaks ............................... 54

Location of areas with the same successional pathway in northeastern
Michigan (Presque Isle, Alpena, Montmorency, Alcona. and Oscoda
Counties). Vegetation codes for early. middle, and late successional stages
are as follows: A = aspen. AB = American elm, BA = black ash, Bas =
basswood, Bee = beech, BF = balsam fir, Bir = white birch, BP = balsam
poplar, BS = black spruce, C = cedar. G = grass species, H = hemlock, JP =
jack pine, LBr = lowland brush, 0 = oak species, RM = red maple, R0 =
red oak, RP = red pine, Shr = shrub, SM = sugar maple, T = tamarack WP =
white pine ............................................................................ 56

Distribution of white-tailed deer fall and winter food potential throughout a
3,132 mi2 area in northeastern Lower Peninsula. Michigan (Presque Isle.
Alpena, Montmorency, Alcona. and Oscoda Counties). F all and winter food
potential categories are classified based on natural breaks. .................. 59

Distribution of white-tailed deer thermal cover potential throughout a 3,132
mi2 area in northeastern Lower Peninsula. Michigan (Presque Isle, Alpena,
Montmorency, Alcona. and Oscoda Counties). Thermal cover potential
categories are classified based on natural breaks ................................ 60

viii

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Distribution of white-tailed deer spring and summer habitat potential
throughout a 3,132 mi2 area in northeastern Lower Peninsula, Michigan
(Presque Isle. Alpena, Montmorency, Alcona. and Oscoda Counties).
Spring and summer habitat potential categories are classified based on
equal intervals ........................................................................ 61

Location of areas with the same successional pathway in Upper Michigan
(Baraga, Marquette. Dickinson, and Iron Counties). Vegetation codes for
early, middle. and late successional stages are as follows: A = aspen, AB =
American elm. BA = black ash, Bas = basswood, Bee = beech. BF = balsam
fir, Bir = white birch, BP = balsam poplar, BS = black spruce, C = cedar. G
= grass species. H = hemlock. Iw = ironwood, JP = jack pine, LBr = lowland
brush. RM = red maple, R0 = red oak, RP = red pine, Shr = shrub, Spg =
sphagmun, SM = sugar maple, T = tamarack, WA = white ash, WP = white
pine, YB = yellow birch ............................................................ 63

Distribution of white-tailed deer fall and winter food potential throughout a
4,696 mi2 area in the Upper Peninsula. Michigan (Baraga, Dickinson, Iron,
and Marquette Counties). F all and winter food potential categories are
classified based on natural breaks ................................................ 65

Distribution of white-tailed deer thermal cover potential throughout a

4,696 mi2 area in the Upper Peninsula. Michigan (Baraga, Dickinson, Iron,
and Marquette Counties). Thermal cover potential categories are

classified based on natural breaks ................................................ 66

Distribution of white-tailed deer spring and summer habitat potential
throughout a 4,696 mi2 area in the Upper Peninsula, Michigan (Baraga,
Dickinson, Iron, and Marquette Counties. Spring and summer habitat
potential categories are classified based on natural breaks .................... 67

Location of areas in northeastern Lower Peninsula. Michigan that have the
potential to support old growth beech-sugar maple forests based on the
spatial output from the habitat type database for the northeastern Lower
Peninsula ............................................................................. 79

Hypothetical example of how managers might use habitat types and habitat
potential modeling to understand temporal changes in forest bird community
distribution. Output from habitat potential models developed for each bird
community could produce suitability curves that would indicate which seral
stages provide habitat for different forest bird communities. Viability of
each bird community decreases below the threshold habitat conditions. . ...82

ix

Appendix 3.

Appendix 7.

Appendix 8.

Appendix 9.

Conceptual diagram for the process of defining and delineating habitat
types. LTAx = land type association x. One LTA was selected at a time for
analysis. Six moisture classes were used: excessively well drained (EWD),
well drained. moderately well drained, somewhat poorly drained, poorly
drained. and very poorly drained (VPD). VEG = vegetation type. HT =
habitat type ........................................................................... 95

White-tailed deer habitat suitability curves for each successional trajectory
in southern Lower Michigan. The curves display habitat dynamics for fall
and winter food, and spring and summer habitat components based on
habitat potential models ........................................................... 102

White-tailed deer habitat suitability curves for each successional trajectory
in northeastern Lower Michigan. The curves display habitat dynamics for
fall and winter food. thermal cover. and spring and summer habitat

components based on habitat potential models ............................... 109 -

White-tailed deer habitat suitability curves for each successional trajectory
in the Upper Peninsula, Michigan. The curves display habitat dynamics for
fall and winter food, thermal cover. and Spring and summer habitat
components based on habitat potential models ............................... 134

INTRODUCTION

In Michigan, the white-tailed deer (Odocoileus virginianus) is an important
natural and economic resource. Consequently. state agencies allocate funds to provide
for research on white-tailed deer population demographic trends and habitat quality
which direct deer management activities. Deer management in Michigan essentially
began in 1895 with the establishment of an official hunting season, bag limit, and a
required hunting license. Wildlife managers realized the importance of regulating and
conserving deer numbers since unregulated commercial market hunting to meet the
human demand for venison in the late 18005 drastically decreased the size of Michigan‘s
deer herd. As a result of the early hunting regulations and an increase in early
successional vegetation subsequent to logging, deer numbers began to increase in the
early 19005. In the early 19005, the bag limit was reduced from 5 deer in 1900 to 3 in
1901, 2 in 1905 and 1 in 1915 (Langenau 1994). In 1921. the “buck law” was
established, which allowed only antlered bucks to be taken by hunters (Langenau 1994).

The Game Division, of the Department of Conservation in Michigan (which later
became the Department of Natural Resources) began using scientific information to form
much of the basis for deer management in the late 19205. More scientific research for
game management was conducted after the Pittman-Robertson Act was established in
193 7. 1n the 19305, evidence of winter starvation, overbrowsing, a low buck:doe ratio.
and an increase in the number of deer-vehicle accidents indicated that deer were abundant
in the state. Biologists suggested that the number of deer was not in balance with the

habitat and deer habitat should be managed for long-term sustainability.

In 1954, harvest of antlerless deer was permitted. Deer habitat quality, however.
was reduced due primarily to forest succession (i.e. loss of early successional stage
forests that provide favorable habitat). but also to heavy browsing by overabundant deer
between 1940-1960. Once again. the size of the herd decreased. In the 19705, managers
realized the importance of managing vegetation communities (i.e. habitat), rather than
just deer numbers in order to maintain sustainable deer populations. The Deer Range
Improvement Program (DRIP). which was established in 1971. allocated a portion of the
revenue from each deer license to be used for deer habitat improvement (Langenau
1994).

According to the Michigan Department of Natural Resources (MDNR), the
improved habitat, a series of mild winters, and artificial feeding of deer by the public
contributed to another increase in the herd size in the 19805. In the late 19805, MDNR
stated that their goal was to maintain an annual October 1 population of 1.3 million deer,
which was based on a 1971 population goal of 1 million deer in the spring herd. The
MDNR restated its commitment to that goal in 1997. Many wildlife managers and
stakeholder groups, however, have questioned the scientific basis of this population goal
(i.e., whether this goal is based on social and biological considerations). Subsequently.
there is a need to re-evaluate Michigan‘s deer population goals with consideration of
habitat suitability and the cultural carrying capacities across the state.

Traditional deer management in Michigan largely focused on regulating the
white-tailed deer population size through hunting (Langenau 1994). Based on large
fluctuations in population size, spatial variations in deer numbers, and skewed buckzdoe

ratios throughout the history of deer management in Michigan, this approach suggests

that it is difficult to obtain population management goals by primarily managing deer
numbers. Perhaps an understanding of how deer habitat changes spatially and temporally
would explain observed spatial and temporal differences in deer population dynamics
and, therefore, aid managers in deer management and planning. Unfortunately, however.
habitat information is generally not a major consideration in deer management because of
the lack of definitive spatial and temporal links between habitat supply and population
response. Michigan’s experience with deer management has provided new insights for
incorporating habitat management and planning into deer management goals.

To better incorporate habitat management into deer management goals, it is
critical to understand which environmental factors (e. g.. biotic and abiotic) influence the
ability of the habitat to support deer and how habitat influences population characteristics
and dynamics. For example, some of the most important underlying factors affecting size

and productivity of white-tailed deer populations are the structure and composition of the
vegetation to provide cover (Verme 1965. Mysterud and Ostbye 1999) and an adequate

and sustainable food source (Smith and Coggins 1984). the spatial arrangement of cover
types within the habitat as determined by soils and geological characteristics (Nixon et al.
1970), and winter severity (Verme 1968). The factors that influence deer population
dynamics may be difficult to analyze within forest stands since deer range widely (e. g..
333-9000+ ac in the northern Lower Peninsula. Garner 2001) to meet their habitat
requirements. Subsequently, an understanding of the relationship between deer and
habitat suitability at a landscape-scale is useful for state-wide deer management planning.
Temporal and spatial scales should be considered when analyzing habitat

suitability. It is difficult to understand the dynamic relationships between wildlife

populations and their habitats without understanding the underlying regulatory
mechanisms within landscapes and the processes by which habitat within landscapes
changes over time. Therefore, it may be useful to use habitat types to evaluate how
habitat suitability changes throughout succession. Habitat types are areas with the same
ecological characteristics that support the same successional trajectory (Daubenmire
1966). The climate, landforms, and soil characteristics such as nutrient content, moisture
regime. and texture influence differences in vegetation structure and composition and
successional patterns within different habitat types. By understanding how the structure
and composition of vegetation changes throughout succession, it is possible to quantify
the potential of the habitat to support deer throughout a successional trajectory. In
addition, identifying and mapping the spatial distribution of habitat types will enable
resource managers to establish long-term management objectives and specific
recommendations based on the ecological potential of landscapes.

The “holistic” approach to land management emerged in the late 1970-19805
when managers started to emphasize the complexity and dynamic nature of ecological
systems at several spatial and temporal scales (Grumbine 1994) instead of primarily
focusing on population density as a basis for management. Clearly, there is a need in
Michigan to take a more holistic or ecosystem-based approach to deer management.
Such an approach should consider the spatial and temporal relationships as well as the
interaction among biotic (e. g., vegetation structure and composition) and abiotic (e. g..
geology and climate) factors within a landscape to understand deer habitat selection

processes and provide a basis for incorporating habitat planning into deer management.

The fundamental purpose of this project was to develop landscape-scale models to
quantify white-tailed deer habitat potential for 3 multi-county areas in Michigan. Habitat
potential is, essentially, an assessment of the degree to which an area meets, or could
potentially meet. white-tailed deer life requisites based on the availability or potential
availability of ecological resources in specific areas of Michigan. The goal of this project
is to contribute to the understanding of the complex relationships between white-tailed
deer populations and their habitat based on the ecological potential of available habitat

types and how they could be managed in Michigan.

Ix)

OBJECTIVES

Assess the spatial patterns of white-tailed deer habitat requirements for 3 multi-

county study areas in Michigan.

Develop spatially explicit landscape-scale models for 3 study areas in Michigan

that quantify white-tailed deer habitat potential.

Suggest an approach for quantifying the relationships between habitat potential

and deer population demographics and spatial structure.

Make recommendations for applying the models to managing wildlife populations

and habitat within a landscape.

STUDY AREA DESCRIPTION

Three study areas were selected in Michigan in coordination with the MDNR
based on the heterogeneity of habitat types and land-use patterns among the 3 areas. The
first site is a 1,872 mi2 area in the southern Lower Peninsula (SLP) consisting of Barry,
Calhoun, and Eaton counties (Figure 1). The second site is a 3,132 mi2 area in the
northeast Lower Peninsula (N LP) consisting of five counties: (Alcona. Alpena,
Montmorency, Oscoda. and Presque Isle). The third Site is a 4.696 mi2 area located in
western Upper Peninsula (UP) and includes Baraga, Dickinson, Iron, and Marquette
counties. Land-use, climate. and land cover differ among the 3 study areas in Michigan.
These area differences are important because they may potentially affect the harshness of
the environment and consequently, the biology and distribution of white-tailed deer
within the areas. Furthermore, deer may require different structural characteristics of
vegetation within each area. For example. thermal cover may be more critical in the
Upper Peninsula during the cold winters than in southern Lower Michigan with mild
winters. In addition, food (i.e.. agricultural crops) may be more readily available in
southern Michigan than in the northern region.

Climate

The SLP site generally has less variable seasonal temperatures. less snowfall, and
a longer growing season than the NLP and UP sites (Table 1). Winters in the Upper
Peninsula are more severe than the winter in the Lower Peninsula with generally more
snowfall and colder temperatures for a longer period of time (Michigan Weather Service

1974, Albert et al. 1986).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 85 170 Miles
:—

Figure 1. Location of the three study areas in Michigan used for modeling white-tailed
deer habitat potential. SLP = southem Lower Peninsula; NLP = northeastern Lower
Peninsula; UP = Upper Peninsula.

Table 1. Climate in Michigan by study area (Sommers 1977). Ranges are the minimum
and maximum of the average values for each specified region. SLP = southern Lower
Peninsula, NLP = northeastern Lower Peninsula, UP = Upper Peninsula.

 

 

Average annual Average Growing Average annual No. days below
Region temp (°F) precip (in) Season (days) Snowfall (in) 0 °F
SLP 46-48 30.8-34.1 141-154 40-60 7-10
NLP 42-44 26.5-29.8 94-156 60-80 10-29
UP 42-44 30.2-30.9 96-1 13 100-180 27-34

 

Vegetation

Cover types in the SLP consist of oak-hickory (scientific names of overstory
vegetation appear in Appendix 1) and beech-maple dominated forests where forest cover
occurs. These are the same cover types that occurred historically, but currently,
agricultural lands also cover much of the area (Sommers 1977, Albert et al. 1986).
Aspen-birch, oak-hickory. and pine-dominated forests occur in the NLP where much of
the coniferous forests that occurred in pre-settlement periods were diminished from
logging (Albert et al. 1986). These pre-settlement forests included jack pine. red pine,
and white pine dominated forests interspersed with hemlock and aspen-birch stands
(Albert et al. 1986). The UP site includes largely maple-beech-hemlock and aspen-birch
dominated forests interspersed with spruce-fir dominated forests (Sommers 1977). Pre-
settlement vegetation included mesic mixed forests (sugar maple, yellow birch. hemlock)
and wet coniferous forests (balsam fir, spruce, tamarack. northern white cedar)

interspersed with dry coniferous forests of red and white pine (Albert et al. 1986).

Land-use and topography

Alfisols, specifically Hapludalfs, in conjunction with the flat or gently rolling
plains in the SLP support productive vegetation (Sommers 1977, Albert et al. 1986).
Field crops such as corn, soybeans. hay. and wheat are grown over 65% of the area
(United States Geological Survey 1999) in the SLP. Acidic sandy soils and spodosols
occur in the NLP (Albert et al. 1986). The NLP can be characterized by lake border
plains along Lake Huron and hilly lands in the inland areas (Sommers 1977). Primary
land-use patterns in the NLP include recreation. such as camping and hunting. due to the
large amount (i.e., approximately 35%) of state and federally owned land (MDNR-
Wildlife Bureau, MDNR-Land and Mineral Services Division. Resources Mapping and
Aerial Photography 2000). as well as privately owned hunt clubs. which exist in the area.
Some agriculture [cropland, (e. g., alfalfa and red kidney beans), and beef and dairy cattle]
occurs in the NLP as well. The UP site consists of hilly uplands (Sommers 1977).
Spodosols and bedrock rich in iron occur over much of the area (Albert et al. 1986).

Forest culture and recreation are primary land-uses that occur in this area.

10

METHODS

Empirical data used throughout this project were obtained by analyzing literature
describing deer habitat requirements. vegetation physiognomy, geological characteristics.
and successional patterns. There is an abundance of valuable ecological information
available in the literature, which is largely underutilized. The extracted information was
used for constructing deer habitat models, habitat type databases, and tables providing
information on the dynamic structure and composition of vegetation throughout

succession. Note: Some of the images in this thesis are presented in color.

Obtaining necessary data

I obtained and reviewed digital coverages (e. g., land cover. soils, landforms,.
ecoregions) from the MDNR and United States Geological Survey (U SGS) to assess their
utility for modeling white-tailed deer habitat potential (Appendix 2). To create different
spatial layers and assess deer habitat potential, I chose data that the MDNR either uses
currently or could easily obtain so that the process and models developed in this project
could be incorporated into the MDNR management planning and evaluation procedures.
Land Cover

Available land cover data differed for each of the 3 study areas. For the SLP, the
most current land cover data available were the 1978 Michigan Resource Information
System (MIRIS) data from the MDNR (MDNR 1999) and the 1992 USGS national land
cover data (U SGS 1999). The MIRIS data included vegetation classification to the

species level for some of the counties. Although the 1992 USGS data are more current.

11

they are more general than the 1978 MIRIS data (Appendix 2). For example, the most
specific category for vegetation classification according to the 1992 USGS data is forest
type such as “evergreen forest” or “deciduous forest”. This coverage was not useful for
habitat suitability analysis due to the generality of the data, but it was helpful, as a double
check on the distribution of certain forest types. For the NLP, available land cover data
included the 1978 MIRIS data. the 1992 USGS national land cover data, and a 1993 Gap
Analysis Program (GAP) Northern Lower Peninsula land cover dataset (MDNR-Wildlife
Bureau and MDNR-Land and Mineral Services Division, Resource Mapping and Aerial
Photography 2000). The 1993 data set was available only for the NLP and classified
vegetation by vegetation type (i.e., mixed lowland hardwood, northern hardwood, or
lowland broad-leaved deciduous shrub). For the UP, available land cover data included
the 1978 MIRIS data and the 1991 Landsat Thematic Mapper coverage (MacLean
Consultants, Ltd. 1995). The 1991 dataset was an inventory of white-tailed deer thermal
cover in the Upper Peninsula (MacLean Consultants, Ltd. 1995).

Presettlement vegetation helped identify potential vegetation. This coverage was
created using notes from the General Land Office Surveys, which describes vegetation
conditions between 1816 and 1856 prior to intensive logging and development (Comer et
al. 1995). Although there is probably bias associated with the presettlement vegetation
map since it was created based on subjective descriptions of vegetation from land
surveyors notes and does not include effects of disturbances and native people on
vegetation composition (Comer et al. 1995). it was a useful tool when combined with soil

and land type association layers to delineate ecological units.

12

Soils

Soils data were at the association level for all of the study areas (United States
Department of Agriculture (USDA) Natural Resources Conservation Service 1998). A
soil association is a natural grouping of related types of soils based on similarities in
climatic or physiographic factors and soil parent materials. Soil moisture regime, soil
type (i.e., sand, loam. clay). and potential dominant vegetation define the soil
associations. Soil characteristics delineated edges between stands with different
structural or compositional characteristics. Specific properties of series within soil
associations such as texture and moisture were obtained from the Iowa State Soils
Information Database (Soil Survey Division. Natural Resources Conservation Service.
USDA 1996).
Land Type Associations

Land type associations (LTAs) are broad regions delineated by large scale
geomorphic features defined by parent material and superficial topography (i.e., outwash
plains, moraines), soil characteristics. and natural overstory vegetation at a scale of
1:60,000 (Cleland et al. 1993). Different LTAs support vegetation types that differ,
either structurally or compositionally. Land type associations existed for the NLP
(DeLain et al. 1999) and the UP, but were not available for southern Michigan. For the
UP, the LTA boundaries we:e\usefiil for delineating ecological units, but the LTA
database did not indicate landfonn types. Possible landform types for LTA boundaries

were identified from the soils database (Soil Survey Division, Natural Resources

Conservation Service, USDA 1996).

13

Ecoregions

Ecoregions at the sub-subsection provided regional landscape ecosystem
classification based on differences in climate, bedrock geology, glacial landfonn, and
soils (Albert 1995). This coverage was available for all 3 study areas. From a deer’s
perspective, ecoregions are probably too broad (i.e., area of ecoregions within study area
ranges from 2896-80,066 ac) to distinguish between or select features from to meet their
life requisites, but the regions do represent areas with distinctive ecological conditions
that affect species composition and productivity. For example, ecoregions in the northem
part of the NLP distinguish areas that support mesic beech-sugar maple and wet
coniferous forests while ecoregions in the southern part of the NLP distinguish xeric jack
pine and oak-pine forests. The composition and arrangement of vegetation in the
northern part of the NLP would provide different deer life requisites than that of the
southern part of the NLP. The ecoregion coverage, therefore. was useful in landscape

classification and delineation of habitat types.

Building a habitat type coverage and quantifying deer habitat potential
Habitat types

A habitat type is a geographic area with the same ecological conditions that
supports the same successional trajectory (Daubenmire 1966). “Habitat typing” was used
to delineate areas with very different ecological conditions that may potentially provide
different habitat conditions for deer. Habitat types were used because they represent the
scale at which deer select annual habitat components. For instance, early successional

stages of northern hardwood habitat types may provide spring and summer food (e. g..

14

Kohn and Mooty 1971), whereas lowland coniferous types may provide thermal cover
requirements (e.g., Venue 1965). One habitat type does not contain all of the vegetation
characteristics that provide annual deer life requisites. Therefore. deer may have to shift
their seasonal home ranges to obtain their life requisites.

Each habitat type is characterized by a specific pattern of compositional change or
successional trajectory. Since communities exhibit measurable, directional change
during succession, it is possible to predict structural and compositional characteristics of
the successional stages within a habitat type and evaluate those characteristics in terms of
deer habitat suitability. Delineation of habitat types was useful to determine the
ecological potential of the landscape to support deer and identify habitat potential (i.e..
the ability of an area to provide habitat for deer based on biological and geological
characteristics). Habitat types were defined by identifying characteristic combinations of
climate, landfonn, soils, and vegetation (Kotar and Coffman 1984. Spies and Bames
1985, Host et al. 1987).

Habitat type themes for the NLP and UP sites were delineated in ArcView 3.2
Geographic Information Systems (GIS. Environmental Systems Research Institute,
Redlands, California) using several existing digital coverages obtained from MDNR:
ecoregions, LTAs, soil associations. and presettlement vegetation. Ecoregions at the sub-
subsection provided the basis for delineating and classifying habitat types since they
define climatic-physiographic boundaries that affect species composition and plant
productivity at a broad scale. At the next level. LTAs, defined primarily by landforms,
were used because overstory composition and successional trends vary among landforms

(Host et a1. 1987). Understanding variation in landforms is useful in understanding plant

15

species recruitment patterns, and direction of compositional change across a landscape
(Host et al. 1987). Soil associations were included in the next level since potential
successional patterns among habitat types vary based on differences in soil texture and
moisture regime (Rowe 1969). It was essential to include underlying geologic factors in
delineating habitat types because forest composition or physiognomy may be affected by
characteristics of deeper soil layers (Spies and Barnes 1985). Presettlement vegetation.
the final layer, described vegetation conditions prior to intensive logging, farming, and
industrial development, which has significantly altered vegetation composition by
favoring establishment and persistence of early successional species.
Delineation of habitat type boundaries

The boundary of a habitat type is. essentially, an intersection of ecoregions.
LTAs, soil associations, and presettlement vegetation. In the process of delineating
habitat types, the ecoregion sub-subsections were initially identified (Appendix 3).
Within each ecoregion, l LTA was selected at a time for analysis. Areas within LTAs
were distinguished based on 6 soil moisture classifications ranging from excessively well
drained to very poorly drained. Within each moisture classification, soil type was
determined (e. g., sand. loam, clay). F inally, potential vegetation was determined from
the presettlement coverage (e.g., northern hardwoods. jack pine). Areas with the same
ecological attributes (e. g., ecoregion and geological characteristics) could potentially
support different vegetation types within a habitat type. If lefi undisturbed. the
vegetation types would succeed to the climax vegetation. Hence, areas with the same
geological characteristics, vegetation types, and climax vegetation define the same habitat

type. In essence, delineation of the habitat types is hierarchical with ecoregions at the

16

coarsest scale and habitat types at the finest scale. Information from each of the 4 layers

used to create the habitat type theme was stored in a database.

Identifying successional pathways of habitat types

Using information from the literature (e. g., Livingston 1905, Harper 1918,
Westveld 1933, Nichols 1935, Ku 1954, Ching 1954, Bryant 1963, Coffman et al. 1980.
Eyre 1980, Kotar and Burger 2000. Burger and Kotar 1999). I identified possible
successional paths for each habitat type based on soil (i.e., moisture and texture) and
landform characteristics. Two or more habitat types may support similar overstory
successional trajectories even if they have slightly different geological characteristics.
They remain as separate habitat types, however, because although the geological
conditions do not differ enough to support different overstory vegetation, the structure of
the overstory may differ between the 2 habitat types. In addition, floristic compositional
differences within the habitat type may be evident in the understory vegetation rather
than the overstory (Barbour et al. 1999). For example, in the UP site, 2 possible habitat
types could support aspen and sugar maple in the early-successional stages; sugar maple.
red maple, yellow birch, and ironwood in the mid-successional stages; and sugar maple
and hemlock in the climax state. The vegetation in both habitat types grows on well to
moderately well-drained soils. but in one habitat type. the soil is more loamy, and in the
other habitat type. the soil is more sandy. The more loamy habitat type supports climax
understory vegetation dominated by spinulose shield fern (Dryopteris austriaca), hairy
solomon’s seal (Polygonatum biflorum). wild lily of the valley (Maianthemum

canadense). lady fem (A thrium filix-femina), and downy violet (Viola sproria). The

17

more sandy habitat type, however, lacks the nutrient-demanding species (e. g., lady fern.

downy violet) that appear on the loamy soils (Coffrnan et al. 1980).

Validation of habitat types

Since the habitat type themes were based on several different coverages, the
boundaries needed to be validated using data that were not incorporated in the process of
delineating the habitat type boundaries. The habitat type theme for the NLP was overlaid
onto a 1993 Landsat TM image (MDNR-Wildlife Bureau and MDNR-Land and Mineral
Services Division, Resource Mapping and Aerial Photography 2000) of land cover, the
habitat type theme for the UP was overlaid onto a 1991 Landsat TM image (MacLean
Consultants, LTD. 1995), and the SLP habitat types were overlaid onto a 1992 satellite
image (U SGS 1999) to determine if the habitat type coverages agreed with actual
vegetation. The datasets selected were used for validation because they are the most
current digital coverages that displays vegetation types and 2 of the coverages are fairly
accurate based on ground truthing and the use of aerial photos: (80.20% for the 1993
Landsat image, MDNR-Wildlife Bureau and MDNR-Land and Mineral Services
Division, Resource Mapping and Aerial Photography 2000; 80-100% for the 1991
Landsat image, MacLean Consultants, Ltd. 1995). The accuracy assessment for the 1991
thematic mapper image (USGS 1999), however, was not complete. Habitat type
boundaries followed the boundaries of current vegetation types closely. The edges of
most vegetation types in the Landsat images were very distinctive. and it was relatively

easy to determine if there were discrepancies.

18

The following procedure was used to validate the NLP habitat types using
the1993 Landsat TM landcover data. The NLP habitat type theme was initially overlaid
onto the landcover data. I selected each habitat type individually, viewed the attributes of
that particular habitat type (i.e., specifically, climax vegetation, soil type and moisture
regime), and determined the vegetation composition according to the 1993 dataset that
occurs within the boundaries of the habitat types. I then compared the 1993 vegetation
composition with the habitat type and determine if the composition could be one
successional stage within the habitat type. For example, the attributes of the habitat type
being evaluated may have consisted of northern hardwoods (e. g.. beech-maple forest),
growing on moderately well drained sandy loam. If the 1993 dataset indicated that aspen
occurred within the boundaries of the habitat type, I would have concluded that the
habitat type was valid. If. however, the 1993 dataset indicated that cedar occurred within
the boundaries of the habitat type, I would have concluded that there was an error
somewhere in the delineation process. This error most likely had occurred in the
presettlement vegetation since it was based on notes from the General Land Office
Survey (Comer et al. 1995). the soil associations. or the 1993 dataset. Seventeen
questionable areas were identified in the NLP and then field checked. [Note: The UP
habitat types were validated with a similar procedure using the 1991 image. The SLP
habitat types were validated with a similar procedure using 1992 satellite image and the
literature (e.g., Bryant 1963, Dodge and Harman 1985, Dodge 1987). The discrepancies
in the UP and SLP were not field checked. but were addressed using literature of

vegetation descriptions of the area]

19

Process for quantifying habitat potential

The literature suggests that white-tailed deer habitat consists of 3 general
requirements: fall and winter food and thermal cover (i.e.. winter habitat), and spring and
summer habitat. Fall and winter foods consist of highly preferred hard mast such as
acorns and beechnuts, and such fruits as black cherries and hawthoms (Rogers et al.
1981). Deer shift their diets from grasses, leaves, and forbs to browse with the first frosts
in autumn (Blouch 1984) and may supplement their diets with agriculture crops when
available (Gladfelter 1984). The quality of fall and winter food affects winter survival as
well as the attainment of fertility in young deer (Morton and Cheatum 1946) and
productivity of older deer (Morton and Cheatum 1946. Cheatum and Severinghaus 1950).

In Michigan where deep snow restricts movements and cold temperatures and
wind stress their energy reserves, deer tend to concentrate in heavy coniferous cover (i.e.,
thermal cover), which blocks snow and provides a more stable microclimate (Ozoga and
Gysel 1972). In areas with severe winters. thermal cover is essential for deer to maintain
energy reserves throughout the winter.

Early successional stages of upland hardwood and upland mixed
hardwood/conifer sites provide spring and summer habitat because they support growth
of herbaceous vegetation such as grasses, sedges, forbs. and ferns (Rogers et al. 1981),
which are important spring and summer food items. Nutritious and easily digestible
spring and summer diets are important for reproduction, body growth of fawns. and antler
development (V erme and Ullrey 1984). Also, Mautz (1978) indicated that high quality

summer deer foods are critical for accumulating a fat layer as their winter energy reserve.

20

To quantify habitat potential within each habitat type, it was necessary to identify
habitat features that describe each life requisite. I constructed a habitat classification key
for each of the 3 requirements that is, essentially, a summary of the information that was
important for development of the habitat potential models. The habitat classification key
includes: 1) variables describing vegetation structurean‘dcompositionvthflat areimportant
for deer habitat, 2) suitability values (e. g., 0 to 1; l = high suitability) for each variable
based on the preferred structure/composition of vegetation to provide deer habitat
requirements. and 3) literature sources describing vegetation structure or white-tailed
deer habitat preferences within the Great Lakes region (Appendices 4-6).

Each study area differs in geological characteristics and vegetation composition.
Also, geological and ecological information that was available in one study area may not
have been available in other study areas. Therefore, it was necessary to construct
different models to quantify habitat potential for different study areas (Table 2).

Habitat requirement: fall and winter food

Information from the habitat type database (e. g., soil characteristics) and
quantitative vegetation structural (e.g., diameter at breast height (dbh), basal area) and
compositional data were included in the fall and winter food potential model. Fall and
winter food potential is a function of woody browse (Equation 1), mast production
(Equations Za or 2b). and site quality (Equations 33 or 3b). Equation 4 was used to

quantify the potential of a habitat type to provide fall and winter food for deer.

21

 

 

 

 

a. $er 5:53 BE .. case is owes: .. secs 885% 5O n ammm as d<
me xowE 3:36 8?. ... xooE c.5835 80:: a. x09: action-=8 328 u mum. AD A5 542
m
D I. m 53
e. 2%”... . Kw“... .. gamma. 3%.“... .N .. eager: so .5 .32
mm Axe—OE mflooam meccotcoov ... 68?: 3b .85.: H 5?: 5:62:28 3.20; 6 A5 #3 Z
3 53 base as .. SEE? case + cats 3.5.3: n BE 33 3m
cm 52; gas av. .. 3585 can; + 353 3.323 .. m: u 3;: as .5 .52
2 £53 eamsé .. cows gee: u 5:56 2a 32 .5 5.5
2 Q2553 scene“: .. case 2228; .. ones gee: n gas 25 ea .32
3:8 owfim .83 :83 ES: H: :82:
am 8 3:8 owfimdosm xmo ... x52 H: xmov mo Begum H 58?: as: 38 A5
mm 3:8 use + :53 Ba nae .. 33 :me .. as? as: u 63 as: as .32 .5m
2 N \ :53 3:32;: 832$ + 33 base 3328. u 53 832m E 3....
Owen— :otazam— 52 3.3..

 

d—zmficom 3&3 u as .2355; .526.— Eoumaozto: n A32 .5355; .525..— Eofizg n
Aim dawEOmZ mo map; 33.0. m E 3322. :3an Loop 823-823 @3536 8 xzom .3 38.538 2.23:3 3 02:0 .N 2an

22

Browse—all study areas
Browse is defined as living twigs S 0.25 inches in diameter and < 7 feet from the
ground (Ryel 1953, Mackey 1996). Suitability of woody browse is a function of the

average of browse quality and browse availability (Equation 1).

(1) Browse index = [(browse quality index) + (browse availability index)] / 2

Where:

Browse quality index: a 0-1 ranking of potential of browse species based on
palatability and nutritional characteristics.

Browse availability index: a 0-1 ranking of the relative availability of browse

within a stand.

Browse qualfl—Preferred browse species generally are more nutritious and
palatable to deer than non-preferred species (Davenport 1941, Ullrey et al. 1964, Ullrey
et al. 1967, Ullrey et al. 1968). Therefore, browse quality indices are 0 to 1 rankings of
potential browse species based on nutritional characteristics (e. g., high protein, low fiber
levels) and palatability. Most preferred browse species include basswood, cedar,
hemlock, red maple, and sugar maple (Davenport 1941) and, therefore, received a browse
quality index of 1.0. Somewhat preferred browse species include black ash, birch, black
cherry, and American elm and received a browse quality index of 0.75. Less preferred
browse species include aspen. beech, oak, jack pine, white pine, and balsam poplar. A

browse quality index of 0.5 was assigned to those foods. “Last resort” browse species

were assigned a browse quality index of 0.25 and include beech, red pine, spruce,
tamarack, and balsam fir (Appendix 4a). The browse quality index for seral stages with
more than one potential browse species was calculated as the average of quality indices
for each dominant browse species present in the seral stage.

Browse availability—In addition to browse quality, the quantity of browse that is
available to deer also influences fall and winter food suitability. Browse availability is
influenced by the age of a stand. For instance, a regenerating aspen stand (0-10 years)
provides an abundance of browse [i.e., many aspen stems provide twigs that are within a
deer’s reach (<6 fi) (Graham et al. 1963)] and received a browse availability index of 1.0.
An old grth (100+ years) maple stand, however, does not provide abundant browse
and received a browse availability index of 0.0.

One of 5 indices assigned to a potential browse species indicated relative browse
availability (Appendix 43). An availability index of 1.0 = high availability; 0.75 =
moderately high availability; 0.5 = medium availability;.0.25 = hardly available; 0 = not
available to deer based on normal browsing patterns (i.e., trace amounts of browse may
be available). These indices were assigned based on the assumption that browse
production is, in part, a function of stand age (Ryel 1953). Deciduous species were
assumed to have high browse availability in age class 1-10 years. High browse
availability of aspen continues during the 10-30 year-old age class. Browse availability
tends to decrease as deciduous trees age because although edible material is produced, it
is largely above the reach of deer (Graham et al. 1963). Therefore, older stands have
lower availability indices. Coniferous species were assumed to have high browse

availability in the age class 20—50 years (Ryel 1953, Verme 1965). The amount of

browse on cedar trees, for example, increases up to 3 inches DBH and then decreases due
to self-pruning of lower branches (Lake States Forest Experimental Station 1940). Data
describing browse production for coniferous species other than cedar (e. g., balsam fir,
spruce, pine) were not found and, therefore, were assumed to follow the same relative
browse production pattern as cedar.

Several potential browse species could be available in one seral stage. For
example, in a 10-30 year old red oak stand (availability index = 0.5), 1-10 year old red
maple (availability index = 1) could be growing in the understory. The browse
availability index for the stand is the highest availability index of each potential browse
species. In the above case, for instance, the browse availability index of the oak-maple
stand would be 1.0.

Mast—southern Lower Peninsula and northeastem Lower Peninsula
In the SLP and NLP, mast is a function of the type of potential mast species, and

basal area and DBH of the dominant mast producing species Oiquation 2a).
(2a) Mast index = [(spp. index) * (DBH index) * (basal area index)] + mast bonus

Where:

Spp. index: a 0-1 ranking of the quality of hard mast producing species.

DBH index: a 0-1 ranking that describes the ideal DBH (in) of mast species for
optimal mast production.

Basal area index: a 0-1 ranking that describes the ideal basal area (ftz/ac) structure

of mast species for optimal mast production.

25

Mast bonus: addition of 0.05 if mast producing species other than oak or beech

are present.

Species indexand mast bonus—Mast is defined as the fruit of forest trees and is
categorized into hard mast (e. g., nuts such as beechnuts and acorns) and soft mast (e. g.,
berries such as black cherries). Mast is essential for providing fall and winter food for
deer (Duvendeck 1964). In Lower Michigan, oaks are the main mast producing species
(Armbruster et al. 1987, as cited in Bender and Haufler 1987). Acorns typically make up
the bulk of all mast consumed by deer (Harlow et al. 1975). Therefore, sites dominated
by oak are assumed to have higher fall and winter food potential than sites dominated by
other mast species (e. g., beech) (Bender and Haufler 1987). Stands dominated by oak
received a mast species index of 1.0 whereas stands dominated by beech received a mast
species index of 0.5. Other mast species received a mast species index of 0.0 (Appendix
4b).

Although stands containing sofi mast species have a mast species index equal to
0.0, they still provide some fall and winter food value to deer. Sofi mast species such as
black cherry may provide a supplement to fall and winter diets and therefore, presence of
soft mast may increase the potential of an area to provide fall and winter food. Stands
that contain soft mast species have an added suitability bonus of 0.05.

Diameter at breast heigfi— Diameter at breast height of the dominant mast
producing species describes the ideal tree size for optimal mast production. Generally,
oaks (i.e., red oak, white oak, black oak) begin to bear fruit at age 25 (3-7 inches DBH),
but not abundantly until around age 50 (7-12 inches DBH) (Rogers 1990, Sander 1990a,

Sander 1990b). Duvendeck (1964) observed that mature oaks > 40 years and 2 12 inches

DBH have relatively high acorn yields. Since oaks have the potential for producing
acorns when they are 3-7 inches DBH, seral stages that contain oaks 3-7 inches DBH
have a DBH index of 0.5. Smaller oaks would not produce acorns and, therefore,
received a DBH index of 0.0. Seral stages that could contain oak with an average DBH 2
12 inches received a DBH index of 1.0 Appendix 4b).

Beech trees initially start to produce mast at age 40 (approximately 4 inches
DBH) but it is not until about 60 years (approximately 8 inches DBH) that they produce
large quantities (Tubbs and Houston 1990). Since beech trees have the potential for
producing mast when they are about 4 inches DBH, seral stages that contain beech 4
inches DBH received an index of 0.5. Anything smaller would not produce nuts and,
therefore, received a DBH index of 0.0. Seral stages that could contain beech with an
average DBH 2 12 inches DBH received a DBH index of 1.0 (Appendix 4b).

Basal area—The final variable for quantifying mast production is basal area
(fiZ/ac) of the dominant hard mast producing species, which describes the ideal structure
of a stand for optimal mast production. Data on basal area of oak and beech for different
size or age classes were limited. Subsequently, the relationship between size/age, and
basal area for habitat types containing oak and beech are assumed based on literature
describing other aspects of oak or beech stand structure (e. g., size and stem density). The
only basal area information for oak in mixed stands was from Host et a1. (1987), who
measured a 42-52 fiz/ac basal area for a 71-73 year-old oak in a mixed oak ecosystem in
northwest Lower Peninsula. In the SLP, oak may occur in “pure” stands [e.g., associates
in the stand comprise no more than 20% of the stocking (Eyre 1980)]. Basal area of oak

averaging 3-6 in DBH in pure stands can reach 40-70 fiz/ac and stands averaging 11-20 in

27

DBH can reach a basal area > 90 fiz/ac (Marty and Rudolph 1970). Basal area for
different age classes of beech was estimated from data describing the average number of
trees per acre in various size classes for beech in southern Michigan (Schneider 1966)
(Appendix 4b).

This model assumes that acorn production is maximized when the basal area of
oak exceeds 32.7 ftZ/ac (Armbruster et al. 1987, as cited in Bender and Haufler 1987) and
beechnut production is maximized when basal area of beech exceeds 17 fiz/ac. 1 could
not find data that describe the ideal basal area of beech for optimal mast production;
therefore, the following describes the process used to generate the ideal structure for
beechnut production: The average maximum size of beech trees in the Lake States is
approximately 20-25 inches DBH (200-250 years) (Tubbs and Houston 1990). Larger
diameter trees generally have larger crowns than smaller diameter trees. Crown size
seems to be an important factor in mast production in general, with larger crowns
producing more mast (Christisen and Korschgen 1955, Goodrum et al. 1971, Sander
1990). Therefore, it can be assumed that the largest size trees have the largest crowns,
and would produce the most mast if the entire canopy is healthy. That is. beech trees
between greater than 20 inches DBH would produce the largest mast crops. Schneider
(1966) estimated that atypical sugar maple-beech stand in southern Michigan supports
approximately 5 trees per acre that are > 20 inches in diameter. The basal area, then, for
a stand containing 5 beech trees per acre with an average DBH > 20 inches is 17 fiZ/ac.

An additional assumption regarding basal area is that a few large trees has the

same potential production as many smaller trees. To have a basal area equal to 17 fiZ/ac,

28

then, a stand where the DBH of beech averages 16.5 inches must have a stem density of
11 trees/ac. This structure is reasonable based on Schneider’s (1966) data.

If a seral stage contained mast producing trees in the 10-30 year age class (mixed
stands) or in the 1-10 year age class (pure stands), the stand was assumed to have a basal
area that had some suitability (0.5) for mast production. It was assumed that any seral
stage containing mast producing trees < 10 years-old does not have the basal area
structure necessary for adequate mast production and, therefore, received a basal area
index of 0.0. The basal area of oaks decreases in late successional stages in the NLP
because there, oaks occur in middle successional stages and as vegetation succeeds, oaks
are gradually replaced by climax vegetation such as red maple or white pine (Burger and
Kotar 1999). In the SLP, oaks may be dominant in the climax stage. Therefore, basal
area suitability remains optimal for mast production in later successional stages
(Appendix 4b).

Mast—Upper Peninsula

 

In the SLP and NLP, oak and beech are dominant components of several habitat
types and therefore structural data (e. g., dbh, basal area) could be associated with
different successional stages within certain habitat types. In the UP, however, oak is
rarely a dominant component of habitat types (Coffman et al. 1980). Subsequently, it
was difficult to assign structural data for oak in different habitat types to quantify oak
mast potential. A model alternative to the one for the NLP was developed for the UP
site. In the UP mast model, mast is a function of potential mast species, habitat type, and

successional stage (Equation 2b).

29

(2b) Mast index = greater of (oak HT rank * oak succ.stage rank) or
(beech HT rank* beech succ. stage rank)

Where:

Oak HT rank: a 0-1 ranking of habitat types, according to the relative degree to
which a habitat type can support oak.

Oak succ. stage rank: a 0-1 ranking of successional stages according to the stage
at which oak would be most prevalent.

Beech HT rank: 3 0-1 ranking of habitat types, according to the relative degree to
which a habitat type can support beech.

Beech succ. stage rank: a 0-1 ranking of successional stages according to the stage

at which beech would be most prevalent.

Oak HT rankiand beech HT rank—The habitat type rank indicates in which
habitat type oak or beech are more likely to occur or represent a larger basal area than
oak or beech in other habitat types. This index ranks a habitat type according to the
relative degree to which it can provide oak or beech mast for deer.

To rank habitat types according to the relative degree to which they can support
oak, the location of oak stands in the study area was identified in ArcView fi'om a 1991
Landsat TM image of the Upper Peninsula (MacLean Consultants, LTD. 1995). This
information was converted into a shapefile, which was spatially joined with the habitat
type shapefile for the UP. Once spatially joined, oak stand density within each habitat
type was determined and distributed in 5 categories based on natural breaks in the data.

Habitat types in each category were assigned suitability values of 0.2, 0.4, 0.6, 0.8, and

30

1.0 fi'om lowest to highest category, respectively. Habitat types such as lowland sites
dominated by swamp conifers that would not typically support oak received a habitat
type rank index of 0.0. This approach assumed that the density of oak stands within
habitat types as determined from a spatial join of information from the Landsat TM
image (MacLean Consultants, LTD. 1995) and habitat type shapefiles is an accurate
indicator of the relative amount of oak mast that a habitat type can potentially provide.
The ranking of the habitat types based on oak stand density corresponded with
descriptions from The Society of American Foresters of oak composition within different
cover types (Eyre 1980).

Beech is an associate in several habitat types in the UP because many of the
habitat types are characterized by well to moderately well drained sandy to loamy soils,
which are conducive to supporting beech (Tubbs and Houston 1990). The 1991 Landsat
TM image did not classify individual beech stands, but rather included beech in the
“northern hardwood” classification (MacLean Consultants, Ltd. 1995). Habitat types
were ranked according to the relative degree to which they can provide beech by first
identifying the habitat types that would not typically support beech [e.g., excessively
drained sandy soils or poorly drained soils (Tubbs and Houston 1990)] and assigning
those habitat types a suitability index of 0.0. Habitat types characterized by drier soils
(e.g., well drained or excessively well drained sand) received a suitability index of 0.8
since beech is sensitive to decreased soil moisture (Tubbs and Houston 1990). Habitat
types characterized by well drained to moderately well drained sandy or loamy soils, the
optimal conditions for beech production (Tubbs and Houston 1990), received a suitability

index of 1.0. Because deer prefer acorns over beechnuts, and acorns make up the bulk of

31

all mast consumed by deer (Harlow et al. 1975), the suitability values assigned to habitat
types for ranking beech mast production were multiplied by 0.75. That is, the highest
ranking that a habitat type could obtain for beech mast production as it relates to deer fall
and winter food suitability is 0.75. This assumes that sites dominated by beech have
lower mast suitability for deer than sites dominated by oak (Bender and Haufler 1987).
Oakgand beech successional stage ranks—The successional stages within each
habitat type were ranked according to the stage at which oak or beech would be most
prevalent. Oak is generally not a climatic species and, therefore, is most prevalent in
middle successional stages (i.e., 30-100 years) before it is out-competed by more shade
tolerant species (Sander 1990a). Middle successional stages, then, received a 1.0 for the
successional stage ranking for oak. Earlier and later stages decreased in suitability for
providing oak mast. Beech, on the other hand, is highly shade tolerant and is recognized
as a climax species (Tubbs and Houston 1990). Late successional stages (i.e., > 100
years) received a 1.0 for the successional stage ranking for beech because they most
likely provide more beech mast than earlier stages. Suitability decreases in earlier stages.
The mast index is the larger product of (the habitat type ranking for oak * the
successional stage ranking for oak) or (the habitat type ranking for beech * the
successional stage ranking for beech). This model assumes that relative mast quality and
production in a given area can be determined through assessment of geological conditions
favorable for supporting oak and beech, and successional stages at which oak and beech

are likely to be most prevalent.

32

Site Quality

Landform characteristics, soil chemical properties (e.g., fertility), and soil
physical properties (e.g., texture and moisture content) affect the growth and productivity
of forests (Merkel 1988). Therefore, an assessment of site quality based on geological
attributes is essential for evaluating the productive potential of forests (Merkel 1988).
Site quality is a function of soil fertility, soil moisture. and landform type in the NLP
(Equation 33) or soil fertility and soil moisture in the SLP and UP (Equation 3b).

Landform information was unavailable for the SLP and UP sites.

(3a) Site quality = [(fertility index) * (moisture index) * (landform index)]/3

(3b) Site quality = [(fertility index) * (moisture index)/2

Where:

Fertility index : a 0-1 ranking of soils based on nutrient holding potential.
Moisture index: a 0-1 ranking of moisture content according to vegetation growth
potential at a particular moisture content.

Landform index: a 0-1 ranking of landform type according to vegetation growth

potential on soils derived from a particular landform.

Soil fertility within habitat types ———lnitially, habitat types with the same overstory

successional trajectory were identified. If different soil types could support the same

successional trajectory, the habitat types were ranked based on soil fertility (For example,

33

beech-maple forests typically occur on loamy-sandy loam soils (Tubbs and Houston
1990). Soil fertility is naturally low in sandy soils whereas loamy and clay soils have a
higher fertility (Williams 1998). Habitat types with sand received a soil fertility index of
0.6, those with sandy loam received an index of 0.8 and those with loam or clay received
an index of 1.0. If, however, a trajectory occurred on only one soil type [e.g., jack pine
communities in the NLP characteristically grow on sand (Nutter 1973, Larsen 1982)],
then the fertility index for that soil type was 1.0 since there is no comparison between
other soil types.

Soil moisture content within habitat types—Since soil moisture content also
influences site productivity (Johnson et al. 1987), it is an important variable for
determining site quality. Moisture content was ranked according to the conditions in
which the habitat types typically occur. For example, cedar does not develop well on
extremely wet or extremely dry sites (Johnston 1990) but does best on sites with slow
drainage (Ku 1954). Therefore, poorly drained or somewhat poorly drained soils
received a higher soil moisture index (e. g., 1.0) for cedar dominated communities than
very poorly drained or well drained and drier sites. Black spruce, tamarack, and black
ash communities grow best on the wettest sites with little or no drainage (Ku 1954) and
received a moisture index of 1.0 on poorly drained or very poorly drained soils. Upland
hardwoods such as sugar maple, beech, or hemlock communities develop best on
moderately well drained or well drained, but not excessively well drained soils
(Livingston 1905) and, therefore, received a moisture index of 1.0 on well drained soils.
Excessively well drained sites received a moisture index of 1.0 for red pine and jack pine

dominated vegetation types (Livingston 1905). White pine, red maple, or oak

34

communities typically occur on well drained soils (Livingston 1905): thus well drained
soils for those types of communities received a moisture index of 1.0.

Landform—Type of landfonn (i.e., parent materials and topography) influences
soil development and plant species composition (Rowe 1969). Landform type, or
variations in parent material composition and topography, may exert a significant
influence on species establishment and growth (Host et al. 1987) and, therefore, is
included in assessing site quality for different habitat types. Twelve landforms as
indicated by the LTA coverage (DeLain et a1. 1999) occurred in the NLP site:
limestone/dolomite, broad moraine ridge, moraine ridge, broad flat outwash plain, narrow
outwash channel, ice contact ridge, flat moraine or till plain, flat lake plain, pitted
outwash plain, deltaic deposit, dune-swale complex, and drumlin field.

Habitat types with the same overstory successional trajectory were identified. If a
trajectory occurred on only one LTA, the landform index for the habitat types with that
specific trajectory was 1.0. If, however, different LTAs could support the same
successional trajectory, the LTAs were ranked according to the degree to which they
provide suitable vegetation growing conditions. For instance, cedar grows best on
limestone-derived soils (Johnston 1990). Therefore, limestone-dolomite LTAs received a
landform index of 1.0. Any other landform received a lower index (e.g., 0.8) because
cedar does not grow best on them. Moraines and till plains typically have moderate to
high fertility (Knapp 1993). Ice contact ridges have little to no nonpedogenic textural
bands, which are important to forest growth whereas outwash plains have no bands and
generally have low fertility (Host et al.1987). Johnson et al. (1987) observed that

productivity of northern hardwoods was lower when they grow on outwash plains than

35

other landform types. Based on the above information, for northern hardwood types,
moraine ridges and till plains were ranked higher (e. g., 1.0) than ice contact ridges (e. g.,
0.8), which were ranked higher than outwash plains (e. g.. 0.6). Flat lake plain rankings
depended on soil types since fertility of soils derived from lake plains depends largely on
soil type [e.g., sands derived from lake plains are largely infertile (e. g., landfonn index =
0.6), whereas loams have higher fertility (e. g., landform index = 0.8 for sandy loams, and
1.0 for loams) (Knapp 1993)]. Not enough information was found to rank deltaic
deposits, dune-swale complexes, or drumlin fields. These LTAs were assumed to have
the same landform index as other LTAs that supported the same trajectory.
Fall and winter food potential

Fall and winter food potential in the NLP and UP is the weighted average of
browse and mast production indices modified by the site quality index (Equation 4a). In
the SLP, fall and winter food potential is the average of browse and mast production

indices modified by the site quality index (Equation 4b)..

(4a) F WFP = [[2 * (browse index) + (mast index)]/3] * site quality index

(4b) FWFP = [[(browse index) + (mast index)]/2| * site quality index

Where:

F WFP: an index of fall and winter food potential on a scale of 0-1; 1 = highest
potential.

Browse index: output from Equation 1.

Mast index: output from Equation 2a or 2b.

36

Site quality index: output from Equation 3a or 3b.

This model assumes that in the NLP and UP, browse is twice as important in
supplying fall and winter food as mast. Throughout the fall and winter in Michigan,
annual fluctuations in mast production and snow cover causes mast availability to be
unpredictable. Therefore, browse is the only food that is continuously available during
the fall and winter (Blouch 1984). Mast is important, however, because a prolonged diet
of only woody browse may cause starvation in deer (Mautz 1978). In the SLP, the
model assumes that browse is not twice as important as mast since mast is more available
in the SLP due to less snowfall than in the northern areas (Sommers 1977).

The model also assumes that better quality sites, as assessed by the site quality
index (Equation 3) support a higher yield and quality of plants (Keorper and Richardson
1980) and, therefore, the vegetation growing on the higher quality sites would provide
better habitat conditions. It is important to include site quality evaluation in a wildlife
habitat assessment because the quality of soils directly affects the quality and yield of
vegetation (i.e., habitat), which may affect presence, abundance, distribution, viability, or
physical condition of wildlife. Research (e. g., Severinghaus and Moen 1983, Ford et al.
1997) suggests that there is a positive relationship between the physical condition of deer
and habitat quality. Since soils directly affect vegetation quality and production (Denney
1944), and the physical condition of wildlife reflects variations in habitat quality (Ford et
al. 1997), it is reasonable to assume that soil fundamentally affects the physical condition
or distribution of wildlife (Albrecht 1944, Denney 1944, Crawford 1950). Studies to

support the argument that soil conditions affect the physical condition of wildlife found

37

that there was a positive relationship between soil fertility or quality and the physical
condition of rabbits and raccoons, (as measured by size) in Missouri (Albrecht 1944,
Denney 1944, Crawford 1950).
Habitat requirement: thermal cover

Information from the habitat type database and quantitative vegetation structural
and compositional data from the literature were also included in the thermal cover
potential model. Thermal cover potential is a function of forest composition (Equation
5), forest structure (Equation 6) and site quality (Equation 3). Equation 7 was used to
quantify the potential of a habitat type to provide thermal cover for deer in the NLP and
UP. Thermal cover was not considered an important deer requirement in the SLP based
on unpublished data by J. Pusateri (M.S. Graduate Research Assistant at Michigan State
Uiversity, Department of Fisheries and Wildlife) for southern Lower Michigan.

Forest Composition

Forest composition is a function of the type of forest (i.e., deciduous or

coniferous) and the type of dominant coniferous species within a seral stage (Equation 5).

(5) Forest composition index = (forest type index) * (coniferous species index)

Where:

Forest type index: a 0-1 ranking of the relative ability of certain types of forests
(e. g., coniferous. deciduous, mixed) to provide thermal cover for deer.

Coniferous species index: a 0-1 ranking of the relative ability of the

dominant coniferous tree species to provide thermal cover for deer.

38

Forest type—Verme (1965) suggested that the type of forest overstory
substantially affects the microclimate and the amount of snow that reaches the ground.
Swamp conifers provide much better protection from severe temperatures and snowfall
than deciduous stands. Stands comprised of a mixture of coniferous and deciduous types
also provide better protection from winter weather than purely deciduous stands (V erme
1965). Seral stages in which the forest type is coniferous received a forest type index of
1.0; mixed coniferous stands received a forest type index of 0.5, and deciduous stages
received a forest type index of 0.0 (Appendix 5).

Coniferous species—Northem white-cedar and hemlock are the optimal species
for thermal cover protection in Michigan due to their excellent ability to protect against
strong winds, block snow, and provide winter forage (V erme 1965, Euler and Thurston
1980, Blouch 1984, Doepker and Ozoga 1991). Therefore, seral stages dominated by
cedar and hemlock receive a coniferous species index of 1.0. Spruce- and fir-dominated
stages are less suitable as thermal cover and received a coniferous species index of 0.8.
Although spruce and fir have adequate snow and wind blocking characteristics, they do
not provide the excellent forage associated with cedar and hemlock (Davenport et al.
1953) and are, therefore, less suitable as wintering areas by deer. Pine stands are
sometimes utilized as thermal cover only if other preferred swamp conifers are absent
(Bender and Haufler 1987). Since the wind and snow interception abilities of pine are
not as suitable as those of cedar, hemlock, spruce, or fir types, stages dominated by pine
received a coniferous species index of 0.4 (Appendix 5).

A reduction function is used to evaluate the forest composition index. If a stand is

deciduous, it has no potential for providing thermal cover. The suitability of coniferous

39

and mixed coniferous/deciduous stands, however, depends on the type of dominant
coniferous species present. Coniferous or mixed stands have the highest potential for
providing thermal cover when cedar or hemlock dominates and the lowest potential when
pine dominates.
Forest Structure

Important structural attributes for determining thermal cover potential include
basal area, percent canopy cover, tree size (diameter), and the age structure of the forest
(e. g., even-aged or uneven-aged) (Equation 6). This model assumes that deer will
initially select stands based on the vertical structure of the vegetation and not the age
structure of the stand. Therefore, the suitability of the vegetation structure, as indicated
by the average of basal area, canopy cover, and tree size indices, is weighted twice as

important in determining thermal cover potential as age structure.

2 a: basal area + canopy + tree size + age structure

 

(6) Forest :ructure = index cover index index index
In ex 3
3
Where:

Basal area index: a 0-1 ranking that describes the ideal basal area (fig/ac) structure
for optimal thermal cover protection in lowland coniferous vegetation
types.

Canopy cover index: a 0-1 ranking that describes the ideal percent canopy cover
for optimal thermal cover protection.

Tree size index: a 0-1 ranking that describes the ideal average tree diameter size

(in) for optimal thermal cover protection.

40

Age structure index: a rating of 1 if the stand is even-aged and a rating of 0 if the

stand is uneven aged.

Basal area—Basal area is an important variable to quantify thermal cover
suitability because it measures the ability of a stand to block wind from the side and
indicates relative degree of canopy cover (Nelson 1951). Basal area categories for habitat
types that could potentially provide thermal cover (e. g., lowland swamp conifers) were
obtained from several sources describing the structure of cedar. [Notez Basal area data
were not available for lowland conifers other than cedar. Therefore, it is assumed that
basal area trends for other species will be similar to those of cedar.] Generally,
increasing age of cedar is correlated with increasing basal area (Gevorkiantz and Duerr
1939), but the maximum basal area that a typical cedar stand obtains is approximately
300 fiz/ac at maturity (100-200 years) (Johnston 1990).

This model assumes that wind blockage for white-tailed deer is maximized when
the basal area is 174-261 ftz/ac (Bender and Haufler 1987). The basal area of typical 50
year-old cedar stands (i.e., stands dominated by cedar but also contain associated species
such as balsam fir, spruce species, or lowland hardwoods) is approximately 200 ftz/ac
(Johnston 1990). It was assumed that cedar stands > 30 years would have the basal area
structure necessary for adequate wind blockage and, therefore, received a basal area
index of 1.0 (Appendix 5). Cedar in the 10-30 year age class may have a basal area
ranging from 30—100 ftZ/ac depending on site conditions (Johnston 1990). Cedar in the

10-30 year age class may provide some thermal protection although it would be less

41

suitable than older age classes. Therefore, cedar 10-30 years of age received a basal area
index of 0.5, and younger cedar stands received a 0.0.

Seral stages within certain habitat types may be dominated by species other than
cedar (e.g., spruce or hemlock), but still potentially provide thermal cover for deer. Since
other conifers such as spruce and hemlock commonly occur in the same seral stages as
cedar, it was assumed that the basal area of stands dominated by those species would still
fall within the basal area categories assigned for cedar.

Canopv cover—Information on the canopy cover structure for different age
classes of stands dominated by lowland tree species was limited. It was assumed that
lowland stands 1-50 years of age had 40-70% canopy cover and anything older had 70-
100% canopy cover. A tight canopy (> 75% cover) appreciably reduces snowfall and
wind within the stand (V erme 1965, Ozoga 1968). Therefore, seral stages with 70-100%
canopy cover received a canopy cover index of 1.0 and stages with 40-70% cover
received an index of 0.5 (Appendix 5).

Tree size—Highest quality deer wintering areas are generally dominated by
mature or older lowland swamp conifers (Verme 1965, Ozoga 1968, Doepker and Ozoga
1991). Larger trees (i.e., pole size or larger) generally provide a more stable
microclimate and better protection from wind and snowfall than smaller trees (i.e.,
saplings) (Verme 1965, Ozoga 1968, Ozoga and Gysel 1972). For example, a deeryard
that supported a large deer herd throughout the winter near Shingleton, upper Michigan
consisted mainly of cedar 6-12 inches in diameter (Verme and Johnston 1986). Cedar 4-
8 inches in diameter dominated a deeryard in northern Lower Michigan (Ozoga and

Gysel 1972).

42

The diameter of mature cedar (approximately 100 years old) can range from pole
size (3-7 inches) to sawtimber size (> 9 inches) depending on site quality (Johnston
1990). Therefore, seral stages that contain lowland swamp conifers that are pole size or
larger received a tree size index of 1.0. Saplings can still provide some thermal cover for
deer. Thus, sapling-size trees (0.5-3.0 inches dbh) received a tree size index of 0.5 and
anything smaller received a 0.0 (Appendix 5).

Age structure—Verme (1965) indicated that deer did not use uneven-aged stands
dominated by large cedar for thermal cover. In uneven-aged stands, much of the snowfall
is not intercepted and accumulated on the ground Werme 1965). In addition, the
understory is relatively drafiy and prone to wide temperature fluctuations, thus providing
little thermal protection for deer (Verme 1965).

Even-aged stands provide the best conditions for thermal protection (V errne
1965). Lowland coniferous stands are typically even-aged until about 150-200 years
when windfall of old trees creates an uneven structure. Therfore, seral stages > 150
years-old received an age structure index of 0.0 and stages < 150 years-old received an
index of 1.0 (Appendix 5).

Therm_al cover potentfi

Thermal cover potential is calculated by the following reduction function

(Equation 7):

(7) TCP = forest composition index * forest structure index * site quality index

Where:

TCP: an index of thermal cover potential on a scale of 0-1; 1 = highest potential.

43

Forest composition index: output from Equation 5.
Forest structure index: output from Equation 6.

Site quality index: output from Equation 3a or 3b.

This model assumes that forest composition, forest structure, and site quality are
equally important in determining the potential of an area to provide thermal cover. Deer
select stands for thermal cover based on the forest composition and forest structure
variables, which are unique to particular stands. Site quality, however, affects the size
and productivity of vegetation and is unique to individual habitat types. Recall that the
structure of stands varies widely. Diameter of mature cedar, for example. can range from
pole size to sawtimber size (Johnston 1990). Ryel (1953) observed that where site
quality differed, the ages of trees of the same diameter also differed. That is, for a given
diameter, the average age of trees from high quality sites (as assessed by ring evaluation
and soil conditions) was lower than the average age of trees from lower quality sites
(Ryel 1953). Therefore, the site quality index indicates that the structure (e. g., diameter,
basal area) of vegetation growing on higher quality habitat types would be on the upper
end of the range and, therefore, would provide better habitat conditions than vegetation
growing on lower quality habitat types.

Habitat requirement: spring and summer habitat

In the NLP and UP, spring and summer habitat potential, a critical requirement

for reproduction, growth of fawns, and antler development (V erme and Ullrey 1984), is

primarily a function of habitat type and successional stage (Equation 8).

44

(8) SSHP = (VT preference index) * (forage avail. index) * (site quality index)

Where:

SSHP: an index of spring and summer habitat potential on a scale of 0-1; 1:
highest potential.

VT preference index: a 0-1 ranking of the relative ability of vegetation types to
provide spring and summer habitat requirements (e.g., food, cover) based
on deer preference for specific vegetation types (Kohn and Mooty 1971,
McCaffery et al. 1974. Mackey 1996).

Forage availability index: a 0-1 ranking that reflects the intensity of summer deer
use based on the relative amounts of forage available in different

successional stages.

VT preference index—During the spring and summer, white-tailed deer appear to
utilize some habitat types more than others. The most important factor governing deer
use of habitat types appears to be availability of preferred forage species (Kohn and
Mooty 1971, McCaffery et al. 1974). Track counts and observations of deer locations in
Minnesota indicated a preference for upland deciduous and upland mixed ((i.e., deciduous
and coniferous) vegetation types over uplanduc’oniferous types, lowland types, and fields
or grassy open areas (Kohn and Mooty 1971). For example, on a tracking survey, Kohn
and Mooty (1971) observed that 70% of deer tracks occurred in upland deciduous and
mixed types, which comprised 52% of the total tracking route. Fifty-four percent of radio

collared deer were located in these types, which accounted for 53% of deer home ranges.

45

This suggests that deer use of upland deciduous and upland mixed types is greater than
availability, and therefore, deer are selecting for those types over other vegetation types
(Kohn and Mooty 1971). Kohn and Mooty (1971) observed that deer used the upland
deciduous and mixed types intensively for morning and evening foraging. Specifically,
deer utilized aspen for summer foraging (Kohn and Mooty 1971). McCaffery et al.
(1974), Stormer and Bauer (1980), and Mackey (1990) reported that white-tailed deer
also intensively used aspen for summer foraging in Wisconsin and Michigan. Since the
upland deciduous and mixed vegetation types provide spring and summer food for deer,
the habitat type preference index for these types is 1.0 (Appendix 6).

Deer use of upland coniferous vegetation types in Minnesota was not as extensive
as deer use of the upland deciduous and mixed types (Kohn and Mooty 1971). For
example, deer tracks were observed in 7.5% of upland coniferous vegetation types, which
accounted for 10% of the tracking route (Kohn and Mooty 1971). Upland coniferous
types comprised an average of 24% of deer home ranges. and radio-collared deer were
located in those areas approximately 24% of the time (Kohn and Mooty 1971). These
data suggest that since use of upland coniferous types is proportional to availability, deer
are not actively selecting for coniferous types. Although coniferous types do not provide
the excellent forage that upland deciduous and mixed types provide during the spring and
summer, Kohn and Mooty (1971) observed that upland coniferous types may be
important for bedding during midday. Therefore, upland coniferous vegetation types
received a 0.4 for the habitat type preference index.

Kohn and Mooty (1971) observed that deer tended to avoid lowland areas during

the spring and summer. Approximately 5% of tracks and 0.6% of deer locations occurred

46

in lowland coniferous, lowland deciduous, and lowland shrub vegetation types. which
comprised 19.1% or the track route and 2.9% of deer home ranges (Kohn and Mooty
1971). Mackey (1990) observed that deer in the Stonington Peninsula of Michigan’s
Upper Peninsula used lowland coniferous vegetation (e. g., cedar) more than expected in
the summer. Deer in other parts of the Upper Peninsula, however, did not use lowland
coniferous types. Since deer use lowland types for spring and summer habitat less than
other vegetation types, these types received a vegetation type preference index of 0.2.
Although lowland areas are generally not utilized extensively in the spring and summer,
the vegetation type preference index is not 0 because lowland areas and wetlands may be
important providers of aquatic emergent plants, which may be an important source of
nutrients for deer (Davenport 1941) (Appendix 6).

Successional stage index—Nutritious and palatable forage is critical in the spring
when deer are nutritionally stressed following the winter, and metabolic rates begin to
increase as deer prepare for fawning and antler grth (V errne 1969). Studies conducted
in Minnesota (Kohn and Mooty 1971), Wisconsin (McCaffery et a1. 1974), and Michigan
(Stormer and Bauer 1980) indicated that in the spring and summer, white-tailed deer tend
to utilize early successional stages (i.e.. 1-30 years) of upland habitat types, which
provide more spring and summer forage than mature (i.e., 30-100 years) or late
successional (i.e., > 100 years) forests.

In Minnesota, deer use of upland forested habitat types was inversely proportional
to the age of the stand since the availability and quality of spring and summer forage
tends to decrease as stands mature (Kohn and Mooty 1971, Stormer and Bauer 1980).

For example, leaves of aspen, and current year’s grth of plants (e. g., aspen, maple,

47

honeysuckle) < 3 m tall comprised approximately 70% of the total food consumed during
the summer in Minnesota (Kohn and Mooty 1971). The most intensive use of stands
within upland deciduous and upland mixed forest types occurred when the trees were <
30 feet tall and < 10 feet apart (Kohn and Mooty 1971), which is characteristic of early
successional stages. In Wisconsin, summer deer distribution among forest types reflected
the distribution of aspen (McCaffery et a1. 1974). Similarly, in Michigan, leaves of
regenerating aspen rated high in preference as summer deer food (Stormer and Bauer
1980)

Deer use of a particular habitat type largely depends on the successional stage of
the forest. For example, Kohn and Mooty (1971) observed that the most intensively used
areas within a deer’s home range were in early successional stages and were less
intensively used after the trees reached maturity. Studies have suggested that over time,
natural plant succession can result in deficiencies in the quantity and quality of spring and
summer deer food (Verme 1969). Since spring and summer deer food is more abundant
and more preferred in early successional vegetation types of upland deciduous and mixed
forests and spring and summer use declines in later successional forests, early
successional stages received a 1.0 for the forage availability index. Middle successional
stages of habitat types dominated by upland deciduous and upland mixed vegetation
received a 0.6 and late successional types received a 0.2 (Appendix 6).

Upland coniferous vegetation may be important as bedding areas and were most
intensively used in Minnesota when trees were 11-29 feet apart and slightly over 50 feet
tall (Kohn and Mooty 1971). Early successional stages of upland coniferous types may

also provide some food and therefore, the successional stage index is 1.0 for early stages

48

and middle stages of habitat types dominated by upland conifers throughout succession.
Suitability declines to 0.5 for mature stages of upland coniferous types (Appendix 6).
Lowland types are assumed to follow the same suitability pattern throughout succession
as upland deciduous and mixed types. Therefore, early successional stages received a
1.0, middle stages received a 0.6, and late stages received a 0.2.
Spring and summer habitat potential

This model assumes that the type of vegetation (e. g., upland deciduous, mixed
deciduous and coniferous, lowland coniferous), successional stage as an index to the
relative amount of forage available, and site quality are equally important in determining
the potential of an area to provide spring and summer habitat for deer. Deer select stands
based on the vegetation composition indicative of the habitat type, and also, the degree of
maturity of the vegetation. Site quality (Equation 3) affects the size and productivity of
vegetation (Merkel 1988) and is unique to different habitat types. It is assumed that
higher quality sites support a higher yield and nutritional quality of plants (Keorper and
Richardson 1980) and therefore, would provide better spring and summer habitat

conditions for deer.

49

RESULTS

Southern Lower Peninsula
Successional pathways

There were 995 polygons delineated in the SLP site with 6 possible successional

pathways (Table 3, Figure 2). The SLP site is divided into 2 broad forest types:

Table 3. Possible successional pathways and proportion of the southern Lower Peninsula
study area within each successional trajectory (Traj.). Dominant overstory vegetation is
identified for early-, mid-, and late-successional stages. A = aspen, AB = American elm,
BA = black ash, Bas = basswood, BC = black cherry, Bee = beech, Bir = white birch, CO
= black oak, Cw = cottonwood, For = forbs, G = grass species, Hck = hickory species,
LBr = lowland brush, RM = red maple, R0 = red oak, Sdg = seedges, Shr = shrub, SM =
sugar maple, SiM = silver maple, SWO = swamp white oak, WA = white ash, Wil =
willow species, WO = white oak, WP = white pine. Open Water includes all lakes and
rivers.

 

 

Traj. Prop. EARLY MID LATE
0 0.021 Open Water
1 0.399 A SM/Bee/Bas/RM/WA SM/Bee/Bas
2 0.028 A/RO/Hck/BC RM/SM/RO/WO/BC RM/Bas/RO/WA
3 0.398 G/Sdg/For G/Shr/BO/WO/BC WO/BO/RO/Hck/RM
4 0.028 G/Sdg/For G/Shr/BO/WO/RO/Hck WO/BO/RO/WP
5 0.109 LBr T/BA/Bir/RM/AE RM/AE/SWO
6 0.017 Wil/Cw SiM/AE/GA/Cw SM/AE/RM/Bas

 

sugar maple-beech in the northern part and oak-hickory in the southern part (Dodge
1987). Nearly 40% of the study area (north) consists of mesic northern hardwood-
dominated habitat types with aspen dominating early successional stages; sugar maple,
beech, basswood, red maple, and white ash dominating middle stages; and sugar maple,

beech and basswood dominating late successional stages (Table 3, Trajectory 1).

50

.003 SE3 n 95 £00 833 u 03 .86on 32:3 n :3 in.“ 833 H <3 £00 323 953m H 03m .039:

002mm u 25 .0308 Ema M 2m 02% u Em .8303 u mcm .030 02 H OM .0308 02 u SE 03.5 00232 n um: £20on 3.8on H
30: .momoomm 88m H O .300.“ n com 6003:0300 n 30 .030 x003 H 00 £85 833 u :m £0002 M 8m $.55 0.053 H on 000333
N mam Ema x85 u <m .EE 505:2. n m< .0003 u 4. 530:8 ma 0.8 momma 3006885 20— 00a .2028 580 00m 3000 nozfiowo>
$200300 scam Ea anon—no .Emv 53:22 80508 E 33880 $006383 08$. 05 ~33 mafia «0 005804 .N 2&5

   

IH
3:: E o o
amiém<§m Bozomsza 39:3 U o
03wm§§ 3302:5203 a: I m
9500:8023 fiEoéoEomEmB 8.385 U 4
3220525903 2:038:50 3%..qu I m

<3\om\am\2¢ 090350.395 82395. <l_ gm
aeoomim <sz§mpom§m
Sty—>834 <D_ o

HH<A an: >AM<H

WHO—

 

 

 

51

Approximately 40% of the study area (south) consists of dry central hardwood-dominated
habitat types with grasses, sedges, and forbes dominating early successional stages;
grasses, shrubs, black oak, white oak, and black cherry dominating middle stages; and
white oak, black oak, red oak, hickories, and red maple dominating late stages (Table 3,
Trajectory 3). Eleven percent of the study area is in lowland deciduous types (Table 3,
Trajectories 5-6). Other habitat types each comprised < 3% of the SLP site (Table 3).
Habitat suitability curves for each successional trajectory

Appendix 7 illustrates how fall and winter food, and spring and summer habitat
suitability for deer changes throughout succession in the SLP. Fall and winter food
suitability is generally highest during middle stages (ages 30-100) when the quality and
abundance of mast and browse is high. Late successional stages generally do not provide
high fall and winter food suitability for deer in the SLP. Fall and winter food suitability
increases, however, in late successional stages habitat types where beech is dominant in
climax stages (e. g., Appendix 7, Trajectory 1) and beech mast is available. Thermal
cover does not seem to be a primary winter habitat requirement for deer in the midwest
oak-hickory region, which includes southern Lower Michigan (Torgerson and Porath
1984). Spring and summer habitat suitability for deer in the SLP is highest in early
successional stages (< 30 years).
Habitat potential

The habitat potential index within a habitat type for each deer habitat requirement
was determined as the highest suitability for the requirement that a habitat type could

attain based on vegetation and geological characteristics. Distinct regional differences

52

in habitat potential are evident in the SLP study site (Figure 3, Figure 4 ). The distinct
differences mirror the distribution of the mesic northern hardwoods and central
hardwoods habitat types. The northern hardwoods do not provide as high quality fall and
winter food or spring and summer habitat as those habitat types dominated by grasses,

sedges, forbes in early stages and oak in later stages.

   

Fall and Winter
Food Potential

Water
E 0033
5;] 0.34-0.67

- 0.68-0.84
- 0.85-0.93

 

0 9 18 Miles
:—

Figure 3. Distribution of white-tailed deer fall and winter food potential throughout a
1,872 mi2 area in southern Lower Peninsula, Michigan (Barry, Calhoun, and Eaton
Counties). Fall and winter food potential categories are classified based on natural
breaks.

. 1 ....

  
 

Spring and Summer
Habitat Potential
Water
Q3 0020
- 0.21-0.81
- 0.82-0.90
- 091-10

0 9 18 Miles

Figure 4. Distribution of white-tailed deer spring and summer habitat potential
throughout a 1,872 mi area in southern Lower Peninsula, Michigan (Barry, Calhoun, and
Eaton Counties). Spring and summer habitat potential categories are classified based on
natural breaks.

Northeastern Lower Peninsula
Successional pathways
Six hundred and thirty-two polygons were delineated in the NLP study site with

24 possible successional pathways (Table 4, Figure 5).

Table 4. Possible successional pathways and proportion of the northeastern Lower
Peninsula study area within each successional trajectory (Traj.). Dominant overstory
vegetation is identified for early-, mid-, and late successional stages. A = aspen, AB =
American elm, BA = black ash, Bas = basswood, Bee = beech, BF = balsam fir, Bir =
white birch, BP = balsam poplar, BS = black spruce, C = cedar, G = grass species, H =
hemlock, JP = jack pine, LBr = lowland brush, O = oak species, RM = red maple, R0 =
red oak, RP = red pine, Shr = shrub, SM = sugar maple, T = tamarack WP = white pine.

Open Water includes all lakes and rivers.

 

 

Traj. Prop. EARLY MID LATE

0 0.017 Open Water

1 0.018 A A/RM/WP WP/RM/Bee/SM
2 0.266 A O/RM/WP WP/RM

3 0.009 A/BP BA/AE RM/SM/BA

4 0.001 A/BP/Bir A/BP/BA BA/C

5 0.037 A/Bir BF/WS/RM C/BF/RM

6 0.001 A/Bir BS/BF/C C/BS/BF

7 0.012 A/Bir RM/Bee/WA/Bas/WP SM/Bee/H

8 0.008 A/Bir RM/Bee/WP WP/RM/Bee/SM/WA
9 0.006 A/Bir SM/Bee/Bas/RM/WA SM/Bee

10 0.082 A/Bir SM/Bee/Bas/RM/WA SM/Bee/Bas

11 0.007 A/Bir SM/O/Bee SM/Bee/H

12 0.003 A/Bir WP/RM/Bee WP/RM/Bee/SM
13 0.016 A/Bir BS/BF/C C/BS/BF

14 0.013 A/JP JP/RP RO/RM/WP

15 0.011 A/O O/RM/WP/BC WP/RM/Bee/SM
16 0.018 LBr BS/T/BF/C/BA BS

17 0.016 JP/WP/RP/O JP/WP/RP/RM/BC WP/RM

18 0.001 LBr/T/BA/Bir T/BA/RM C/BS/BF

19 0.009 LBr BS/T/BF/C/WP BS/T

20 0.098 LBr/A/Bir/BP/BA C/BS/BF C/H

21 0.056 LBr/Bir/T/BA C/BS/BF C/BS/BF

22 0.025 LBr/Bir/T/BA T/BA/RM C/BS/BF

23 0.233 Shr/G JP/RP/WP/O WP/RP/O/JP
24 0.037 T/BA/Bir/RM T/BA/Bir/RM/C C/BS/BF

 

55

.050 0003 u 0? 0.0000008 n 0. .000000 0090 M 2m .0300 H 00w .050 000 n 0M .0000

000 u OM .2008 000 U SE .020000 000 n O .0020 000.32 u and .0000 000“ n 00. £02800 n I 0000000 00% n O .0000 u U 000.000
0.003 H mm 000000 0000—00 M 0m .0003 0:03 .1. 0m .00 0000000 H am 00000 n 00m 00030000 n 00m .000 00003 u <m .000 00000E<

u m< .00000 n < 030:8 00 000 00000 0000000000 000— 000 .2028 .3000 000 00000 0000uow0> 0000000 000000 000 .0000?
.0002000002 .0000? .000 0000005 00000002 00000000000 00 303500 000000000000 0800 05 £03 00000 00 0000004 .m 0000i

I“.

2000203' 0.0

 

        

 

832 a 00000 0203020?
0000003 00350.: 000' 8
00000 20203 20000000' 00
00000 00000 2000.0000' .0
:0 00000 2000002000 8
000 03000000 00' 2
00000 20203 00200000' M:
2003 092000.300 000030-00 :
00 20000000 .00

200803003 00.03205 02' 2
032000 .00.: 002' E
00000 00000 002' 0.
200003003 803003 002' 2
00.8030 80520 M02' :
0.008000 2520308030 002B 2
80:20 2302000000030 M02 0
02320303003 038002... :02 0
000030 032023000020 :02 0.
00000 00000 :02! 0
20000 200300 ..02' m
020 20002 00002- 4
«0000000 0220 002' m
2003 0.5205 <1 0

20003003 030202 20“

hO>—M—\0¥3
HH<1~ GE

 

>Hv—

56

Approximately 27% of the study area consists of an uplands dry-mesic hardwoods
habitat type with aspen dominating early successional stages; red oak, red maple, and
white pine dominating middle successional stages; and white pine and red maple
dominating late successional stages (Table 4, Trajectory 2). Approximately 23% of the
study area consists of excessively dry mixed hardwood-conifer habitat type with shrubs
and grass species dominating early stages; jack pine, red pine, white pine, and oak
species dominating middle stages; and white pine, red pine, oak species, and jack pine
dominating late successional stages (Table 4, Trajectory 23). Only 10% of the study area
is dominated by lowland brush, aspen, balsam poplar, and black ash in early stages;
cedar, spruce, and fir in the middle stages, and cedar with hemlock and other conifer
species as minor components in late successional stages (Table 4, Trajectory 20). Other
habitat types each comprise < 10% of the NLP study area (Table 4).

Habitat suitability curves for each successional trajectory

Appendix 8 illustrates how the suitability of vegetation to provide fall and winter
food, thermal cover, and spring and summer habitat for deer changes throughout
succession in the NLP. Fall and winter food suitability generally is highest during early
and middle successional stages (< 30-100 years) when browse and mast is most
abundant. Late successional stages generally do not provide high fall and winter food
suitability for deer. Habitat types in the NLP that have thermal cover potential (e. g.,
poorly drained sites that support lowland conifers at some time during succession)
generally provide best thermal protection for deer in the middle to late successional

stages (> 100 years). Suitability tends to decrease in late stages (> 150 years) when

57

lowland conifers become uneven-aged. Spring and summer habitat suitability for deer is
highest in early successional stages (< 30 years).
Habitat potential

Distinct regional differences in habitat potential are evident throughout the NLP
study area (Figures 6-8). The distinct differences tend to follow habitat types which
distinguish vegetation in the northern portion (i.e., predominantly poorly drained,
coniferous types), from vegetation in the southern portion (i.e., predominantly dry-mesic
deciduous or mixed coniferous-deciduous types). It is evident that, according to the
model, habitat types in the southern portion of the study area have higher fall and winter
food potential than those in the northern portion (Figure 6).

The reverse pattern is evident for the distribution of thermal cover (Figure 7).
High thermal cover potential mirrors the distribution of habitat types characterized by
poorly drained soils and dominated by conifers, which occur in the northern portion of
the study area. Habitat types dominated by hardwood forests, mixed coniferous-
deciduous forests, and upland coniferous sites in the southern portion of the study area
will never provide high thermal cover protection for deer.

Habitat types that support upland mixed coniferous-deciduous and upland
deciduous vegetation throughout succession largely occur in the northern portion of the

study area and have high potential for providing spring and summer habitat (Figure 8).

58

Fall and Winter
Food Potential

Lfir ] Water

3' 1 0052

5’" f3 0.53-0.60
i; .43 0.61-0.68
- 0.69-0.75
- 0.76-0.95

24 Miles

 

Figure 6. Distribution of white-tailed deer fall and winter food potential throughout a
3,132 mi2 area in northeastern Lower Peninsula, Michigan (Presque Isle, Alpena,
Montmorency, Alcona, and Oscoda Counties). Fall and winter food potential categories

are classified based on natural breaks.

Thermal Cover
Potential

[:] Water
0

0050

- 0.51-0.80
- 0.81-0.93
- 0.94-1.o

 

0 12 24 Miles
:—

Fi e 7. Distribution of white-tailed deer thermal cover potential throughout a 3,132
mi area in northeastern Lower Peninsula, Michigan (Presque Isle, Alpena,
Montmorency, Alcona, and Oscoda Counties). Thermal cover potential categories are

classified based on natural breaks.

60

Spring and Summer
Habitat Potential

E Water

[ " ' |o-o.2o
[”1021-040
E3717] 0.41-0.60
iii; 0.61-0.80
- 0.8l-l.0

24 Miles

 

Figure 8. Distribution of white-tailed deer spring and summer habitat potential
throughout a 3,132 mi2 area in northeastern Lower Peninsula, Michigan (Presque Isle,
Alpena, Montmorency, Alcona, and Oscoda Counties). Spring and summer habitat
potential categories are classified based on equal intervals.

6l

Upper Peninsula

Successional pathways

There were 1,848 polygons delineated in the UP study site with 16 possible

successional pathways (Table 5, Figure 9). Approximately 53% of the study area

Table 5. Possible successional pathways and proportion of the Upper Peninsula study
area within each successional trajectory (Traj.). Dominant overstory vegetation is
identified for early-, mid-, and late-successional stages. A = aspen, AB = American elm,
BA = black ash, Bas = basswood, Bee = beech, BF = balsam fir, Bir = white birch, BP =
balsam poplar, BS = black spruce, C = cedar, G = grass species, H = hemlock, [W =
ironwood, JP = jack pine, LBr = lowland brush, RM = red maple, R0 = red oak, RP = red
pine, Shr = shrub, Spg = sphagnum, SM = sugar maple, T = tamarack, WA = white ash,
WP = white pine, YB = yellow birch. Open Water includes all lakes and rivers.

 

 

Traj. Prop. EARLY MID LATE

O 0.024

1 0.001 A SM/RM/Iw/Bas/Y B SM/H

2 0.006 A/BP WA/BA/AE WA/RM/SM/BA
3 0.011 A/BP/Bir BF/WS/C/BS H/RM/SM
4 0.035 LBr/A/BP/Bir BF/WS/C/RM H/RM/SM
5 0.030 A/Bir SM/WP/RM/YB/RO/BF H/RM

6 0.283 A/Bir SM/YB/RM/Iw/BF/WS SM/H

7 0.022 A/Bir WP/RP/BF/WS RM/RO
8 0.058 A/SM SM/Bas/RM/Bee SM

9 0.016 A/SM SM/Bas/RM/YB/lw H/SM

10 0.222 A/SM SM/RM/YB/Iw SM/H

11 0.030 A/SM/WA SM/Bas/Iw/Bee/WA/AE SM

12 0.031 LBr LBr/Bir/RM/BP BS/C/H
13 0.094 LBr T/BA/Bir/RM C/BS/BF
14 0.066 LBr/Bir/BP C/BS/BF C/H

15 0.049 Shr/G JP JP/RP

16 0.023 Spg/LBr BS/T/BF/C BS/T

 

£83 326» n M? can 833 n A? 58 823 n <3 Jog—=8 n H .039: Ema M 2m .Escwanm n mam 53%

n Em .059 to“ H mm £8 “.2 n OM .292: on: M 2m £33 was—32 u and .33 xoam u A: .8958: M 3m £02an H 3 £20on
55% H O .53 n U 602% x83 H mm .333 83:5 H .5 £83 823 n :m ..E .5559 .1. mm .3089 n com .vooBmmmn n mam 53
x83 u <m .83 58582. n m< £33 n < 526:8 mm 0.8 momma 555883 82 EB .0328 .330 com mecca :oufiowo> .Amozazoo
. no: 98 £03305 632532 .5955 newanz 5an 5 5358 1335083 08$ 05 53> 935 we cos—804 d 2sz

I“
3:2 cm

   

 

xuoeuom vomoaxm I C

 

tmm uimEmm 522ml 2
.25 E 95' 2
:5 .5556 535:5 I z
.5553 5:535 54 I m.
:65 535554 .3 .2
2m m<23\uom\a=5\2m 2352' :
55 352225 52' o.
:52 35232225 52 a

5 8922395 52 a
9:5 maimamaa 52' N.
:25 min—$255525 52' o
2%: 555525335 52 m
2352 2.20535 5:525; v
292%: mEQmBEm 5252a n
5393223 522523 52' N
:25 mEQEgszm <fl _
5533 o

HH<A 9:2 >AM<H

 

MSU—

63

consists of mesic northern hardwood-dominated habitat types with aspen dominating
early successional stages; sugar maple, red maple, ironwood, and basswood dominating
middle stages (with various species representing minor components); and sugar maple
and hemlock dominating late successional stages (i.e., Table 5, Trajectories 6, 9, 10).
Approximately 16% of the study area is dominated by poorly drained cedar-spruce-fir
forests (Table 5, Trajectories l3, 14). Other habitat types each comprise < 6% of the UP
study site (Table 5).
Habitat suitability curves for each successional trajectory

Appendix 9 illustrates how fall and winter food, thermal cover, and spring and
summer habitat suitability for deer changes throughout succession in the UP. Fall and
winter suitability is generally highest during middle stages (Ages 50-100) when the
quality and abundance of browse and mast is high. Late successional stages generally do
not provide high fall and winter food suitability for deer in the UP. Habitat types in the
UP that have thermal cover potential generally provide the best thermal protection for
deer in the middle to late stages (> 100 years). Suitability tends to decrease in stages >
15 years when lowland coniferous stands become uneven-aged due to windfall or other
disturbances. Spring and summer habitat suitability for dee in the UP is highest in early
successional stages (< 30 years).
Habitat potential

No distinct regional differences in habitat potential are evident in the UP study
site (Figures 10-12 as were evident in the NLP (Figures 6-8). Habitat types in the UP are
more “patchy” or interspersed than in the NLP due to geological fragmentation caused by

glacial activity. For example, areas characterized by mesic loamy soils that support

64

northern hardwoods are frequently interspersed within areas characterized by shallow
underlying bedrock that cause high water tables and support swamp conifers. Generally,
fall and winter food potential is highest in Iron, southern Baraga, and central Marquette
counties in habitat types that provide conditions conducive for supporting oak and beech
(e.g., mesic northern hardwoods) (Figure 10). Drier, more sandy areas in Marquette and
poorly drained areas throughout the study area do not provide high fall and winter food

potential.

Fall and Winter
Food Potential
[fl Water
0-0.7I
[Wig 0.72-0.85

“0.86-0.92
-o.93-1.o

30 Miles

 

Figure 10. Distribution of white-tailed deer fall and winter food potential throughout a
4,696 mi2 area in the Upper Peninsula, Michigan (Baraga, Dickinson, Iron, and
Marquette Counties). Fall and winter food potential categories are classified based on
natural breaks.

65

Habitat types that provide high thermal cover potential are interspersed
throughout the study area and support lowland swamp conifers (Figure 11). Several
habitat types in the UP study site have high spring and summer habitat potential for deer
because they support aspen in early successional stages, which deer highly prefer during

the spring and summer (Kohn and Mooty 1971, Mackey 1996) (Figure 12).

Thennal Cover
Potential

-0-0~7|
-0.72-0.80 .1 "
-0.8l-l.0

 

Fi ure 11. Distribution of white-tailed deer thermal cover potential throughout a 4,696
mi area in the Upper Peninsula, Michigan (Baraga, Dickinson, Iron, and Marquette
Counties). Thermal cover potential categories are classified based on natural breaks.

66

Spring and Summer
Habitat Potential

:1 Water

L . .10
{#16040
-0.41—o.9o

-o.91-1.0

30 Miles

 

Figure 12. Distribution of white-tailed deer spring and summer habitat potential
throughout a 4,696 mi2 area in the Upper Peninsula, Michigan (Baraga, Dickinson, Iron,
and Marquette Counties. Spring and summer habitat potential categories are classified
based on natural breaks.

67

DISCUSSION

Consideration of agriculture in habitat potential models

In some areas of Michigan, white-tailed deer have evolved some dependence on
agriculture (Gladfelter 1984, Braun 1996). For example, in the Midwest agriculture
region, which includes southern Michigan, corn and soybeans comprise a major portion
of a deer‘s fall diet (Gladfelter 1984). It is important, then, to understand how agriculture
affects white-tailed deer habitat if one’s objectives were to quantify white-tailed deer
habitat suitability at a specified time.

Agriculture, as a habitat variable, was not included in the habitat potential models
for the following reasons: First, the suitability of agriculture fields as a deer food source
depends on the size of the field, the crop species cultivated, the type of tillage practice
(e. g., fall plowed, no-till), and the distance to protective cover (Bender and Haufler
1984). These variables may change annually or seasonally.

Second, deer utilization of agriculture crops for food will differ in different areas
of Michigan. For example, in the SLP, 65% of the area is in agriculture (as calculated
with data from U.S. Geological Survey 1999); in the NLP, 10% of the area is in
agriculture (as calculated with data from MDNR-Land and Mineral Services Division,
Resource Mapping and Aerial Photography 2000); and in the UP, only 5% of the study
area is in agriculture (as calculated with data from MacLean Consultants, Ltd. 1995).
Deer utilization of agriculture crops has been studied in the NLP (e. g., Braun 1996), but

has not been quantified in other areas. The presence of agriculture would most likely

68

enhance food suitability if the fields are planted in preferred food types or if adequate
cover is still available.

Third, factors such as agriculture and other land use are always going to modify
the current habitat conditions for deer. White-tailed deer habitat potential, however, is an
inherent characteristic, or a property, of the land. The habitat potential models
constructed for this thesis are intended to serve as a foundation for analyzing white-tailed
deer habitat within landscapes, and the relationship between landscape characteristics and

deer demographics or distribution.

Value of habitat typing

Traditionally, natural resource classification or planning has been based on
existing vegetation conditions or biotic attributes and developed for specific uses. Land
cover classification, for example, is used widely by wildlife managers for habitat
planning. Land cover databases indicate the type of material that covers the earth’s
surface at a specific location at a specific time. They do not consider, however, how
abiotic factors might influence the distribution of wildlife and other natural resources.
Consequently, when using land cover classification to evaluate wildlife species responses
to management or to ecosystem change, managers must make assumptions about
potential vegetation and successional dynamics, which could lead to unrealistic
predictions.

Ecological classifications, an alternative to land cover classification, is based on
ecological potential and is necessary to effectively manage natural systems. Habitat

typing, a type of ecological classification, uses natural vegetation trajectories (i.e., biotic

69

attributes) and geological features (i.e., abiotic attributes) to separate large heterogeneous
areas into smaller, ecologically homogeneous units that are useful for management
(Stocker et al. 1977, Kotar and Burger 2000). Each habitat type possesses unique
biological and geological characteristics and is different from the rest. The habitat type
database developed in this project incorporated information from other habitat typing
guides (e.g., Coffinan et al. 1980, Burger and Kotar 1999, Kotar and Burger 2000), soils
associations, land type associations, ecoregions, and presettlement and current vegetation
to recognize areas with the same ecological conditions that support the same successional
trajectory (Figures 2, 5, 9). This habitat typing procedure provides a basis for better
understanding the relationship between ecosystem structure, processes, composition, and
the distribution of natural resources than solely examining land cover. For example, this
project demonstrates the value of using habitat typing to describe the spatial and temporal
distribution of deer habitat and how managers can use that information for deer
management and planning throughout the state.

Habitat typing provides a logical separation of ecological units that are essential
for effective natural resource management. An important key for wildlife management,
for instance, is an understanding of the ecological and geological characteristics that
drive the distribution of habitat. Several researchers have used land cover to map the
distribution of deer habitat. which could determine the location for potential habitat
improvement programs. But perhaps those areas identified for habitat improvement
could never provide optimal conditions to provide certain deer habitat requirements based
on inherent characteristics. For example, in the NLP, habitat types that support aspen,

balsam poplar, and birch in early stages; aspen, balsam poplar, black ash in middle

70

stages; and black ash, cedar in late stages have medium fall and winter food potential
(0.65), medium thermal cover potential (0.50), and low spring and summer habitat
potential (0.20). Programs to “improve” habitat for deer in these areas would still not
result in providing excellent habitat conditions for deer (Appendix 8, Figure A8-4).
Distinguishing habitat types that could potentially provide deer habitat from those that
may not provide certain requirements would help identify target areas for habitat
management. This project has clearly demonstrated the applicability of using an
ecological classification system (i.e., habitat typing) coupled with landscape-scale models
to identify areas for possible habitat management by recognizing those habitat types that
have potential for providing specific white-tailed deer habitat components.

Since managers are facing increasing pressure to manage for the habitat needs of
a growing list of species, many agencies are shifting their management approach from
focusing on single species to maintenance of the structure, composition, and processes
within entire systems over a broad planning landscape (i.e., an ecosystem-based
management approach). Ecological classification systems will also aid managers in
developing ecosystem-based management approaches. Ecosystem-based management
implies incorporation of ecological processes (e.g., succession), and properties of
landscapes at a broad spatial scale. The foundation for incorporating those ecosystem-
based management concepts is habitat typing. By identifying boundaries of habitat types,
managers can better understand the ecological processes such as succession, which occur
within the habitat type. By analyzing several habitat types together, managers can
understand community processes (e.g., dynamics of poorly drained coniferous forests).

And by analyzing many habitat types at a broad spatial scale, managers can begin to

71

understand spatial distribution of communities and species, spatial interrelations, and
processes that occur throughout a landscape. With the emergence of new perspectives for
management including ecosystem-based strategies. the habitat typing process will
become increasingly important to managers interested in conserving ecosystem processes

and ecological integrity.

Applications of habitat potential models

The MDNR has indicated that there is an urgent need to revisit Michigan’s deer
population goals and develop a deer management strategy that is scientifically based and
considers the ecological relationship between population characteristics, dynamics, and
distribution, and habitat dynamics and distribution (W. Moritz, MDNR. Wildlife
Division, personal communication). The use of habitat potential models is the
fundamental key for meeting the MDNR's need.

In natural resource management, habitat analyses based on existing land cover
databases are often used to explain the distribution and structure of populations. If it is
true that the distribution and structure of wildlife populations is determined only by the
existing structure and composition of vegetation, it would be pointless to investigate
anything other than that. However, no previous studies have investigated the relationship
between habitat potential and the variations in population demographics. Perhaps
inherent characteristics of habitat types can explain the distribution of deer or the
occurrence of certain demographics. Perhaps the existing structure and composition of
vegetation alone does not drive the distribution of deer but vegetation physiognomy, in

combination with soil and geological conditions, does. Or perhaps habitat potential, a

72

measure of the ability of an area to provide suitable deer habitat, explains trends in the
distribution of deer or spatial trends in the distribution of deer demographics (Felix et al.
In Press). For example, Verme (1973) observed that the dispersal tendencies of deer in
the Upper Peninsula, Michigan were not different than those that occurred 30 years
earlier, despite appreciable changes in habitat. It is evident that existing vegetation
conditions were not the only factors influencing deer dispersal patterns because the
habitat changed. Perhaps underlying characteristics or the area, or habitat potential, does
play a role in determining dispersal patterns.

The boundaries and dynamics of habitat types do not change, unless they undergo
natural or anthropogenic disturbances beyond conditions defined by their historical range
of variability. Land cover, however, does change. If indeed there is a quantifiable
relationship between habitat potential within habitat types and deer distribution or
demographic patterns, managers can start to make more realistic predictions and
recommendations for deer management. F irst, managers will be able to determine which
areas are likely to support specific deer demographics such as larger antler beams, higher
antler points and high fawnzdoe ratios. Second, managers will be able to predict the
distribution or movements of deer based on the location of habitat types that provide deer
habitat requirements and the existing vegetation type of the respective habitat type.
Third, managers will be able to define biologically meaningful deer management unit
boundaries.

Understanding the relationship between habitat potential and deer demographics

Felix et al. (In Press) began initial investigations regarding the relationship

between habitat potential and deer population demographics. Based on analysis of

73

variance and Tukey-Kramer multiple comparison tests, we concluded that there was a
significant (P < 0.001) negative relationship between fall and winter food potential and
the distribution of yearling average antler beam diameters collected within each township
from 1999 hunter harvest data in the NLP study area. Although the relationship was not
positive, it was evident that distinct regional patterns existed both in the distribution of
habitat components and deer demographics. We are further investigating the relationship
between fall and winter food potential, thermal cover potential, and spring and summer
habitat potential with other demographic parameters as well (Felix et al.. unpublished
data).

To provide meaningful results, the scale at which population demographic data
are collected must be compatible with that of habitat types. In the Felix et al. (In Press)
study, we used 1999 data collected at the township scale. As such, fall and winter food
potential had to be re-scaled to that of the township. We re-scaled fall and winter food
potential by calculating a weighted average of fall and winter food potential based on the
proportion of each habitat type within each township. Re—scaling had to be done to
conduct statistical analyses. The problem with re-scaling habitat potential to fit within
township boundaries is that resolution is lost at the ecological scale that is important for
deer. It would be ideal if hunter harvest data could be collected within habitat types, but
if that is not feasible, at least it should be collected by townships (or section, if possible).
Data collected by county is useless for this type of analysis because county boundaries
are too large to describe deer demographic or movement patterns. There is a pressing

need for a dataset that is collected consistently for several years at a scale that can be

74

compatible with habitat types in order to effectively quantify the relationship between
habitat potential and trends in deer distribution and demographics.
Understanding the relationship between habitat potential and deer spatial structure

Research (e. g., Felix et al., unpublished data) has also begun assessing the utility
of white-tailed deer habitat potential models to understand deer spatial structure and
subsequently define biologically meaningful deer management units (DMUs).
Researchers (e.g., Verme 1973, VanDeelen et al. 1998) have argued that information on
dispersal patterns would facilitate delineation of realistic management units for various
herds. I would argue that information on dispersal patterns in conjunction with how they
relate to habitat potential would facilitate establishment of deer management units based
on similar ecological properties and similar potentials to provide deer life requisites.

We overlaid winter and summer home ranges of migratory deer (i.e., seasonal
home ranges > 1 km apart) and non-migratory deer (i.e., overlapping seasonal home
ranges) in the northeastern Lower Peninsula (Garner 2001) on the maps depicting habitat
potential, which produced a product that indicated what proportion of each home range
was in high or low habitat potential (Felix et al. 2002). It was evident that the
juxtaposition of habitat types is useful in determining if or where deer move. For
example, lowland swamp conifers are important for providing winter habitat
requirements and early successional stages of northern hardwoods are important for
providing summer habitat. It is apparent that habitat types cannot provide both winter
and summer habitat requirements for deer. Therefore, it may be necessary for deer to
shift their seasonal home ranges to obtain their annual life requisites (e. g., migratory

deer), unless habitat types that provide seasonal life requisites occur adjacent to each

75

other, thus allowing deer to obtain habitat requirements without moving (e. g., non-
migratory deer).

The results of the analysis implied that habitat potential models can be used to
explain spatial movement patterns of deer based on knowledge of the distribution of
habitat types that can provide deer life requisites. An understanding of how deer
movement patterns relate to habitat potential, in conjunction with identification of areas
with similar population management goals, may aid in establishing effective and

biologically meaningful deer management units.

Recommendations for applying habitat potential models to managing wildlife
populations and habitat

The future of natural resources lies in maintaining the composition of native
ecosystems through management that considers spatial and temporal patterns of
ecological processes, the role of abiotic factors within systems. and how they influence
ecosystem structure, function, and composition. Managers have begun to adopt this
framework, but in doing so, have identified information needs at different levels of
biological organization. First, knowledge of the spatial arrangement of ecosystems
within landscapes allows managers to understand landscape structure, which, in turn,
provides predictive ability on how properties of landscapes influence processes at the
ecosystem and species levels (Haber 1990). Second, there is a pressing need to identify
the composition and structure of species within ecological communities so managers can
determine if inherent ecosystems exist and if ecosystems are functioning within their

historical range of variability (Haufler et al. 2002). Haufler et al. (2002) indicated that

76

such information is critical because an understanding of ecosystems is fundamental to
making planning decisions for biodiversity. Third, there is a need for definitive spatial
and temporal links between habitat supply and wildlife species response (e. g., viability).
The use of habitat typing and the concept of habitat potential as a framework for
managing systems has several applications at each of these levels of biological
organization.
Landscape-level

Haufler et al. (2002) suggest that landscape-level planning should consider the
amount of area needed to provide all successional stages of all native ecosystems to
maintain species and processes associated with each ecosystem. The habitat type
databases and spatial projections developed in this project are tools to aid in landscape—
level planning since they identify boundaries of ecosystems (e.g., habitat types) and
describe the successional pathway within habitat types. Using the habitat type database
as a tool, managers can, first of all, identify the boundaries and properties (e. g., soils, land
type associations, and potential vegetation) of ecosystems within the landscape.
Secondly, through the spatial projections, managers can assess the structure of landscapes
by quantifying the heterogeneity, size, arrangement. and connectivity of ecosystems
(Noss 1990). These landscape characteristics affect the relationship between
communities within different ecosystems and the distribution and dynamics of species
(N oss 1990). Third, managers can compare the arrangement of habitat types and their
ecological potentials with current conditions to assess if all stages of all habitat types are
represented within the landscape. Haufler et al. (2002) indicated that representation of all

potential ecological conditions within a planning landscape would allow for the

77

maintenance of ecosystem integrity and biological diversity. Information identified at the
landscape-level can then be linked to management at the ecosystem and species levels.
Ecosystem-level

Ecosystems are dynamic and require ecological processes to fluctuate within the
historical range of variability for long-term functioning and sustainability. Natural
variability within ecosystems allows them to withstand a range of disturbances without
changing the processes that control its behavior (Haufler et al. 2002). Human use of
resources tends to reduce the variability of ecosystems so that a steady flow of goods and
services can be produced to satisfy human demands (Gunderson et al. 1995). Where once
the threat of species extinction was perceived as the only problem resulting from the
reduction of variability and diversity within systems. the loss of entire ecosystems is now
a very real threat (Grumbine 1994, Noss et al. 1997). Management at the ecosystem:
level, however, is beginning to be conducted to maintain ecological sustainability and
consider natural processes that contribute to the variability in composition. structure, and
function within systems (Vogt et al. 1997). Tools such as habitat type databases and
spatial projections of habitat types could help managers maintain ecological integrity by
identifying the ecological boundaries that delineate different systems, and describing
structure, composition, and natural processes that are characteristic of each system (e. g.,
within habitat types).

Noss et al. (1997) indicated that managers lack information on the composition of
potential vegetation, which is essential for understanding temporal changes within
ecosystems. Habitat type maps and databases, such as those developed for this study, can

be used to identify ecosystems that can potentially provide unique or rare characteristics

78

and subsequently be managed to maintain those conditions. For example, Michigan has
lost nearly all of its mature-old grth successional stages of oak, white pine-red pine,
and beech-maple communities (Noss et al. 1997). Some areas that once provided these
conditions cannot be realistically restored because they have been converted to cities or
other developments. However, habitat type maps can identify other areas that could
potentially be restored to those old growth communities. Figure 13, for instance,
illustrates the location of habitat types that have the potential to provide old growth
beech-sugar maple communities. The majority of these sites currently support early

successional stages of northern hardwoods (44%) or agriculture (20%).

 

 

 

0 12 24 miles

I:—

Figure 13. Location of areas in northeastern Lower Peninsula, Michigan that have the
potential to support old growth beech-sugar maple forests based on the spatial output
fi'om the habitat type database for the northeastern Lower Peninsula developed by.

79

The habitat type databases can also be used for planning at the ecosystem-level.
For example, the Civilian Conservation Corps established red pine plantations throughout
Michigan beginning in the 19305. Managers within the MDNR have indicated they
would like to manage state forests for broad ecological objectives and would like to
restore certain areas to natural vegetation by cutting red pine plantations on sites that
would not naturally support red pine. They are, however, unsure what vegetation would
regenerate subsequent to cutting red pine (B. Mastenbrook, MDNR, personal
communication). Managers can use the habitat type database to identify the geological
characteristics and successional paths inherent to those sites selected for cutting.
Subsequently, managers can understand how the vegetation composition will change on
those sites following the cutting of red pine.

Currently, the habitat type databases developed for this project contain
information on successional trajectories, geological characteristics. and vegetation
structural characteristics for each habitat type delineated. The databases only provide
vegetation structural information for variables important for quantifying white-tailed deer
habitat potential. They could, however, be the foundation for including additional
vegetation structural characteristics or other information unique to different habitat types
(e.g., the historical range of variability of conditions found within habitat types or
frequency of processes driving those conditions). With knowledge of the historical range
of conditions within habitat types and comparison of those to current conditions,
managers can devise plans to either maintain the system if it is functioning within the
historical range or restore systems that are not operating within the historical range of

variability. An understanding of the historical range of variability within ecosystems can

80

also provide insights to the needs of species that depend on certain vegetation types
within ecosystems, and subsequently management can be linked between the ecosystem-
and species-levels (Haufler et al. 2002).

Managers can use the habitat type database in conjunction with the habitat
potential model process at the ecosystem-level to develop habitat potential models for
communities that utilize the same resources. For example, research has indicated that
management of forest birds has traditionally occurred within stands but perhaps would be
more meaningful and effective if managers understood how habitat relationships,
distributions, composition, and dynamics of bird species changes as vegetation changes
temporally within ecosystems (Franzreb et al. 2000). Habitat potential models can be
developed for bird communities that utilize the same ecosystem (e. g., northern
hardwoods). Potential habitat for assemblages of birds can then be projected throughout
space and time, giving managers an indication of which areas can provide vegetation
conditions that are important for providing habitat requirements for groups of birds and
when those vegetation conditions would be provided (Figure 14). Comparison of this
potential with current distributions of species or assemblages of species would help
managers identify areas that are not providing habitat for certain bird assemblages and
should be targeted for forest bird management.

Species-level

Certain species may be targeted for management because of special interest (e. g.,
game species), conservation concern (e. g.. restored or endangered species), or because
they indicate trends or conditions in ecosystems (Haufler et al. 2002). Typical

management questions for species are how, where, or when should managers manipulate

81

— Bird community A
Bird community B
...g... Bird community C

 

  

 

 

 
   

 

>.
3:
.5
cc
.2
:3
{I}
H
as
.21
..o
a
E 1\

Threshold

conditions

wiltll?‘ if. . . , ,, , ..
Aspen red oak red maple/white pine

Successional Stage

Figure 14. Hypothetical example of how managers might use habitat types and habitat
potential modeling to understand temporal changes in forest bird community distribution.
Output from habitat potential models developed for each bird community could produce
suitability curves that would indicate which seral stages provide habitat for different
forest bird communities. Viability of each bird community decreases below the threshold
habitat conditions.

populations or habitat to obtain management objectives? Answers to these questions
requires an understanding of how the distribution of ecological and geological
characteristics drives the distribution of habitat. The ecological and geological

information contained in the habitat type database and the modeling process used in this

82

study can be used to develop habitat potential models for other species, thereby
addressing the need for better defined relationships between wildlife species and the
biotic and abiotic factors that define their habitat.

For example, moose (A Ices alces) have generated the attention of Michigan
biologists because they have not attained expected growth rates and population size since
their re-establishment into the western Upper Peninsula in the mid 19808. Michigan
DNR biologists are concerned that uncertainties about moose population numbers and
factors affecting population dynamics will hinder moose management efforts (Dodge
2002). Recent research has revealed that low moose reproductive rates have contributed
to slower-than-expected population growth, but no studies have analyzed the relationship
of habitat characteristics and components to moose productivity (Dodge 2002).
Thompson and Stuart (1998) indicated that attempting to manage moose without
knowledge of the relationships between the capability of an area to support moose and
moose productivity is costly and ineffective, wastes time and resources. and may
jeopardize moose populations. Clearly. there is a need to understand what underlying
mechanisms influence elements of moose population dynamics (e. g., mortality,
reproduction, movement) and demographic parameters (e.g.. density, sex ratio, antler
size).

Current research indicates that habitat potential can explain patterns in the
distribution of white-tailed deer with specific demographic characteristic (Felix et al. In
press) and spatial movement patterns (Felix et al. unpublished data). Therefore, it is
reasonable to suggest that managers could use the same process to address similar

population management questions for moose; specifically, what is causing low

83

productivity? An evaluation of moose habitat potential throughout the western Upper
Peninsula would entail identifying minimum winter and summer habitat requirements
(e. g., food quantity and nutritional qualities, thermal cover) for moose, determining the
types of soils and landforms on which vegetation providing moose habitat is most
productive and assessing changes in suitability over space and time. Information on the
spatial and temporal distribution of moose habitat, in conjunction with information on
population dynamics (e. g., Dodge 2002) will allow managers to determine factors
influencing observed low productivity rates. Subsequently, managers can target
management activities on those specific factors influencing low production to sustain
moose populations in Michigan that are in balance with the potential of the land.
Different species utilize habitat types at different scales. Deer and moose, for
example, utilize many habitat types to obtain their life requisites. Smaller species.
however, could obtain their life requisites within 1 stand or within regions of 1 stand. For
example, red-backed salamanders (Plethodon cinereus) inhabit stands of northern
hardwood habitat types that have appropriate soil properties and vegetation structure
(Hanaburgh 2001). Sometimes, species such as salamanders are selected as indicators of
specific ecological conditions such as soil type or specific stand characteristics such as
amount of down woody debris (Hanaburgh 2001 ). Use of habitat type databases and
maps and habitat potential models for indicators to assess the distribution and potential
availability of specific ecological conditions is an important technique for biodiversity
management. For example, a manager could use habitat type maps and databases to
identify areas that could provide stand and soil conditions necessary for the salamanders.

Microsite conditions within habitat types specific to salamanders or other small species

84

could be incorporated into the database. The process used to quantify habitat potential
for deer in this study could be adapted to quantify habitat potential at smaller scales for
salamanders to produce a product that identifies which areas have the potential to provide
specific ecological conditions necessary to sustain salamanders and other species
associated with those conditions. -

Often, habitat is not considered as much as it could be in species management, but
rather, managers emphasize population management. Or, if habitat is considered in
species management, conditions are usually assessed from existing land cover conditions
and does not consider how abiotic factors might influence habitat composition and the
distribution of wildlife. Subsequently, when predicting species response to management
activities or ecosystem change, managers must make assumptions about species-habitat
relationships, potential vegetation, and spatial and temporal changes in habitat. Making
assumptions, in turn, could lead to unrealistic predictions.

Databases that contain information on spatial and temporal changes in vegetation
structure and composition within habitat types and models that quantify habitat potential
for various species are important tools for wildlife mangers to make realistic predictions
based on: 1) an understanding of the spatial and temporal distribution of habitat
requirements for a species; 2) quantified relationships between habitat potential and
population demographics; and 3) identified factors that drive distributions of populations
and productivity. Regardless of the reason for management of selected wildlife species
(i.e., special interest, management concern, or as indicators), an understanding of habitat
potential-population relationships is essential for effective, efficient, sustainable, and

scientific species management.

85

Validation

The purpose of modeling is to represent a simplified version of a real-life system
and provide an opportunity to explore ideas regarding ecological systems (Jackson et al.
2000). One of the challenges in modeling is evaluating the behavior of the model as it
relates to real systems and with respect to how it is intended to be used (i.e., validation).
The habitat type themes and the white-tailed deer habitat potential models developed in
this thesis have yet to be validated.

The first step in validation is to assess whether or not the structure and functional
relationships of the models and habitat type themes describe the composition and
dynamics of deer habitat and how deer respond to it. The second step is to evaluate the
correspondence between model behavior, expected patterns, and data from real systems.
The following are some suggestions to validate the habitat type theme and habitat
potential models.

Habitat type themes

The habitat type themes can be used to identify the location, composition, and
successional dynamics within habitat types. The distribution of habitat types according
to the habitat type themes developed in this thesis does seem reasonable and was verified,
in part, through comparisons with satellite images of vegetation cover and land use (see
methods section “Validation of habitat types” p. 18). Only some of the habitat types in
the northern Lower Peninsula have been ground-truthed. The location of habitat types
should be ground-truthed to validate the spatial arrangement of habitat types. Habitat

types would need to be identified in the field through identification of soil properties and

86

plants indicative of specific habitat types (e.g.. Coffman et al. 1980, Burger and Kotar
1999).

Validation of successional pathways and structural characteristics of vegetation
with in habitat types could be accomplished with a field study that used exclosures
established in different vegetation types within habitat types. Exclosures would allow the
investigator to observe how the composition and structure of vegetation changes in the
absence of browsing.

White-tailed deer habitat potential models

The habitat potential models were constructed to describe inherent land
characteristics and classify habitat types in a way that is ecologically meaningful to deer
(Figures 3, 4, 6. 7. 8, 10, 11, 12). This type of information. when assessed with
population demographic characteristics and movement data will help managers determine
where it is necessary to establish different deer management goals or re-define
biologically meaningful deer management units based on deer biology and the capability
of areas to provide habitat components.

The model variables selected and the structure of the models to quantify white-
tailed deer habitat potential seem logical and justified based on analysis of literature
describing white-tailed deer habitat (see “Process for quantifying habitat potential” p.
14). Some variables or the model structure may need additional adjustment as more data
become available. For example. white-tailed deer habitat components have not been
quantified in southern Lower Michigan. Some of the information used to construct the
models was based on preliminary data from initial studies with radio collared deer (e. g..

J. Pusateri, graduate research assistant, Michigan State University, Department of

87

Fisheries and Wildlife). The models should be validated in the series of steps described
below.

First, a relationship between habitat suitability to population viability metrics
(e. g., fawnzdoe ratios, yearling antler beam diameters, femur fat, winter survival) would
have to be determined for each habitat requirement. For example, winter survival may be
a suitable metric to evaluate thermal cover suitability, whereas, yearling antler beam
diameters may be a suitable index to evaluate the relationship between population
conditions and spring and summer habitat. Some assumptions, however, may be critical
when investigating the relationship between habitat suitability and population viability.
That is, density dependent factors may affect the observed relationship. For instance,
within a population, antler beam diameter, weight, and reproductive rates of white-tailed
deer vary with the nutritional level within a population (Severinghaus and Moen 1983).
The nutritional level within a population is a function of population density as well as
habitat quality and soil properties. To accurately quantify a relationship between habitat
conditions and population viability, however, the population would need to be below the
threshold that would adversely affect deer conditions and productivity. The second step
in validating the habitat potential models is to assess current habitat conditions in
successional stages of different habitat types. Then it can be determined where current
conditions fall along the habitat suitability curves for each habitat type. Finally, research
could examine the correspondence between the suitability of different habitat components
and population metrics. It would not be necessary to evaluate parts of the models
separately since the variables in the model are assumed to together represent the ability of

vegetation types to provide deer habitat components.

88

Future directions

The MDNR has assumed a leadership role implementing ecosystem management
to conserve, protect, and manage Michigan’s natural resources for current and future
generations (MDNR 2002). As such. the MDNR has begun to appoint management
teams for different ecoregions of Michigan to develop strategies to sustain and enhance
representative ecosystems within Michigan’s ecoregions. The tools developed in this
thesis (e. g., habitat type databases and themes, habitat potential modeling process) will
help aid managers in accomplishing ecosystem management goals that sustain functional
ecological systems while maintaining economic success and social acceptance. For
example, the habitat type themes and databases indicate the location and properties of
habitat types within different ecoregions of Michigan. They can help identify specific
habitat types that should be maintained in a certain ecological state (e. g., for endangered
species) or that should be managed to sustain specific public benefits (e.g., hunting or
other recreation) and opportunities (e. g., education) into the future. In addition. the
habitat potential models can serve as a framework for managing species or communities
or for developing models that simulate possible future outcomes of potential habitat
management situations.

When a house is being built, the carpet is not installed first. The foundation is
poured and the framework is constructed before the finishing touches are added. The use
of habitat typing and the habitat potential models developed in this project are pieces of
the foundation and framework for ecosystem management. The foundation and
framework must be solid (i.e., validation is critical) and then the finishing touches be

added and fine-tuned to meet different ecological objectives. I encourage managers to

89

validate, develop, modify, incorporate additional pieces of information, and apply the
concepts presented in this thesis to help guide ecosystem management goals, protect
Michigan’s natural resources, and sustain Michigan’s leadership role in ecosystem

management.

90

APPENDICES

91

Appendix 1. Scientific names of overstory vegetation that was mentioned in the text of
this thesis.

 

Common name

Scientific name

 

American elm
Aspen
Balsam fir
Balsam poplar
Basswood
Beech

Black ash
Black cherry
Black oak
Black spruce
Cedar
Cottonwood
Green ash
Hemlock
Hickory
Ironwood
Jack pine
Red maple
Red oak

Red pine
Silver maple
Sugar maple
Swamp white oak
Tamarack
White ash
White birch
White oak
White pine
White spruce
Willow
Yellow birch

Ulmus americana
Populus tremuloides/grandidentata
A bies balsamea
Populus spp.

Tilia americana

F agus grandifolia
F raxinus nigra
Prunus serotina
Quercus velutina
Picea mariana

T huja occidentalis
Populus deltoides

F raxinus pennsylvanica
T suga canadensis

C arva spp.

Ostrya virginianus
Pinus banksiana
Acer rubrum
Quercus rubra
Pinus resinosa

Acer saccharinum

A cer saccharum
Quercus bicolor
Larix laricina

F raxinus americana
Betula papyrifera
Quercus alba

Pinus strobus

Picea glauca

Salix spp.

Betula alleghaniensis

 

92

 

.399 .38.. we 5.89.56 2.. co xoozo 20:8 a me new: on. .92).
detmoEmmfiu .Eocow he, ”maze—gee: 8.3.2. he. .38: b3 82

.83. 2a. a. 8... @832 5.5%?) .23 3.3.3..

.399 5.23%? ..o mow—22.0 .3.QO
.e £93.25 was...“ was 8%. coufiowg 635:0 3532. 3.86:.

.39.. .955. 5252. 8.5.252. wanna—:6“. .5. .383 2.8
Bobbfi so .09» >9: 22.8%?» ..e 229:5. 2: .395 cozflowo.»
688:0 3:66 toga .38 233.08% Enact... ...w:e£.<

.8320 .38.. getteeflmzozgooe
..wd. 8.5.82 $20 .38.. 235 .58.. commas. 0.882..
Emacs. Eém .5 women. 5.32.63? .38 was. 725.3%

.88. as 2.. a 535:2
5 528%?» ..e 230:5 ecu «5:38:80 32.3me

.850
was .8250 2.. .3 03.6.... .5232. 3.26:8 @628
.5 women 535:). c. 5.50%; .3 5:32.58 89.880

dogma» 285:: can
655.2: use 23x2 :8 Enact? 3.586 335.083
..em .m:e_§oomma ..em :88ng 5.5» 895 $805.00

dozfiow? .822 .3222.

560 east.
..2 mOmD 30,.

£52 who.

Eon—0.335
ce.§owo>

mzeafiemmxx ..em

 

.:e.§omo> Ea .m..¢m can $5828. .383 cc $889. .593 8a 80:... mse..m.oemm<
$83.65. ea comma. mow—«ESE. 7336.30 message—o 8.. .33: $53235 25w oEafiwe.m>~.&o:mE..o E5...» @832 2.»... .28..
dozaoEmmmB .352. .e 2880... 2.. a. .o>o.
50.685 2.. m. m7; .mEoeeQEeo 0.8...“ a. mm :03 3 83083 2.8 can. Jacobean.
can 0.56.5.5. a. 5&6 8... 3.5. 98. .5852 cc $08 08... .m.on.w :32on 108.62. 688.3 E moococotB .8
58.3.3 53%—.35 e. .83 .5. owes. 03 8a 938 82.. :w:e£.< woman. 5.82.638 82388 2.835. 3.8%»: 8325. 323.com. Mtg?
mace—show: Eiflbmon— owaageu

 

.893 .859. .0 2883.326 2: .5. 32%..ng maesatomow was 839.3% o...m..m>m .0 En. .m x..o:o&<

93

94

 

.:0..20w0> 50.50
w:.m8m8 0:0 8...». .232. .0 5.5.2.3 5.. 3.83 .:0..20w0>
.2520: 0.8.0:. .0: 80.. 0:0 5.2208.» 50.50 8....08D

.25.... .253 :. 52.59:. 8%. :0..20w0> 8.5.0:.

880.0 .85.. :. 55.00.00...
0:0.w 8.00% .0 8.00% 02.20.. 830 .20.. EH .0205...
:0 .0080 A.4 .80....5: :. :0..20w0> .0 5......08Q SF .085... 30..

.20.. SF .0205... :0 .0080 0.358..
.00.... 0:. :. .050 .055... .00.. 02.22.03 .0 E0526. EH .8053 .3.

 

80:558.

5......08Q 0mm.0>0U

 

880 .N 55......

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N Om;
0 hm. _ . Om; l >430 T
N Om> _
m H: _I_ 3201. as; .
_ . Om> - .
— N Om; T— Q2<m r
v ...I _ . Om; 7
[ 03.... :0.w0._00m.
_ N Om> _L
m .5. _ 1 >430
, _ 00> . .1
N Om> _l .
NE _ 1 $23 .4 _ 030. Al
. . Om; - .
_ N Om;
. H: T Qz<m
_ . Om;
09¢ .2303 0a.»... :0..0.0w0> 2.»... ..0m 0E.w0y. 0.5.0.03. «3.... :0.w0.00m.

0%. .250: H ...E .09.. :0..0.0w0> u Om> .Ema 00:.0:0 3:00.. 50> 0:0 .00:.0:0 3.000.. .00:.0:0 3.000.. 8:30:80 .00:.0:0
..03 3.0.0.0008 .00:.0:0 ..03 .2555 00:.0..0 ..03 3.030808 ”000: 0.03 880.0 0.0.0.0.: x.m 0.03.0:0 00.. 06.. 0 .0 00.00.00 003
0...... 0:0 .x :0..0.0080 0%. 0:0. M Jr... .02.». .850: w:..00:..00 0:0 w:.:_..00 ..0 0800:. 0... 00.. E0:w0.0 3:300:00 .m x.0:0n.n.<

95

Appendix 4a. Habitat classification key describing fall and winter food habitat
requirements (browse component), suitability indices (SI) on a scale of 0-1 (1 = high and
0 = poor), and sources.

Component: Browse

 

 

 

Variable Options SI Literature
Browse quality cedar 1.00 Davenport 1941
hemlock 1.00 Ullrey et al. 1964
red maple 1.00 Ullrey et al. 1967
sugar maple 1.00 Ullrey et al. 1968
silver maple 1.00 Rogers et al. 1981
basswood 0.75
black ash 0.75
birch (white and yellow) 0.75
black cherry 0.75
American elm 0.75
willow 0.75
aspen 0.50
balsam poplar 0.50
ironwood 0.50
jack pine 0.50
oak 0.50
white pine 0.50
beech 0.25
balsam fir 0.25
cottonwood 0.25
red pine 0.25
spruce 0.25
tamarack 0.25
Browse availability aspen Graham et al. 1963
Age (for all age classes)
l-10 1.00
10-30 1.00
30-50 0.50
50-100 0.00
100+ 0.00

96

Appendix 4a. Cont’d

 

 

Variable Options SI Literature
Browse avail. (com) other deciduous
Age
1-10 1.00
10-30 0. 50
30-50 0.00
50- 100 0.00
100+ 0.00
coniferous Ryel 1953
Age Verme ] 965
1-10 0.75
10-30 1 .00
30-50 0.75
50-100 0.50
100+ 0.00
lowland brush 1 .00

 

Appendix 4b. Habitat classification key describing fall and winter food habitat
requirements (mast component), suitability values (SI) on a scale of 0-1 (1 = high and 0 =

poor), and sources.

Component: Mast

 

 

 

 

Variable Options SI Literature
Mast species {me hard mast Harlow et al. 1975
oak (black, red, white) 1.00
beech 0.75
other 0.00
Mast bonus sofi mast + 0.05 Rogers et al. 1981
Basal area oak in mixed stands Bender and Haufler 1987
Stand age fiz/ac Host et al. 1987
1-10 no data 0
10-30 no data 0.5
30-50 42-52 1.00
50-100 no data 1.00
100+ no data 0.25

97

Appendix 4b. Cont’d

 

 

Variable Options SI Literature
Basal Area (cont) oak in pure stands Marty and Rudolph 1970
Stand age fiz/ac
1-10 20-40 1.00
10-30 40-70 1.00
30-50 60—90 1.00
50-100 90+ 1.00
100+ 90+ 1.00
beech Tubbs and Houston 1990
Stand age ftz/ac
1-10 0 0.00
10-30 1-10 0.50
30-50 14-35 1.00
50-100 40-70 1.00
100+ 70-75 1.00
DBH oak Sander 1990a, b
Stand age inches Duvendeck 1964
1-10 0-2 0.00
10-30 3-6 0.50
30-50 7-12 1.00
50-100 11-20 1.00
100+ 20+ 1.00
beech Tubbs and Houston 1990
Stand age inches
1-10 0 0.00
10-30 0-1 0.00
30-50 2-7 0.50
50-100 8—13 1.00
100+ 14-24 1.00

 

98

Appendix 5a. Habitat classification key describing thermal cover habitat requirements
(forest composition component), suitability indices (SI) on a scale of 0-1 (1 = high and O
= poor), and sources.

Component: Forest Composition

 

 

 

Variable Options SI Literature

Forest Type Coniferous 1.00 Verme 1965
Mixed 0.50
Deciduous 0.00

Coniferous Species Cedar 1.00 Davenport et al. 1953
Hemlock 1.00 Verme 1965
Spruce 0.80 Euler and Thurston 1980
Fir 0.80 Bender and Haufler 1987
Pine 0.40 Blouch 1984

 

Appendix 5b. Habitat classification key describing thermal cover habitat requirements
(forest structure component), suitability suitability indices (SI) on a scale of 0-1 (1 = high
and 0 = poor), and sources.

Component: Forest Structure

 

 

Variable Options SI Literature
Basal Area for Stand age fiz/ac Bender and Haufler 1987
lowland conifers 1-10 0-30 0.00 Johnston 1990
10-30 30-100 0.50

30-50 174-261 1.00
50-100 174-261 1.00
100+ 300 max 1.00

 

 

Canopy Cover Stand age % Verme 1965
1-50 40-70 0.50 Ozoga 1968
50+ 70-100 1.00
Tree Size (DBH) Stand age inches Verme 1965
1-10 0 0.00 Ozoga 1968
10-30 0-0.5 0.00 Ozoga and Gysel 1972
30-50 0.5-3 0.50 Johnston 1990
50-100 3-9 1.00
100+ 9+ 1.00

99

Appendix 5b. (Cont’d)

 

Variable Options SI Literature

 

Age Structure Stand age structure
< 150 even-aged 1.00 Verme 1965
> 150 uneven-aged 0.00

 

100

Appendix 6. Habitat classification key describing spring and summer habitat
requirements, suitability indices (SI) on a scale of 0-1 (1 = high and 0 = poor), and

 

 

 

sources.
Variable Options SI Literature
Vegetation Type Openings 1.00 Kohn and Mooty 1971
Preference Upland deciduous 1.00 McCaffery et a1. 1974
Upland mixed 1.00 Stormer and Bauer 1980
Upland coniferous 0.40 Mackey 1996
Lowland shrub/brush 0.30
Lowland deciduous 0.20
Lowland mixed 0.20
Lowland coniferous 0.20
Forage Availability
—for upland deciduous, Early successional 1.00 Verme 1969
upland mixed, and Early-mid successional 0.80 Kohn and Mooty 1971
lowland types Mid successional 0.60 McCaffery et al. 1974
Mid-late successional 0.40 Stormer and Bauer 1980
Late successional 0.20 Mackey 1996
-for upland coniferous Early successional 1.00
Early-mid successional 1.00
Mid successional 1.00
Mid-late successional 0.50
Late successional 0.50

 

101

Appendix 7. White-tailed deer habitat suitability curves for each successional trajectory
in southern Lower Michigan. The curves display habitat dynamics for fall and winter
food, and spring and summer habitat components based on habitat potential models
developed in this document.

102

 

 

E
g
.0
Q
8
1 2 3 4 5 6
Seral Stage
B.
é
t
g
.9
£1
3

 

Seral Stage

Figure A7-1. Suitability curves for fall and winter food (A), and spring and summer
habitat (B), for areas that support aspen in early successional stages (1-2; ages < 30
years), sugar maple-beech-basswood-red maple-white ash in middle stages (3-4; ages 30-
100 years), and sugar maple-beech-basswood in late stages (5-6; ages > 100 years).

103

 

 

Suitability Potontiai

 

 

1 2 3 4 5 6

 

Suitability Potential

 

Seral Stage

Figure A7-2. Suitability curves for fall and winter food (A), and spring and summer
habitat (B) for areas that support aspen-red oak-hickory-black cherry in early
successional stages (1-2; ages < 30 years), red maple-sugar maple-red oak-white oak-
black cherry in middle stages (3-4: ages 30-100 years), and red maple-basswood-red oak-
white ash in later stages (5-6; ages > 100 years).

104

Sultablllty Potential

Suitability Potential

 

1 2 3 4 5 6
Seral Stage

Figure A7-3. Suitability curves for fall and winter food (A), and spring and summer
habitat (B) for areas that support grass-sedges—forbs in early successional stages (1-2;
ages < 30 years), grass-shrubs-black oak-white oak-black cherry in middle stages (3-4;
ages 30-100 years), and white oak-black oak-red oak-hickory-red maple in later stages
(5-6; ages > 100 years).

105

Suitability Potential

 

 

B.
3
i:
i
g
.n
.8.
iii
1 2 3 4 5 6
Seral Stage

Figure A7-4. Suitability curves for fall and winter food (A), and spring and summer
habitat (B) for areas that support grass-sedges-forbs in early successional stages (1-2;
ages < 30 years), grass-shrubs-black oak-white oak-red oak-hickory in middle stages (3-
4; ages 30-100 years), and white oak-black oak-red oak-white pine in later stages (5-6;
ages > 100 years).

106

Suitability Potential

 

 

Suitability Potential

 

1 2 3 4 5 6
Seral Stage

Figure A7-5. Suitability curves for fall and winter food (A), and spring and summer
habitat (B) for areas that support lowland brush in early successional stages (1-2; ages <
30 years), tamarack-black ash-birch-red maple-American elm in middle stages (34; ages
30~100 years), and red maple—American elm-swamp white oak in later stages (5-6; ages >
100 years).

107

 

Suitability Potential

 

Seral Stage

 

 

 

Suitability Potential

 

1 2 3 4 5 6
Seral Stage

Figure A7-6. Suitability curves for fall and winter food (A), and spring and summer
habitat (B) for areas that support willow-cottonwood in early successional stages (1-2;
ages < 30 years), silver maple-American elm-green ash-cottonwood in middle stages (3-
4; ages 30—100 years), and sugar maple-American elm-red maple-basswood in later stages
(5-6; ages > 100 years).

108

Appendix 8.. White-tailed deer habitat suitability curves for each successional trajectory
in northeastern Lower, Michigan. The curves display habitat dynamics for fall and
winter food, thermal cover, and spring and summer habitat components based on habitat
potential models developed in this document.

109

P“
p
Q.

Suitebllltylndex
O O O O O
h in b '0 b

999
‘06“

 

 

Successional Stage

 

 

 

 

 

 

Suitability Index

 

 

 

 

 

 

 

 

Sultabillty Index
3

 

0.1 -i
o I 'T T V
1 2 3 4 5 B

Successional Stage

 

 

 

Figure A8-1. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen in early successional stages
(1-2; ages < 30 years), aspen-red maple-white pine in middle stages (34; ages 30-100
years), and white pine-red maple-beech-sugar maple in late stages (5-6; ages > 100
years).

110

Suitability Index

 

Successional Stage

Suitability Index

 

Successional Stage

 

 

 

 

 

 

 

 

Suitability Index
0
b.

 

 

 

1 i 5 A 5 a
Successional Stage
Figure A8-2. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen in early successional stages

(1-2; ages < 30 years), oak-red maple-white pine in middle stages (3-4; ages 30-100
years), and white pine-red maple in later stages (5-6; ages > 100 years).

111

9.0999
CGNOQ-fl

SuitabIlIty Index
a o o
L: 'u h

9
..x

 

1 2 3 4 5 6
Successional Stage

 

 

 

Sultablllty Index

 

 

 

 

 

 

 

 

 

 

SultabllIty Index
a
In

 

 

 

1 5 .4. Q g s
Succeeslonal Stage
Figure A8-3. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-balsam poplar in early
successional stages (1-2; ages < 30 years), black ash-American elm in middle stages (3-4;

ages 30—100 years), and red maple-sugar maple-black ash in later stages (5-6; ages > 100
years).

112

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

0.9 —- — — _, —~4

 

 

 

 

0.7 ——
0.6 .
0.5 « er 7
0.4 <
0.3 4 , ,,
0.2 1‘ 7 ~
0-1 4 i' \
1 2 3 4 6 B
Successional Stage

SuitabIIity Index

 

 

 

Figure A84. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-balsam poplar-birch in early
successional stages (1-2; ages < 30 years), aspen-balsam poplar-black ash in middle
stages (3-4; ages 30-100 years), and black ash-cedar in later stages (5-6; ages > 100
years).

113

 

SuitabIIIty Index

 

 

1 2 3 4 5 6
Succeaeional Stage

SuitabIIIty Index

 

 

1 2 3 4 5 6
Successional Stage

 

 

C. 1

0.9 \
oa -— \
0.7 «
0.0 4
0.5 «
0.4 «
0.3 ..
0.2 4 , - ~
0.1 .

 

 

Suitability index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-5. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), balsam fir-white spruce-red maple in middle stages (3-4;
ages 30-100 years), and hemlock-cedar-balsam fir-red maple in later stages (5-6; ages >
100 years).

“4

 

 

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

Suitability Index

 

Successional Stage

 

0.9 -
0.9 -— ____
0.7 ,
0.9
0.5«
0.4 9 ,
0.3 « A v—A
0.2 A i

0.1 4 7, ~ ~~~~~~

 

 

 

 

Suitability Index

 

 

 

 

Successional Stage

Figure A8-6. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), black spruce-balsam fir-cedar in middle stages (3-4; ages
30-100 years), and cedar-black spruce-balsam fir in later stages (5-6; ages > 100 years).

115

Suitability Index

 

 

1 2 3 4 5 6
Successional Stage

 

 

Suitability Index

 

 

 

1 2 3 4 5 6
Successional Stage

 

0.9
0.9
0.7 -
0.9 « ,9
0.5 «. 7
0.44
0.3 9 - i
0.2 ~~ . ,0 , , g g
0.1 - - x _-
0 . . . ,
1 2 3 4 5 s

Successional Stage

 

 

Suitability Index

 

 

 

 

Figure A8-7. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), red maple-beech-white ash-basswood-white pine in middle
stages (34; ages 30-100 years), and sugar maple-beech-hemlock in later stages (5-6; ages
> 100 years).

116

 

 

Suitability Index

 

 

1 2 3 4 5 6
Successional Stage

 

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

0.9 \

Suitability Index
o
in

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-8. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), red maple-beech-white pine in middle stages (34; ages 30-
100 years), and white pine-red maple-beech-sugar maple-white ash in later stages (5-6;
ages > 100 years).

117

 

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

Suitability Index

 

Succeeaional Stage

C. 1

0.9 \ ——i

0.8 1———- \ -
0.7 ~ ,
0.6 .
0.5 «~—
0.4 1 ~
0.3 « A 7
0.2 . ,
0.1 1 , . ,,

 

 

Suitability Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-9. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), sugar maple-beech-basswood-red maple-white ash in
middle stages (3-4; ages 30-100 years), and sugar maple-beech in later stages (5-6; ages >
100 years).

118

Suitability Index

 

1 2 3 4 5 6
Sueeeeeional Stage

Suitability Index

 

1 2 3 4 5 6

Succeeeional Stage

 

 

0.9
0.8
0.7 1 ~
0.6 .
0.5 -
0.4 1 ,
0.3 .
0.2 .
0.1 ~

 

Suitability Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-lO. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), sugar maple-beeeh-basswood-red maple-white ash in
middle stages (34; ages 30-100 years), and sugar maple-beech-basswood in later stages
(5-6; ages > 100 years).

119

Suitability Index

 

1 2 3 4 5 6
Successional Stage '

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

Suitability index

 

 

 

 

Successional Stage

Figure A8-11. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), sugar maple-oak-beech in middle stages (3-4; ages 30-100
years), and sugar maple-beech-hemlock in later stages (5-6; ages > 100 years).

120

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

 

0.9
0.8
0.7 ~-
0.61-~
0.5-
0.4-
0.3 ”a n
0.2.-. a -. ,7 ,, 7
0.1-1 -*—- , 7* ,, ~~ , ,2". *7 1A,-

 

J

Suitability Index

 

 

 

 

Successional Stage

Figure A8-12. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen—birch in early successional
stages (1-2; ages < 30 years), white pine—red maple-beech in middle stages (34; ages 30-
100 years), and white pine-red maple-beech-sugar maple in later stages (5-6; ages > 100
years).

121

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

Suitability Index

 

1 2 3 4 5 6

Succeeeional Stage

 

 

 

07‘ S
O

00»
C

04»
0.3.,-,,--._,. , ._ ,2 -1,

01 A -7 w

0.1+ W» — A- - ~ ,9

Suitability index
O
CI
1
l

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-l3. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), black spruce-balsam fir-cedar in middle stages (3-4; ages
30-100 years), and cedar-black spruce-balsam fir in later stages (5-6; ages > 100 years).

122

 

Suitability Index

 

 

1 2 3 4 5 6
Successional Stage

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

 

 

0.9
0.7 - ,

0.9«» - 7.
0.5« — S a
0.44
0.3 4»
0.2 .._. s

0.19 a- S-

 

Sultablllty Index

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-14. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-jack pine in early successional
stages (1-2; ages < 30 years), jack pine-red pine in middle stages (3-4; ages 30-100
years), and red oak-red maple-white pine in later stages (5-6; ages > 100 years).

123

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

Suitability index

 

1 2 3 4 5 6
Successional Stage

 

 

 

 

0.9
0.9 —-
0.7 4
0.0 4
0.54
0.44
0.3 44 -
0.24 4- 4 44
0.1 4 ~ - 44 4 4 -

 

Suitability Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-15. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-oak in early successional
stages (1-2; ages < 30 years), oak-red maple-white pine-black cherry in middle stages (3-
4; ages 30-100 years), and white pine-red maple-beech-sugar maple in later stages (5-6;
ages > 100 years).

124

Suitability Index

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6
Successional Stage
B.
X
3
5
g
3
a
a
in
1 2 3 4 5 6
Successional Stage
C. 1
0.9 _
0.8
g 0.74
.. 0.644 4 4 a
$0.54-- a
a cb—vrr- --——————_—+~‘__.t4-—A. . v . , A, ,2
g 0.4
m 0.3 44 ._,_ - e __- . a, s
0.2 4~~—~4-444 4 44 u“ to SE. ”A, - naught
0.14 #_ 4. -2“
0 .
1 2 3 4 5 6

Successional Stage

Figure A8-l6. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush in early successional
stages (1-2; ages < 30 years), black spruce-tamarack-balsam fir-cedar-black ash in middle
stages (3-4; ages 30-100 years), and black spruce in later stages (5-6; ages > 100 years).

Sultablllty Index

 

1 2 3 4 5 6
Successional Stage

Sultablllty Index

 

1 2 3 4 5 6

Successional Stage

 

 

0.9
0.8
0.74;» _ , ~

0.6.. _,__ -..7. -7 ,, 7,, .7 _ , . .7 7. .
o.5~»—n~—-- m m ——~~—A fl , ~ «g
0.4 - ~ g -,
0.3.,__-._ -_,u_ WW W
0.2 .a ._. -_- _ *7 -- i «A, ,, A
0.1%— a» ,- >7 ~ MM... ”Hm a

 

 

SuItabIIIty Index

 

 

 

 

Successional Stage

Figure A8—l7. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support jack pine-white pine-red pine-oak in
early successional stages (1-2; ages < 30 years), jack pine-white pine-red pine-red maple-
black cherry in middle stages (3-4; ages 30-100 years), and white pine-red maple in later
stages (5-6; ages > 100 years).

126

SultabIIIty Index

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6
SucceeaionaIStage
B.
K
O
a
.5
5
3
:2
3
tn
1 2 3 4 5 6
Successional Stage
C. 1
0.9 —
0.8
x y__ ,, 7
«S 0.7
E 0.6« 777—7- 7 7 7i 7 7
£05.77 77 7 7 7 7 77 77
n
g 0.4« -——77 7 7 7 7 77 77 ~
g 0,3._n-.~_.._ _-_- _. .-.._ - -__ . ._ _
0.2 W 777fi— 77v ~777*-~7 v
0.1 ~~ _ 7* ~ 77 7 ~77 7 7m
0 .
1 2 3 4 5 6

Successional Stage

Figure A8-18. Suitability curves for fall and winter food (A), thermal cover m), and
spring and summer habitat (C) for areas that support tamarack-black ash-birch in early
successional stages (1-2; ages < 30 years), tamarack-black ash-red maple in middle stages
(3-4; ages 30-100 years), and cedar-black spruce-balsam fir in later stages (5-6; ages >
100 years).

127

 

Suitability Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

 

 

 

 

 

 

B.
X
0
1:
5
g
3
£
a
m
1 2 3 4 5 6
Successional Stage
C. 1
0.97———-
0.8 77-4
g 0.747
5 .
g .
3
g e
3 a
m .
0 7 .
1 2 3 4 5 6

Successional Stage

Figure A8-19. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush in early successional
stages (1-2; ages < 30 years), black spruce-tamarack-balsam fir—eedar-white pine in
middle stages (3-4; ages 30-100 years), and black spruce-tamarack in later stages (5-6;
ages > 100 years).

128

SultabIIIty Index

 

1 2 3 4 5 6
Successional Stage

SultabIIIty Index

 

1 2 3 4 5 6
Successional Stage

 

 

 

 

 

0.3 ~_——————--—-
0.7 7 7

0.67
0.5 7
0.4 . 7-
0.37 77 7

SuItabIlIty Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-20. Suitability curves for fall and winter food (A), thermal cover (B). and
spring and summer habitat (C) for areas that support lowland brush-aspen-birch-balsam
poplar-black ash in early successional stages (1-2; ages < 30 years), cedar-black spruce-
balsam fir in middle stages (3-4; ages 30-100 years), and cedar-hemlock in later stages
(5-6; ages > 100 years).

129

Suitability index

 

1 2 3 4 5 6
Successional Stage

 

Suitability Index

 

 

 

1 2 3 4 5 6
Successional Stage

 

0.9 7 7— —————~
0.3 7 - *_W__,
0.7-
0.6-
0.5«
0.47
0.37
0.2
0.1 «

 

 

 

 

Suitability Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-21. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush-birch-tamarack-black
ash in early successional stages (1-2: ages < 30 years), cedar-black spruce-balsam fir in
middle stages (3-4; ages 30-100 years), and cedar-black spruce-balsam fir in later stages
(5-6; ages > 100 years).

130

Suitability Index

 

1 2 3 4 5 5
Successional Stage

Suitability index

 

1 2 3 4 5 5
Successional Stage

 

 

0.9 _
on
0.7 7
0.6»7
ML
0.47 _ ,

0.3.w A- ”A--- ._ S _ i g

 

Suitability Index

 

0.1- v H7 -7- a -, ., ,-

 

 

 

Successional Stage

Figure A8-22. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush-birch-tamarack-biack
ash in early successional stages (1-2; ages < 30 years), tamarack-black ash-red maple in
middle stages (3-4; ages 30-100 years), and cedar-black spruce-balsam fir in later stages
(5-6; ages > 100 years).

131

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

X
3
.E
.4?
3
5
3
(D
SucceeeionalStage
B.
X
0
1:
E
E
3
9!
3
m
1 2 3 4 5 5
Successional Stage
C. 1

0.9 _.

0.8 ——_____ __-___
x .1
'8 0.7
5 0.6- 7
50.57 7 , , _,
a
g 0.4
30;» -7 7 , -, ,,

0.2«~7~—77 ., ,

0.1477~ - , W- A

o 1
1 2 3 4 5 5

Successional Stage

Figure A8-23. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support shrubs—grass in early successional
stages (1-2; < 30 years), jack pine-red pine-white pine-oak in middle stages (3-4; 30-100
years), and white pine-red pine-oak-jack pine in later stages (5-6; > 100 years).

132

 

 

0.9
0.8 777-7- - —
0.7
0.6 77— 77—77—7-
o.5 \ “hm
0.4 77 777
0.3 7 __
0.2 7

0.1 77

 

 

 

 

 

 

Suitability Index

 

 

 

 

 

1 2 3 4 5 6
Successional Stage

 

 

 

Suitability index

 

1 2 3 4 5 6

Successional Stage

 

 

 

 

037-7—777—77777-7
0.77
0.67
0.57
0.47
0.3777 777 77 77 7

0.2 77777777 777
0.17777

Suitability Index

 

 

 

 

1 2 3 4 5 6
Successional Stage

Figure A8-24. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support tamarack-black ash-birch-red maple
in early successional stages (1-2; < 30 years), tamarack-black ash-birch-red maple-cedar
in middle stages (3-4; 30-100 years), and cedar-black spruce-balsam fir in later stages (5-
6; > 100 years).

133

Appendix 9. White-tailed deer habitat suitability curves for each successional trajectory
in the Upper Peninsula, Michigan. The curves display habitat dynamics for fall and
winter food, thermal cover, and spring and summer habitat components based on habitat

potential models developed in this document.

134

Suitability Potential

 

Suitability Potential

 

1 2 3 4 5 6
Seral Stage

 

 

 

 

 

 

Suitability Potential
0
0|

 

 

 

Seral Stage

Figure A9-1. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen in early successional stages
(1-2; ages < 30 years), sugar maple-red maple-ironwood-basswood-yellow birch in
middle stages (3-4; ages 30-100 years), and sugar maple-hemlock in late stages (5-6; ages
> 100 years).

135

 

Suitability Potential

 

 

 

 

Suitability Potential

 

 

 

0.9 ——7-77 7
0.817—-— a, .
0.7 77 7 ,7 _-;*,__-d_ .__._
0.6 4
0.5 «
0.4 ~

 

 

 

 

 

Suitability Potential

02 7- 7 77 77
0.17 7 .. _-__

 

 

 

 

 

Seral Stage

Figure A9-2. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-balsam poplar in early
successional stages (1-2; ages < 30 years), white ash-black ash-American elm in middle
stages (3-4; ages 30-100 years), and white ash-red maple-sugar maple-black ash in later
stages (5-6; ages > 100 years).

136

Suitability Potential

 

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6
SeralStage
B.
Tu
z
3
e
a.
5
3
.§
3
U)
1 2 3 4 5 6
SeraIStaue
C. 1
0.9 W
_ 0.8 _
§ 0.7 W
g 0.67
0.57
E
3 0.4- 7 7 7
£0.3777 7777— A 7 ,_ , r r-
3
in 0.2 77 777 77 7777 77 7 77
0.177 7 7 -- .7 ,-
0 .
1 2 3 4 5 6
SeralStage

Figure A9—3. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-balsam poplar-birch in early
successional stages (1-2; ages < 30 years), balsam fir-white spruce-cedar-black spruce in
middle stages (34; ages 30-100 years), and hemlock-red maple-sugar maple in later
stages (5-6; ages > 100 years).

137

Suitability Potential

 

 

 

 

 

 

 

 

 

 

 

B.
E
a
8
g
.9
£
5
C.
=7:
3
8
g
a
.5. .
3 .
'o .
1 2 3 4 5 e

Seral Stage

Figure A974. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush-aspen-balsam poplar-
birch in early successional stages (1-2; ages < 30 years), balsam fir-white spruce-cedar-
red maple in middle stages (3-4; ages 30-100 years), and hemlock-red maple-sugar maple
in later stages (5-6; ages > 100 years).

138

 

 

 

 

 

 

 

 

 

 

 

'2'
§
g a
3
Q
5
SeraiStage
B.
‘8 .
a.
E a
75 .
£ .
5 .
1 2 3 4 5 6
SeraIStaae
C.
t .
a.
g e
.e
a. .
.3
.0 . . . .
1 2 3 4 5 6

Seral Stage

Figure A9-5. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), sugar maple-white pine-red maple-yellow birch-red oak-
balsam fir in middle stages (34; ages 30-100 years), and hemlock-red maple in later
stages (5-6; ages > 100 years).

139

Suitability Potential

 

 

Suitability Potential

 

 

 

 

Suitability Potential

  

 

 

 

1 2 3 4 5 6
Seral Stage

Figure A9-6. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), sugar maple-yellow birch-red maple-ironwood-balsam fir-
white spruce in middle stages (34; ages 30-100 years), and sugar maple-hemlock in later
stages (5-6; ages > 100 years).

140

 

 

 

 

 

 

 

 

 

 

T-
a
C
t
a.
g
3
£
3
m
1 2 3 4 5 6
B.
T-
a:
C
t
E
3
£
3
in
1 2 3 4 5 6
SeraIStacle
C.
3
C
t
it
5
.‘5
a
3
0.1 77 7 7
0 7
1 2 3 4 5 6

Seral Stage

Figure A9-7. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-birch in early successional
stages (1-2; ages < 30 years), white pine-red pine-balsam fir-white spruce in middle
stages (3-4; ages 30-100 years), and red maple-red oak in later stages (5-6; ages > 100
years). 7

141

Suitability Potential

 

Seral Stage

 

 

Suitability Potential

 

 

 

 

Suitability Potential

  

 

 

 

1 2 3 4 5 6
Seral Stage

Figure A9-8. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-sugar maple in early
successional stages (1-2; ages < 30 years), sugar maple-basswood-red maple-beech in
middle stages (34; ages 30-100 years), and sugar maple in later stages (5-6; ages > 100
years).

142

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Til
a
C
t
a.
g
3
t9.
3
to
1 2 3 4 5 6
Seral Stage
B.
TI
8
C
t
a.
S
3
§
3
m
C.
3
1:
S
a.
g
3
£
3
in
0 .
1 2 3 4 5 6

Seral Stage

Figure A9-9. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-sugar maple in early
successional stages (1-2; ages < 30 years), sugar maple-basswood-red maple—yellow
birch-ironwood in middle stages (3-4; ages 30-100 years), and hemlock-sugar maple in
later stages (5-6; ages > 100 years).

143

Suitability Potential

 

 

Suitability Potential

 

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6
SeralStaoe
C 1
0.9 \
— 0.6 \
i 0.7
g 0.67 , 7 7
g 0.5 "
3 04*
t! 03777
3
e) 0.2- 77 7 7 77
0.1 7 77 7 7 7 7 7
0 . . . .
1 2 3 4 5 6

Seral Stage

Figure A9-10. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-sugar maple in early
successional stages (1-2; ages < 30 years), sugar maple-red maple-yellow birch-ironwood
in middle stages (34; ages 30-100 years), and sugar maple-hemlock in later stages (5-6;
ages > 100 years).

144

 

Suitability Potential

 

 

 

 

 

 

 

 

 

 

SeraIStage
B.
'3
“E
‘3
a.
g
3
.8.
3
it)
C.
T:
33
t
a.
E
3 a
Q . 7
3
m .
0 .
1 2 3 4 5 6

Seral Stage

Figure A9-11. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support aspen-sugar maple-white ash in
early successional stages (1-2; ages < 30 years), sugar maple-basswood-ironwood—beech-
white abs-American elm in middle stages (3-4; ages 30-100 years), and sugar maple in
later stages (5-6; ages > 100 years).

145

 

Suitability Potential

 

 

 

 

 

 

 

 

 

 

 

 

 

B.
T:
a
C
3
O
a.
s
3
£
3
in
SeralStaee
C. 1
Ta
3
C
’5
a.
1.?
3
£
3
m -. ,_ ”Whoa - m , -
0.1+“ 7 777777
0 I
1 2 3 4 5 6

Seral Stage

Figure A9-12. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush in early successional
stages (1-2; ages < 30 years), lowland brush-birch-red maple-balsam poplar in middle
stages (3-4; ages 30-100 years), and black spruce-cedar-hemlock in later stages (5-6; ages
> 100 years).

146

 

 

 

 

 

 

 

 

 

 

 

T-
:3
é
a.
E
3
2
3
to
SeraIStage
B.
3
e
t
a
E
3
Q
3
to
SeralStaee
C. 1
0.9
7' 0.8
=3 0.7
$057777 7 777 7 7 7 77 77 77 777 71
£05777 -—~-- 7
3 0.477777777777-7 77 77 7 7
50.3 77 7 -7-“ -_. - b ,
3
.. .1...._\ .. M S
0.1777 777- 7777 777 7 77 _ 7 , , 5-3.
0 .
1 2 3 4 5 6
SeraIStage

Figure A9-13. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush in early successional
stages (1-2; ages < 30 years), tamarack-black ash-birch-red maple in middle stages (3-4;
ages 30-100 years), and cedar-black spruce-balsam fir in later stages (5-6; ages > 100
years).

147

 

 

 

 

 

 

 

 

 

 

 

 

T-
a
C
S
o
a
E
3
£
a
to
B.
3
a
c
‘3
m
E
3
£
a
in
1 2 3 4 5 6
Seral Stage
C. 1
3
c
S
a.
E
3 ,
g _ B
a
to _ W
0 7 7
1 2 3 4 5 6

Seral Stage

Figure A9-l4. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support lowland brush-birch-balsam poplar
in early successional stages (1-2; ages < 30 years), cedar-black spruce-balsam fir in
middle stages (3-4; ages 30-100 years), and cedar-hemlock in later stages (5-6; ages >
100 years).

148

Suitability Potential

 

Seral Stage

Suitability Potential

 

Seral Stage

 

 

0.9
0.8
0.7
0.6. ._ _ _. . .. .-__ _ __ _ _.

0.5 7777777- 77777777777777 7 7 77 7
0.4
0.377 7
0.277 77 77
0.1.. _ .2 ._ ._ __ __ . .. ._,.- a..-_.._

 

 

 

Suitability Potential

 

 

 

 

Seral Stage

Figure A9-15. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support shrubs-grass in early successional
stages (1-2; ages < 30 years). jack pine in middle stages (3-4; ages 30-100 years), and
jack pine-red pine in later stages (5-6; ages > 100 years).

149

 

 

 

 

 

 

 

 

 

 

 

 

3 a
E .
a .
g o
3 a
g .
3
to
1 2 3 4 5 6
SeralStage
B.
ii
a:
5
e
at
E
3
a
3
in
1 2 3 4 5 6
SeralStaee
C. 1
T-
a
E
e
a
E
3
£ .
3
w .
0 7 . 7
1 2 3 4 5 6

Figure A9-16. Suitability curves for fall and winter food (A), thermal cover (B), and
spring and summer habitat (C) for areas that support sphagnum-lowland brush in early
successional stages (1-2; ages < 30 years), black spruce-tamarack-balsam fir-cedar in
middle stages (3-4; ages 30-100 years), and black spruce-tamarack in later stages (5-6;
ages > 100 years).

150

LITERATURE CITED

Albert, D. A. 1995. Regional landscape ecosystems of Michigan, Minnesota, and
Wisconsin: a working map and classification. Gen. Tech. Rep. NC-l 78. St. Paul,
MN: U.S. Department of Agriculture, Forest Service, North Central Forest
Experiment Station. 250pp.

Albert, D. A., S. R. Denton, B. V. Barnes. 1986. Regional Landscape ecosystems of
Michigan. School of Natural Resources, University of Michigan, Ann Arbor,
Michigan. 32pp.

Albrecht, W. A. 1944. Soil fertility and wildlife—cause and effect. North American
Wildlife Conference. 9:19-28.

Armbruster, M. J ., R. S. Lunetta, J. B. Haufler, R. J. Seppala, and R. A. Steinbach. 1987.
A white-tailed deer habitat model for agricultural areas in the Saginaw River

Basin. Drafi Tech. Rep.. Waterways Exp. Stat., Corps of Engineers. Vicksburg,
MS.

Barbour, M.G., J. H. Burk, W.D. Pitts. F. S. Gilliam, and M. W. Schwartz. 1999.
Terrestrial plant ecology, 3rd ed. Benjamin/Cummings. Menlo Park, CA. 649pp.

Bender, LC, and J .B. Haufler. 1987. A white-tailed deer HSI for the Upper Great Lakes
region. Michigan State University, Department of Fisheries and Wildlife, East
Lansing. Michigan, unpubl.

Blouch, R. I. 1984. Northern Great Lakes States and Ontario forests. Pages 391-410 in
Halls, L. K., R. E. McCabe, and L. R. Jahn (eds.). White-tailed deer ecology and
management. Stackpole Books. Harrisburg. 870pp.

Braun, K. F. 1996. Ecological factors influencing white-tailed deer damage to
agricultural crops in northern Lower Michigan. MS. Thesis, Department of
Fisheries and Wildlife, Michigan State University, East Lansing, MI. 258pp.

Bryant, R. L. 1963. The lowland hardwood forests of Ingham County, Michigan: their
structure and ecology. Ph.D. Dissertation, Department of Forestry, Michigan State
University, East Lansing, MI. 213pp.

Burger T. L. and J. Kotar. 1999. Field guide to forest habitat types of northern lower
Michigan, Department of Forest Ecology and Management, University of
Wisconsin-Madison, Madison, WI.

Cheatum, E. L. and C. W. Severinghaus. 1950. Variations in fertility of white-tailed

deer related to range conditions. Transactions of the North American Wildlife
Conference 15: 170-190.

151

Ching, K. K. 1954. Forest succession on the well-drained soil in the Higgens Lake area
of Michigan. PhD. Dissertation, Department of Forestry, Michigan State College,
East Lansing, MI. 143pp.

Christisen, D. M. and L. J. Korschgen. 1955. Acorn yields and wildlife usage in
Missouri. North American Wildlife Conference 20: 337-356.

Cleland, D. T., J. B. Hart. G. E. Host, K. S. Pregitzer, and C. W. Ram. 1993. Field
Guide: ecological classification and inventory system of the Huron-Manistee
National Forests. U. S. Department of Agriculture, Forest Service.

Coffman, M. S., E. Alyanak, and J. Ferris. 1980. Field guide to habitat type

classification system for Upper Peninsula of Michigan. School of Forestry and
Wood Products, Michigan Technological University, Houghton, MI. 140pp.

Comer, P. J ., D. A. Albert, H. A. Wells, B. L. Hart, J. B. Raab, D. L. Price, D. M.
Kashian, R. A. Corner, and D. W. Schuen. 1995. Michigan’s presettlement
vegetation. as interpreted from the General Land Office Surveys 1816-1856.
Michigan Natural Features Inventory, Lansing. MI. digital map and database.

Crawford, B. T. 1950. Some specific relationships between soils and wildlife. Journal of
Wildlife Management 14:115-121.

Daubenmire, R. 1966. Vegetation: Identification of typal communities. Science
151:291-298.

Davenport, L. 1941. Deer nutrition studies. Michigan Department of Natural Resources,
Wildlife Division. Unpubl. data.

Davenport, L. A., D. F. Switzenberg, R. C. vanEtten, and W. D. Burnett. 1953. A study
of deeryard carrying capacity by controlled browsing. North American Wildlife
Conference. 1 8: 581-596.

DeLain, C. J ., R. A. comer, D. A. Albert, and M. B. Austin. 1999. Landtype association
of northern Michigan Section VII. Michigan Natural Features Inventory, Lansing,
MI. Digital map and database.

Denney, A. H. 1944. Wildlife relationships to soil types. North American Wildlife
Conference. 9: 316-323.

Dodge, S. 1987. Presettlement forest of south-central Michigan. The Michigan Botanist
26:139-153

Dodge, S. and J. R. Harman. 1985. Woodlot composition and successional trends in
south-central Lower Michigan. The Michigan Botanist 24:43-54.

152

Dodge, W. B. 2002. Population dynamics of moose in the western Upper Peninsula of
Michigan 1999-2001. M.S. Thesis, Department of Fisheries and Wildlife,
Michigan State University, East Lansing, MI.134pp.

Doepker, R. V. and J. J. Ozoga. 1991. Wildlife values of northern white-cedar. Pages
15-34 in D. O. Lantagne (ed.). Northern white-cedar in Michigan. Workshop
Proceedings, Ag. Exp. Sta. Res. Rep. 512.

Duvendeck, J. P. 1964. The value and prediction of acorn crops for deer. Ph.D.
Dissertation, Department of Fisheries and Wildlife, Michigan State University,
East Lansing, MI. 83pp.

Euler, D. and L. Thurston. 1980. Characteristics of hemlock stands related to deer use in
East-Central Ontario. Journal of Applied Ecology 17: 1-6.

Eyre. F. H. 1980. Forest cover types of the United States and Canada. Society of
American foresters, Washington, D. C. l48pp.

Felix, A. B., H. Campa, 111, K. F. Millenbah, S. L. Panken, S. R. Winterstein, and W. E.
Moritz. Applications of using landscape-scale models to quantify white-tailed
deer habitat potential in Michigan, U.S.A. Proceedings from the 25th Congress of
the lntemational Union of Game Biologists. In Press.

Felix, A. B., B. Hughey. M. S. Garner. H. Campa, 111, K. F. Millenbah. S. R. Winterstein,
and W. E. Moritz. 2002. Implications for using landscape-scale habitat potential
models to understand white-tailed deer spatial structure. TWS 9th Annual
Conference, Bismarck, ND. Abstract.

Ford, W. M., A. S. Johnson, P. E. Hale, and J. M. Wentworth. 1997. Influence of forest
type, stand age, and weather on deer weights and antler size in the southern
Appalachians. Southern Journal of Applied Forestry 21: 11-18.

Franzreb, K. E, F inch, D. M., P. B. Wood, and D. E. Capen. 2000. Management
strategies for the conservation of forest birds. 2000. Pages 1-10 in Bonney, R.,
D. N. Pashley, R. J. Cooper, and L. Niles (eds.). Strategies for bird conservation:
the Partners in Flight planning process. Rocky Mountain Research Station,
Ogden, UT. 281pp.

Garner, M. S. 2001. Movement patterns and behavior at winter feeding and fall baiting
stations in a population of white-tailed deer infected with bovine tuberculosis in
the northeastern Lower Peninsula of Michigan. Ph.D. Dissertation, Department of
Fisheries and Wildlife, Michigan State University, East Lansing, Mich. 270pp.

Gevorkiantz, S. R. and W. A. Duerr. 1939. Volume and yield of northern white cedar in
the Lake States. Lake States Forest Experiment Station. 55pp.

153

Gladfelter, H. L. 1984. Midwest agricultural region. Pages 427-440 in Halls, L. K., R. E.
McCabe, and L. R. J ahn (eds.). White-tailed deer ecology and management.
Stackpole Books. Harrisburg. 870pp.

Goodrum, P.D., V. H. Reid, and C. E. Boyd. 1971. Acorn yields, characteristics, and
management criteria of oaks for wildlife. Journal of Wildlife Management 35:
520-532.

Graham, S. A., R. P. Harrison, Jr., and C. E. Westell, Jr. 1963. Aspens, phoenix trees of
the Great Lakes Region. University of Michigan Press, Ann Arbor, MI. 272pp.

Grumbine, R. E. 1994. What is ecosystem management? Conservation Biology 8: 27-
38.

Gunderson, L. H., C. S. Holling, S. S. Light. 1995. Barriers and bridges to the renewal
of ecosystems and institutions. Columbia University Press, New York. 593pp.

Haber, W. 1990. Using landscape ecology in planning and management. Pages 217-232
in Zonneveld, I. S. and R. T. T. Forrnan (eds.). Changing landscapes: an
ecological perspective. Springer-Verlag, New York. 286pp.

Hanaburgh, C. 2001. Modeling the effects of management approaches on forest and
wildlife resources in northern hardwood forests. Ph.D. Dissertation, Department
of Fisheries and Wildlife, Michigan State University, East Lansing, MI. 269pp.

Harlow, R. F., J. B. Whelan, H. S. Crawford, and J. E. Skeen. 1975. Deer foods during
years of oak mast abundance and scarcity. Journal of Wildlife Management 39:
330-336.

Harper, R. M. 1918. The plant population of northern Lower Michigan and its
environment. Bulletin of the Torrey Botanical Club 45: 23-42.

Haufler, J. B., R. K. Baydack, H. Campa, III, B. J. Kemohan, C. Miller, L. J. O’Neil, L.
Waits. 2002. Performance measures for ecosystem management and ecological

sustainability. Wildlife Society Technical Review 02-1, 33pp.

Host, G. E., K. S. Pregitzer, C. W. Ramm, J. B. Hart, and D. T. Cleland. 1987.
Landform-mediated differences in successional pathways among upland forest
ecosystems in northwestern Lower Michigan. Forest Science 33: 445-457.

Jackson, L. J ., A. S. Trebitz, and K. L. Cottingham. 2002. An introduction to the
practice of ecological modeling. BioScience 50:694-706.

Johnson, J. E., C. L. Haag, J. G. Bockheim, and G. G. Erdmann. 1987. Soil-site
relationships and soil characteristics associated with even-aged red maple (A cer

154

rubrum) stands in Wisconsin and Michigan. Forest Ecology and Management 21:
75-89.

Johnston, W. F. 1990. Thuja occidentalis L., Northern white-cedar. Pages 580-589 in
Burns, R. M. and B. H. Honkala, tech coords. Silvics of North America:
Hardwoods. Agriculture Handbook 654. U.S. Department of Agriculture, Forest
Service, Washington, DC. Vol 1, 877pp.

Knapp, B. D. 1993. Soil survey of Presque Isle County, Michigan. U.S. Department of
Agriculture, Soil Conservation Service. 252 pp.

Koerper, G. J. and C. J. Richardson. 1979. Biomass and net annual primary production
regression for Populus grandidentata on three sites in northern Lower Michigan.
Canadian Journal of Forest Research 10: 92-101.

Kohn, B. E. and J. J. Mooty. 1971. Summer habitat of white-tailed deer in north-central
Minnesota. Journal of Wildlife Management 476-487.

Kotar, J. and T. L. Burger. 2000. Field guide to forest habitat type classification for
north central Minnesota. Terra Silva Consultants, Madison, WI.

Kotar, J. and M. Coffman. 1984. Habitat-type classification system in Michigan and
Wisconsin. Pages 100-103 in Bockheim, J .G. (ed.). Forest land classification:
experiences, problems, perspectives: proceedings of the symposium. University of
Wisconsin, Madison, WI. 276pp.

Ku, T. 1954. Forest succession on the poorly drained soils in the Higgins Lake area of
Michigan. Ph.D. Dissertation, Department of Forestry, Michigan State College,
East Lansing, MI. 196 pp.

Lake States Forest Experiment Station. 1940. White cedar for deer food. U.S.
Department of Forestry, St. Paul, Minn. Tech. Note 159.

Langenau, E. E. 1994. 100 years of deer management in Michigan. Michigan
Department of Natural Resources, Wildlife Division Report No. 3213. 15pp.

Larsen, W. C. 1982. Structure, biomass, and net primary productivity for an age-
sequence of jack pine ecosystems. Ph.D. Dissertation, Department of Botany and
Plant Pathology, Michigan State University, East Lansing, IVII. 140pp.

Livingston, B. E. 1905. The relation of soils to natural vegetation in Roscommon and
Crawford Counties, Michigan. Botanical Gazette 39: 22-41.

155

Mackey, T. 1996. White-tailed deer movements, habitat use, and browsing effects on
vegetation in the Upper Peninsula of Michigan. M.S. Thesis, Department of
Fisheries and Wildlife, Michigan State University, East Lansing, MI 130pp.

MacLean Consultants, Ltd. 1995. Inventory of deer habitat in Michigan’s Upper
Peninsula using landsat thematic mapper data. MacLean Consultants, Ltd.,
Houghton, MI. digital map.

Marty, R. and V. Rudolph. 1970. A guide to oak management. Society of American
Foresters, Lower Michigan Chapter, Department of Forestry, Michigan State
University, East Lansing, MI. 22pp.

Mautz, W. W. 1978. Sledding on a brushy hillside: the fat cycle in deer. Wildlife
Society Bulletin 6:88-90.

McCaffery, K. R., J. Tranetzki, and J. Piechura. 1974. Summer foods of deer in northern
Wisconsin. Journal of Wildlife Management 38: 215-219.

Merkel, D. M. 1988. Soil nutrients in glaciated Michigan landscapes: distribution of
nutrients and relationships with stand productivity. Ph.D. Dissertation.
Department of Forestry, Michigan State University, East Lansing, MI. 126pp.

Michigan Department of Natural Resources. 1999. 1978 Landuse/Cover Arc Coverage.
Michigan Department of Natural Resources, Lansing, MI. digital map.

Michigan Department of Natural Resources. 2002. Ecosystem based management.
www.mighigan.gov/dnr/0,l607,7-153-10366_1 1865-24426-,00.htm1. Accessed
DecemberlO, 2002.

Michigan Department of Natural Resources-Wildlife Bureau, Michigan Department of
Natural Resources-Land and Mineral Services Division, Resource Mapping and
Aerial Photography. 2000. Gap Northern Lower Peninsula Land Cover.
Michigan Department of Natural Resources, Lansing, MI. digital map.

Michigan Weather Service. 1974. Climate of Michigan by stations, 2“‘1 ed. Michigan
Department of Agriculture, Michigan Weather Service, cooperative with NOAA-
National Weather Service, U.S. Department of Commerce. East Lansing, MI.

Morton, G. H. and E. L. Cheatum. 1946. Regional differences in breeding potential of
white-tailed deer in New York. Journal of Wildlife Management 10: 242-248.

Mysterud. A. and E. Ostbye. 1999. Cover as a habitat element for temperate ungulates:
effects on habitat selection and demography. Wildlife Society Bulletin 27: 385-
394.

156

Nichols, G. E. 1935. The hemlock-white pine-northern hardwood forest region of
eastern North America. Ecology 16: 403-422.

Nixon, C. M., M. W. McClain, and K. R. Russell. 1970. Deer food habits and range
characteristics in Ohio. Journal of Wildlife Management 34: 870-886.

Noss, R. F. 1990. Indicator for monitoring biodiversity: a hierarchical approach.
Conservation Biology 4:355-364.

Noss, R. F., E. T. LaRoe, 111, J. M. Scott. 1997. Endangered ecosystems of the United
States: a preliminary assessment of loss and degradation. National Biological
Service, Technical Report BSR9501, Washington D. C. 80pp.

Nutter, M. D. 1973. Soil texture and forest change in northern Lower Michigan. M.A.
Thesis, Department of Geography, Michigan State University, East Lansing, MI.

85pp.

Ozoga, J. J. 1968. Variations in deeryard microclimate. Journal of Wildlife
Management 32: 574-585.

Ozoga, J. J. and L. W. Gysel. 1972. Response of white-tailed deer to winter weather.
Journal of Wildlife Management. 36: 892-896.

Rogers, L. L., J. J. Mooty, and D. Dawson. 1981. Foods of white-tailed deer in the

Upper Great Lakes Region—a review. USDA For. Serv. Gen. Tech. Rep. NC-
65.24pp.

Rogers, R. 1990. Quercus alba. White oak. Pages 605-613 in Burns, R. M. and B. H.
Honkala, tech. Coords. Silvics of North America: 2. Hardwoods. Agriculture
Handbook 654. U. S. Department of Agriculture, Forest Service, Washington,
DC. vol 2, 877pp.

Rowe, J. S. 1969. Plant community as a landscape feature. Pages 63-81 in Essays in
plant geography and ecology. Symposium in terrestrial plant ecology, Francis
Xavier University, Oct. 1966. Nova Scotia Museum, Halifax, NS.

Ryel, L. A. 1953. Effects of forest age and grth on the availability of forage for deer
in white cedar swamps. M.S. Thesis, Department of Fisheries and Wildlife,
Michigan State College, East Lansing, MI. 99pp.

Sander, I. L. 1990a. Quercus rubra. Northern red oak. Pages 727-733 in Burns, R. M.
and B. H. Honkala, tech. Coords. Silvics of North America: 2. Hardwoods.
Agriculture Handbook 654. U. S. Department of Agriculture, Forest Service,
Washington, D. C. vol 2, 877 pp.

157

Sander, I. L. 1990b. Quercus velutina. Black oak. Pages 744-750 in Burns, R. M.
and B. H. Honkala, tech. Coords. Silvics of North America: 2. Hardwoods.
Agriculture Handbook 654. U. S. Department of Agriculture, Forest Service,
Washington, D. C. vol 2, 877 pp.

Schneider, G. 1966. A 20-year investigation in a sugar maple-beech stand in southern
Michigan. Michigan State University Agricultural Experiment Station. Research
Bulletin 15. 6lpp.

Severinghaus, C. W. and A. N. Moen. 1983. Prediction of weight and reproductive rates
of a white-tailed deer population from records of antler beam diameter among
yearling males. New York Fish and Game Journal. 30:30-38.

Smith, R. L. and J. L. Coggins. 1984. Basis and role of management. Pages 571-600 in
Halls, L. K., R. E. McCabe, and L. R. Jahn (eds.). White-tailed deer: ecology and
management. Stackpole Books. Harrisburg. 870pp.

Soil Survey Division, Natural Resources Conservation Service, United States Department
of Agriculture. 1996. Official Soil Series Descriptions [online www]. Available
url: http://www.statlab.iastate.edu/soils/osd/. Accessed June 2000-October 2002.

Sommers, L. M. (ed.). 1977. Atlas of Michigan, Michigan State University Press, East
Lansing, Michigan. 242pp.

Spies, T. A., and B. V. Barnes. 1985. A multifactor ecological classification of the
northern hardwood and conifer ecosystems of Sylvania Recreation Area, Upper
Peninsula, Michigan. Canadian Journal of Forest Research 15: 949-960.

Stocker, M., F. F. Gilbert, and D. W. Smith. 1977. Vegetation and deer habitat relations
in southern Ontario: classification of habitat types. Journal of Applied Ecology
14:419-432.

Stormer, F. A. and W. A. Bauer. 1980. Summer forage use by tame deer in northem
Michigan. Journal of Wildlife Management 44: 98-106.

Thompson, 1. D. and R. W. Stewart. 1998. Management of moose habitat. Pages 377-
389 in A. W. Franzman and C.C. Schwartz (eds.). Ecology and management of
the North American moose. Smithsonian Institution Press, Washington. 733pp.

Torgerson, O. and W. R. Porath. 1984. Midwest oak-hickory forest. Pages 411-426 in
Halls, L. K., R. E. McCabe, and L. R. Jahn (eds.). White-tailed deer ecology and
management. Stackpole Books. Harrisburg. 870pp.

Tubbs, CH. and D. R. Houston. 1990. Fagus grandifolia. American beech. Pages 325-
332 in Burns, R. M. and B. H. Honkala, tech. coords. Silvics of North America:

158

2. Hardwoods. Agriculture Handbook 654. U. S. Department of Agriculture,
Forest Service, Washington, D. C. vol 2, 877pp.

Ullrey, D.E., W.G. Youatt, H.E. Johnson, L.D. Fay, and BE. Brent. 1967. Digestibility
of cedar and jack pine browse for the white-tailed deer. Journal of Wildlife
Management 31: 448-454.

Ullrey, D.E., W.G. Youatt, H.E. Johnson, L.D. Fay, B.E. Brent, and KB. Kemp. 1968.
Digestibility of cedar and balsam fir browse for the white-tailed deer. Journal of
Wildlife Management 32: 162-171.

Ullrey, D.E., W.G. Youatt, H.E. Johnson, P.K. Ku, and L.D. Fay. 1964. Digestibility of
cedar and aspen browse for the white-tailed deer. Journal of Wildlife Management
28: 791-797.

United States Department of Agriculture, Natural Resources Conservation Service. 1998.
Soil Survey Geographic (SSURGO) database for Michigan. USDA, NRCS, Fort
Worth, Texas. digital map.

United States Geological Survey. 1999. National Land Cover Data. U.S. Department of
the Interior. Sioux Falls, SD. digital map.

VanDeelen, T. R., H. Campa, III, M. Hamady, and J. B. Haufler. 1998. Migration and
seasonal range dynamics of deer using adjacent deeryards in northern Michigan.
Journal of Wildlife Management 62: 205-213.

Verme, L. J. 1965. Swamp conifer deeryards in northern Michigan: their ecology and
management. Journal of Forestry 63: 523-529.

Verme, L. J. 1968. An index of winter weather severity for northern deer. Journal of
Wildlife Management 34: 870-886.

Verme, L. J. 1969. Reproductive patterns of white-tailed deer related to nutritional
plane. Journal of Wildlife Management 33: 881-887.

Verme. L. J. 1973. Movements of deer in Upper Michigan. Journal of Wildlife
Management 37:545-552.

Verme, L. J. and W. F. Johnston. 1986. Regeneration of northern white cedar deeryards
in upper Michigan. Journal of Wildlife Management 50: 307-313.

Verme, L. J. and D. E. Ullrey. 1984. Physiology and nutrition. Pages 91-118 410 in

Halls, L. K., R. E. McCabe, and L. R. Jahn (eds.). White-tailed deer ecology and
management. Stackpole Books. Harrisburg. 870pp.

159

Vogt, K. A., J. C. Gordon, J. P. Wargo, D. G. Vogt, H. Asbjomsen, P. A. Palmiotto, H. J.
Clark, J. L. O’Hara, W. S. Keaton, T. Patel-Weynand, E. Witten. 1997.
Ecosystems: balancing science with management. Springer-Verlag, New York.

470pp.

Westveld, R. H. 1933. The relation of certain soil characteristics to forest growth and
composition in the northern hardwood forest of northern Michigan. Michigan
Agricultural Experiment Station Technical Bulletin 135: 5 1 pp.

Williams, T. E. 1998. Soil Survey of Alcona County, Michigan. U.S. Department of
Agriculture, Natural Resources Conservation Service, and Forest Service. 409pp.

160