 

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 RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRCIDatoDuesz-sz

AN EXAMINATION OF POPULATION-LEVEL QUALITY INDICES As A
MEASURE OF WHITE-TAILED DEER (ODOCOILEUS VIRGINIANUS) HERD
CONDITION IN MICHIGAN.

By

Sarah Laurel Panken

A THESIS

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

MASTER OF SCIENCE

Department of Fisheries and Wildlife

2002

ABSTRACT
AN EXAMINATION OF POPULATION-LEVEL QUALITY INDICES AS A
MEASURE OF WHITE-TAILED DEER (ODOCOILEUS VIRGINIANUS) HERD
CONDITION IN MICHIGAN
By

Sarah Laurel Panken

One of the best methods to examine the condition of white-tailed deer is to
evaluate the physical growth of yearling deer. Antler development, consisting of average
beam diameter and number of points, as well as lactation status were identiﬁed as
measurable population-level quality indices for yearling deer in Michigan. The condition
of the deer herd, as measured by these quality indices, was examined temporally and
spatially both within and among 3 distinct regional study sites in Michigan. Each of the
population-level quality indices increased along a north to south regional gradient, which
suggested that deer in the southern part of Michigan were in relatively better condition
than their counterparts in the more northern regions. This trend in deer herd condition
may be attributed to winter severity, population density, and habitat quality differences
among the 3 distinct regional study sites in Michigan. The temporal and spatial trends in
each of these factors were evaluated both within and among the 3 distinct regional study
sites, as was the relationship between these factors and yearling antler development.
Multiple factors affected deer herd condition within each regional study site; in the
northern part of the state weather gave way to a mixture of weather and density, and in

the southern-most part of the state, density had the most inﬂuence on herd condition.

T 0 Max.
1 could not have done it without your love, support, and patience.

iii

ACKNOWLEDGEMENTS

Funding for this project was provided by the Michigan Department of Natural
Resources (MDNR), Wildlife Division and carried out in conjunction with the
Partnership for Ecosystem Research and Management (PERM) between the MDNR and
Michigan State University (MSU), Department of Fisheries and Wildlife.

I want to extend my gratitude to all of my master’s committee members. To Dr.
Rique Campa and Dr. Chuck Nelson, thank you for the guidance you gave me. To Dr.
Bill Moritz and Brent Rudolph of the MDNR, Wildlife Division, thank you for your
contributions to this project. To my major professor, Dr. Scott Winterstein, who gave me
the opportunity to develop my conﬁdence and expertise.

This project could not have been completed without the assistance of Autumn
Larkins, Mark Monroe, and Alexis Newitt who spent hours in front of a computer doing
both qualitative and quantitative data analysis. I must also express my deepest gratitude
to Ali Felix and Sarah Mayhew for Showing me the ropes and helping me with so many
different aspects of this project. Thank you to Nikki Lamp and Katie Kahl for reading
through parts of this thesis. Finally, I am indebted to all of my fellow graduate students
who have answered my questions and provided me with the support I needed to complete

this project.

iv

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... viii
LIST OF FIGURES ........................................................................................................ xiv
GENERAL INTRODUCTION ......................................................................................... 1
Project Overview ...................................................................................... 1
STUDY AREA ................................................................................................................. 4
LITERATURE CITED ......................................................................................... 10

CHAPTER 1: A REGIONAL COMPARISON OF WHITE-TAILED DEER HERD
CONDITION IN MICHIGAN

INTRODUCTION ................................................................................................ 1 1
Objectives ................................................................................................. 17
METHODS ........................................................................................................... 18
Data Sources ............................................................................................. 18
Biodata .......................................................................................... 18

Summer Driving Transects ........................................................... 19

Sample Size ............................................................................................... 21
Yearling Age-Class ................................................................................... 22
Population Quality Indices ........................................................................ 22

Fall Recruitment ............................................................................ 22

Lactation Status ............................................................................. 23

Antler Measurement ...................................................................... 24

Within Regions ......................................................................................... 25
Among Regions ........................................................................................ 26
Relationship Among Indices ..................................................................... 27
RESULTS ............................................................................................................. 29
Sample Sizes ............................................................................................. 29

UP ................................................................................................. 29

NLP ............................................................................................... 32

SLP ................................................................................................ 36

Fall Recruitment ........................................................................................ 39
Comparison within Study Sites ................................................................. 46

UP ................................................................................................. 46

NLP ............................................................................................... 60

SLP ................................................................................................ 72

Among Region Comparisons .................................................................... 87
Temporal Comparisons ................................................................. 87

Spatial Comparisons ..................................................................... 93

Relationship Among Quality Indices ........................................................ 97
DISCUSSION ..................................................................................................... 104
Fall Recruitment ...................................................................................... 104

Land Unit Classiﬁcation ......................................................................... 105
Condition Indices .................................................................................... 106
Temporal and Spatial Trends .................................................................. 107
Sample Sizes and Aggregation of Data .................................................. 109

Biases ...................................................................................................... 110
Biodata ........................................................................................ 110

Summer Driving Transect ........................................................... 113

Bias Reduction ............................................................................ 114

Assumptions ............................................................................................ 1 15
Alternative Indices .................................................................................. 116
Conclusion .............................................................................................. 1 18
LITERATURE CITED ....................................................................................... 120

CHAPTER 2: AN EXAMINATION OF UNDERLYING MECHANISMS THAT MAY
INFLUENCE WHITE-TAILED DEER HERD CONDITION IN MICHIGAN

INTRODUCTION .............................................................................................. 1 23
Objectives ............................................................................................... 1 28
METHODS ......................................................................................................... 129
Data Sources ........................................................................................... 129
Quality Indices ............................................................................ 129

Population Density ...................................................................... 129

Temperature ................................................................................ 1 30

Snowfall ...................................................................................... 132

Winter Severity ........................................................................... 132

Habitat Potential .......................................................................... 134

Within Regions ....................................................................................... 136
Among Regions ...................................................................................... 139
RESULTS ........................................................................................................... 140
Comparisons within Study Sites ............................................................. 140

UP ............................................................................................... 140

NLP ............................................................................................. 154

SLP .............................................................................................. 177

Comparisons Among Regions ................................................................ I92

Deer Numbers ............................................................................. 192

Temperature ................................................................................ l 99

Snowfall ...................................................................................... 202

Winter Severity Index ................................................................. 205

vi

DISCUSSION .................................................................................................... 209

Data Limitations ...................................................................................... 211
Incomplete Data .......................................................................... 211

Data Aggregation ........................................................................ 212

Biased Data ................................................................................. 212

Assumptions ............................................................................................ 2 13
Additional Approaches ........................................................................... 21 4
Conclusion .............................................................................................. 21 5
LITERATURE CITED ....................................................................................... 216
CONCLUSIONS AND MANAGEMENT IMPLICATIONS ............................ 220
Recommendations ................................................................................... 223

vii

LIST OF TABLES

CHAPTER 1: A REGIONAL COMPARISON OF WHITE-TAILED DEER HERD
CONDITION INDICES IN MICHIGAN

Table 1.1. Sample size of yearling deer in each county in the Upper Peninsula study site
for (a) males that had an average beam diameter measured, (b) males that had number of
points counted, and (0) females that were checked for lactation status ............................ 30

Table 1.2. Sample size of yearlings in each county of the Northern Lower Peninsula
study site for (a) males that had average beam diameter measured, (b) males that had the
number of points counted, and (c) females that were checked for lactation status .......... 33

Table 1.3. Sample size of yearling deer in each county in the Southern Lower

Peninsula study site for (a) males that had an average beam diameter measured, (b)
males that had the number of points counted, and (c) females that were checked for
lactation status ................................................................................................................... 37

Table 1.4. Summer of 2000 total, average, and maximum fawn-to-doe ratio based
on both morning and evening deer counts for each regional route in July, August,
and September (b,c,d) and September only (a). ............................................................... 40

Table 1.5. Total deer numbers observed based on the summer 2000 driving transect deer
counts for each regional route in July, August, and September (b,c,d) and September
only (a) .............................................................................................................................. 42

Table 1.6. Summer of 2001 total, average, and maximum fawn-to-doe ratios based
on evening deer counts only for each regional route in August and September (c,d),
September and October (b) , and September only (a) ...................................................... 44

Table 1.7. Total deer numbers observed based on the summer 2001 deer counts for each
regional route in August and September (c,d), September and October (b), and September
only (a) .............................................................................................................................. 45

Table 1.8. The lactation status of yearling does harvested in the Upper Peninsula
study site each year between 1993 and 2000, as well as the percent of yearling does
lactating (HM+) and not lactating (HM-) each year. ........................................................ 47

Table 1.9. The mean average beam diameter in mm for male yearlings harvested in

each county in the Upper Peninsula study site for each year between 1987 and 2000

and the standard deviation, as well as the yearly averages for the entire study site

and overall averages for each county. ............................................................................... 50

viii

Table 1.10. The minimum and maximum average beam diameter in mm collected
from male yearlings harvested in each county in the Upper Peninsula study site for
each year between 1987 and 2000. ................................................................................... 50

Table 1.11. Mean pair-wise comparisons of average beam diameter for male
yearlings harvested in the Upper Peninsula study site on a yearly basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 53

Table 1.12 Mean pair-wise comparisons of average beam diameter for male
yearlings harvested in the Upper Peninsula study site on a county basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 53

Table 1.13. The mean point count for male yearlings harvested in each county in

the Upper Peninsula study site for each year between 1987 and 2000 and the

standard deviation, as well as the yearly averages for the entire study site and

overall averages for each county ....................................................................................... 56

Table 1.14. The minimum and maximum number of points counted on male
yearlings harvested in each county in the Upper Peninsula study site for each year
between 1987 and 2000. ................................................................................................... 56

Table 1.15. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Upper Peninsula study site on a yearly basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05. ................................................................ 58

Table 1.16. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Upper Peninsula study site on a county basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05. ................................................................ 58

Table 1.17. The lactation status of yearling does harvested in the Northern Lower
Peninsula study site each year between 1993 and 2000, as well as the percent of
yearling does lactating (HM+) and not lactating (HM-) each year ................................... 61

Table 1.18. The mean average beam diameter in mm for male yearlings harvested

in each county in the Northern Lower Peninsula study site for each year between

1987 and 2000 and the standard deviation, as well as the yearly averages for the

entire study site and overall averages for each county. .................................................... 63

Table 1.19. The minimum and maximum average beam diameter in mm collected
from male yearlings harvested in each county in the Northern Lower Peninsula
study site for each year between 1987 and 2000. ............................................................. 63

Table 1.20. Mean pair-wise comparisons of average beam diameter for male

yearlings harvested in the Northern Lower Peninsula study site on a yearly basis.
The x denotes a signiﬁcant difference at an alpha level of 0.05. ...................................... 66

ix

Table 1.21. Mean pair-wise comparisons of average beam diameter for male
yearlings harvested in the Northern Lower Peninsula study site on a county basis.
The x denotes a signiﬁcant difference at an alpha level of 0.05. ...................................... 66

Table 1.22. The mean point count for male yearlings harvested in each county in

the Northern Lower Peninsula study site for each year between 1987 and 2000 and

the standard deviation, as well as yearly averages for the entire study site and

overall averages for each county ....................................................................................... 69

Table 1.23. The minimum and maximum number of points counted on male
yearlings harvested in each county in the Northern Lower Peninsula study site for
each year between 1987 and 2000. ................................................................................... 69

Table 1.24. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Northern Lower Peninsula study site on a yearly basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 71

Table 1.25. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Northern Lower Peninsula study site on a county basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 71

Table 1.26. The lactation status of yearling does harvested in the Southern Lower
Peninsula study site each year between 1993 and 2000, as well as the percent of
yearling does lactating (HM+) and not lactating (HM-) each year ................................... 75

Table 1.27. The mean average beam diameter in mm for male yearlings harvested

in each county in the Southern Lower Peninsula study site for each year between

1987 and 2000 and the standard deviation, as well as the yearly averages for the

entire study site and overall averages for each county. .................................................... 77

Table 1.28. The minimum and maximum average beam diameter in mm collected
from male yearlings harvested in each county in the Southern Lower Peninsula
study site for each year between 1987 and 2000. ............................................................. 77

Table 1.29. Mean pair-wise comparisons of average beam diameter for male
yearlings harvested in the Southern Lower Peninsula study site on a yearly basis.
The x denotes a signiﬁcant difference at an alpha level of 0.05. ...................................... 80

Table 1.30. Mean pair-wise comparisons of average beam diameter for male
yearlings harvested in the Southern Lower Peninsula study site on a county basis.
The x denotes a signiﬁcant difference at an alpha level of 0.05. ...................................... 80

Table 1.31. The mean point count for male yearlings harvested in each county in

the Southern Lower Peninsula study site for each year between 1987 and 2000 and

the standard deviation, as well as yearly averages for the entire study site and

overall averages for each county ....................................................................................... 82

Table 1.32. The minimum and maximum number of points counted on male
yearlings harvested in each county in the Southern Lower Peninsula study site for
each year between 1987 and 2000. ................................................................................... 82

Table 1.33. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Southern Lower Peninsula study site on a yearly basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 85

Table 1.34. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Southern Lower Peninsula study site on a county basis. The x
denotes a signiﬁcant difference at an alpha level of 0.05. ................................................ 85

CHAPTER 2: AN EXAMINATION OF UNDERLYING MECHANISMS THAT MAY
INFLUENCE WHITE-TAILED DEER HERD CONDITION IN MICHIGAN

Table 2.1. The total number of deer, as estimated from the sex-age-kill estimate
method, in each county in the Upper Peninsula study site between 1987 and 2000;

the yearly average total number of deer and standard deviation (StDev) for the

entire study site, and the average total number of deer for each county. ........................ 141

Table 2.2. The cumulative mean monthly temperature (summed from January

through September) in °C for each county in the Upper Peninsula study site from

1987 to 2000; the yearly average cumulative mean monthly temperature and

standard deviation (StDev) for the entire study site, and the average cumulative

mean monthly temperature for each county .................................................................... 145

Table 2.3. The cumulative total monthly snowfall (summed from January through

April) in cm for each county in the Upper Peninsula study site from 1987 to 2000;

the yearly average cumulative total monthly snowfall and the standard deviation

(StDev) for the entire study site, and the average cumulative total monthly snowfall

for each county ................................................................................................................ 148

Table 2.4. The mean corrected Winter Severity Index for each year between 1986
and 1999 in the Upper Peninsula study site. ................................................................... 152

Table 2.5. Results of the ANOVA model run by county in the Upper Peninsula

study site to determine which underlying mechanisms that have the potential to

inﬂuence male yearling average beam diameter as a measure of deer herd quality in
Michigan were signiﬁcant at an alpha level of 0.10. ...................................................... 155

xi

Table 2.6. Results of the ANOVA model run by county in the Upper Peninsula

study site to determine which underlying mechanisms that have the potential to

inﬂuence male yearling point count as a measure of deer herd quality in Michigan

were signiﬁcant at an alpha level of 0.10. ...................................................................... 156

Table 2.7. The total number of deer, as estimated from the sex-age-kill estimate

method, in each county in the Northern Lower Peninsula study site between 1987

and 2000; the yearly average total number of deer and the standard deviation for the
entire study site, and the average total number of deer for each county ......................... 158

Table 2.8. The cumulative mean monthly temperature (summed from January

through September) in °C for each county in the Northern Lower Peninsula study

site from 1987 to 2000; the yearly average cumulative mean monthly temperature

and the standard deviation (StDev) for the entire study site, and the average

cumulative mean monthly temperature for each county. ................................................ 162

Table 2.9. The cumulative total monthly snowfall (summed from January through
April) in cm for each county in the Northern Lower Peninsula study site from 1987

to 2000; the yearly average cumulative total monthly snowfall and the standard
deviation (StDev) for the entire study site, and the average cumulative total monthly
snowfall for each county. ................................................................................................ 165

Table 2.10. The mean corrected Winter Severity Index for each year between 1986
and 1999 in the Northern Lower Peninsula study site. ................................................... 169

Table 2.11. The potential of the fall and winter food, spring and summer food, and
thermal cover to provide suitable habitat for white-tailed deer in each county in the
Northern Lower Peninsula study site (0 is the lowest suitability and 1 is the highest
suitability). ...................................................................................................................... 1 72

Table 2.12. Results of the AN OVA model run by county in the Northern Lower
Peninsula study site to determine which underlying mechanisms that have the

potential to inﬂuence male yearling average beam diameter as a measure of deer

herd quality in Michigan were signiﬁcant at an alpha level of 0.10. .............................. 174

Table 2.13. Results of the regression analyses conducted to determine the
relationship between each habitat requirement and male yearling average beam
diameter in the Northern Lower Peninsula study site at an alpha level of 0.10. ............ 175

Table 2.14. Results of the ANOVA model run by county in the Northern Lower
Peninsula study site to determine which underlying mechanisms that have the

potential to inﬂuence male yearling point count as a measure of deer herd quality in
Michigan were signiﬁcant at an alpha level of 0.10. ...................................................... 176

xii

Table 2.15. Results of the regression analyses conducted to determine the
relationship between each habitat requirement and male yearling point count in the
Northern Lower Peninsula study site at an alpha level of 0.10. ..................................... 178

Table 2.16. The total number of deer, as estimated from the sex-age-kill estimate
method, in each county in the Southern Lower Peninsula study site between 1987
and 2000; the yearly average total number of deer and the standard deviation for the
entire study site, and the average total number of deer for each county. ........................ 179

Table 2.17. The cumulative mean monthly temperature (summed from January
through September in °C for each county in the Southern Lower Peninsula study
site from 1987 to 2000; the yearly average cumulative mean monthly temperature
and the standard deviation (StDev) for the entire study site, and the average
cumulative mean monthly temperature for each county. ................................................ 183

Table 2.18. The cumulative total monthly snowfall (summed from January through
Apri l) in cm for each county in the Southern Lower Peninsula study site from 1987

to 2 000; the yearly average cumulative total monthly snowfall and the standard
deviation (StDev) for the entire study site, and the average cumulative total monthly
snowfall for each county. ................................................................................................ 187

Table 2.19. The mean corrected Winter Severity Index for each year between 1986
and l 999 in the Southern Lower Peninsula study site. ................................................... 190

Table 2.20. Results of the ANOVA model run by county in the Southern Lower
Peninsula study site to determine which underlying mechanisms that have the
POtential to inﬂuence male yearling average beam diameter as a measure of deer

herd quality in Michigan were signiﬁcant at an alpha level of 0.10. .............................. 193

Table 2.21. Results of the ANOVA model run by county in the Southern Lower
Penlnsula study site to determine which underlying mechanisms that have the
poFential to inﬂuence male yearling point count as a measure of deer herd quality in
Mlchigan were signiﬁcant at an alpha level of 0.10. ...................................................... 194

Table 2.22. The mean comparison of (a) deer population density, (b) temperature,
(c) Snowfall, and (d) the winter severity index among the 3 regional study sites at an
8L1Dha level of 0.10 ........................................................................................................... 198

xiii

LIST OF FIGURES

STUDY AREA

Figure 1.1. The three distinct regions of Michigan based on Albert et a1. 1986
classiﬁcation of ecological regions. . ................................................................................ 5

Figure 1.2. Ecological regions (ecoregions) of Michigan based on homogeneity of
localized climate and physiography. Classiﬁed according to Albert et al. 1986 ............. 6

Figure 1.3. The three designated study sites. One site was located within each of

the three ecological regions of Michigan and consisted of several counties. ................... 7
CHAPTER 1: A REGIONAL COMPARISON OF WHITE-TAILED DEER HERD
CONDITION INDICES IN MICHGAN

Figure 1.4. The data sheet used for the summer driving transects for both 2000 and
2001 ................................................................................................................................... 20

Figure 1.5. Lactation status of yearling does harvested in the Upper Peninsula
study site for each year between 1993 and 2000. ............................................................. 47

Figure 1.6. The percent of yearling does harvested in the Upper Peninsula study
site that were lactating during each year between 1993 and 2000. ................................... 49

Figure 1.7. The temporal change in average beam diameter in mm for male
yearlings harvested in the Upper Peninsula study site between 1987 and 2000 ............... 51

Figure 1.8. Mean yearling average beam diameter by county in the Upper
Peninsula study site for 1987-2000. .................................................................................. 54

Figure 1.9. The temporal change in average point count for male yearlings
harvested in the Upper Peninsula study site between 1987 and 2000. ............................. 57

Figure 1.10. Yearling average point count by county in the Upper Peninsula study
site for 1987-2000. ............................................................................................................ 59

Figure 1.11. The lactation status of yearling does harvested in the Northern Lower
Peninsula study site each year between 1993 and 2000. .................................................. 61

Figure 1.12. The percent of yearling does harvested in the Northern Lower
Peninsula study site that were lactating during each year between 1993 and 2000. ........ 62

xiv

Figure 1.13. The temporal change in average beam diameter in mm for male
yearlings harvested in the Northern Lower Peninsula study site between 1987 and
2000 ................................................................................................................................... 65

Figure 1.14. Mean yearling average beam diameter by county in the Northern
Lower Peninsula study site for 1987-2000. ...................................................................... 67

Figure 1.15. The temporal change in average point count for male yearlings
harvested in the Northern Lower Peninsula study site between 1987 and 2000. ............. 70

Figure 1.16. Yearling average point count by county in the Northern Lower
Peninsula study site for 1987—2000. .................................................................................. 73

Figure 1.17. The lactation status of yearling does harvested in the Southern Lower
Peninsula study site each year between 1993 and 2000. .................................................. 75

Figure 1.18. The percent of yearling does harvested in the Southern Lower
Peninsula study site that were lactating during each year between 1993 and 2000. ........ 76

Figure 1.19. The temporal change in average beam diameter in mm for male
yearlings harvested in the Southern Lower Peninsula study site between 1987 and
2000 ................................................................................................................................... 78

Figure 1.20. Mean yearling average beam diameter by county in the Southern
Lower Peninsula study site for 1987-2000. ...................................................................... 81

Figure 1.21. The temporal change in average point count for male yearlings
harvested in the Southern Lower Peninsula study site between 1987 and 2000. ............. 84

Figure 1.22. Yearling average point count by county in the Southern Lower
Peninsula study site for 1987-2000. .................................................................................. 86

Figure 1.23. Yearly trends in the percentage of yearling does harvested in each of
the 3 regional study sites that were lactating from 1993-2000. ........................................ 88

Figure 1.24. Yearly trends in the average beam diameter collected from male

yearlings harvested in each of the 3 regional study sites from 1987-2000. The *

denotes which years the average beam diameter for male yearlings was signiﬁcantly
different (p-value < 0.05) among the UP and NLP study sites. ........................................ 90

Figure 1.25. Yearly trends in the point count collected from male yearlings

harvested in each of the 3 regional study sites from 1987-2000. The * denotes

which years the point count for male yearlings was signiﬁcantly different (p-value <
0.05) among the UP and NLP study sites. ........................................................................ 92

XV

Figure 1.26. The regional comparison of the percentage of does lactating in each of
the 3 study sites during 1993-2000. .................................................................................. 94

Figure 1.27. The regional comparison of the average beam diameter for male
yearlings harvested in each of the 3 study sites during 1987-2000. ................................. 96

Figure 1.28. The regional comparison of the average point count for male yearlings
harvested in each of the 3 study sites during 1987-2000. ................................................. 98

Figure 1.29. The relationship between the percent of yearling does lactating and the
mean average beam diameter for male yearlings in the (a) UP study site, the (b)
NLP study site, and the (c) SLP study site ...................................................................... 100

Figure 1.30. The relationship between the percent of yearling does lactating and the
average point count for male yearlings in the (a) UP study site, the (b) NLP study
site, and (c) the SLP study site. ....................................................................................... 101

Figure 1.31. The relationship between the mean average beam diameter for male
yearlings and male yearling average point count in the (a) UP study site, (b) NLP
study site and (c) the SLP Study site. ............................................................................. 103

CHAPTER 2: AN EXAMINATION OF UNDERLYING MECHANISMS THAT MAY
INFLUENCE WHITE-TAILED DEER HERD CONDITION IN MICHIGAN

Figure 2.1. The cumulative mean monthly temperature for Alpena County in 2000.131
Figure 2.2. The cumulative total monthly snowfall for Marquette County in 2000. ..... 133

Figure 2.3. The corrected cumulative winter severity index (WSI) for each of the 3
regional study sites in 2000 ............................................................................................. 135

Figure 2.4. The habitat types for the Northern Lower Peninsula study site as

deﬁned by Felix et al. (unpublished data) and the habitat potential for fall and

winter food, spring and summer food, and thermal cover based on the habitat types
delineation. ...................................................................................................................... 1 37

Figure 2.5. The weighted habitat potential for fall and winter food, spring and
summer food, and thermal cover for each county in the Northern Lower Peninsula
study site. ........................................................................................................................ 138

Figure 2.6. The temporal trend in the average total number of deer, as estimated by

the sex-age-kill estimate method, for the Upper Peninsula study site except Baraga
County between 1987 and 2000 ...................................................................................... 142

xvi

Figure 2.7. The average total number of deer by county in the Upper Peninsula
study site for 1987 through 2000. ................................................................................... 143

Figure 2.8. The temporal trend in the average cumulative temperature (summed
from January through September) in °C for the Upper Peninsula study site between
1987 and 2000. ................................................................................................................ 146

Figure 2.9. Cumulative mean monthly temperature (summed from January through
September) in °C by county in the Upper Peninsula study site from 1987 through
2000 ................................................................................................................................. 147

Figure 2.10. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Upper Peninsula study site between 1987 and
2000 ................................................................................................................................. 149

Figure 2.11. Cumulative total monthly snowfall (summed from January through
April) in cm by county in the Upper Peninsula study site from 1987 through 2000. ..... 151

Figure 2.12. The temporal trend in the corrected WSI for the Upper Peninsula
study site between 1987 and 2000. ................................................................................. 153

Figure 2.13. The temporal trend in the average total number of deer, as estimated
by the sex-age-kill estimate method, for the Northern Lower Peninsula study Site
between 1987 and 2000. ................................................................................................. 159

Figure 2.14. The average number of deer by county in the Northern Lower
Peninsula study site from 1987 through 2000 ................................................................. 160

Figure 2.15. The temporal trend in the average cumulative temperature (summed

from January through September) in °C for the Northem Lower Peninsula study site
between 1987 and 2000. ................................................................................................. 163

Figure 2.16. Cumulative mean monthly temperature (summed from January

through September) in °C by county in the Northern Lower Peninsula study site
from 1987 through 2000. ................................................................................................ 164

Figure 2.17. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Northern Lower Peninsula study Site between
1987 and 2000. ................................................................................................................ 167

Figure 2.18. Cumulative total monthly snowfall (summed from January through
April) in cm by county in the Northern Lower Peninsula study site from 1987
through 2000. .................................................................................................................. 168

Figure 2.19. The temporal trend in the corrected WSI for the Northern Lower
Peninsula study site between 1987 and 2000 .................................................................. 170

xvii

Figure 2.20. The temporal trend in the average total number of deer, as estimated
by the sex-age-kill estimate method, for the Southern Lower Peninsula study site
between 1987 and 2000. ................................................................................................. 180

Figure 2.21. The average total number of deer by county in the Southern Lower
Peninsula study site from 1987 through 2000 ................................................................. 181

Figure 2.22. The temporal trend in the average cumulative temperature (summed
from January through September) in °C for the Southern Lower Peninsula study site
between 1987 and 2000. ................................................................................................. 184

Figure 2.23. Cumulative mean monthly temperature (summed from January
through September) in °C by county in the Southern Lower Peninsula study site
from 1987 through 2000. ................................................................................................ 185

Figure 2.24. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Southern Lower Peninsula study site between
1987 and 2000. ................................................................................................................ 188

Figure 2.25. Cumulative total monthly snowfall (summed from January through
April) in cm by county in the Southern Lower Peninsula study site from 1987
through 2000. .................................................................................................................. 189

Figure 2.26. The temporal trend in the corrected WSI for the Southern Lower
Peninsula study site between 1987 and 2000 .................................................................. 191

Figure 2.27. The comparison of yearly trends in the average total number of deer, as
estimated by the sex-age-kill estimate method, for each of the 3 regional study sites
between 1987 and 2000. ................................................................................................. 195

Figure 2.28. The regional comparison of the density of deer present in each of the 3
regional study sites averaged across years from 1987 to 2000. ...................................... 197

Figure 2.29. The comparison of yearly trends in the average cumulative
temperature (summed from January through September) for each of the 3 regional
study sites between 1987 and 2000 ................................................................................. 200

Figure 2.30. The regional comparison of the cumulative temperature (summed
from January through September) in °C averaged across years from 1987 to 2000
among each of the 3 regional study sites. ....................................................................... 201

Figure 2.31. The comparison of yearly trends in the average cumulative snowfall

(summed from January through April) for each of the 3 regional study sites between
1987 and 2000. ................................................................................................................ 203

xviii

Figure 2.32. The regional comparison of the cumulative snowfall (summed from
January through April) in cm averaged across years from 1987 to 2000 among the 3
regional study sites. ......................................................................................................... 204

Figure 2.33. The comparison of yearly trends in the corrected WSI for each of the 3
regional study sites between 1987 and 2000 ................................................................... 206

Figure 2.34. The regional comparison of the corrected WSI averaged across years
from 1987 to 2000 among the 3 regional study sites. ..................................................... 207

xix

GENERAL INTRODUCTION

Project Overview
The research discussed in this thesis was developed as part of a larger project that
attempts to examine deer management using an ecosystem paradigm. This “umbrella”
project is more holistic in nature than traditional approaches to deer management because
it attempts to address the ecological, social, and biological aspects of deer management at
a landscape level. To quantify the ecological aspect of deer management, habitat models,
based on the physiography of the land and vegetation types, were used to predict the
potential for deer habitat requirements in speciﬁed areas of the Michigan. To quantify
the social aspect of deer management, surveys were conducted to understand how people
in Michigan view deer and deer management. To quantify the biological aspect of deer
management, the condition of the deer herd, as measured by quality indices, was
examined in different regions of Michigan. Upon completion of the 3 different sections of
the “umbrella” project, the ecological, social, and biological aspects will be incorporated
into a landscape-based model to aid managers in developing statewide scientiﬁcally
sound deer management strategies.
The overall objectives of the study are:
1) Assess whether a hierarchical ecological classiﬁcation system effectively
delimits land units exhibiting differing deer reproductive potential, expressed
in the form of recruitment to fall hunting populations and measured in fawn-

to-doe ratios in check station data.

2) Determine whether reproductive potential, as measured by fawn-to-doe ratios
in check station data, can be used to estimate biological carrying capacity of
ecological units.

3) Determine whether antler measurements from harvested bucks in check
station data can be substituted for reproductive recruitment to determine
biological carrying capacity, especially in areas where antlerless harvests are
absent or limited.

4) Assess whether a habitat-based model of deer population estimation can be
used to calculate population goals and compare with models based on
reproductive potential.

5) Estimate cultural carrying capacity in study areas using primary stakeholder
groups based on farming and hunting participation.

6) Integrate animal-based models, habitat-models, and cultural carrying capacity
models into a statewide deer management strategy.

This thesis concentrates on the biological aspect of the landscape level approach to deer
management in Michigan by looking at the differences in deer herd condition in the 3
main regions of the state and the underlying mechanisms that have the potential to
inﬂuence herd condition. This thesis is divided into 3 sections:

Chapter 1: A Regional Comparison of White-Tailed Deer Herd Condition Indices

in Michigan.

Chapter 2: An Examination of Underlying Mechanisms That May Inﬂuence

White-Tailed Deer Herd Condition in Michigan.

Conclusions and Management Implications.

In Chapter 1, I used p0pulation-level quality indices to examine the temporal and
spatial trends in the condition of the white-tailed deer herd in Michigan from 1987
through 2000. Chapter 2 addresses two main topics. First, I explored temporal and
spatial trends in some of the underlying mechanisms that have the potential to inﬂuence
the condition of white-tailed deer in Michigan. Second, I determined the relationship
between those factors that may inﬂuence deer herd condition and male yearling antler
measurements as indices of white-tailed deer herd quality. In the ﬁnal section, I discuss
overall conclusions and management implications based on the results from Chapters 1

and 2.

STUDY AREA

Michigan is divided into 3 distinct regions based on climate and physiography
(Figure 1.1; Albert et a1. 1986). These 3 regions were then further classiﬁed according to
Albert et a1. (1986) into smaller ecological regions (ecoregions) based on the
homogeneity of the localized climate and physiography, and, to some degree, vegetation
(Figure 1.2). We designated 3 study sites. Each site was located within a different region
of Michigan (Region 1, Region 2, and Region 3) and consisted of several distinct
ecoregions (Figure 1.3). These study sites were selected based on the heterogeneity of
habitat, the existence of available information and the variation in deer numbers. In
addition, these study sites were deﬁned by political, not ecological, boundaries (i.e.,
counties) because the data used in this project were collected within political boundaries

that serve as a convenient way to deﬁne an area unit for deer management.

Upper Peninsula (UP)

Region 1 (Figure 1.1) is the Upper Peninsula of Michigan. Northern hardwood forest
ecosystems are common in this region, as are pine forest ecosystems. Soils are primarily
loam and sandy textures. This region has little agriculture and the least amount of urban
development of the 3 regions. The climate in Region 1 is the harshest of all of the
regions in Michigan; winters tend to last longer and generally have more snowfall and
colder temperatures. Total precipitation in this region is 800-900 mm and the annual
average temperature is between 4 °C and 5 °C. Generally higher than the elevation in the

other 2 regions, the elevation in Region 1varies from 184-604 m, but is predominately

   

Region 1.
Upper Peninsula

 

 

 

 

 

 

 

 

 

 

 

Region 2. .i
Northern Lower {J
Peninsula 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Region 3.
Southern Lower Z Y
Peninsula .

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.1. The three distinct regions of Michigan based on Albert et al. 1986
classiﬁcation of ecological regions.

 

Figure 1.2. Ecological regions (ecoregions) of Michigan based on homogeneity of
localized climate and physiography. Classiﬁed according to Albert et a1. 1986.

 

 

     
 
   
  

Presque
Isle

Mont- Alpena \
morenc' ;‘ ’

 

 

 

E Upper Peninsula Study Site

Northern Lower Peninsula Study Site
Southern Lower Peninsula Study Site i

  

 

 

 

 

 

 

 

 

 

 

  
      

 

Ir
II III

       

 

 

 

 

 

Figure 1.3. The three designated study sites. One site was located within each of the
three ecological regions of Michigan and consisted of several counties.

between 366-518 In (Albert et a1. 1986). Four counties, Baraga, Dickinson, Iron and

Marquette, were selected within this region (Figure 1.3).

Northern Lower Peninsula (NLP)

Region 2 (Figure 1) is the mid-northem region of Michigan. The vegetation in
this region is characterized by northern hardwood forest and pine forest ecosystems, and
is not as diverse as Region 1. Oak-pine (Quercus sp.-Pinus sp.) forests are common in
addition to pine forests that consist of white spruce (Picea glauca), balsam ﬁr (Abies
balsamea), and northern white cedar (T huja occidentalis L.).

The soils in Region 2 tend to be predominately sandy. Agriculture and urban
development exist in this region but these land uses are not as prominent as in the
southem-most region. Therefore, there has not been as much permanent alteration of the
vegetation communities present. The climate for the NLP is more variable than the
climate in the Southern Lower Peninsula; winters are cold and there is more snowfall.
The total precipitation for Region 2 is about 770 mm and the annual average temperature
is 6.2 0C. The highest elevations in the Lower Peninsula of Michigan, which range up to
526 m, occur in Region 2 (Albert et a1. 1986). In addition, Bovine TB has recently been
discovered in region 2. This, in part, was the impetus for the selection of speciﬁc
counties within this region. Five counties, Alcona, Alpena, Montmorency, Oscoda, and

Presque Isle, were selected (Figure 3).

Southern Lower Peninsula (SLP)

Region 3 (Figure l) is the southern-most region of Michigan. The vegetation is
characterized by hardwood forest ecosystems, such as beech (F agus grandifolia), sugar
maple (Acer saccharum), oak (Quercus sp.), and hickory forests (Carya sp.), that exist
primarily on loam and clay soil textures. Although sandy soils are not as common in this
region, there are some areas where this soil type exists. There is a high occurrence of
agriculture and urban development that has greatly reduced the once diverse plant
communities in this region. The climate for the southern region is more mild than in the
more northern regions; ﬂuctuations in temperature are less severe and typically there is
less snowfall and a longer growing season (Michigan Weather Service 1974, Ozoga et a1.
1994). Total annual precipitation for this region is approximately 9.4 mm and the annual
average temperature is approximately 19 °C. This region is ﬂatter when compared with
the more northern regions. Elevations range between 178-390 m, but are predominately
below 305 m (Albert et a1. 1986). Three counties, Barry, Calhoun, and Eaton, were

selected within this region (Figure 1.3).

LITERATURE CITED

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

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

Ozoga, J. J ., R. V. Doepker, and M. S. Sargent. 1994. Ecology and mangement of white-

tailed deer in Michigan. Michigan Department of Natural Resources, Wildlife
Division Report 3209.

10

CHAPTER 1

A REGIONAL COMPARISON OF WHITE-TAILED DEER HERD
CONDITION IN MICHIGAN

INTRODUCTION

In Michigan, traditional white-tailed deer (Odocoileus virginianus) management
strategies manipulated the number of deer in an area. Unchecked harvest, as well as the
alteration of habitat due to human activities, has historically inﬂuenced deer numbers.
By the late 18003, Michigan deer numbers were low because of unregulated market
hunting, and the public called for hunting regulations. Early measures to restrict hunting
were weak, and it was not until 1895 that legislation was passed to establish an ofﬁcial
hunting season, a bag limit, and a required license to hunt deer (Langenau 1994). These
restrictions on hunting, in addition to forest ﬁre control and a subsequent increase in
available deer browse, contributed to a slow population increase in the early 19005.
However, concern about slow growth of the deer herd led to further hunting restrictions
in the early 19005, such as reducing the number of days hunters could take deer as well as
the bag limit. The “buck law” was passed in 1921, which allowed only one buck to be
taken per hunter per year (Jenkins and Bartlett 1959).

About this time, the newly created State Department of Conservation, now the
Michigan Department of Natural Resources (MDNR), began to promote scientiﬁc deer
management. In 1928, the Game Division within the Department of Conservation was
created, which encouraged specialized personnel to gather scientiﬁc data in an organized
manner to use as the basis for management decision-making. This approach was

intensiﬁed alter the 1937 Pittman-Robertson Act was passed. Concurrently, the deer herd

ll

in Michigan rebounded to approximately 1 million deer and by 1930, winter starvation
and over-browsing indicated that the herd had exceeded carrying capacity. Since this
approach to deer management was not creating the desired results, the Department of
Conservation decided to alter its approach to deer management and reduce deer numbers
as well as improve long-term suitable habitat. Despite these efforts, the herd continued to
grow, habitat was severely damaged, and hunters became concerned with low buck-to-
doe ratios (Jenkins and Bartlett 1959).

In 1941, the Department of Conservation again allowed antlerless hunting in
selected areas in an effort to control population growth. The antlerless regulations were
experimental and in 1952, after what many described as a slaughter of antlerless deer, the
Department initiated an area and quota system to regulate the harvest of antlerless deer.
During this period of increased antlerless hunting, resources for deer became scarcer in
the 19605 due to forest succession that produced mature forest and prevented the
availability of browse. The result was another decrease in the Michigan deer herd. By
the 19705, managers refocused on improving deer habitat and in 1971 , the Deer Range
Improvement Program (DRIP) was established. This program allocated a certain portion
of the revenue generated from hunting license sales to be used for land acquisition, as
well as improving and maintaining deer habitat (Langenau 1994).

Deer habitat improvements were facilitated by more logging in northern Michigan
and an increase in agriculture in southern Michigan. These changes in land-use coupled
with mild winters and artiﬁcial feeding contributed to a 1989 peak in deer numbers. This
led the MDNR to set a goal of 1.3 million deer in the fall herd and speciﬁed that 35% of

the fall herd be antlered bucks. To advance this goal, an increased harvest of antlerless

12

deer was encouraged by the MDNR. In addition, block permits were issued to allow
landowners to harvest nuisance deer. The MDNR restated its commitment to this
population goal in 1997 and has been initiating regulations to help reach this target
population size since that time. Many managers and stakeholders, however, have
questioned the scientiﬁc basis for this population goal. Therefore, there is a need to re-
evaluate Michigan’s population goal from a biological and social context (i.e., public
attitudes toward deer in Michigan).

Traditionally, public attitudes about white-tailed deer focused on deer numbers.
Deer were not only a food source but the skins could be traded for money or other goods.
Therefore, in general, the public perception was that higher deer numbers were better.
Currently, attitudes about deer in Michigan depend on which sector of the public you are
dealing with. Overall, hunters still seem to be more concerned with deer numbers,
whereas the general public appears to be more concerned with the condition of the herd.
The public often uses the number of deer seen as roadkill and the number of deer-vehicle
accidents to form their perception of deer over— or under-abundance. Many individuals
attribute the perceived lack of deer as a direct result of over-harvest and are not aware of
the underlying habitat factors that could also inﬂuence deer population dynamics (W.
Moritz, MDNR, Wildlife Division, pers. commun.).

Public attitudes have inﬂuenced the current management strategy and facilitated
the MDNR’S shift in management focus from primarily managing for the size of the herd
to managing for the quality of the herd. This shift in management focus is evident
nationwide, as management objectives in several states have shifted from the promotion

of herd growth to population stabilization or herd reduction (Foster et a1. 1997).

13

The recent discovery of Bovine Tuberculosis (Myobacterium bovis) (TB) in
northeastern Michigan has also served as an impetus for the MDNR to examine herd
quality. Bovine TB is spread through aerosols and close contact of infected individuals is
necessary to transmit the disease. Poor habitat along with high deer population densities
have resulted in an unhealthy herd in which TB has been able to spread and become self-
sustaining (Schmitt et a1. 1997). In light of the emergence of this disease, and large
population ﬂuctuations, and habitat damage resulting from traditional deer management
strategies (McCabe and McCabe 1997), the MDNR realized the need to develop an
alternative management approach that would incorporate ecological as well as social and
biological objectives.

The current approach is more holistic than traditional management practices. It is
ecosystem-based and incorporates landscape-level planning into deer management
instead of primarily manipulating deer numbers to achieve management goals.
Ecosystem-based management emerged in the 19705 when managers began to
acknowledge that to develop realistic management strategies, “living systems” should be
regarded as complex and dynamic, varying at many spatial and temporal scales (Johnson
and Agee 1988, Grumbine 1994). Implementation of ecosystem-based management
requires a broad knowledge base that synthesizes the relationship among the structure and
function of ecosystem processes and composition, species demographics, and the
economic and social values of users of these ecosystems.

A hierarchical land classiﬁcation can provide the framework for developing
ecosystem-based deer management strategies by delineating areas of differing ecological

potential at various spatial scales (Albert et a1. 1986). Using these ecological units to

14

minimize spatial variation of habitat potential, the biological and cultural carrying
capacity can be estimated. Understanding these data allows the MDNR to establish deer
management goals for ecologically and culturally homogenous areas that integrate
biological and social signiﬁcance.

There is an established relationship among the productivity and physical condition
of deer and the quality of the habitat (Ford et al. 1997). Therefore, 1 way to achieve a
more accurate estimate of the biological carrying capacity, is to consider herd quality, or
health, because the physical condition of the deer herd can affect the maximum number
of deer that can be supported at equilibrium (McCullough 1982). Nutritionally stressed
deer are generally in poor physical condition and, thus, tend to experience low
reproductive rates, low survival rates, and low recruitment (Huot 1988). However, before
considering herd health in deer management efforts, managers need to have a deﬁnition
of quality or condition of white-tailed deer (Rasmussen 1985). According to Hamilton et
al. (1995) a quality deer herd is deﬁned by deer that are in good physical condition and
that are in balance with the existing habitat.

One of the best methods to measure the condition of deer is to evaluate the
physical growth of deer (Severinghaus 1955, Cowan and Long 1962, McCullough 1982,
Severinghaus and Moen 1983, Rasmussen 1985). Antler development, consisting of
average beam diameter and number of points, lactation status and fall recruitment have
been identiﬁed as measurable population-level quality indices (Hill et al. 1981, Cook and
Winterstein 2000). These population-level quality indices can reﬂect the possible
reproductive potential of the deer, which is important in deﬁning an appropriate scale at

which to measure the quality of the deer herd.

15

In addition to having a clear deﬁnition of herd quality, it is important to examine
spatial and temporal patterns in herd quality. Studying temporal and spatial variability in
herd condition may help isolate some of the underlying mechanisms (e. g., weather) that
cause shifts in the biological carrying capacity of the deer population. Spatial and
temporal trends may also be beneﬁcial for examining how population-level quality
indices correspond with certain ecological and social aspects of management. A
complete deﬁnition of a high quality deer population, coupled with ecological and social
considerations allows managers to develop sound management strategies as well as

inform the public about how and why management decisions are made.

16

Objectives
This study will deﬁne indices of herd quality for white-tailed deer in Michigan by
evaluating measurable characteristics over spatial and temporal scales to determine which
are appropriate indicators of herd health. Speciﬁcally, the following objectives will be
met:
1) Use experimental road survey methods to determine if adequate data can be
obtained to determine fall recruitment, as measured by fawn-to-doe ratios;
2) Determine the appropriate landscape scale at which to examine differences in
deer herd condition;
3) Assess whether population-level indices, such as antler measurements or
lactation status, can be used to examine differences in deer herd condition;
4) Use population-level indices to explore temporal and spatial patterns in the
condition of the deer population;
By meeting these objectives, this study will help the MDNR to develop an ecosystem-
based management approach for white-tailed deer in Michigan and, if successful, will

provide a framework for managing other species in a holistic manner.

17

METHODS

Data Sources
Biodata

Information on the biological characteristics of the deer herd is necessary to
evaluate population quality indices and examine how they reﬂect the biological carrying
capacity of white—tailed deer in different areas of the state. The MDNR provides this type
of biophysical data, known as the biodata, which have been collected every year since
1951 from deer that hunters voluntarily bring into MDNR check stations each deer
season. Currently, the MDNR maintains 4 highway check stations that are located along
3 major southbound highways and at the Mackinac Bridge. Other check stations are
located at 75 ﬁeld ofﬁces, state game areas, and recreation areas. These are dispersed
throughout the state to minimize a bias in the spatial distribution of the data (Cook and
Winterstein 2001).

The check stations are run by MDNR employees or trained volunteers. They
collect information on the location (county, deer management unit, township and range
coordinates, and private or public land), season (archery, ﬁrearm, muzzleloader, early
ﬁrearm, late ﬁrearm) and composition (sex and age) of the kill. This is in addition to the
information collected on physical condition which includes antler development of bucks
(number of points, left and right beam diameter), lactation status of does, and preliminary
TB status based on chest cavity observations (Cook and Winterstein 2001). Since 1987,
these data have been compiled into a database in Statistics Package for the Social

Sciences (SPSS). This study encompasses the data collected between 1987 and 2000.

18

Summer Driving T ransects

The biodata provide information about biological characteristics that are potential
quality indices, but other indices of herd condition needed to be examined. In addition to
antler development and lactation, fall recruitment, as assessed by fawn-to-doe ratios, is a
measure of deer reproductive potential (Ozoga et a1. 1994). To gather information on
fawn-to-doe ratios, an experimental protocol was developed to maximize the number of
deer observed in different habitat settings.

Summer driving transects were ﬁrst conducted in the summer of 2000 and based
on the ﬁndings, were modiﬁed in the summer of 2001. In the summer of 2000, driving
surveys were conducted by driving a ﬁxed route (transect) each morning and evening for
3 consecutive days during July, August, and September in the SLP and the NLP. The
summer driving transect in the UP, however, was conducted only in September due to
logistical constraints. There were a total of 4 driving transects, 1 in the SLP, 2 in the
NLP (designated NLPl and NLP2) and 1 in the UP, each of which consisted of a
different distance. To achieve complete driving route circuits, the starting point was
alternated between a clockwise and counterclockwise direction each morning and
evening. Each route in the SLP and NLP was driven a total of 6 times each month, 3
morning and 3 evening. The route in the UP, however, was driven 3 times, 1 morning and
2 evenings over a 2-day period. Two observers recorded the number of does, fawns and
bucks observed at the speciﬁc mileage along the route, in addition to the side of the road
(left, right or in), and the habitat type (woods, opening, row crop, pasture or

development) where deer were observed (Figure 1.4).

.Som new ooom Son Com $0855 wE>Eo 588% 05 .8.“ wow: Suﬁ 8% egg. 4..— Paar—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

qu_>o 22mg no.0 26m mcEmaO muoo>> umom c930 4..
3:95:00 552.. Co 25. Co 62m «25 Esau moo bangs. c2895
”$3.5 mcﬁcm ”2:08:50 5533
USP 9.3an AmreZomnO
as: Scam

”59:32 $28

an Nnjz F52 14w ”c050”. H2mm—

20

Beginning summer 2001, the July driving routes were eliminated so that survey
efforts could be concentrated on the latter part of August and September, and in 1 case
early October (N LPl ), to maximize the number of deer observed. Additionally, the
morning transects were eliminated because higher fawn-to-doe ratios and overall doe
counts existed in the evening. In the summer of 2001 , driving routes were conducted for
4 consecutive nights, alternating clockwise and counterclockwise directions, to achieve 3

complete circuits and maximize the number of does observed.

Sample Size

The data used for this project were historical in nature. Once the data were
collected, there was not an opportunity to increase the amount of data; sample sizes were
contingent on how many hunter-killed deer were checked each year. Therefore,
preliminary analyses were conducted to determine where sufﬁcient sample sizes existed
so that statistical analyses could be performed. Minimum required sample sizes were

calculated with the following equation:

n = z-’.VZp(I-p))/EZ

Where n is the required sample size, za/z is the value on the standard normal curve that
corresponds to (1 -a) percent conﬁdence, p is the proportion of interest, and E is the
desired margin of error (Wackerly 1996). In areas where sample sizes were not sufﬁcient
to conduct statistical analyses, these data were used to guide the aggregation of the data,

as well as the interpretation of the results.

21

Yearling Age-Class

When the project was initiated, population quality indices for all age classes of the
deer harvested were to be examined. After exploring the literature and the biodata, it was
determined that population quality indices from yearling deer were the best measure of
overall herd condition. The exception, however, was fall recruitment because it was
based on fawn-to-doe ratios from the summer driving routes, and it was impossible to
determine the age of the deer observed beyond the category of fawn or adult. Thus,
except for fall recruitment, the examination of the quality indices was focused on the
yearling age class for 2 reasons. First, this age class had the largest sample sizes.
Secondly, biological information collected from yearlings is more telling of the overall
herd quality because males in this age class have the added burden of completing body
grth while developing antlers (Ullrey 1983, Rasmussen 1985). Therefore, male
yearlings are more likely to have larger antlers with more points and wider average beam
diameter if they are in good condition. Similarly, female yearlings are more likely to

produce fawns if they are in good condition.

Population Quality Indices
Fall Recruitment

Information obtained from the summer driving routes was used to examine
differences in deer reproductive potential, expressed as fall recruitment, and measured by
fawn-to-doe ratios. Based on the number of adult does and the number of fawns

observed each time that a route was driven, a fawn-to-doe ratio was determined.

22

Number of Fawns Observed

Number of Adult Does Observed

 

FawnzDoe Ratio =

To detect differences in reproductive potential, fawn-to-doe ratios were compared
qualitatively at the regional level. The total, average, and maximum deer count per region
were calculated to examine the differences, as were the total, average, and maximum
fawn-to-doe ratio for each region. This was partially due to extremely small sample sizes
at smaller scales (i.e., county level). In addition, there was not enough information to

conduct any sort of quantitative statistical analyses.

Lactation Status

F awn-to-doe ratios are not the only possible way to estimate fall recruitment.
Another method is to examine the lactation status of does. This is appropriate because,
theoretically, the number of lactating does reﬂects the number of surviving fawns (Cook
and Winterstein 2000). In the biodata, lactation status reﬂects whether or not there was
milk present on a doe either when the deer was brought into the check station or when the
deer was ﬁeld dressed by the hunter. Does were classiﬁed as lactating or having milk
(HM+), or as not lactating or not having milk (HM-). Lactation status is a biophysical
index that has been collected only since 1993 and due to small sample sizes, these data
were aggregated to the regional level, even though these data were originally collected at

the county level. In addition, lactation status data are nominal-level data; they only deﬁne

23

the presence or absence of lactation in female deer. Therefore, only basic statistical

analyses such as descriptive statistics were employed.

Antler Measurements

Reproductive potential as measured by fawn-to-doe ratios is one possible method
for estimating deer herd condition (Cook and Winterstein 2000). Another possible
method is to evaluate antler measurements from harvested bucks that were recorded in
check station data and determine whether these indices can be substituted for
reproductive potential to determine the condition of deer. There were 2 types of antler
measurements examined: average beam diameter and point count. The average beam
diameter measurement was obtained when the right and left beam of a set of antlers were
each measured to the nearest millimeter with calipers approximately 1 inch above the
base of the skull and then averaged together. Point counts were obtained using the
eastern point count system, in which all of the points were counted on both the right and
left beam of a set of antlers. The antler measurement data are ratio-level data. Therefore,
unlike the lactation data, statistical analyses were appropriate for these data.

For each of the 3 study sites, differences in yearling average beam diameter and
point count were examined at a regional, county, and, for 1999 and 2000, township level
to determine the appropriate spatial scale at which to work. Tests for normalcy were
conducted and after the data were found to be normally distributed, analysis of variance
(ANOVA) was conducted to test for differences at different spatial scales. It was
established that even though there were statistically signiﬁcant differences among both

average beam diameter and point count at the township level, realistically the MDNR

24

would not manage deer at this spatial scale. Therefore, the county level was determined
to be the most appropriate spatial scale at which to examine differences in antler

measurements .

Within Regions

Examining whether speciﬁc quality indices can be used to estimate the biological
carrying capacity has limited usefulness. To truly understand what underlying
mechanisms cause shifts in herd health, it is necessary to explore how the physical
condition of deer changes temporally and spatially. Temporal trends give insight as to
how the physical condition of the deer population has changed over time, whereas spatial
trends give insight as to how the physical condition of the deer population changes over
space.

Since it was not appropriate to conduct analyses at the county level, all analyses
performed on the lactation data were conducted at the regional level. Within each region,
the number of does lactating each year from 1993-2000 was compared to the total
number of does checked for lactation in the same year to determine the percentage of
does lactating and the ratio of does lactating (HM+) to does not lactating (HM-). These
yearly percents were then plotted and compared to one another. Ranking and frequency
methods were used because it was not appropriate to employ quantitative methods on the
lactation data.

Unlike the lactation data, the analyses for average beam diameter and point count
were conducted at the county level within each study site. Preliminary analyses were

conducted to determine how the data were distributed. Frequency distributions were used

25

to determine the range of the average beam diameter and points for yearling male deer on
a yearly basis from 1987-2000. In addition, the mean average beam diameter and mean
point count per county were calculated on a yearly basis from 1987-2000. After initial
descriptive analyses, an ANOVA was conducted to examine whether there were
statistical differences among counties for either average beam diameter per county or
point count per county for each year from 1987-2000. Mean comparisons were then
examined using a Tukey-Kramer multiple comparison test with an alpha level of 0.05.

Temporal trends in both average beam diameter and point count were spatially
analyzed using a geographic information system (GIS; ArcView 3.2, Environmental
Systems Research Institute). This showed the spatial variability in the physical condition
of the deer herd, and allowed for an eventual overlay of the spatial distribution of quality
indices with a landscape-level habitat potential model that was developed as another part
of this project (A. Felix, unpublished data). This model minimized spatial variation using
Albert’s hierarchical land classiﬁcation, and provided a framework to assess the

biological carrying capacity within ecologically homogeneous areas.

Among Regions

The long-term percent of lactating does from 1993 through 2000 was calculated
for each of the 3 regions. This was accomplished by averaging the percent of does
lactating over the 8 years that these data were available for this study, which resulted in 1
average value for each region. These long-term regional averages were then plotted on a
line graph to examine trends among all 3 regions. Using the same method, the percent of

does lactating was also compared among the 3 regions on a yearly basis from 1993-2000.

26

The long-term mean for both average beam diameter and point count from 1987
through 2000 for each of the 3 regions was also calculated. These overall means were
calculated by averaging all of the values for each antler measurement over the 14 years of
this study, which resulted in 1 average value for each antler measurement for each region.
Regional averages were then compared by performing an ANOVA to determine if there
were statistically signiﬁcant differences in antler measurements among regions at an
alpha level of 0.05. If an overall difference among regions was detected, mean
comparisons were conducted using a Tukey-Kramer multiple comparison test at an alpha
level of 0.05. In addition, yearly comparisons were conducted for both average beam
diameter and point count for 1987-2000 among the 3 regions. This was accomplished by
conducting an ANOVA and mean comparisons were again examined using a Tukey-

Kramer multiple comparison test at an alpha level of 0.05.

Relationship Among Indices

Once qualitative and statistical differences in the quality indices were established
at the appropriate spatial scale, and the quality indices were compared both within and
among study sites, the potential relationships among the quality indices were examined.
Scatter plots were used to investigate whether a relationship existed between lactation
status and average beam diameter or point count.

The relationship between average beam diameter and point count, as well as the
relationship between each of the antler measurements and lactation status was examined.

Regression analyses were used to determine the relationship between point count and

27

average beam diameter. Scatter plots were also used to examine if there was a correlation

between average beam diameter or point count and lactation status of does.

28

RESULTS

Sample Sizes
UP

Of the 3 study sites, sample sizes of yearling deer in the UP study site were
generally smaller than those in the NLP study site, but larger than those in the SLP study
site. This trend was consistent for yearling males that had antler measurements taken, but
not for yearling does that were checked for lactation. In the UP study site, the number of
male yearlings for which an average beam diameter measurement was calculated ranged
between 146 in 1997 and 1464 in 1994 (Table 1.1a). When the sample size was broken
down by county, Marquette and Iron County had the largest overall number of male
yearlings for which an average beam diameter measurement was calculated (n=3157 and
n=2850, respectively) (Table 1.1a). Of the 4 counties, Baraga and Dickinson County had
the least total number of male yearlings for which an average beam diameter
measurement was calculated (n=2112 and n=1910, respectively) (Table 1.1a).

The number of male yearlings for which a point count was generated followed the
same trend as the number of male yearlings for which an average beam diameter
measurement was calculated. In the UP study site, the number of male yearlings for
which a point count was generated ranged between 145 in 1997 and 1474 in 1994 (Table
1.1b). When the sample size was broken down by county, Marquette and Iron County had
the largest overall number of male yearlings for which a point count was generated
(n=3182 and n=2922, respectively) (Table 1.1b). Baraga and Dickinson County had the
least overall number of male yearlings for which a point count was generated (n=2109

and n=1960, respectively) (Table 1.1b).

29

Table 1.1. Sample size of yearling deer in each county in the Upper Peninsula study site
for (3) males that had an average beam diameter measured, (b) males that had number of
points counted, and (0) females that were checked for lactation status.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) Avera 5e Beam Diameter

BARAGA DICKINSON IRON MARQUETTE Total
1987 144 184 248 232 808
1988 160 205 247 239 851
1989 88 147 168 215 618
1990 106 161 171 227 665
1991 203 1 17 190 239 749
1992 145 69 89 174 477
1993 165 96 118 209 588
1994 329 269 3 82 484 1464
1995 240 131 276 297 944
1996 61 45 62 67 235
1997 17 26 51 52 146
1998 121 94 266 205 686
1999 157 160 316 247 880
2000 176 206 266 270 918
Total 2112 1910 2850 3157 10029

(b) Point Count

BARAGA DICKINSON IRON MARQUETTE Total
1987 142 180 247 227 796
1988 157 212 257 241 867
1989 86 154 173 219 632
1990 108 178 193 236 715
1991 203 1 15 184 234 736
1992 145 76 101 181 503
1993 167 100 121 211 599
1994 327 271 392 484 1474
1995 241 136 282 309 968
1996 61 49 65 68 243
1997 17 26 50 52 145
1998 120 92 266 207 685
1999 157 159 315 240 871
2000 178 212 276 273 939
Total 2109 1960 2922 3182 10173

 

 

 

 

 

 

 

 

30

Table 1.1 (con’t)

(c) Lactation

 

 

 

BARAGA DICKINSON IRON MARQUETTE Total

1993 0 4 5 2 1 l
1994 2 5 9 7 23
1995 9 l 5 52 27 103
1996 3 l4 8 10 3 5
1997 0 0 2 6 8
1998 0 7 2 l 10
1999 4 l 3 23 l 0 50
2000 6 27 43 19 95
Total 24 85 144 82 33 5

 

 

 

 

 

 

 

 

31

The number of yearling does that were checked for lactation followed a slightly
different trend than for the number of male yearlings for which an average beam diameter
or point count were calculated (Table 1.1c). In the UP study site, the number of yearling
does checked for lactation ranged between 8 in 1997 and 103 in 1995 (Table 1.1c). When
the sample size was broken down by county, Iron and Dickinson County had the largest
overall number of yearling does checked for lactation (n=144 and n=85, respectively)
(Table 1.1c). Marquette and Baraga County had the least overall number of yearling does

that were checked for lactation (n=82 and n=24, respectively) (Table 1.1c).

NLP

Based upon the biodata, the largest sample sizes of yearling deer were located in
the NLP study site. This included the sample size of yearling deer for each of the
population-level quality indices: yearling average beam diameter, point count, and
lactation status. In the NLP study site, the number of male yearlings for which an
average beam diameter measurement was calculated ranged between 887 in 1992 and
2313 in 1998 (Table 1.2a). When the sample size was broken down by county, Alcona
and Alpena County had the largest overall number of male yearlings for which an
average beam diameter measurement was calculated (n=5337 and n=5294, respectively)
(Table 1.2a). Oscoda County had slightly fewer male yearlings for which an average
beam diameter measurement was calculated (n=4233) (Table 1.2a). Montmorency and
Presque Isle County had the least overall number of male yearlings for which an average
beam diameter measurement was calculated (n=3962 and n=3244, respectively) (Table

1.2a).

32

Table 1.2. Sample size of yearlings in each county of the Northern Lower Peninsula
study site for (a) males that had average beam diameter measured, (b) males that had the
number of points counted, and (0) females that were checked for lactation status.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a) Average Beam Diameter

ALCON A ALPENA MONTMORENCY OSCODA PRESQUE ISLE Total
1987 438 422 318 377 246 1801
1988 353 352 294 290 252 1541
1989 184 239 180 228 165 996
1990 424 284 373 426 21 1 1718
1991 402 265 335 415 182 1599
1992 203 172 178 179 155 887
1993 189 262 160 141 161 913
1994 392 396 299 363 214 1664
1995 400 333 298 368 229 1628
1996 475 541 277 309 158 1760
1997 460 410 305 303 260 1738
1998 503 659 392 369 390 2313
1999 448 564 252 208 276 1748
2000 466 395 301 257 345 1764
Total 5337 5294 3962 4233 3244 22070

(b) Point Count

ALCONA ALPENA MONTMORENCY OSCODA PRESQUE ISLE Total
1987 430 418 313 367 244 1772
1988 361 366 321 297 255 1600
1989 185 238 187 229 165 1004
1990 427 280 380 438 213 1738
1991 399 256 335 415 180 1585
1992 222 174 179 181 156 912
1993 190 264 164 157 160 935
1994 384 385 303 358 216 1646
1995 415 334 306 372 232 1659
1996 536 589 300 311 167 1903
1997 461 414 312 304 266 1757
1998 560 659 407 395 407 2428
1999 508 564 300 251 294 1917
2000 460 397 302 259 351 1769
Total 5538 5338 4109 4334 3306 22625

 

 

 

 

 

 

 

 

33

 

 

Table 1.2 (Con’t)

(c) Lactation

 

 

 

 

 

 

 

 

 

 

ALCONA ALPENA MONTMOREN CY OSCODA PRESQUE ISLE Total
1993 16 24 15 l3 13 81
1994 12 8 5 l4 8 47
1995 18 23 l3 19 18 91
1996 39 51 44 26 13 173
1997 55 43 47 48 15 208
1998 259 171 179 158 68 835
1999 108 154 97 1 12 82 553
2000 137 88 73 87 68 453
Total 644 562 473 477 285 2441

 

34

 

The number of male yearlings for which a point count was generated followed the
same trend as the number of male yearling for which an average beam diameter
measurement was calculated. In the NLP study site, the number of male yearlings for
which a point count was generated ranged between 912 in 1992 and 2428 in 1998 (Table
1.2b). When the sample size was broken down by county, Alcona and Alpena County had
the largest overall number of male yearlings for which a point count was generated
(n=5538 and n=553 8, respectively) (Table 1.2b). Oscoda County had slightly fewer male
yearlings for which a point count was generated (n=4334) (Table 1.2b). Montmorency
and Presque Isle County had the least overall number of male yearlings for which a point
count was generated (n=4109 and n=3306, respectively) (Table 1.2b).

The number of yearling does that were checked for lactation also followed the
same trend as the number of male yearlings for which an average beam diameter or point
count were calculated (Table 1.2c). In the NLP study site, the number of yearling does
checked for lactation ranged between 47 in 1994 and 835 in 1998 (Table 1.2c). When the
sample size was broken down by county, Alcona and Alpena County had the largest
overall number of yearling does checked for lactation (n=644 and n=562, respectively)
(Table 1.2c). Oscoda County had slightly fewer yearling does checked for lactation
(n=477) (Table 1.2c). Montmorency and Presque Isle County had the least overall
number of yearling does that were checked for lactation (n=473 and n=285, respectively)

(Table 1.2c).

35

SLP

In general, the SLP study site had the smallest sample sizes of yearling deer
among the 3 study sites, with the exception of the number of yearling does that were
checked for lactation, which was larger than the number of yearling does checked for
lactation in the UP study site. Also, the yearly sample sizes were more consistent in this
study site than the other 2 sites. In the SLP study site, the number of male yearlings for
which an average beam diameter measurement was calculated ranged between 273 in
1987 and 394 in 1995 (Table 1.3a). When the sample size was broken down by county,
Barry County had the largest overall number of male yearlings for which an average
beam diameter measurement was calculated (n=3280), whereas Eaton County had the
smallest (n=530) (Table 1.3a). In Calhoun County, the number of male yearling for
which an average beam diameter measurement was calculated fell in between the number
of male yearlings for the other 2 counties (n=530) (Table 1.3a).

The number of male yearlings for which a point count was generated followed the
same trend as the number of male yearlings for which an average beam diameter
measurement was calculated. In the SLP study site, the number of male yearlings for
which a point count was generated ranged between 257 in 1991 and 387 in 1995 (Table
1.3b). When the sample size was broken down by county, Barry County had the largest
overall number of male yearlings for which a point count was generated (n=3170), and
Eaton County had the least overall number of male yearlings for which a point count was
generated (n=514) (Table 1.3b). In Calhoun County, the number of male yearlings for
which a point count was generated fell in between the number of male yearlings for the

other 2 counties (n=656) (Table 1.3b).

36

Table 1.3. Sample size of yearling deer in each county in the Southern Lower Peninsula
study site for (a) males that had an average beam diameter measured, (b) males that had
the number of points counted, and (c) females that were checked for lactation status.

 

 

 

 

 

 

 

 

 

 

 

 

 

(a) Average Beam Diameter

BARRY CALHOUN EATON Total
1987 220 37 16 273
1988 215 43 30 288
1989 248 34 29 311
1990 230 53 25 308
1991 211 43 21 275
1992 223 39 17 279
1993 212 41 26 279
1994 261 48 82 391
1995 264 65 65 394
1996 264 42 43 349
1997 243 62 39 344
1998 241 56 36 333
1999 229 59 52 340
2000 219 55 49 323
Total 3280 677 530 4487

(b) Point Count

BARRY CALHOUN EATON Total
1987 210 35 16 261
1988 210 38 29 277
1989 238 34 26 298
1990 223 45 23 291
1991 195 41 21 257
1992 219 37 15 271
1993 203 37 25 265
1994 247 52 80 379
1995 259 64 64 387
1996 260 43 43 346
1997 236 61 38 335
1998 234 56 36 326
1999 224 59 49 332
2000 212 54 49 315
Total 3 170 656 5 l 4 4340

 

 

 

 

 

 

 

37

Table 1.3 (Con’t)

(c) Lactation

 

 

 

 

 

 

 

 

BARRY CALHOUN EATON Total
1993 22 13 2 37
1994 35 21 5 61
1995 26 19 6 51
1996 39 29 10 78
1997 64 41 12 117
1998 58 33 8 99
1999 46 15 7 68
2000 53 20 8 81
Total 343 191 58 592

 

38

 

 

The number of yearling does that were checked for lactation followed a slightly
different trend than did the number of male yearlings for which an average beam
diameter or point count were calculated (Table 1.3c). In the SLP study site, the number
of yearling does checked for lactation ranged between 37 in 1987 and 117 in 1997 (Table
3c). When the sample size was broken down by county, Barry County had the largest
overall number of yearling does checked for lactation (n=343) (Table 1.3c), whereas
Eaton County had the least overall number of yearling does that were checked for
lactation (n=5 8) (Table 1.3c). In Calhoun County, the number of yearling does checked
for lactation fell in between the number of yearling does for the other 2 counties (n=191)

(Table 1.3c).

Fall Recruitment

Throughout the 2000 summer, there were, in general, more deer observed during
the evening transects than in the morning transects, regardless of which month the survey
was conducted (Table 1.4). The average number of fawns observed in the morning
ranged from 10 to 21, whereas the average number of fawns observed in the evening
ranged from 12 to 57 (Table 1.4). Similarly, the average number of does observed in the
morning ranged from 24 to 52, and the average number of does observed in the evening
ranged from 26 to 108 (Table 1.4). Fawn-to-doe ratios, on the other hand, were not
consistently higher in the evening. The average fawn-to-doe ratio in the UP site and
NLP2 site was higher in the evening (0.70 and 0.65, respectively) than in the morning
(0.59 and 0.45, respectively) (Table 1.4a&c). However, the average fawn-to-doe ratio in
the NLPl and SLP was higher in the morning (0.28 and 0.58, respectively) than in the

evening (0.27 and 0.49 respectively) (Table 1.4b&d).

39

Table 1.4. Summer of 2000 total, average, and maximum fawn-to-doe ratio based on
both morning and evening deer counts for each regional route in July, August, and
September (b,c,d) and September only (a) .

(a) The Upper Peninsula

Area Month Fawn Doe Ratio Fawn Doe
E E
UP 1 7 29 0.59 5 7 81

Total 17 29 0.59 57 81 0.70
A 17 29 0.59 81 0.70
Max 17 29 0.59 57 81 0.70

(b) The Northern Lower Peninsula Route #1

Area Month Fawn Doe Ratio Fawn Doe Ratio
E E
NLPI Jul 17 64 0.27 15 120 0.13
A 15 55 0.27 28 119 0.24
1 1 37 0.30 38 85 0.45

 

Total 43 0.28 81 324 0.25
A 14 0.28 27 108 0.27
Max 17 0.3 38 120 0.45

(c) The Northern Lower Peninsula Route #2

Area Month Fawn Ratio Fawn Doe Ratio
E E
NLP2 Jul 9 28 0.32 11 49 0.22
A 8 16 0.50 15 16 0.94
14 27 0.52 11 14 0.79

 

Total 31 71 0.44 37 79 0.47
A 10 24 0.45 12 26 0.65
Max 14 28 0.52 15 49 0.94

 

(d) The Southern Lower Peninsula

Area Month Fawn Doe Ratio Fawn Doe Ratio
E E
SLP Jul 60 0.12 83 0.11
A 35 0.83 46 0.78

34 0.79 89 0.58

0.49 0.44
0.58
0.83

 

40

Deer numbers and fawn-to-doe ratios also differed by month, except in the UP
site where transects were only conducted in September (Table 1.5a). For the NLPl site,
deer numbers were highest in July and August (n=271 and n=273, respectively), with a
slight decline in September (n=216) (Table 1.5b). In the NLP2 site, the highest deer
numbers were observed in July (n=120) and there was a sharp decline in deer numbers
for August and September (n=71 and n=76, respectively) (Table 1.5c). Deer numbers in
the SLP were, conversely, lower in July and August (n=198 and n=194, respectively)
than in September (n=252) (Table 1.5d). The number of does observed, for the most part,
declined as the summer progressed, but the number of fawns observed tended to increase
(Tables 1.4&1.5). This may explain why generally there were higher fawn-to-doe ratios
in August and September than in July (Table 1.4). For the NLPl site, the fawn-to-doe
ratio steadily increased throughout the 3 months for both the morning and evening
transects; morning fawn-to-doe ratios increased ﬁom 0.27 to 0.30 and evening fawn-to-
doe ratios increased from 0.13 to 0.45 (Table 1.4b). The morning fawn-to-doe ratios for
the NLP2 site increased from 0.32 in July to 0.52 in September and the evening fawn-to-
doe ratios increased from 0.22 in July to 0.94 in August then declined slightly to 0.79 in
September (Table 1.40). The morning fawn-to-doe ratios in the SLP site increased from
0.12 in July to 0.83 in August then declined slightly to 0.79 in September (Table 1.4d).
The evening fawn-to-doe ratios in the SLP site followed the same trend; fawn-to-doe
ratios increased from 0.11 in July to 0.78 in August then declined to 0.58 in September

(Table 1.4d).

41

Table 1.5. Total deer numbers observed based on the summer 2000 driving transect deer
counts for each regional route in July, August, and September (b,c,d) and September

only (a).

(a) The Upper Peninsula

Area Month Fawn
UP 74

Total 74

 

(b) The Northern Lower Peninsula Route #1
Area Month Fawn
NLPl Jul 32
A 43
49

 

Total 1 24

(c) The Northern Lower Peninsula Route #2
Area Month Fawn
NLP2 Jul 20
A 23
25

Total 68 25

 

(d) The Southern Lower Peninsula

 

 

 

 

 

 

Area Month Fawn Doe Buck nknowri Total
SLP July 16 143 21 18 198
August 65 81 21 27 194
September 79 123 21 29 252
Total 1 60 347 63 74 644

 

 

 

 

 

 

 

 

 

 

42

Of the deer observed along each transect, the majority were does, regardless of
region, followed by the number of fawns, bucks, and unknown deer observed (Table 1.5).
There were more deer observed in each category in the NLP] site than in the other 3 sites

(total n=760) (Table 1.5b). The number of deer observed in each category in the SLP
site was slightly lower than in the NLP] site (total n=644) (Table 1.5d), whereas the
number of deer observed in each category in the NLP2 site was much lower than in the
NLPI site (total n=267) (Table 1.5c). The fewest deer in each category were observed in
the UP site (total n=195) (Table 1.5a), this may be a reﬂection of the fact that deer were
only observed during 1 month (September).

In the summer of 2001, deer numbers in the NLPl site were highest in September
(n=164), and declined in October (n=109) (Table 1.7b). In the NLP2 site, the highest
deer numbers were observed in August (n=204) with a sharp decline in deer numbers for
September (n=135) (Table 1.7c). Deer numbers in the SLP were also higher in August
(n=121) than in September (n=90) (Table 1.7d). In addition, the number of does
observed, for the most part, declined as the summer progressed, as did the number of
fawns (Tables 1.6&l.7).

The fawn-to-doe ratios were not only more consistent but also higher in 2001
when compared to the fawn-to-doe ratios in the summer of 2000. There did not appear to
be large differences in the fawn-to-doe ratios among months or among transect sites,
although in general the fawn-to-doe ratios increased in September and October, with the
exception of the NLP2 site (Table 1.6). The fawn-to-doe ratio in the NLP] site increased
from 1.06 in September to 1.23 in October (Table 1.6b), whereas the fawn-to-doe ratio in

the NLP2 site decreased from 1.00 in August to 0.88 in September (Table 1.6c). Fawn-

43

Table 1.6. Summer of 2001 total, average, and maximum fawn-to-doe ratios based on
evening deer counts only for each regional route in August and September (c,d),
September and October (b) , and September only (a).

(a) The Upper Peninsula

Area anth Fawn
UP 35

Total 35
35
Max 35

 

(b) The Northern Lower Peninsula Route #1

Area Month Fawn
MP1 73
October 53

Total 126
63
Max 45

 

(c) The Northern Lower Peninsula Route #2

Area Ninth Fawn
NIP2 83
53

Total 136
68
What 83

 

(d) The Southern Lower Peninsula
Area anth Fawn Doe Ratio
SIP 45 59
36 37 0.97

81 0.84

 

96
40.5 48
45 59

0.97

44

Table 1.7 . Total deer numbers observed based on the summer 2001 deer counts for each
regional route in August and September (c,d), September and October (b), and September

only (a).

 

 

 

 

(a) The Upper Peninsula
Ara anth Fawn Doe Iilck Unlcmwn Total
UP Sqltarba 35 31 17 32 115
Total 35 31 17 32 115

 

 

 

 

 

 

 

 

 

 

(b) The Northern Lower Peninsula Route #1

 

 

 

 

 

Area Month Fawn Doe Buck Unknown Total
NIP] Septerrber 73 69 18 4 164
October 53 43 1 1 2 109
Total 126 1 12 29 6 273

 

 

 

 

 

 

 

 

 

 

(c) The Northern Lower Peninsula Route #2

 

 

 

 

 

Area Month Fawn Doe Buck Unlmown Total

NIP2 Auglst 83 83 29 9 204

Sthen'ber 53 60 17 5 135

Total 136 143 46 14 339

 

 

 

 

 

 

 

 

 

 

(d) The Southern Lower Peninsula
Area Month Fawn Doe Buck Urllmown Total
SIP August 45 59 11 121

Sglterrba 36 37 ll 6 90

Total 81 96 22 12 211

 

 

O\

 

 

 

 

 

 

 

 

 

 

 

 

 

45

to-doe ratios in the SLP site increased from 0.76 in August to 0.97 in September (Table
1.6d). The direction of change in the UP site was unknown because transects were
conducted only in September (Table 1.6a).

The sex and age composition of deer observed along each transect were similar in
2000 and 2001. The majority were does, regardless of region, followed by the number
fawns, bucks, and unknown deer observed (Table 1.7). There were, however, fewer
unknown deer in 2001 when compared with 2000 (Tables 1.5&1.7). There were more
deer observed in each category in the NLP2 site than in the other 3 sites (total n=339)
(Table 1.7c). The number of deer observed in each category in the NLP] site was lower
than in the NLP2 site (total n=273) (Table 7b), whereas the number of deer observed in
each category in the SLP site was slightly lower than in the NLPl site (total n=211)
(Table 1.7d). The fewest deer in each category were observed in the UP site (total
n=115) (Table 1.7a), this may be a reﬂection of the fact that deer were only observed

during 1 month (September).

Comparisons within Study Sites
UP

Of the total number of yearling does that were checked for lactation in the UP
study site, the majority were not lactating (HM-) (Table 1.8). The percent of yearling
does that were not lactating ranged between 81.8% and 100%, whereas the percent of
yearling does that were lactating (HM+) ranged between 0% and 19.4% (Table 1.8).
Overall, the number of does that were checked for lactation in the UP study site was

relatively small (Figure 1.5). In fact in only 1 year (1995) did the number of yearling does

46

Table 1.8. The lactation status of yearling does harvested in the Upper Peninsula study
site each year between 1993 and 2000, as well as the percent of yearling does lactating
(HM+) and not lactating (HM-) each year.

Number of Yearling Does

90

 

 

 

 

 

 

 

 

 

 

Year HM+ HM- Total %HM+ %HM-
1993 2 9 1 1 18.2 81.8
1994 3 20 23 13 87
1995 20 83 103 19.4 80.6
1996 5 30 35 14.3 85.7
1997 0 8 8 0 100
1998 0 10 10 0 100
1999 4 46 50 8 92
2000 10 85 95 10.5 89.5
HM- I HM+

 

 

1993

1 994

1995

1996

 

 

1997

Year

1998

1999

 

Figure 1.5. Lactation status of yearling does harvested in the Upper Peninsula study site
for each year between 1993 and 2000.

47

checked for lactation exceed 100 (Table 1.8; Figure 1.5). Years (e.g., 2000) in which a
relatively large number of does were checked for lactation in the UP study did not
necessarily translate into a higher percent of lactation (Table 1.8; Figure 1.6). There were,
however, years (e. g., 1993) in which small numbers of yearling does checked for
lactation in the UP study site did translate into a higher percent of lactation (Table 1.8;
Figure 1.6). In addition, there were several years in which, of the yearling does checked
for lactation, there were no does lactating (e.g., 1997 and 1998) (Table 1.8; Figure 1.6).

In the UP study site, the overall mean average beam diameter for male yearlings
across the 14 years of this study was 16.8mm and this index ranged between 16.1mm in
1995 and 17.5mm in 1987 for the 14 years of the study (Table 1.9). The mean average
beam diameter for male yearlings in the UP study site peaked in 1987, 1991, and 1998;
this index tended to oscillate around 17mm, but never fell below 16mm (Figure 1.7). The
minimum average beam diameter for male yearlings was 6mm in Marquette County in
1994 and 1995 as well as in Dickinson County in 1989, although most of the low
averages ranged between 8mm and 11mm (Table 1.10). The maximum average beam
diameter for male yearlings was 31.5mm in Baraga County in 1992 and Iron County in
1987; however, the majority of the high averages ranged between 25mm and 29mm
(Table 1.10).

The results from the ANOVA model run with male yearling average beam
diameter indicated that the year variable was statistically signiﬁcant (F13, 9973 = 15.03; P <
0.0001). Also, the results from the Tukey-Kramer multiple comparison test illustrated

which years the male yearling average beam diameter was signiﬁcantly different from

48

+ Yearly Average - - Long-Term Average
25.0

 

20.0

15.0

10.0

Percent of Does Lactating

5.0 «e

 

 

   

0.0

 

1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.6. The percent of yearling does harvested in the Upper Peninsula study site that
were lactating during each year between 1993 and 2000.

49

Table 1.9. The mean average beam diameter in mm for male yearlings harvested in each
county in the Upper Peninsula study site for each year between 1987 and 2000 and the
standard deviation, as well as the yearly averages for the entire study site and overall
averages for each county.

 

 

 

Year Baraga Dickinson Iron Marquette Yearly Average StDev
1987 19.0 16.8 17.7 16.9 17.5 0.99
1988 17.4 16.4 17.1 16.6 16.8 0.46
1989 18.9 15.5 16.9 15.5 16.4 1.62
1990 18.1 16.1 16.8 16.2 16.6 0.90
1991 18.1 16.6 16.9 17.1 17.2 0.63
1992 18.2 15.1 16.7 16.9 17.0 1.26
1993 17.0 16.6 17.2 16.1 16.7 0.46
1994 16.5 16.6 17.0 16.2 16.5 0.31
1995 16.0 15.3 16.5 16.1 16.1 0.49
1996 18.0 14.7 15.7 15.9 16.2 1.36
1997 15.6 16.6 16.0 16.6 16.3 0.49
1998 18.1 16.9 17.1 17.3 17.3 0.54
1999 17.3 17.5 17.0 17.2 17.2 0.17
2000 17.8 16.8 16.4 17.2 17.0 0.63
Total 17.5 16.4 16.9 16.6 16.8 0.47

 

 

 

 

 

 

 

 

 

Table 1.10. The minimum and maximum average beam diameter in mm collected from
male yearlings harvested in each county in the Upper Peninsula study site for each year
between 1987 and 2000.

 

 

 

Baraga Dickinson Iron Marc uette

Min Max Min Max Min Max Min Max
1987 12 26.5 9 25 8 31.5 9.5 26.5
1988 7.5 25 7 28 10 26 8 30
1989 10 37 6 23.5 10.5 25 7.5 24
1990 10 30.5 8.5 25 8.5 26 7 26
1991 12 26 10.5 25 10 25 9.5 27.5
1992 10 31.5 7.5 25 10 25 10 26
1993 9.5 26.5 9 25.5 11 26 9 24.5
1994 9 25 8.5 29.5 9.5 25.5 6 31
1995 9 25.5 7.5 28.5 8.5 24.5 6 27
1996 9.5 29 7.5 24.5 7 25 1,1 23
1997 11.5 21 8 22.5 10.5 24 11.5 27.5
1998 10 26.5 8.5 28 7 25 9 26
1999 11.5 29 10.5 27.5 9 26 10 25.5
2000 7.5 28 6 26 7 25 9.5 25.5

 

 

 

 

 

 

 

 

 

 

 

50

+Yearly Average " " Long-Term Average

 

17.75

17.5 -- e ~ ~ 7

17.25

17

16.75

Average Beam Diameter

16.5

16.25

 

16

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.7. The temporal change in average beam diameter in mm for male yearlings
harvested in the Upper Peninsula study site between 1987 and 2000.

51

one another (Table 1.11). The Tukey-Kramer multiple comparison test results showed
that the male yearling average beam diameter in each of 1987 and 1995 was signiﬁcantly
different (P S 0.05) from the male yearling average beam diameter in most of the other
years in the study (Table 1.11). Also, male yearling average beam diameter in each of
1998, 1999, and 2000 was signiﬁcantly different (P S 0.05) from the male yearling
average beam diameter in the middle years of the study as well as from each other (Table
1.11). There were additional years in which the male yearling average beam diameter
was signiﬁcantly different (P S 0.05) from one another, although there was not as much
of a distinct pattern (Table 1.11).

When the mean average beam diameter was examined by county, male yearlings
in Baraga and Iron County had the largest average beam diameter measurements in the
UP study site (1? = 17.5mm and 56 = 16.9mm, respectively) (Figure 1.8). Marquette and
Dickinson County had the smallest average beam diameter measurements in the UP study
site ( J? = 16.6mm and f = 16.4mm, respectively) (Figure 1.8). The results from the
ANOVA model also indicated that the county variable was statistically signiﬁcant (F3,
9973 = 44.42; P < 0.0001). According to the Tukey-Kramer multiple comparison test, the
average beam diameter for male yearlings in Baraga County was signiﬁcantly different (P
S 0.05) from the average beam diameter in each of the other 3 counties (Table 1.12). In
addition, the average beam diameter for male yearlings in Dickinson County was also
signiﬁcantly different (P S 0.05) from the average beam diameter in each of the other 3
counties (Table 1.12).

The overall point count for male yearlings in the UP study site was 3.086 and the

average point count for male yearlings ranged between 2.88 in 1990 and 3.34 in 1998

52

Table 1.11. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Upper Peninsula study site on a yearly basis. The x denotes a signiﬁcant
difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Table 1.12. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Upper Peninsula study site on a county basis. The x denotes a signiﬁcant
difference at an alpha level of 0.05.

Dickinson Iron

Dickinson
Iron
uette

 

53

 

  

Mean Yearling Average Beam Diameter

l:l ﬁckhson (16.4 mm)
erqlette (16.6 mm)
- I'on (16.9 mm)

- Baraga (17.5 mm)

Figure 1.8. Mean yearling average beam diameter by county in the Upper Peninsula
study site for 1987-2000.

54

(Table 1.13). The yearly average point count for male yearlings in the UP study site
peaked in 1987, 1991, and 1998; this index oscillated around 3, but never fell below 2.8
(Figure 1.9). The minimum number of points for male yearlings in this study site was 2,
which was consistent among years and counties (Table 1.14). The maximum number of
points for male yearlings reached 20 in Marquette County in 1994; however, most of the
maximum number of points ranged between 7 and 10 (Table 1.14).

The results of the ANOVA model run with the male yearling points indicated that
the year variable was statistically signiﬁcant (F13,10117 = 8.83; P < 0.0001), and the results
from the Tukey-Kramer multiple comparison test illustrated which years the male
yearling point count was signiﬁcantly different from one another (Table 1.15).
According to the Tukey-Kramer multiple comparison test, the point count for male
yearlings in each of 1987, 1998, and 1999 was signiﬁcantly different (P S 0.05) from the
point count for the majority of the other years in the study, especially those years between
1989 and 1996 (Table 1.15). The point count for male yearlings in other years were
signiﬁcantly different (P S 0.05) from one another; however, there was not a distinct
pattern (Table 1.15).

When the average point count was examined by county, male yearlings with the
most number of points were found in Baraga and Marquette County (f = 3.34 and 56 =
3.08, respectively) (Figure 1.10). On the other hand, male yearlings in Iron and
Dickinson County had the fewest number of points (if = 3.01 and 5:" = 2.93, respectively)
(Figure 1.10). The results from the ANOVA model also indicated that the county variable
was statistically signiﬁcant (F3, 1m 17 = 20.78; P < 0.0001). The Tukey-Kramer multiple

comparison test results showed that the point count for male yearlings in Baraga County

55

Table 1.13. The mean point count for male yearlings harvested in each county in the
Upper Peninsula study site for each year between 1987 and 2000 and the standard

deviation, as well as the yearly averages for the entire study site and overall averages for

 

 

 

each county.
Year Baraga Dickinson Iron Marquette Yearly Avergg StDev
1987 4.14 2.85 3.25 3.24 3.32 . 0.55
1988 3.34 3.04 3.08 3.11 3.13 0.13
1989 3.36 2.69 3.01 2.89 2.94 0.28
1990 3.53 2.76 2.75 2.79 2.88 0.38
1991 3.70 2.86 2.88 3.24 3.22 0.40
1992 3.59 2.62 2.78 2.90 3.03 0.43
1993 3.18 3.04 2.97 3.03 3.06 0.09
1994 2.99 2.82 3.00 3.09 2.99 0.11
1995 3.08 2.71 2.78 2.94 2.90 0.16
1996 3.13 2.57 2.88 2.94 2.90 0.23
1997 2.47 3.27 2.90 3.15 3.01 0.35
1998 3.69 3.20 3.35 3.17 3.34 0.24
1999 3.17 3.40 3.26 3.39 3.30 0.11
2000 3.21 3.16 2.79 3.18 3.07 0.20
Total 3.34 2.93 3.01 3.08 3.09 0.18

 

 

 

 

 

 

 

 

 

Table 1.14. The minimum and maximum number of points counted on male yearlings
harvested in each county in the Upper Peninsula study site for each year between 1987
and 2000.

 

 

 

 

 

 

 

 

 

 

Baraga Dickinson Iron Marc uette
Min Max Min Max Min Max Min Max

1987 2 9 2 8 2 9 2 8
1988 2 1 3 2 8 2 8 2 8
1989 2 8 2 7 2 8 2 8
1990 2 9 2 8 2 8 2 8
1991 2 8 2 7 2 8 2 8
1 992 2 1 5 2 6 2 8 2 9
1993 2 9 2 8 2 7 2 8
1994 2 8 2 8 2 8 2 20
1995 2 8 2 8 2 8 2 8
1996 2 8 2 6 2 8 2 6
1997 2 5 2 6 2 8 2 8
1998 2 10 2 8 2 10 2 7
1999 2 1 1 2 8 2 8 2 8
2000 2 10 2 16 2 7 2 8

 

 

 

 

56

—-O- Yearly Average " ' Long-Term Average

 

3.4

3.3» , ~ ~

3.2 —

3.1 i

Average Point Count

2.9

 

 

2.8

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.9. The temporal change in average point count for male yearlings harvested in
the Upper Peninsula study site between 1987 and 2000.

57

Table 1.15. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Upper Peninsula study site on a yearly basis. The x denotes a signiﬁcant
difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

1987
1988
1989
1990

1992
1993
1994
1995
1996
1997
1998
1999
2000

 

Table 1.16. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Upper Peninsula study site on a county basis. The x denotes a signiﬁcant
difference at an alpha level of 0.05.

Dickinson Iron

Dickinson
Iron
uette

 

58

 

  

Yearling Average Point Count
[:1 Dickinson (2.93)
Iron (3.01)

- Marquette (3.08)
- Baraga (3.34)

Figure 1.10. Yearling average point count by county in the Upper Peninsula study site
for 1987-2000.

59

was signiﬁcantly different (P S 0.05) from the point count in each of the other 3 counties
(Table 16). In addition, the point count for male yearlings in Dickinson County was also
signiﬁcantly different (P S 0.05) from the male yearling point count in Marquette County

(Table 16).

NLP

In each year that yearling does were checked for lactation in the NLP study site,
there were more does that were not lactating (HM-) then were lactating (HM+) (Table
1.17). The percent of yearling does not lactating ranged between 83% and 93.8%,
whereas the percent of yearling does lactating ranged between 6.2% and 17% (Table
1.17). However, the percent of yearling does that were or were not lactating was
consistent among years (Table 1.17). In general, the number of does that were checked
for lactation in the NLP study site was relatively large in comparison to the number of
yearling does checked for lactation in the other 2 study sites. In fact, there were several
years in which the number of yearling does checked for lactation exceeded 100 (Table
1.17; Figure 1.11). Years (e. g., 1998) in which large numbers of yearling does were
checked for lactation in the NLP study did not necessarily translate into a higher percent
of does lactating (Table 1.17; Figure 1.12). There were, however, years (e.g., 1994) in
which small numbers of yearling does checked for lactation in the NLP study site did
translate into a higher percent of lactation (Table 1.17; Figure 1.12).

In the NLP study site, the overall mean average beam diameter for male yearlings
was 17.3mm and ranged between 16.7mm in 1997 and 18.036 mm in 2000 for the 14

years of the study (Table 1.18). The mean average beam diameter for male yearlings in

60

Table 1.17. The lactation status of yearling does harvested in the Northern Lower

Peninsula study site each year between 1993 and 2000, as well as the percent of yearling

does lactating (HM+) and not lactating (HM-) each year.

750
700
650
600
550

450
400
350
300

Number of Yearling Does

200

150 r
100 ~ ~

50

 

 

 

 

 

 

 

 

 

 

 

500 ~

250 .

 

   
   

 

 

Year HM+ HM- Total % HM+ °/oHM-
1993 5 76 8 1 6.2 93.8
1994 8 39 47 1 7 83
1995 14 80 91 15.4 84.6
1996 26 147 173 15 85
1997 22 186 208 10.6 89.4
1998 124 707 835 14.9 85.1
1999 57 496 553 10.3 89.7
2000 50 283 453 l 1 89
El HM- I HM+
1993 1994 1995 1996 1997 1998 1999 2000
Year

Figure 1.11. The lactation status of yearling does harvested in the Northern Lower
Peninsula study site each year between 1993 and 2000.

61

 

+ Yearly Average ' " Long-Term Average

 

20.0

17.5. L, e ,, , e e , ,

15.0

  
 

12.5 *r

10.0 #- -

Percent of Docs lactating

7.5 ~

 

5.0

 

 

1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.12. The percent of yearling does harvested in the Northern Lower Peninsula
study site that were lactating during each year between 1993 and 2000.

62

Table 1.18. The mean average beam diameter in mm for male yearlings harvested in
each county in the Northern Lower Peninsula study site for each year between 1987 and
2000 and the standard deviation, as well as the yearly averages for the entire study site
and overall averages for each county.

 

 

 

 

 

 

 

 

 

 

 

Year Alcona Alpena Montmorencey Oscoda Presque Isle Yearly Average StDev
1987 16.5 18.4 17.0 16.2 17.8 17.1 0.91
1988 16.8 18.4 16.9 16.6 17.4 17.2 0.72
1989 16.0 17.8 16.5 15.8 17.1 16.7 0.82
1990 17.5 18.2 17.6 17.2 17.0 17.5 0.45
1991 17.5 17.3 18.1 17.0 17.7 17.5 0.42
1992 16.8 18.0 16.5 16 17.4 16.9 0.78
1993 17.4 17.7 16.9 16.5 17.2 17.2 0.46
1994 17.6 18.0 16.8 16.8 17.4 17.4 0.52
1995 18.4 17.6 16.8 16.9 17.7 17.5 0.65
1996 17.0 17.5 16.8 16.3 17.6 17.1 0.54
1997 16.5 17.0 16.8 16.5 16.9 16.7 0.23
1998 17.0 18.7 17.5 16.4 18.6 17.8 1.00
1999 16.8 17.4 15.9 15.4 17.8 16.9 1.01
2000 18.2 18.3 17.3 17.8 18.4 18.0 0.45
Total 17.2 17.9 17.0 16.6 17.7 17.3 0.52

 

Table 1.19. The minimum and maximum average beam diameter in mm collected from
male yearlings harvested in each county in the Northern Lower Peninsula study site for
each year between 1987 and 2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Alcona Alpena Montmorency Oscoda Pregue Isle

Min Max Min Max Min Max Min Max Min Max

1987 6 26 8 28 9 27 7 27 10 29.5
1988 6.5 27 8.5 26 9.5 25.5 7 27.5 6.5 25
1989 7 25 8.5 26.5 8 28 8.5 29.5 10 24

1990 7.5 28.5 10.5 28.5 5 27.5 8 33 7.5 26.5

1991 7 35.5 8 27 11 33 6.5 29.5 10 26.5
1992 8 29.5 10 27.5 8 27 8 26.5 8 25

1993 7 28.5 9 25.5 10.5 29.5 9 26 8 25.5
1994 6.5 27.5 9 28 8.5 25 8 32 9.5 31
1995 8.5 31 8 33 10.5 28.5 9.5 28 8.5 29

1996 8.5 32 8.5 27 8.5 26 7.5 27 10.5 31.5
1997 8 28 9.5 24 9.5 25.5 8 29 9.5 26
1998 6.5 30 6 29 6 27 6.5 25.5 9.5 29

1999 6 31 6.5 32 6 29 6.5 24 7.5 28.5
2000 5 28.5 8.5 32 7 31 9 28 9 29

 

63

 

 

the NLP study site increased in 1990, 1995, 1998, and 2000. Although this index tended
to oscillate around 17mm, it never fell below 16.5mm (Figure 1.13). The minimum
average beam diameter for male yearlings was 5mm in Alcona County in 2000. Most of
the minimum averages, however, ranged between 7mm and 10mm (Table 1.19). The
maximum average beam diameter for male yearlings was 35.5mm in Alcona County in
1991; however, the majority of the high averages ranged between 25mm and 31mm
(Table 1.19).

The results from the ANOVA model run with male yearling average beam
diameter indicated that the year variable was statistically signiﬁcant (F13, 22000 = 22.14; P
< 0.0001). The results from the Tukey-Kramer multiple comparison test also illustrated
which years the male yearling average beam diameter was signiﬁcantly different from
one another (Table 1.20). According to the Tukey-Kramer multiple comparison test, the
average beam diameter for male yearlings in each of 1989, 1997, 1998, 1999, and 2000
was signiﬁcantly different (P S 0.05) from the male yearling average beam diameter in
the majority of the other years in the study (Table 1.20). In addition, the average beam
diameter for male yearlings in other years was signiﬁcantly different (P S 0.05) from one
another (Table 1.20). The male yearling average beam diameter was signiﬁcantly
different (P S 0.05) between 1987 and 1988. This was also the case for the male yearling
average beam diameter in 1992 and 1996 (Table 1.20).

When the mean average beam diameter was examined by county, male yearlings
in Alpena and Presque Isle County had the largest average beam diameter measurements
in the NLP study site (3? = 17.9mm and f = 17.7mm, respectively) (Figure 1.14).

Alcona, Montmorency, and Oscoda County had the smallest average beam diameter

64

—O—Yearly Average ' " Long-Term Average

 

18.25 .

18a , ~ A -,, e

17.75—- —- -- -- —--—

17.5 e 777 ,, A i- -- 3 a

17.25 e A -_ , 1,- A» ._.- -S. ,_ , L,

Average Beam Diameter

16.75 L.“ _. i . , i ,,7,_# ”a _ l_ ,,

 

16.5 . r . .

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.13. The temporal change in average beam diameter in mm for male yearlings
harvested in the Northern Lower Peninsula study site between 1987 and 2000.

65

Table 1.20. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Northern Lower Peninsula study site on a yearly basis. The x denotes a
Signiﬁcant difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Table 1.21. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Northern Lower Peninsula study site on a county basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

 

66

  

Mean Yearling Average Beam Diameter

m Oscoda (16.6 mm)

-_. Mlntmorency (17.0 mm)
- Alcona (17.2 mm)

- Presque Isle (17.7 mm)
- Alpena (17.9 mm)

 

Figure 1.14. Mean yearling average beam diameter by county in the Northern Lower
Peninsula study site for 1987-2000.

67

measurements in the NLP study site (3? = 17.2m, 1? = 17.0mm, f =16.6, respectively)
(Figure 1.14). The results from the ANOVA model also indicated that the county
variable was statistically signiﬁcant (F4, 22000 = 107.02; P < 0.0001). The Tukey-Kramer
multiple comparison test results showed that the average beam diameter for male
yearlings in Alpena, Oscoda, and Presque Isle County was signiﬁcantly different (P S
0.05) from the male yearling average beam diameter in each of the other 4 counties
(Table 1.21). In addition, the average beam diameter for male yearlings in each of
Alcona and Montmorency County was signiﬁcantly different from the average beam
diameter in 3 out of the other 4 counties, and the average beam diameter for male
yearlings in those counties was not signiﬁcantly different from each other (Table 1.21 ).

The overall point count for male yearlings in the NLP study site was 3.16 and the
average point count for yearlings ranged between 3.01 in 1992 and 3.35 in 2000 (Table
1.22). The yearly average point count for male yearlings in the NLP study site increased
in each of 1991, 1994, 1998, and 2000; this index oscillated around 3 (Figure 1.15). The
minimum number of points for male yearlings in this study site was 2, which was
consistent among years and counties (Table 1.23). The maximum number of points for
male yearlings reached 23 in Alpena County in 1988, although most of the maximum
number of points ranged between 8 and 10 (Table 1.23).

The results of the ANOVA model run with male yearling points indicated that the
year variable was statistically signiﬁcant (F4, 22555 = 10.58; P < 0.0001). The results from
the Tukey-Kramer multiple comparison test illustrated which years the male yearling
point count was signiﬁcantly different from one another (Table 1.24). According to the

Tukey-Kramer multiple comparison test, the male yearling point count in 1994 and 2000

68

Table 1.22. The mean point count for male yearlings harvested in each county in the

Northern Lower Peninsula study site for each year between 1987 and 2000 and the
standard deviation, as well as yearly averages for the entire study site and overall
averages for each county.

 

 

 

 

 

 

 

 

 

 

 

Year Alcona Alpena Montmorency Oscoda Presque Isle Yearly Average StDev
1987 3.02 3.34 3.06 2.74 3.28 3.08 0.24
1988 3.24 3.45 2.99 2.80 3.09 3.13 0.25
1989 3.03 3.48 2.88 2.66 3.05 3.03 0.30
1990 3.15 3.36 2.97 2.98 2.96 3.08 0.17
1991 3.49 3.27 3.34 2.99 3.19 3.26 0.19
1992 3.15 3.05 2.90 2.96 2.92 3.01 0.11
1993 3.28 3.27 3.02 2.85 3.05 3.12 0.18
1994 3.41 3.46 2.97 3.06 3.49 3.28 0.24
1995 3.49 3.45 2.83 2.90 3.31 3.20 0.31
1996 3.02 3.03 2.99 2.77 3.01 2.98 0.11
1997 2.96 3.17 3.05 2.95 3.13 3.05 0.10
1998 3.17 3.42 3.20 2.78 3.65 3.26 0.33
1999 3.18 3.35 2.65 2.62 3.45 3.12 0.39
2000 3.68 3.37 3.14 3.21 3.62 3.44 0.24
Total 3.24 3.32 3.01 2.88 3.29 3.16 0.19

 

Table 1.23. The minimum and maximum number of points counted on male yearlings

harvested in each county in the Northern Lower Peninsula study site for each year

between 1987 and 2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Alcona Alpena Montmorency Oscoda Presque Isle
Min Max Min Max Min Max Min Max Min Max
1987 2 8 2 8 2 12 2 13 2 8
1988 2 12 2 23 2 20 2 10 2 8
1989 2 8 2 10 2 8 2 9 2 8
1990 2 . 8 2 10 2 9 2 l3 2 9
1991 2 9 2 8 2 9 2 9 2 8
1992 2 8 2 9 2 6 2 8 2 8
1993 2 8 2 9 2 9 2 8 2 8
1994 2 9 2 19 2 8 2 9 2 10
1995 2 8 2 1 1 2 8 2 8 2 9
1996 2 9 2 9 2 1 3 2 7 2 8
1997 2 8 2 8 2 9 2 8 2 10
1998 2 10 2 1 1 2 9 2 13 2 1 3
1999 2 9 2 8 2 8 2 8 2 8
2000 2 9 2 10 2 9 2 9 2 8

 

69

 

 

+Yearly Average ' " Long-Term Average

 

3.45

3.35 »- , v , ,, 7 W 7 7 7 7 7 -7 ,7 S __ __

3.25- .- ,,

3.15 "'"""-'—"--'-7--..--.-.---_--___-._._-.,-

Average Point Count

3.05 - -. - . .._- .. _ _

 

 

2.95 I
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Year

Figure 1.15. The temporal change in average point count for male yearlings harvested in
the Northern Lower Peninsula study site between 1987 and 2000.

70

Table 1.24. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Northern Lower Peninsula study site on a yearly basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Table 1.25. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Northern Lower Peninsula study site on a county basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

Alcona Oscoda ue Isle
Alcona

Montmo
Oscoda
ue Isle

 

71

was signiﬁcantly different (P S 0.05) from the male yearling point count in most of the
other years in the study (Table 1.24). Male yearling point count was signiﬁcantly
different (P S 0.05) between 1991 and 1998 (Table 1.24). Furthermore, the male yearling
point count in each of 1989, 1992, 1997, and 1999 was signiﬁcantly different (P S 0.05)
from the male yearling point count in each of the same years (Table 1.24). There were
additional years in which the male yearling point count was signiﬁcantly different (P S
0.05) from other years; however, there was not a distinct pattern (Table 1.24).

When the average point count was examined by county, male yearlings with the
most points were found in Alpena and Presque Isle County (3? = 3.32 and f = 3.29,
respectively) (Figure 1.16). On the other hand, male yearlings in Alcona, Montmorency,
and Oscoda County had the fewest number of points (a? = 3.24, f = 3.01, and J? = 2.88,
respectively) (Figure 1.16). The results from the ANOVA model also indicated that the
county variable was statistically signiﬁcant (F4, 22555 = 61.72; P < 0.0001). The results of
the Tukey-Kramer multiple comparison test showed that the point count for male
yearlings in Montmorency County was signiﬁcantly different (P S 0.05) from the male
yearling point count in each of the other 4 counties, as was the male yearling point count
in Oscoda County (Table 1.25). In addition, the point count for male yearlings in Alcona
County was also signiﬁcantly (P S 0.05) from the point count for male yearlings in

Alpena County (Table 1.25).

SLP
In each year that yearling does were checked for lactation in the SLP study site,

there were more does that were not lactating (HM-) then were lactating (HM+) (Table

72

  

Yearling Average Point Count

E Oscoda (2.88)

Montmorency (3.01)

Alcona (3.24)

- Presque Isle (3.29)

- Alpena (3.32)

  

Figure 1.16. Yearling average point count by county in the Northem Lower Peninsula
study site for 1987-2000.

73

1.26). The percent of yearling does not lactating ranged between 56.8% and 80.8%,
whereas the percent of yearling does lactating ranged between 19.2% and 43.2% (Table
1.26). The percent of yearling does that were lactating in this study site was the highest
out of the 3 study sites. Overall, the number of does that were checked for lactation in
the SLP study site was relatively small, in fact there was only 1 year (1997) in which the
number of yearling does checked for lactation exceeded 100 (Table 1.26; Figure 1.17).
Years (e. g., 1997) in which large number of yearling does were checked for lactation in
the SLP study did not necessarily translate into a higher percent of does lactating (Table
1.26; Figure 1.18). There were, however, years (e.g., 1993) in which small numbers of
yearling does checked for lactation in the SLP study site did translate into a higher
percent of lactation (Table 1.26; Figure 1.18).

The overall mean average beam diameter for male yearlings in the SLP study site
was 21.91mm and ranged between 21.1mm in 1996 and 22.7 mm in 1989 for the entire
14 years of the study (Table 1.27). Although there has been a steady decrease in the male
yearling average beam diameter in this study site over the past 14 years, the mean
average beam diameter for male yearlings in the SLP study site increased in 1989, 1991,
and 1995 (Figure 1.19). The mean average beam diameter for male yearlings in this
study site tended to oscillate around 22mm and never fell below 21mm (Figure 1.19).
The minimum average beam diameter for male yearlings was 6mm in Barry County in
1993. Most of the minimum average beam diameter measurements, however, ranged
between 10mm and 15mm (Table 1.28). The maximum average beam diameter for male
yearlings was 34mm in Barry County in 1990; however, the majority of the high averages

ranged between 28mm and 33mm (Table 1.28).

74

Table 1.26. The lactation status of yearling does harvested in the Southern Lower
Peninsula study site each year between 1993 and 2000, as well as the percent of yearling
does lactating (HM+) and not lactating (HM-) each year.

 

 

Year HM+ HM- Total %HM+ %HM-
1993 16 21 37 43 2 56 8
1994 17 44 61 27 9 72 1
1995 15 36 51 29 4 70 6
1996 15 63 78 19 2 80 8
1997 35 82 117 29 9 70 1
1998 26 73 99 26 3 73 7
1999 22 46 68 32 4 67.6
2000 17 64 81 21 79

 

 

 

 

 

 

 

 

HM- I HM+

 

90

Number of Yearling Does

 

 

 

1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.17. The lactation status of yearling does harvested in the Southern Lower
Peninsula study site each year between 1993 and 2000.

75

+ Yearly Average ' " long-Term Average

 

45.0

35.0

D."
o

Percent of Docs lactating
w
p
O

 

20.0

 

 

 

15.0
1993 1994 1995 1996 1997 1998 1999 20(1)

Year

Figure 1.18. The percent of yearling does harvested in the Southern Lower Peninsula
study site that were lactating during each year between 1993 and 2000.

76

Table 1.27 . The mean average beam diameter in mm for male yearlings harvested in
each county in the Southern Lower Peninsula study site for each year between 1987 and
2000 and the standard deviation, as well as the yearly averages for the entire study site
and overall averages for each county.

 

 

 

Year Barry Calhoun Eaton Yearly Average StDev
1987 22.1 22.9 23.1 22.3 0.53
1988 22.8 21.7 23.1 22.6 0.74
1989 22.6 23.3 22.4 22.7 0.47
1990 22.4 20.9 22.9 22.2 1.04
1991 22.6 22.5 22.5 22.6 0.01
1992 22.1 21.6 22.6 22.0 0.54
1993 21.5 21.9 23.4 21.7 1.03
1994 21.4 20.7 22.5 21.5 0.90
1995 22.3 21.8 23.0 22.3 0.61
1996 21.1 20.7 21.7 21.1 0.53
1997 21.2 21.7 22.8 21.5 0.80
1998 21.8 21.5 21.4 21.7 0.20
1999 21.2 21.6 21.2 21.3 0.22
2000 21.3 21.2 22.9 21.5 0.99
Total 21.9 21.6 22.5 21.9 0.43

 

 

 

 

 

 

 

 

Table 1.28. The minimum and maximum average beam diameter in mm collected from
male yearlings harvested in each county in the Southern Lower Peninsula study site for
each year between 1987 and 2000.

 

 

 

Barry Calhoun Eaton

Min Max Min Max Min Max
1987 12.5 32.5 17.5 30 20.5 26
1988 15 30.5 10 28.5 15.5 31.5
1989 10 33 15.5 32.5 11 29.5
1990 14.5 34 14 29 15.5 30.5
1991 14 32.5 18 28.5 18 28
1992 11.5 30.5 13 30 17 27
1993 6 33.5 15.5 26.5 13 35.5
1994 12.5 29 14.5 26 15.5 30
1995 15 30.5 12 30.5 10.5 30.5
1996 9.5 29.5 12.5 29.5 14.5 28
1997 12 36.5 13.5 28.5 12.5 33.5
1998 13 31 14.5 30.5 13.5 31.5
1999 12 30.5 15 29.5 15 32
2000 8.5 28.5 13.5 28 15.5 30

 

 

 

 

 

 

 

 

 

77

22.75 -

22.5

22.25

22

Average Beam Diameter

21.5

21.25

21

—+—Yearly Average " - Long-Term Average

 

21.75

 

 

I I I T

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.19. The temporal change in average beam diameter in mm for male yearlings
harvested in the Southern Lower Peninsula study site between 1987 and 2000.

78

 

The results from the ANOVA model run with male yearling average beam
diameter indicated that the year variable was statistically signiﬁcant (F2, 4445 = 4.18; P <
0.0001) in the SLP study site. The results from the Tukey-Kramer multiple comparison
test also showed which years the male yearling average beam diameter was signiﬁcantly
different from one another (Table 1.29). The Tukey-Kramer multiple comparison test
results showed that the male yearling average beam diameter in 1996 and 1999 was
signiﬁcantly different (P S 0.05) from the male yearling average beam diameter in the
early years of the study (Table 1.29). The male yearling average beam diameter in 1989
was also signiﬁcantly different (P S 0.05) from the male yearling average beam diameter
in 1994 (Table 1.29).

When the mean average beam diameter was examined by county, male yearlings
in Eaton County had the largest average beam diameter measurements in the SLP study
site (i = 22.5mm), whereas male yearlings in Calhoun County had the smallest ( J? =
21 .6mm) (Figure 1.20). The mean average beam diameter for male yearlings in Barry
County fell in between the mean average beam diameter for male yearlings in the other 2
counties (56 = 21 .9mm) (Figure 1.20). The results from the ANOVA model also
indicated that the county variable was statistically signiﬁcant (F 1 3, 4445 = 10.01; P <
0.0001). The results of the Tukey-Kramer multiple comparison test showed that the
average beam diameter for male yearlings in Baton County was signiﬁcantly different (P
S 0.05) from the average beam diameter in each of the other 2 counties (Table 1.30).
The overall average point count for male yearlings in the SLP study site was 5.38 and the
average point count for male yearlings ranged between 5.08 in 1996 and 5.81 in 1991

(Table 1.31). Although the average point count for male yearlings in the SLP study

79

Table 1.29. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Southern Lower Peninsula study site on a yearly basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Table 1.30. Mean pair-wise comparisons of average beam diameter for male yearlings
harvested in the Southern Lower Peninsula study site on a county basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

Calhoun Eaton

Calhoun
Eaton

 

80

 

Mean Yearling Average Beam Diameter
1:] Calhoun (21 .6 mm)

Barry (21.9 mm)

- Eaton (22.5 mm)

 

Figure 1.20. Mean yearling average beam diameter by county in the Southern Lower
Peninsula study site for 1987-2000.

81

Table 1.31. The mean point count for male yearlings harvested in each county in the
Southem Lower Peninsula study site for each year between 1987 and 2000 and the
standard deviation, as well as yearly averages for the entire study site and overall
averages for each county.

 

 

 

Year Barry Calhoun Eaton Yearly Average StDev
1987 5.54 5.94 6.44 5.65 0.45
1988 5.50 6.05 5.62 5.59 0.29
1989 5.66 5.94 5.96 5.72 0.17
1990 5.53 5.67 5.74 5.57 0.10
1991 5.68 6.15 6.33 5.81 0.34
1992 5.27 5.14 6.47 5.32 0.73
1993 5.16 5.95 6.12 5.36 0.51
1994 5.11 5.19 5.19 5.14 0.05
1995 5.61 5.94 5.75 5.69 0.16
1996 5.03 5.26 5.23 5.08 0.13
1997 4.88 5.93 5.42 5.13 0.53
1998 5.02 5.18 5.42 5.09 0.20
1999 5.20 5.46 5.39 5.27 0.13
2000 5.01 4.74 6.04 5.12 0.69
Total 5.29 5.58 5.65 5.38 0.19

 

 

 

 

 

 

 

 

Table 1.32. The minimum and maximum number of points counted on male yearlings
harvested in each county in the Southern Lower Peninsula study site for each year
between 1987 and 2000.

 

 

 

Barry Calhoun Eaton
Min Max Min Max Min Max
1987 2 10 2 9 2 10
1988 2 9 2 9 2 9
1989 2 13 2 10 3 10
1990 2 10 2 12 2 9
1991 2 10 2 10 2 10
1992 2 19 2 10 3 11
1993 2 9 2 10 2 10
1994 2 10 2 10 2 10
1995 2 10 2 10 2 10
1996 2 10 2 10 2 9
1997 2 10 2 10 2 9
1998 2 8 2 8 2 10
1999 2 10 2 10 2 9
2000 2 ll 2 8 2 10

 

 

 

 

 

 

 

 

 

82

site has been steadily decreasing over the past 14 years, the yearly average point count for
male yearlings increased sharply in 1991 and 1995 (Figure 1.21). In addition, the yearly
average point count for male yearlings in this study site oscillated around 5.4, but never
fell below 5.0 (Figure 1.21). The minimum number of points for male yearlings in this
study site was 2, which was consistent among most years and counties even though there
were a few years in Eaton county in which the minimum number of points was 3 (Table
1.32). The maximum number of points for male yearlings reached 19 in Barry County in
1992, although most of the maximum number of points ranged between 8 and 10 (Table
1.32).

The results of the ANOVA model run with male yearling points indicated that the
year variable was statistically signiﬁcant (F13, 4293 = 3.84; P < 0.0001), and the results
from the Tukey-Kramer multiple comparison test illustrated which years the male
yearling point count was signiﬁcantly different from one another (Table 1.33).
According to the Tukey-Kramer multiple comparison test, the point count for male
yearlings in 1991 was signiﬁcantly different (P S 0.05) from the point count for male
yearlings in each of 1994, 1996, 1998, and 2000 (Table 1.33). In addition, the point count
for male yearlings in other years was signiﬁcantly different (P S 0.05) from one another
(Table 1.33).

When the average point count was examined by county, male yearlings with the
most points were found in Baton County (3? = 5.65) (Figure 1.22). On the other hand,
male yearlings in Barry County had the fewest amount of points (i = 5.29) (Figure 1.22).
The average point count for male yearlings in Calhoun County fell in between the

average point count for male yearlings in the other 2 counties ()7 = 5.58) (Figure 1.22).

83

+Yearly Average " ' Regional Average

 

5.9

5.7-- — - , ~ ~ - w

5.5 .- —- -- - —~ —-

5.4 ~ - ~~ , e

Average Point Count

5.3 e _ we ~

5.2 - ~ 5 ~

 

 

5 ' i I 7
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Year

Figure 1.21. The temporal change in average point count for male yearlings harvested in
the Southern Lower Peninsula study site between 1987 and 2000.

84

Table 1.33. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Southern Lower Peninsula study site on a yearly basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05.

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Table 1.34. Mean pair-wise comparisons of average point count for male yearlings
harvested in the Southern Lower Peninsula study site on a county basis. The x denotes a
signiﬁcant difference at an alpha level of 0.05 .

Calhoun Eaton

Calhoun
Eaton

 

85

 

 

 

Yearling Average Point Count

[:] Barry (5.29)
Calhoun (5.58)

- Eaton (5.65)

 

Figure 1.22. Yearling average point count by county in the Southern Lower Peninsula
study site for 1987-2000.

86

The results from the ANOVA model also indicated that the county variable was
statistically signiﬁcant (F2, 4298 = 16.11; P < 0.0001). According to the Tukey-Kramer
multiple comparison test, the point count for male yearlings in Barry County was
signiﬁcantly different (P S 0.05) from the point count in each of the other 2 counties

(Table 1.34).

Among Region Comparisons
Temporal Comparisons

Among the 3 study sites, the percent of yearling does that were lactating was the
highest in the SLP study site for each year between 1987 and 2000 (Figure 1.23). The
percent of yearling does that were lactating in the SLP study site was consistently above
15%, whereas the percent of yearling does lactating in the NLP study site was
consistently above 5% but lower than 20% (Figure 1.23). In the UP study site the percent
of yearling does lactating was the lowest of the 3 study sites and consistently ranged
between 0% and 20% (Figure 1.23). The percent of yearling does that were lactating in
the SLP study site was highest in 1993 and 1999; however, in the UP study site, the
percent of yearling does that were lactating was highest in 1995 and in the NLP study
site, lactation was highest in 1994 (Figure 1.23).
The temporal trend in the percent of yearling does lactating in the SLP and UP study sites
tended to be similar. In general, in years that the percent of yearling does lactating in the
SLP increased, so the did the percent of yearling does lactating in the UP with the
exception of those years in which 0% yearling does were lactating (Figure 1.23). The

temporal trend in the percent of yearling does lactating in the NLP, for the most part,

87

' e»- UP +NLP +SLP

 

50.0

45.0

40.0

'/. Does Lactating

 

W
, a:

 

 

 

 

 

1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.23. Yearly trends in the percentage of yearling does harvested in each of the 3
regional study sites that were lactating from 1993-2000.

88

tended to be opposite of the trends in the percent of yearling does lactating in the SLP
and UP study sites (Figure 1.23). As the percent of yearling does that were lactating in
the NLP study site increased, in general, the percent of yearling does lactating decreased
in the other 2 study sites (Figure 1.23). In addition, there seemed to be more variation in
the percent of yearling does lactating in the SLP study site and the UP study site while
the percent of yearling does lactating in the NLP study site appeared to be more
consistent from year to year (Figure 1.23).

The mean average beam diameter for male yearlings was also much larger in the
SLP study site than in the UP or NLP study sites for each year between 1987 and 2000
(Figure 1.24). The male yearling average beam diameter in the SLP study site was
consistently above 21mm, whereas the male yearling average beam diameter in the NLP
study site was consistently above 16.5mm but lower than 18.5mm (Figure 1.24). In the
UP study site the male yearling average beam diameter was the lowest out of the 3 study
sites and consistently ranged between 16mm and 17.5mm (Figure 1.24). The mean
average beam diameter for male yearlings in the SLP study site was highest in 1990;
however, in the UP study site, the mean average beam diameter for male yearlings was
highest in 1987 and in the NLP study site, it was highest in 2000 (Figure 1.24). Despite
the differences in the mean average beam diameter for male yearlings among the 3
regional study sites, this population quality index showed the same temporal trends
among the 3 study sites. In general, if the mean average beam diameter for male
yearlings increased in 1 study site, for the most part, it also increased in the other 2 study

sites (Figure 1.24).

89

UP —I—— NLP + SLP
23.0

22.5 J
22.0
21.5
21.0
20.5
20.0
19.5
19.0 ~
18.5
18.0
17.5
17.0 ,
16.5 . ‘
16.0 i
15.5
15.0 .

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Average Beam Diameter

   

 

 

 

Year

Figure 1.24. Yearly trends in the average beam diameter collected from male yearlings
harvested in each of the 3 regional study sites from 1987-2000. The * denotes which
years the average beam diameter for male yearlings was signiﬁcantly different (p-value <
0.05) among the UP and NLP study sites.

90

Results of the AN OVA model run to determine if there were signiﬁcant
differences in the mean average beam diameter among the 3 regional study sites indicated
that the interaction variable (region*year) was statistically signiﬁcant F 25, 35544 = 11.72; P
< 0.0001). Although the mean average beam diameter for male yearlings in the SLP
study site has decreased slightly during this study, it was still signiﬁcantly different (P <
0.0001) from the mean average beam diameter for male yearlings in both the UP and
NLP study sites for each year of the study (Figure 1.24). The mean average beam
diameter for male yearlings in the UP and NLP study sites was similar, but there were
certain years in which there was a signiﬁcant difference in the mean average beam
diameter for male yearlings between the 2 study sites (Figure 1.24). The results from the
Tukey-Kramer multiple comparison test showed that a signiﬁcant difference in the mean
average beam diameter for male yearlings in the UP and NLP study sites occurred in
1990 (P < 0.0001), 1994 (P < 0.0001), 1995(P < 0.0001), 1996(P = 0.0434), and 2000 (P
< 0.0001) (Figure 1.24).

As with the percent of yearling does lactating and the average beam diameter for
male yearlings, the average point count for male yearlings was highest in the SLP study
site (Figure 1.25). The male yearling average point count in the SLP study site was
consistently above 5, whereas the male yearling average point count in the NLP and UP
study sites was consistently above 3 but lower than 4 (Figure 1.25). The point count for
male yearlings in the SLP study site was highest in 1991; however, in the UP study site,
the mean average beam diameter for male yearlings was highest in 1998 and in the NLP
study site, it was highest in 2000 (Figure 1.25). Like the male yearling average beam

diameter, the average point count for male yearlings among the 3 regional study sites

91

4 UP +NLP +SLP

 

 

Average Point Count

 

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 1.25. Yearly trends in the point count collected from male yearlings harvested in
each of the 3 regional study sites from 1987-2000. The * denotes which years the point
count for male yearlings was signiﬁcantly different (p-value < 0.05) among the UP and
NLP study sites.

92

followed similar temporal trends; if the average point count for male yearlings increased
in 1 study site, in general, it also increased in the other 2 study sites (Figure 1.25).

The results of the ANOVA model run to determine if there were signiﬁcant
differences in the average point count for male yearlings among the 3 regional study sites
indicated that the interaction variable (region*year) was statistically signiﬁcant (F25, 37095
= 8.68; P < 0.0001). The average point count for male yearlings in the SLP study site has
remained constant, and was signiﬁcantly different (P < 0.0001) from the average point
count for male yearlings in both the UP and NLP study sites for each year of the study
(Figure 1.25). The average point count for male yearlings in the UP and NLP study sites
was very similar, although there were certain years in which there was a signiﬁcant
difference in the average point count for male yearlings between the 2 study sites (Figure
1.25). The results from the Tukey-Kramer multiple comparison test showed that a
signiﬁcant difference in the average point count for male yearlings in the UP and NLP
study sites occurred in 1994 (P = 0.0004), 1995 (P = 0.0012), and 2000 (P < 0.0001)

(Figure 1.25).

Spatial Comparisons

Comparing the long-term average for each region illustrated that the percent of
yearling does lactating in each study site increased along a north to south regional
gradient in Michigan (Figure 1.26). The UP study site had the lowest percent of yearling
does lactating at 10.4%, and the SLP study site had the highest percent of yearling does
lactating at 28.7% (Figure 1.26). The percent of yearling does lactating in the NLP study

site fell in between that of the other 2 study sites with 12.5%. Although an increase in the

93

 

 

 

 

 

 

10.4%

 

 

 

 

12.5%

 

Figure 1.26. The regional comparison of the percentage of yearling does lactating in

each of the 3 study sites during 1993-2000.

94

 

percent of yearling does lactating existed moving from north to south, the increase was
much greater between the NLP and SLP study sites then between the UP and NLP study
sites (Figure 1.26). This spatial trend suggested that the percent of yearling does lactating
in the UP and NLP study sites was more similar while the percent of yearling does
lactating in the SLP study site was dissimilar to the percent of yearling does lactating in
the other 2 study sites.

The mean average beam diameter for male yearlings also increased along a north
to south regional gradient in Michigan (Figure 1.27). The UP study site had the smallest
mean average beam diameter for male yearlings at 16.8mm, and the SLP study site had
the largest mean average beam diameter for male yearlings at 21 .9mm (Figure 1.27).
Male yearlings in the NLP study site had a mean average beam diameter that fell in
between the mean average beam diameter for male yearlings in the other 2 study sites at
17.3mm (Figure 1.27). The increase in the mean average beam diameter for male
yearlings increased only slightly between the UP and NLP study sites, but increased
substantially between the NLP and SLP study Sites (Figure 1.27). This spatial trend
suggested that the mean average beam diameter for male yearlings in the UP and NLP
study sites was similar, but the mean average beam diameter for male yearlings in the
SLP study site was dissimilar to the mean average beam diameter for male yearlings in
the other 2 study sites. Despite that fact that the mean average beam diameter for male
yearlings appeared more similar in the UP and NLP study sites, the results from the
Tukey-Kramer multiple comparison test showed that there was a signiﬁcant difference in
the mean average beam diameter for male yearlings among the 3 regional study sites (P <

0.0001).

95

 

 

 

 

16.8mm

 

 

17.3mm

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

21.9mm

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.27 . The regional comparison of the average beam diameter for male yearlings
harvested in each of the 3 study sites during 1987-2000.

96

Similar to the other 2 population-level quality indices, the average point count for male
yearlings increased along the north to south regional gradient in Michigan (Figure 1.28).
The average point count for male yearlings was lowest in the UP study site at 3.09 and
highest in the SLP study site at 5.38 (Figure 1.28). At 3.16, the average point count for
male yearlings in the NLP study site fell in between the average point count for male
yearlings in the other 2 study sites (Figure 1.28). Following the same trend as the other 2
quality indices, the increase in the average point count for male yearlings increased only
slightly between the UP and NLP study sites, but increased substantially between the
NLP and SLP study sites (Figure 1.28). This spatial trend suggested that the average
point count for male yearlings in the UP and NLP study sites was similar, and the average
point count for male yearlings in the SLP study site was dissimilar to the average point
count for male yearlings in the other 2 study sites. Despite that fact that the average point
count for male yearlings appeared more similar in the UP and NLP study sites, the results
from the Tukey-Kramer multiple comparison test showed that there was a signiﬁcant
difference in the average point count for male yearlings among the 3 regional study sites.
The average point count for male yearlings in the SLP study site was signiﬁcantly
different from the average point count for male yearlings in both the UP and NLP study
sites (P < 0.0001). In addition, the average point count for male yearlings in the UP was

signiﬁcantly different from male yearling average point count in the NLP (P = 0.0041).

Relationship Among Quality Indices

There did not appear to a strong relationship between the percent of yearling does

lactating and the mean average beam diameter of male yearlings in any of the 3 regional

97

 

3.09

 

 

 

Figure 1.28. The regional comparison of the average point count for male yearlings
harvested in each of the 3 study sites during 1987-2000.

 

SLP

 

 

 

3.16

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5.38

 

 

 

 

 

 

 

 

 

 

 

 

98

 

study sites (Figure 1.29). In the UP study site, it looked as if there was a very slight
negative relationship between the 2 quality indices, but that was more than likely an
artifact of the 2 years in which no yearling does were lactating (R2 = 0.20; P = 0.26)
(Figure 1293). The relationship between the percent of yearling does lactating and the
mean average beam diameter for male yearlings in the NLP and SLP study sites, on the
other hand, looked as if was slightly positive; however, there was not a signiﬁcant trend
between the 2 quality indices in either regional study site (R2 = 0.06; P = 0.55 and R2 =
0.10; P = 0.44, respectively) (Figure 1.29b&c).

The relationship between the percent of yearling does lactating and the male
yearling average point count was similar to the relationship between the percent of
yearling does lactating and the mean average beam diameter for male yearlings; there did
not appear to be a strong relationship between the 2 quality indices (Figure 1.30). In the
UP study site, there appeared to be a slight negative relationship between the percent of
yearling does lactating and the average point count for male yearlings (R2 = 0.36; P =
0.12); although, this trend was probably due to the fact that for 2 years there were no
lactating yearling does in this study site (Figure 1.30a). Conversely, there appeared to be
a slight positive relationship between the percent of yearling does lactating and the male
yearling average beam diameter in the SLP study site (R2 = 0.21; P = 0.26) (Figure
1.30c). In addition, there appeared to be no relationship between the percent of yearling
does lactating and the average point count for male yearlings in the NLP study site (R2 =
0.02; P = 0.73) (Figure 1.30b).

Unlike the relationship between the percent of yearling does lactating and the 2

antler measurements, there was a strong positive relationship between the mean average

99

(a) UP

 

17.4
0
17.2 ’

17 °
16.8
16.6
16.4

0
16.2 .

16 .
0.0 5.0 10.0 15.0 20.0 25.0

 

 

 

(b) NLP

 

18.2
18 °
17.8 .
17.6 ,
17.4 .
17 °
0
O

16.81

16.6 T
0.0 5.0 10.0 15.0 20.0

 

 

 

Male Yearling Average Beam Diameter

(c) SLP

 

22.4
22.24
22'
21.80 .
21.6
21.4
21.2 .
21

 

 

 

0.0 10.0 20.0 30.0 40.0 50.0

Percent of Yearling Does Lactating

Figure 1.29. The relationship between the percent of yearling does lactating and the
mean average beam diameter for male yearlings in the (a) UP study site, the (b) NLP
study site, and the (c) SLP study site.

100

(a) UP

 

3.4

3.3 °

3.2 -

3.1 “
3 1’ 9

2.9 * 9 °

2.8

 

 

 

0.0 5.0 10.0 15.0 20.0 25.0

(b) NLP

 

3.4
3.3 .

3.2 *I I

3.1 ~ ’

 

 

 

2.9 .
0.0 5.0 10.0 15.0 20.0

Male Yearling Average Point Count

(0) SLP

 

5.8
5.7 e
5.6
5.5 *
5.4
5.3
5.2
5.1 0 O
5

O.

 

 

 

0.0 10.0 20.0 30.0 40.0 50.0

Percent of Yearling Does Lactating

Figure 1.30. The relationship between the percent of yearling does lactating and the
average point count for male yearlings in the (a) UP study site, the (b) NLP study site,
and (c) the SLP study site.

101

beam diameter for male yearlings and the male yearling average point count (Figure
1.31). In each of the 3 regional study sites, there was a signiﬁcant correlation between
the 2 antler measurement quality indices; although, the relationship between the mean
average beam diameter for male yearlings and male yearling average point count was
more signiﬁcant in the UP and NLP study sites (R2 = 0.78; P < 0.0001 and R2 = 0.69; P =

0.0003, respectively) than in the SLP study site (R2 = 0.81; P < 0.0001).

102

(a) UP

 

3.3
3.2 '
3.1
3
2.9 ’

2.8 ’ .
16 16.5 17 17.5 18

 

 

 

 

(b) NLP

 

3.5

3.3 ’1

 

 

 

 

16.5 17 17.5 18 18.5

Male Yearling Average Point Count
DJ

(c) SLP

 

5.9
5.8 *
5.7
5.6
5.5
5.4 7.
5.3
5.2
5.1 '
5
4.9

 

 

 

 

21 21.5 22 22.5 23

Male Yearling Mean Average Beam Diameter

Figure 1.31. The relationship between the mean average beam diameter for male
yearlings and male yearling average point count in the (a) UP study site, (b) NLP study
site and (c) the SLP Study site.

103

DISCUSSION

Fall Recruitment

The summer driving transects were developed to determine if road survey
methodology for deer could be used to obtain fall recruitment, as measured by fawn-to-
doe ratios. The main goal of the summer driving transects was to maximize the number
of does observed along each route so that there would be a greater likelihood that fawns
would also be seen in the same area; therefore, translating into relatively accurate fawn-
to-doe ratios, as a measure of fall recruitment. In the summer of 2000, however, only
during the evening transects throughout the entirety of July and August in the NLPl route
did the number of does exceed 100; there were no months in which during the morning
transects the number of does exceeded 100. Even though the transect procedures were
modiﬁed for the summer of 2001 , there were no months in which the number of does
observed exceeded 100. This was problematic because the predetermined sample size of
100 does was not met and although fawn-to-doe ratios could be calculated, they were
probably not very accurate due to the fact that the sample sizes of does were so small.
Because there was some question as to how accurate the fawn-to-doe ratios were, they
were not translated into fall recruitment. Thus, deer reproductive potential, expressed as
fall recruitment, and measured by fawn-to-doe ratios was not examined as a deer

condition index in this study.

104

Land Unit Classification

Although this study originally sought to utilize a hierarchical land classiﬁcation
system to develop an ecosystem-based deer management strategy based on ecological
units in which the spatial variation was minimized, the way in which the data were
collected and, subsequently, aggregated only allowed the difference in the quality indices
to be compared across broad ecological regions. All of the data used in this study were
not, by design, collected within deﬁned ecological boundaries, or ecoregions, but were
collected within political boundaries. Even though these data do fall within differing
ecoregions it is simply an artifact of the political boundaries overlapping the ecological
boundaries. Therefore, in theory, the data that are collected within the political
boundaries can be transformed into ecological boundaries by determining in which
ecoregion the data exists.

The problem in this study, however, was that the lactation data were examined at
a regional level and the antler measurements were examined at a county level by
necessity of the data collection and sample sizes. The scales used were much too broad
to transform the data from political to ecological boundaries because the political
boundaries overlapped multiple ecoregions, which made it difﬁcult to assign a region or
county to only 1 ecoregion. Since 1999, the biodata have been collected at a township
level, which is a smaller scale than either the regional or county level. Data that exists at
smaller scales means that there will not be as much multiple overlap between political
boundaries and ecoregions. Therefore, collecting data at smaller scales over the long—
terrn suggests that eventually it will be more feasible to transform data from political

boundaries to ecological boundaries.

105

Condition Indices

It has been established through previous research that population-level quality
indices such as lactation status and antler measurements can be used to assess the
physical condition of deer (Severinghaus 1955, Cowan and Long 1962, McCullough
1982, Severinghaus and Moen 1983, Rasmussen 1985). Of the quality indices examined
in this study, the lactation status of yearling does provided the least insight into the
condition of the Michigan deer herd. This was, in part, due to the fact that there were
relatively few yearling does checked for lactation at MDNR check stations. Thus, it was
difﬁcult to draw conclusions based on such small sample sizes. Also, it has been shown
that as fall progresses, there are fewer does that still show signs of lactation; thus,
although the ﬁrearm season in late November provides the largest amount of biophysical
data, the lactation rates probably do not accurately reﬂect the true number of yearlings
that produced offspring (Cook and Winterstein 2000).

The antler measurement indices, on the other hand, seemed to be better indicators
of herd condition. This was, in part, due to the fact that antler measurements were
obtained and recorded from a relatively large number of yearling males checked at
MDNR check stations. Also, average beam diameter has been correlated with weight,
which is 1 of the best indicators of deer condition (Moen and Severinghaus 1981). This
suggests that the average beam diameter of yearlings is even a better index of herd
condition because this age-class has the added burden of body grth while growing
their ﬁrst set of antlers; therefore, a yearling that has a larger average beam diameter
indicates that he is in better physical condition. In addition, the number of points that

exist on a set of antlers is directly correlated with the average beam diameter; thus, this

106

suggests that the number of points, especially for yearling deer, is also a good indicator of
physical condition.

Although the antler measurements were better indicators of herd condition than
lactation status, all of these indices could be used as relative measures of herd quality in
Michigan. The broad trends that exist in these indices among the different regions of the
state suggest they could be used to compare the condition of deer in 1 area of the state
relative to the condition of deer in another area of the state, but that these indices should
not be used as deﬁnitive measures of herd condition or quality. In addition, it may be
difﬁcult to translate the population-level quality indices as relative measures of herd
condition into a deﬁnitive quantiﬁcation of white-tailed deer survival or reproduction,
even though both are related to herd quality. For example, it would be difﬁcult to
determine if a yearling buck with a relatively large average beam diameter and point
count would have a greater chance of surviving a harsh winter than a yearling buck that
had a relatively small average beam diameter and point count. Theoretically, a deer in
better physical condition should have an improved chance to survive adverse weather and
time periods when resources are scarce. There are, however, multiple factors that
inﬂuence the survival of deer or their ability to reproduce; therefore, it would be difﬁcult
to determine if survival or reproduction was a factor of physical condition or some other

inﬂuence.

Temporal and Spatial Trends

Within each of the 3 regional study sites, all of the quality indices tended to track

one another. For example, if the average beam diameter for male yearlings increased in

107

the UP study site, in general, the point count for male yearlings and the percent of does
lactating also increased in the same study site. Over the 14 years that comprised this
study, it was apparent that the quality indices in some of the study sites were increasing
while in other study sites they were decreasing. In the UP and NLP study sites, the male
yearling average beam diameter and point count, as well as the percent of does lactating
has generally increased over the 14 years of this study. The increase in the quality
indices of deer in the UP and NLP study sites may be due to the fact that since 1998 there
has been a larger number of deer checked at MDNR stations. This was probably a result
of the discovery of Bovine TB in Michigan because hunting pressures were increased to
reduce deer density and, hence, the transmission of the disease in the NLP study site.
Also, there has been an increase in the general willingness of hunters to check their deer
out of concern for Bovine TB in all parts of the state.

Although the male yearling average beam diameter, point count, and the percent
of does lactating has remained high in the SLP study site, these indices have, in general,
decreased over the 14 years of this study. Deer densities in this area of the state have
traditionally been high because the agriculture that exists supplements the natural food
sources. It is possible that the decrease in the quality indices of deer in the SLP study site
may be attributed to deer densities that have become so high that the competition for
resources has increased so much that the resources utilized by deer in that area have
become more difﬁcult to ﬁnd and the deer herd cannot be sustained in such good
condition.

Among the 3 regional study sites, only the antler measurements tended to follow

the same trends. For example, if the male yearling average beam diameter increased in

108

the UP study site, in general it also increased in the NLP and SLP study sites. Overall,
the male yearling point count followed the same sort of pattern, but the percent of
yearling does lactating did not. One reason the percent of yearling does lactating did not
follow this trend may be attributed to the small sample sizes; it is possible that if the
sample sizes had been larger that this index would follow similar trends in all 3 regional
study sites.

The long-term averages of the male yearling average beam diameter, male
yearling point count, and percent of does lactating revealed that all 3 of these quality
indices followed a regional gradient. As you move from north to south in Michigan, each,
of the condition indices increased. This suggests that deer in the southern part of the state
are in better physical condition than their more northern counterparts. In addition, even
though some of the long-term averages do not appear to be very different, especially
between the UP and NLP study sites, both antler measurements were statistically
different among the 3 regional study sites, which implies that in actuality the condition of

deer in these areas are distinct.

Sample Sizes and Aggregation of Data

There were problems with ensuring an adequate sample size to conduct data
analyses when the data were broken down into components, such as spatial scale. The
biodata could not be aggregated below a county level in most instances due to the low
number of deer checked. This resulted in small sample sizes, which were inadequate to
conduct data analyses. Therefore, the dynamics of the quality indices at ﬁner scales were

not examined in this study. Also, the lactation status data were really only useful at the

109

regional level because the sample sizes were too small when these data were broken into
smaller spatial components (e.g., county). Likewise, the fawn-to-doe ratio data provided
only questionable sample sizes at the regional level. The number of does observed was,
in general, very low and never exceeded 90 in any 1 transect. This was also the case for
the nmnber of fawns observed in any 1 transect. According to the MDNR and other
studies conducted in the transect survey area, the density of deer is very high in the SLP
study site (I. Pusateri, unpublished data). Therefore, there is some question as to how
well the deer observed along the transects represent the deer composition and density that
actually exists in those areas. In addition, the transect data were probably inﬂuenced by
the existing vegetation types along each of the routes. This suggests that if a route
consisted of vegetation types not preferred by deer, then it was less likely that deer would
be observed. These limitations need to be considered when performing data analyses;
otherwise, the results may not accurately reﬂect the actual condition of the deer herd.
This may be problematic if managers then use these results to develop deer management

strategies.

Biases
Biodata

Ideally, the biodata would accurately represent the true composition of the white-
tailed deer herd in Michigan. This, however, is not likely due to many potential biases
that may be inherent within the data. First, the biodata consist of information collected
only from deer that were harvested and checked by hunters. Therefore, bias is introduced

because the sample of deer collected is based on only one segment of the population.

110

Furthermore, many hunters may select for a speciﬁc type of deer. According to Bull and
Peyton (2000), hunters would rather take large bucks than does or fawns. Males may
also be more susceptible to harvest due to larger home ranges and greater movements
(Roseberry and Klimstra 1974). Another potential bias is introduced because the
information collected on where deer were harvested does not necessarily reﬂect where
they ranged throughout the rest of the year.

Other biases are introduced because the biodata are collected from deer checked
voluntarily; thus, the checked deer may not be a random representation of the population
of harvested deer. In actuality, approximately 7% of the deer harvested are checked, and
older, larger and antlered deer are more likely to be checked than younger, smaller or
antlerless deer (Bull and Peyton 2000). Also, hunters are more likely to check deer when
it is convenient for them, in terms of location and the amount of time it takes to check a
deer.

In addition, the percent of deer checked varies between those harvested on public
and private land, which also may bias the biodata. Theoretically, hunters are more likely
to check deer harvested on public land. This is due to the fact that hunters hunting on
public land leave the property and are more likely to pass a check station while
transporting their deer (Cook and Winterstein 2001). In some areas of the state, however,
there is very little public land available for hunting. So, in actuality, more deer are
harvested and checked from private lands. Plus, many hunters believe that the
composition of deer differs on public and private land. Therefore, hunters that are

selecting for a speciﬁc type of deer may be more inclined to hunt in a particular area.

111

Not only do biases exist in the biodata due to the voluntary nature of deer
checking, but biases also exist due to the spatial distribution of check stations (Cook and
Winterstein 2001). There are certain counties that account for more or less of the biodata
than is reﬂected from the estimated harvest. Most counties in the UP and SLP are under-
represented; this is most likely due to inaccessibility to or lack of check stations. Several
counties in the NLP, however, are over-represented due to hunters checking their deer for
Bovine TB, which is a self-sustaining disease in the deer herd in that region (Schmitt et
a1. 1997). Also, there are a greater number of check stations, overall, within the NLP
region. So, when analyses are conducted, the results may be heavily biased to this region
of the state.

The seasonal distribution of the hunting seasons may also affect the biological
information collected from the deer for the biodata. Over 80% of the biodata are
collected during ﬁrearm season, with the majority of the non-ﬁrearm season biodata
being collected during archery season. The composition of the harvest differs among the
various hunting seasons, which suggests that the time of year inﬂuences biological
characteristics. Also, the age structure of harvested bucks as well as the sex structure
differs among the hunting seasons (Cook and Winterstein 2001). Therefore, if data
analyses are conducted using a combination of the hunting seasons, the results could be
heavily biased toward the data collected during the ﬁrearm season.

The biodata also may not represent the true composition of the deer population
because the deer that are checked are subject to potential measurement error. Even
though check station volunteers are trained on how to age deer, agers tend to under-age

older deer (over 4 years old) and to over-age younger deer (Cook and Winterstein 2001).

112

The measurement error for yearlings, however, was low (Cook and Winterstein 2001).
Measurement error could also be introduced when the volunteers measure the beam
diameter (mm) on bucks. Since the measurement units are so small, there is some
question as to whether the volunteers are able to accurately read the calipers. There is

also an issue of how consistent the volunteers are when taking the beam measurements.

Summer Driving T ransects

Data collected from the summer driving transects also have potential biases. The
routes themselves may introduce bias to the data collected. Since the routes are driven,
the only deer that are observed are those that can be seen from the road. This may skew
the data collected on fawn—to-doe ratios because the deer observed are only a small subset
of the deer present in an area and may not be representative of the entire deer population.
Therefore, it may appear that deer near roads have either more or fewer fawns than what
is, in reality, a true representation of number of fawns in the overall deer population.
Although, the routes were chosen to maximize the number of deer observed, they were
limited by driving accessibility and the need to encompass a variety of habitat types. So,
there may have been sections of each route that were not ideal for deer observation and,
thus, the number of deer observed while driving the route may have been much lower
than the number of deer actually present.

The time of day that the data were collected may have also biased the fawn-to-doe
transect data. Deer are usually most visible from the road during the early morning and
in the evenings because they tend to be more active during these times of day. Based on

year 1 results, in year 2 the routes were only driven in the evening, which may have

113

affected the number and type of deer observed. It is difﬁcult to see deer at sunset. Thus,
there are probably a lot of deer that do not get counted simply because they cannot be

seen, and if they are observed, it is difﬁcult to differentiate what age and/or sex they are.

Bias Reduction

The overall quality of the biodata could potentially be improved and the effects of
biases reduced if the number of points or average beam diameter values were weighted
by the amount of harvest within a geographic area or by season (Cook and Winterstein
2001). Other ways to reduce biases are to alter the logistics of the check stations. If the
MDNR added check stations in certain areas of the state (e.g., SLP and UP), the biodata
would be a better representation of the deer population throughout the state. Plus, adding
stations may make checking deer more convenient for hunters, and possibly more deer
would be checked (Cook and Winterstein 2001). Enticing hunters to check their
harvested deer would also reduce biases. Having some sort of reward program, or an
education program aimed to inform hunters of the importance of checking their deer
could increase the number of deer checked. Another option may be for the MDNR to
condsider making the deer check mandatory.

Given that we stratiﬁed the state into 3 regions, the geographical biases that exist
in the biodata were not problematic for this study. The nature of the data deﬁned how
they should be aggregated for analyses. Since there are spatial and seasonal biases in the
biodata, when conducting analyses, it made sense to break the data down into spatial
and/or seasonal components. To avoid results biased heavily on biological information

taken from deer in the Northern Lower Peninsula, the data were broken down into

114

regions of the state and initial statistical analyses were conducted at this regional level.
Similarly, to avoid results biased heavily on the biological information taken from deer in
the ﬁrearm season, the biodata can be broken down into the different hunting seasons in

order to conduct data analyses.

Assumptions

Despite all of these potential biases, the biodata are the largest, most
comprehensive, and most readily available data on the composition and physical
condition of the Michigan deer herd, and were used to approximate the true population
parameters of the deer population. Likewise, the fawn-to-doe transects, in spite of biases,
also provided valuable information on the deer herd in Michigan because these data
complemented the biological information collected for the biodata. These data, however,
can only be used appropriately if the biases and limitations of the data are recognized and
understood. So, for the most part, the biodata and the fawn-to-doe transect data are best
used to provide an indication of general trends among the entire population instead of
true estimates of population parameters.

Most data collected have some sort of associated bias; therefore, making
assumptions about the data was necessary to perform analyses. According to Cook and
Winterstein (2001), the biological information contained within the biodata generally
followed the expected trend of deer population parameters in the 3 distinct regions of the
state. Thus, it was assumed that, even though the Biodata consists of information
collected from a portion of the deer harvested, it was more or less representative of the

entire deer population. So, if it can be assumed that the biodata were representative of

115

the entire deer population, it can also be assumed that if measurement errors occurred,
they were negligible.

There were also assumptions made about the fawn-to-doe transect data. Fawns
are large enough to move around, and thus have a greater likelihood of being seen from
the road during the latter part of the summer. So, even though transects are performed
only during late August and September, it was assumed that this time of year provides the
best Opportunity to observe fawns and does together. It was also assumed that the
number of deer observed in the evening were probably representative of the number of
deer that would be observed in the morning. Deer have home ranges; therefore, it was
assumed that the docs and fawns observed in a particular area in the evening would most

likely be in the same vicinity during a morning transect.

Alternative Indices

There are several population-level quality indices that were not used in this study,
but may still be useful in determining white-tailed herd quality in Michigan. One
alternative is to determine the physical condition of white-tailed deer by collecting weight
(Severinghaus 1979, Moen and Severinghaus 1981). Growth and maintenance of
individuals requires a certain amount of food intake and nutrition. If these needs are not
met, weight, along with other characteristics that indicate physical condition will be
substandard (Rasmussen 1985). For example, there is a close relationship between
weight, reproductive rates and antler grth (Hesselton and Sauer 1973). Females that

are underweight will conceive fewer fawns, if any, and bucks will produce inferior

116

quality antlers with fewer points, smaller spread and diameter (Severinghaus et a1. 1950,
Severinghaus and Moen 1983).

One method of collecting weight data is to weigh hunter-killed deer on standard
platform scales at deer harvest check stations during the fall harvest. However, there are
some considerations that must be recognized when using weight as an index of physical
condition. The ﬁrst is age. Weight increases until a deer reaches its prime and then
levels off. Thus, it is important to record an accurate age along with weight
(Severinghaus 1955, Severinghaus 1979). Another is tarsal length. Hind foot
measurements should be taken to compensate for size (height). In addition, season
should be taken into account because ﬂuctuations in body weight occur throughout the
year (Fowler et a1. 1967).

It is not always feasible to set up a weigh station at a harvest check station.
Therefore, another way to assess physical condition of white-tailed deer is to examine fat
deposits. If range conditions are poor, resulting in low food intake, deer lose fat deposits
(Harris 1945). Thus, another physical condition index that can used to determine the
quality of white-tailed deer is measurement of the fat reserves in the animals body,
traditionally femur marrow fat (Nichols and Pelton 1972, Purol, et al. 1977, Watkins et
a1. 1991). This, however, is a highly invasive procedure that requires dissection of the
hindquarter. The result is a substantial loss of edible venison, which makes a less invasive
method more desirable.

The fat content of the mandibular cavity tissue (MCT) is an alternative method to
the traditional femur marrow fat assessment (Purol et a1. 1977). Mandibles are collected

from hunter-killed deer, and can be removed without defacing the trophy, when

117

necessary. Usually, the tissue samples are dried and ether extraction is used to estimate
the fat concentration of marrow (Baker and Lueth 1966). However, marrow dry matter
percentage of the mandible can also be used as an index of body fat concentration, and is
determined by oven-drying or reagent methods (Watkins et al. 1991).

There is an indication that MCT fat has limited usefulness as an index because it
is not mobilized until body fat declines to approximately 12%. This suggests that a deer
with a high percentage of body fat may have the same percentage of manibular fat as a
deer with a low percentage of body fat (Nichols 1974, Watkins et al. 1991).
Consequently, marrow fat cannot be used to discriminate among animals with high body
fat concentrations, but is more useful for deer in poor condition (Baker and Lueth, 1966).
On the other hand, if this index has a relatively good correlation with other factors related
to body condition, such as antler growth or lactation rates; it may be very useﬁll to build a

more complete index of herd quality.

Conclusion

This project evolved based on the nature of the data and the results from analyses.
Initially, the scope of this project only encompassed determining whether population
quality indices (e. g. point counts, average beam diameter and lactation status) could be
used to deﬁne white-tailed deer herd quality in Michigan. After examining the data,
however, it was apparent that the spatial and temporal trends in deer population
parameters could be important to determine ﬂuctuations in herd condition. The spatial
and temporal trends in herd condition may also be beneﬁcial for examining how deer

herd condition corresponds with the ecological and social aspects of deer management.

118

Studying temporal and spatial variation in herd quality may also help to isolate some of
the underlying mechanisms (e. g., habitat quality, deer density and weather) that cause

shifts in the biological carrying capacity of the deer population.

119

LITERATURE CITED

Albert, D. A., S. R. Denton, and B. V. Barnes. 1986. Regional landscape ecosystems of
Michigan. School of Natural Resources, University of Michigan, Ann Arbor, MI.

Baker, M. F., and F. X. Lueth. 1966. Mandibular cavity tissue as a possible indicator of
condition in deer. Proceedings of the Annual Conference of the South East
Association of Game and Fish Commissioners 20:69-74.

Bull, P. and B. Peyton. 2000. Pilot study report: The 1999 Michigan deer check station
survey. Michigan Department of Natural Resources, Wildlife Division, Lansing
MI.

Cook, S. L. and S. R. Winterstein. 2000. The evaluation of the MDNR’S white-tailed
deer lactation data.

. 2001. The evaluation of the MDNR’s biophysical

 

deer check station data.

Cowan, R. L. and T. A. Long. 1962. Studies on antler growth and nutrition of white-
tailed deer. Proceedings of the ﬁrst National White-Tailed Deer Disease
Symposium.

Ford, W. M., A. S. Johnson, and P. E. Hale. 1997. Inﬂuences of forest type, stand age,
and weather on deer weights and antler size in the Southern Appalachians.
Journal Society of American Foresters 21 :1 1-18.

Foster, J. R., J. L. Roseberry, and A. Woolf. 1997. Factors inﬂuencing efﬁciency of
white-tailed deer harvest in Illinois. Journal of Wildlife Management 61 :1091-
1097.

Fowler, J. F., J. D. Newsom, and H. L. Short. 1967. Seasonal variation in food
consumption and weight gain in male and female white-tailed deer. Proceedings
of the Annual Conference of the South East Association of Game and Fish
Commissionsers 21 :24-31 .

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

Hamilton, J., W. M. Knox, and D. C. Guynn, Jr. 1995. How quality deer management
works. Pages 7-18 in K. V. Miller and R. L. Marchington, eds. Quality
Whitetails: The How of Quality Deer Management. Stackpole Books,
Mechanicsburg, PA.

120

Harris, D. 1949. Symptoms of malnutrition in deer. Journal of Wildlife Management
9(4)319-322.

Hesselton, W. T., and P. R. Sauer. 1973. Comparative physical condition of four deer
herds in New York according to several indices. New York Fish and Game
Journal 20(2):77-107.

Hill, H. R., J. Meister, and J. Pohl. 1981. Deer checking station data. Michigan
Department of Natural Resources, Wildlife Division Report 2924.

Huot, J. 1988. Review of methods for evaluating the physical condition of wild
ungulates in northern environments.

Jenkins, D. H. and I. H. Bartlett. 1959. Michigan Whitetails. Michigan Department of
Natural Resources, Wildlife Division Report 96-R.

Johnson, D. R. and J. K. Agee. 1988. Introduction to ecosystem management. Pages 3-
14 in J. K. Agee and D. R. Johnson, eds. Ecosystem management for parks and
wilderness. University of Washington Press, Seattle, WA.

Langenau, E. 1994. 100 Years of deer management in Michigan. Michigan Department
of Natural Resources, Wildlife Division Report 3213.

McCabe, T. R. and R. E. McCabe. 1997. Recounting Whitetails past. Pages 11-26 in W.
J. McShea, H. B. Underwood, and J. H. Rapploe, eds. The Science of
Overabundance: Deer Ecology and Population Management. Smithsonian
Institution Press, Washington, DC.

McCullough, D. R. 1979. The George Reserve deer herd. University of Michigan Press.
Ann Arbor, MI.

. 1982. Antler characteristics of George Reserve white-tailed deer.
Journal of Wildlife Mangement 46:821-826.

 

Moen, A. N., and C. W. Severinghaus. 1981. The annual weight cycle and survival
of white-tailed deer in New York. New York Fish and Game Journal 28(2):162-
177.

Nichols, R. G. 1974. Fat in the manibular cavity as an indicator of condition in deer.
Proceedings of the Annual Conference of the South East Association of Game and
Fish Commissionsers 28:540-548.

Nichols, R. G., and M. R. Pelton. 1972. Variations in fat levels of manibular cavity
Tissue in white-tailed deer (Odocoilues virginianus) in Tennessee. Proceedings of
the Annual Conference of the South East Association of Game and Fish
Commissionsers 26:57-68.

121

Ozoga, J. J ., R. V. Doepker, and M. S. Sargent. 1994. Ecology and mangement of white-
tailed deer in Michigan. Michigan Department of Natural Resources, Wildlife
Division Report 3209.

Purol, D. A., J. N. Stuht, and G. E. Burgoyne. 1977. Mandibular cavity tissue fat as an
indicator of spring physical condition of female white-tailed deer in Michigan.
Michigan Department of Natural Resources, Wildlife Division Report 2792 pp. 1 -
7.

Rasmussen, G. P. 1985. Antler measurements as an index to physical condition and
range quality with respect to white-tailed deer. New York Fish and Game Journal
32:97-113.

Roseberry, J. L. and W. D. Klimstra. 1974. Differential vulnerability during a controlled
deer harvest. Journal of Wildlife Management 38:499-507.

Schmitt, S. M., S. D. Fitzgerald, T. M. Cooley, C. S. Bruning-Fann, L. Sullivan, D.
Berry, T. Carlson, R. B. Minnis, J. B. Payeur, and J. Sikarskie. 1997. Bovine

tuberculosis in free-ranging white-tailed deer from Michigan. Journal of Wildlife
Diseases 33:749-758.

Severinghaus, C. W. 1955. Deer weights as an index of range conditions on two
wilderness areas in the Adirondack Region. New York Fish and Game Journal

2:154-160.

. 1979. New York Fish and Game Journal 26(2):162-187.

 

Severinghaus, C. W., H. F. Maguire, R. A. Cookingham, and J. E. Tanck. 1950.
Variations by age class in the antler beam diameters of white-tailed deer related to

range conditions. Transactions of the North American Wildlife Conference
15:551-5 70.

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(1):30-38.

Ullrey, D. E. 1983. Nutrition and antler development in white-tailed deer. Pages 49-59
in R. D. Brown, editor. Antler development in cervidae. Caesar Kelberg Wildlife
Research Institute, Kingsville, TX.

Wackerly, D. D., W. Mendenhall, and R. L. Scheaffer. 1996. Mathematical Statistics
with Applications. Wadsworth Publishing Company, Belmont CA.

Watkins, B. E., J. H. Witham, D. E. Ullrey, D. J. Watkins, and J. M. Jones. 1991. Body
composition and condition evaluation of white-tailed deer fawns. Journal of
Wildlife Management 55(1):39-50.

122

CHAPTER 2

AN EXAMINATION OF UNDERLYING MECHANISMS THAT MAY
INFLUENCE WHITE-TAILED HERD CONDITION IN MICHIGAN

INTRODUCTION

One of the best methods to determine white-tailed deer herd quality is to evaluate
the physical growth of deer (Cowan and Long 1962, Severinghaus and Moen 1983,
Rasmussen 1985). Some of the more easily obtainable measurements of herd condition
in Michigan include average beam diameter, point count, and lactation status of does
from harvest data (Hill et al.1981, McCullough 1982, Cook and Winterstein 2000).
These population-level quality indices vary by age-class (e.g., in general older deer have
larger antler measurements) but in yearling deer they are more pronounced because this
age-class has the added burden of body grth while growing their ﬁrst set of antlers or
producing their ﬁrst offspring (French et a1. 1956, Cowan and Long 1962, Ullrey 1983).
Therefore, population-level quality indices for yearling deer such as average beam
diameter, point count, and lactation status can be used as a measure of the overall deer
herd quality in Michigan (chapter 1).

In Michigan, male yearling average beam diameter, male yearling point count,
and lactation status of yearling does increases as you move along a north to south
regional gradient. It has been recognized for some time that these quality indices
increase in the more southern regions of the state (Ullrey 1983, Ozoga et a1. 1995, Cook
and Winterstein 2000, chapter 1). This suggests that the overall condition of deer in the
southern part of the state is better than the condition of deer in the northern part of the

state. The real question, however, is why does this north to south trend in the population-

123

level quality indices exist? From the literature, there have been 3 major underlying
mechanisms identiﬁed that have the potential to inﬂuence white-tailed deer herd quality
in Michigan: population density, winter severity, and habitat quality (Blouch 1984,
Matschke et a1., 1984, Ozoga et a1. 1995). These 3 factors vary among the 3 main regions
in the state and may, at least partially, explain the regional trends in deer population-level
quality indices in Michigan.

Herd density, or the number of deer that exist in a speciﬁed area, can affect the
ability of deer to either ﬁnd or utilize suitable resources, which can, subsequently
inﬂuence the physical condition of deer. Higher population densities, in general, lead to
a decrease in deer growth, survival, and reproduction because the competition for
resources can become so great that not all deer are able to obtain adequate food and
shelter (Johnson 1937, Leberg and Smith 1993, Jacobson and Guynn, 1995).
Speciﬁcally, high deer densities can affect forage production and the availability of cover
as well as increase stress-levels (Cheatum and Severinghaus 1950, Sams et a1. 1998).

When deer densities exceed the capacity of the land to sustain a healthy deer herd,
there are noticeable affects on reproduction: younger does breed later in life and newborn
fawn mortality increases (Payne 1970, Miller et al. 1995). There are also detrimental
affects on antler development; a larger percentage of yearling deer have spike antlers
(less than 3 inches) and deer in other age-classes may not produce antlers to their full
potential (Ozoga et al. 1995). In addition, overcrowding is often associated with the
outbreak of disease because these deer are more likely to be in poorer physical condition
and there is more opportunity for disease to be spread because of close physical contact

(Leopold 1933, Jacobson and Guynn 1995).

124

Winter severity, which includes temperature and snowfall, inﬂuences deer
mobility, productivity, and mortality (Blouch 1984). As temperatures drop, deer move
into wintering yards and once snowfall accumulates, the movement of deer may be
restricted, making it difﬁcult to ﬁnd suitable food or shelter (Venue 1968, Marchington
and Hirth 1984). The temperature and amount of snowfall can also affect the ability of
deer to ﬁnd and utilize resources because these factors can not only change the growth
and seasonal availability of food but also can cover nourishing food, in some cases
making this resource unavailable to the deer (Cheatum and Severinghaus 1950, Matschke
et al.1984). The unavailability of resources, in turn, can cause starvation and weaken the
deer so that mortality increases and a subsequent decline in deer numbers occur (Ozoga
1968, Kams 1980).

In addition, as temperature decreases and snowfall increases, deer expend more
energy trying to ﬁnd suitable resources, especially if subzero temperatures, harsh winds,
and large amounts of snowfall are sustained throughout the winter (Verrne 1968, Ozoga
and Gysel 1972, Ozoga et a1. 1995). This can greatly affect the physical condition of
deer because as body fat is depleted, body weight decreases and if adequate resources are
not acquired eventually there may be detrimental effects on other population
characteristics such as reproduction and antler development (Severinghaus et al. 1950).

There is a direct relationship between the productivity of deer and the quality of
the habitat they utilize (Ford et al. 1997). Deer require proper nutrition from forage and
browse to obtain optimal physical condition. As the habitat quality increases, in general,
the level of nutrition also increases because high quality deer habitat provides optimum

levels of nutrients required for growth, reproduction, and, hence, survival (Harlow 1984,

125

Leberg and Smith 1993). This suggests that when deer are found in high quality habitat,
they are more likely to be in better physical condition. Indeed, there have been numerous
studies that show a detrimental effect on deer growth and overall physical condition when
the deer range consists of poor quality habitat that provides poor nutrition (Cheatum and
Severinghaus 1950, French et al. 1956, Shea and Osborne 1995). Speciﬁcally, deer that
range in poor quality habitat are likely to have below average weight and antler
development, both the number of points and the average beam diameter (Kie et al. 1983,
Sams et al. 1998). Reproductive performance is also related to habitat quality, a doe that
ranges in better quality habitat has the ability to produce not only more offspring but
healthier offspring than a doe that ranges in poorer quality habitat (Cheatum and
Severinghaus 1950, Leberg 1993, Shea and Osborne 1995).

The quality of forage is important in determining the physical condition of deer,
but so is the quality of thermal cover in the winter. Deer migrate to areas of thermal
cover when the temperature starts to turn cold (Marchington and Hirth 1984). Yarding in
quality thermal cover can greatly enhance the ability of deer to survive long, harsh
winters, especially as body weights and fat reserves decrease to render deer in relatively
poor physical condition (Severinghaus 1947, Verrne 1965). In some cases, deer will
often forgo higher quality forage and browse to bed in higher quality thermal cover
throughout the winter (Ozoga and Gysel 1972).

There is, obviously, a great deal of interaction among the 3 major underlying
mechanisms that have the potential to inﬂuence white-tailed deer herd condition. High
population densities affect the quantity of forage and browse that is available for deer

consumption and since deer prefer high quality forage, high deer densities can also mean

126

that the better quality habitat is consumed more quickly (Cheatum and Severinghaus

195 0, Jacobson and Guynn 1995). This, in turn, makes the overabundant deer population
more susceptible to “die-offs” during cold winters with a large amount of snowfall
(Severinghaus 1955, Verme 1968). If deer densities are low, however, more resources
may be available to a greater number of deer. Thus, harsh winters may not necessarily
signify that there will be severe reductions in deer density if adequate resources are
available, whereas an overabundance of deer coupled with a nutritional shortage may
decrease the ability of deer to withstand long, harsh winters (Johnson 1937). These
interactions suggest that a combination of deer density, winter severity, and habitat

quality may in fact be at work in inﬂuencing deer herd condition.

127

Objectives
This study will explore why a north to south regional pattern exists in the white-
tailed deer herd quality in Michigan by evaluating underlying mechanisms that have the
potential to inﬂuence deer herd quality in relation to indices of herd condition.
Speciﬁcally, the following objectives will be met:
1) Examine the temporal and spatial trends in the underlying mechanisms that
have the potential to inﬂuence white-tailed deer herd quality in Michigan;
2) Determine if these underlying mechanisms can account for the regional
differences in the indices of herd quality.
By meeting these objectives, this study will help the MDNR to develop scientiﬁcally
sound deer management strategies that take into consideration the factors that may

inﬂuence deer distribution or demographics.

128

METHODS

Data Sources
Quality Indices

The lactation status of yearling does was not examined in relation to underlying
mechanisms that have the potential to inﬂuence deer herd condition because small sample
sizes for this quality index meant that any conclusions based on these data were broad in
nature and further analyses may not be overly meaningﬁll at this time. Male yearling
average beam diameter and point count measurements, however, were examined in
relation to underlying mechanisms that have the potential to inﬂuence deer herd
condition. Male yearling average beam diameter was chosen because this measurement
is one of the best indices of deer herd quality, and male point count was chosen because
this measurement is highly correlated with male yearling average beam diameter. These
measurements were obtained from the biophysical data (biodata) collected from hunter-
killed deer that are voluntarily brought into check stations throughout Michigan (refer to
Chapter 1 for complete details). The mean average beam diameter for male yearlings and
the mean point count for male yearlings were calculated for every county in the 3 study

sites from 1987 through 2000.

Population Density

Population density estimates were calculated based on the sex-age-kill estimate
method as described by Mattson-Hansen (1998). This method uses biological
information collected from hunter-killed deer during the ﬁrearm season to estimate the

deer population size before the hunting season and can be used to project future

129

 

population numbers. There are 3 main components to the sex-age-kill estimate method.
The ﬁrst component is the sex ratio. Only yearling bucks and docs are used to determine
the sex ratio because it is assumed that the harvest data reﬂect each age group
proportionally to its abundance in the population (Severinghaus and Maguire 1955). The
second component is the age ratio, in which the number of yearlings is divided by the
total number of known age adult deer for both bucks and does. The third component is
the annual population survival rate, which for deer in Michigan is 0.90. This accounts for
natural mortality that occurs throughout the year. We used information from the biodata
and the MDNR mail harvest survey to calculate population density estimates. The
number of bucks, does, and fawns was calculated separately, then added together to get a
total population density estimate for each county in each study site from 1987 to 2000
except for Baraga County in the UP study site. A density estimate could not be
calculated for this county because some of the data required to use the SAK estimate

method were not available.

Temperature

The mean monthly temperature data were downloaded from the National Oceanic
and Atmospheric Administration (NOAA) website in tenths of a degree Fahrenheit for
every county in the 3 study sites from 1987 to 2000 (http://www.ncdc.noaa. gov). The
mean monthly data were summed from January through September in each county in the
3 study sites from 1987 to 2000 to get 1 cumulative temperature reading for each county
in each year. These data were then converted from Fahrenheit into centigrade. Figure 2.1

is an example of how the cumulative mean monthly temperature in Alpena County

130

 

 

 

 

 

u 210

O

.5 190-

2

a 1703

a

h

8. 150

E

i- 130

:-

.-=.. 110

=

O

E 90 ~

5

o 70

2

so

15

‘5 30

e

5 10
-10

JAN FEB MAR APR MAY JUN JUL AUG SEP

Month

Figure 2.1. The cumulative mean monthly temperature for Alpena County, Michigan in
2000.

131

increased over the 9-month period in 2000. The mean monthly temperature data were
aggregated from January through September because by late summer (i.e., September),
antler grth has subsided and temperature no longer has an impact on antler

development (Weeks 1995).

Snowfall

The total monthly snowfall data were also downloaded from the NOAA website
in tenths of an inch for each county in all 3 study sites from 1987 to 2000. These
monthly data were then summed from January through April in each county in the 3
study sites from 1987 to 2000 to get 1 cumulative total snowfall reading for each county
in each year. These data were then converted from inches to centimeters. Figure 2.2 is
an example of how the cumulative total monthly snowfall in Marquette County increased
over the 4-month period in 2000. The total monthly snowfall data were aggregated from
January through April because that is when the majority of snow falls in Michigan

regardless of region (Blouch 1984).

Winter Severity

Winter severity is based on a combination of temperature and snowfall. In 1968
Verme devised a method in which winter severity could be quantitatively measured based
on these 2 factors. To obtain temperature and snowfall information, atmospheric chill
and snowfall levels were collected at MDNR stations. For this study, these data were
available in the UP and NLP study sites starting in the winter of 1986-1987 (winters will

from now on be identiﬁed from the ﬁrst year), but in the SLP study site these data were

132

 

180

 

 

 

 

4.4
A 160

E

e

-_- 140

a

E

u 120

m

2»

e 100

E

O

E 80

a

l3

,, 6O

.2

3’

'5 40-

a

5

U 20 J

0 .
JAN FEB MAR APR

Month

Figure 2.2. The cumulative total monthly snowfall for Marquette County, Michigan in
2000.

133

only available since 1988. Atmospheric chill and snowfall data were combined into a
weekly value, and over the course of the winter these weekly values were cumulated to
form a winter severity index (WSI) for each station where data were collected. The WSI
for each station within a speciﬁc region were then averaged together to obtain an overall
WSI for that region.

A corrected WSI was developed by Cook et al. in 2001. This index is based on
the same technique as the regular WSI, except only data collected from approximately
day 42 (mid-December) to day 168 (mid-April) are used in its development. The
corrected WSI was used in this study because the data used for the regular WSI did not
cover the same time period in each of the 3 regions, whereas the time period for
calculating the corrected WSI was standardized among all 3 regions. Therefore, based on
the corrected WSI, an average WSI for each region was calculated; in the UP and NLP
study sites a corrected WSI was calculated for each year from 1986 to 1999, and in the
SLP study site a corrected WSI was calculated for each year from 1988 to 1999. Figure
2.3 illustrates how the weekly corrected cumulative WSI increased in each region in

2000.

Habitat Potential

Habitat potential is essentially a measure of the ability of habitat types to provide
fall and winter food, spring and summer food, and thermal cover requirements for white-
tailed deer. Habitat potential models for white-tailed deer were developed by Felix et al.
(in press), and currently are only available for the NLP study site. To build habitat

potential models, habitat types must ﬁrst be deﬁned. Habitat types were delineated by

134

e-UP +NLP +SLP

 

Corrected Cumulative Average WSI

 

 

 

   

‘— b— 3— b-

8 8 g g g 8 a a: o. o.

D Q '7 '7 LL. LL. 2 E < <

«'3 r\' o '5’ 1" "'- 55 c': rv'i :4

-— N "‘ N N N '—
Date

Figure 2.3. The corrected cumulative winter severity index (WSI) for each of the 3
regional study sites in 2000.

135

overlaying maps of ecoregions, land-type associations, soil associations, and
presettlement vegetation (Figure 2.4). Based on the resulting 24 number of habitat types,
models were developed to predict the suitability for each white-tailed deer habitat
requirement throughout succession, 0 being the lowest suitability and 1 being the highest
suitability (Figure 2.4). To make the habitat potential data compatible with the
demographic data, the suitability for each deer habitat requirement was rescaled from a
smaller spatial scale (i.e., habitat types) to a larger spatial scale (i.e., county boundaries).
A weighted average for fall and winter food, spring and summer food, and thermal cover
potential was calculated based on the area of each potential value. This resulted in 1
habitat potential for each habitat requirement in every county in the NLP study site

(Figure 2.5).

Within Regions

In each study site, an analysis of variance (ANOVA) was conducted to examine
the relationship between male yearling average beam diameter or male yearling point
count and population density, temperature, and snowfall. Regression analyses were
conducted to determine the relationship between male yearling average beam diameter or
male yearling point count and the corrected WSI in all 3 regions. The long-term average
value of the yearling average beam diameter and point count in each county was also
regressed against the habitat potential for fall and winter food, spring and summer food,
and thermal cover in the NLP study site only. To determine in which counties the antler
measurements were related to a speciﬁc underlying mechanism, both the ANOVA and

regression analyses were run by county. Signiﬁcance levels were determined at an alpha

136

 

I >0.78
I 0.71-0.78
I oar-0.70
I

I

U

0.55-0.60
0.00-054
Lake

 

   
        

I >030
I 0.61-0.80
I 0.41-0.60
I 021-040
D 0.00-020
[1 Lake
Spring & Summer
HabitatTypes
I >033
I 0.81093
I 0.51-0.80
' I 0.01-0.50
3: 0.00
___- ___L. Lake

Thermal Cover

Figure 2.4. The habitat types for the Northern Lower Peninsula study site as deﬁned by
Felix et al. (unpublished data) and the habitat potential for fall and winter food, spring
and summer food, and thermal cover based on the habitat types delineation.

137

I: Presque Ble (0.581)
Alpena (0.596)
Mmtmorency (0.629)
- Alcona (0.659)

- Oscoda (0.676)

 

Fall and Winter Food

D Oscoda (0.512)

~ Alpena (0.536)
- Mmtmorency (0.558)
- Presque Ble (0.566)
- Alcona (0.628)

 

Spring and Summer Food

[:1 Oscoda (0.033)

-- Mintmorency (0.17)
- Alcona (0.182)
- Presque Ble (0.39)
- Alpena (0.551)

 

Thermal Cover

Figure 2.5. The weighted habitat potential for fall and winter food, spring and summer
food, and thermal cover for each county in the Northern Lower Peninsula study site.

138

 

level of 0.10. Yearly trends in deer population density, temperature, snowfall, and the
corrected WSI were also examined in each of the 3 regional study sites by calculating the
yearly average of each underlying mechanism in each region and plotting these data in a
line graph. In addition, a long-term average was calculated for each of the underlying
mechanisms (except habitat potential) in the regional study sites. This accomplished by
averaging all of the values for each underlying mechanisms over the 14 years of the

study, which resulted in 1 average value for each underlying mechanism for each region.

Among Regions

Each of the underlying mechanisms that have the potential to inﬂuence male
yearling antler measurements, with the exception of the habitat potential mechanism,
were compared by performing an AN OVA to determine if there were statistically
signiﬁcant differences among these factors among the 3 regional study sites at an alpha
level of 0.10. Mean comparisons of the underlying mechanisms among regions were also
examined using a Tukey-Kramer multiple comparison test at an alpha level of 0.10.
Yearly comparisons were also made among the underlying mechanisms in the 3 regions
by calculating the yearly average of each underlying mechanism in each region and
plotting these data in a line graph to examine trends. This method was necessary due to
the fact that quantitative analyses (e. g., Tukey-Kramer multiple comparison test) could
not be conducted because there were not enough data points and, thus, not enough

degrees-of-freedom (df).

139

RESULTS

Comparisons within Study Sites

UP
The number of deer, as estimated by the sex-age-kill estimate method, in the UP

study site ranged between 26,411 in 1997 and 62,131 in 1991; the long-term average
number of deer from 1987 through 2000 was estimated to be 46,332 (Table 2.1). In
general, the number of deer tended to be higher from 1987 to 1993 due to the fact that in
only 1 year (1988) was the number of deer below the long-term regional average (Figure
2.6). Conversely, the number of deer tended to be lower from 1994 to 2000; in only 2 out
of the 5 years was the number of deer above the long-term regional average. The deer
numbers were at the lowest in the study from 1997 through 2000 (Figure 2.6).

Iron and Marquette Counties, the larger counties in the UP study site, had larger
overall average numbers of deer (f = 50,084 and f = 50,762, respectively) than did
Dickinson County, the smaller county (5? = 40,823) (Figure 2.7). Although Marquette
County had the highest number of deer, in actuality the deer density in this county was
the lowest in the UP study site at 11 deer/km2 (28 deer/miz). The number of deer in Iron
County was slightly below that in Marquette County, but this county had the highest
density of deer in the UP study site at 17 deer/km2 (43 deer/miz). Dickinson County had
the lowest number of deer, but the density of deer fell in between that of Marquette and
Iron Counties at 14 deer/km2 (35 deer/miz). Even though there were counties in the UP
study site that had higher deer densities, in actuality, the difference in deer density among

the counties did not exceed approximately 6 deer/km2 (15 deer/miz).

140

Table 2.1. The total number of deer, as estimated [Tom the sex-age-kill estimate method,
in each county in the Upper Peninsula study site between 1987 and 2000; the yearly
average total number of deer and standard deviation (StDev) for the entire study site, and
the average total number of deer for each county.

 

 

 

 

 

 

 

 

 

Baragg Dickinson Iron Marquette Yearly Avera e StDev
1987 n/a 48529 35262 83279 55690 24797
1988 n/a 48684 49986 28376 42348 121 18
1989 n/a 52866 55623 48784 52424 3441
1990 n/a 42445 48076 49700 46740 3807
1991 n/a 62883 61236 62273 62131 833
1992 n/a 56885 69391 51492 59256 9182
1993 n/a 30266 n/a 86156 58211 39520
1994 n/a 25683 54499 35895 38692 14610
1995 n/a 52481 41638 69396 54505 13989
1996 n/a 35079 55507 61815 50800 13975
1997 n/a 34602 n/a 18221 2641 1 1 1583
1998 n/a 18043 39749 n/a 28896 15348
1999 n/a 30075 38803 26481 31787 6336
2000 n/a 32993 51241 3 8041 40758 9423
Total n/a 40823 50084 50762 46332 5553

 

 

*n/a denotes that the data were not available

141

 

 

+Yearly Average " " Long-Term Average

 

 

65000

60000

55000 a

50000 e

45000 4

40000

35000

30000 *

 

Average Total Number of Deer

25000 ‘

20000

15000

 

 

 

10000
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.6. The temporal trend in the average total number of deer, as estimated by the
sex-age-kill estimate method, for the Upper Peninsula study site except Baraga County
between 1987 and 2000.

142

 

  

Total Deer Numbers by County

I:] Ba ra g a (n/a)
E ch kin s o 11 (40,823)
Iro 11 (50,084)
- Ma rqu e tte (50,762)

Figure 2.7. The average total number of deer by county in the Upper Peninsula study
site for 1987 through 2000.

143

 

The overall cumulative mean monthly temperature for the UP study site was
198.63 °C and the yearly averages ranged between 185.50 °C in 1996 and 217.44 °C in
1987 (Table 2.2). The temporal trend in the cumulative mean monthly temperature
appeared to be cyclical, and oscillated around the long-term regional average (Figure
2.8). In the majority of years, the cumulative mean monthly temperature was at or
slightly below the long-term average. However, in those years in which the cumulative
mean monthly temperature was above the long-term regional average, it was within 20
°C (Figure 2.8).

The highest temperature readings occurred in Marquette and Dickinson Counties,
the eastem-most counties in the UP study site (2? = 213.09 °C and if = 209.18 °C,
respectively) (Figure 2.9). Even though the cumulative mean monthly temperature was
higher in Marquette County than in Dickinson County, the overall average for each
county was similar. Baraga and Iron Counties, on the other hand, the westem-most
counties in the UP study site had the lowest cumulative mean temperature readings (f =
183.47 °C and 56 = 188.77 °C, respectively) (Figure 2.9). The cumulative mean monthly
temperature in these 2 counties with the lower temperature readings was similar.

The cumulative total monthly snowfall in the UP study site ranged between
98.43cm in 1991 and 289.05cm in 1996, with an overall total monthly snowfall for the
region of 169.73cm (Table 2.3). In 9 out of the 14 years of this study, the average
cumulative total monthly snowfall per year was below the regional average. For
example, in 1991, the snowfall was approximately 70cm below the regional average;
although, in the majority of years that the snowfall was below the regional average, it was

within 30cm of the regional average (Figure 2.10). In the majority of years in which the

144

 

 

 

Table 2.2. The cumulative mean monthly temperature (summed from January through
September) in °C for each county in the Upper Peninsula study site from 1987 to 2000;
the yearly average cumulative mean monthly temperature and standard deviation (StDev)
for the entire study site, and the average cumulative mean monthly temperature for each
county.

 

 

 

 

 

 

 

 

 

 

 

Baraga Dickinson Iron Marquette Yearly Average StDev
1987 219.30 214.94 208.11 227.39 217.44 8.08
1988 136.89 208.33 191.06 211.50 186.94 34.56
1989 195.56 200.00 182.39 203.00 195.24 9.09
1990 207.61 210.67 194.83 214.44 206.89 8.51
1991 208.22 215.83 170.33 216.22 202.65 21.86
1992 194.61 204.72 183.28 206.83 197.36 10.80
1993 158.33 203.39 182.22 204.22 187.04 21.68
1994 191.61 200.33 179.11 197.61 192.17 9.44
1995 208.61 213.50 193.17 213.89 207.29 9.72
1996 180.00 194.33 173.06 194.61 185.50 10.74
1997 182.00 202.61 182.44 207.94 193.75 13.49
1998 106.28 230.06 211.06 233.89 195.32 60.19
1999 210.72 217.72 198.06 224.78 212.82 11.39
2000 168.89 212.1 1 193.72 226.94 200.42 25.03
Total 183.47 209.18 188.77 213.09 198.63 14.69

 

145

 

270.00

250.00

230.00

210.00

190.00

170.00

Average Cumulative Temperature in “C

150.00

+Yearly Average ' " 'Long-Tenn Average

 

 

 

 

 

f

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.8. The temporal trend in the average cumulative temperature (summed from
January through September) in °C for the Upper Peninsula study site between 1987 and

2000.

146

 

A
I
l-

 

 

 

Cumulative Mean Monthly Temperature
E Baraga (183.47°C)

, Iro 11 (188.77 °C)

- ch kins o n (209.18°C)
- Marquette (213.09 °C)

 

  

Figure 2.9. Cumulative mean monthly temperature (summed from January through
September) in °C by county in the Upper Peninsula study site from 1987 through 2000.

147

 

" a

Table 2.3. The cumulative total monthly snowfall (summed from January through April)
in cm for each county in the Upper Peninsula study site from 1987 to 2000; the yearly
average cumulative total monthly snowfall and the standard deviation (StDev) for the
entire study site, and the average cumulative total monthly snowfall for each county.

 

 

 

ll

 

 

 

 

 

 

 

 

 

Baraga Dickinson Iron Marquette Yearly Average StDev
1987 246.38 51.56 60.45 144.53 125.73 90.69
1988 494.79 76.71 90.17 212.34 218.50 194.03
1989 322.58 112.27 111.76 190.25 184.21 99.34
1990 226.06 47.24 53.09 182.88 127.32 90.85
1991 204.47 61.21 0.00 128.02 98.43 87.93
1992 213.36 112.01 107.95 203.71 159.26 57.06
1993 222.25 93.73 105.41 173.74 148.78 60.37
1994 196.85 97.54 95.25 159.77 137.35 49.67
1995 358.14 69.09 107.95 186.94 180.53 128.16
1996 518.16 171.96 175.26 290.83 289.05 162.43
1997 505.46 123.95 138.43 263.65 257.87 176.57
1998 205.74 140.21 85.09 156.97 147.00 49.76
1999 188.98 92.46 96.77 162.56 135.19 48.11
2000 265.94 137.41 98.04 166.62 167.01 71.69
Total 297.80 99.10 94.69 187.34 169.73 95.45

 

 

148

 

 

+ Yearly Average " " Long-Term Average

 

290.00
270.00

 

250.00 *
230.00
210.00
190.00
170.00 - -------------—----—---_.
150.00 a
130.00 '
110.00 .

Average Cumulative Snowfall (cm)

 

90.00 -
70.00 *

 

 

 

50.00 . . . . . , , , ,
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.10. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Upper Peninsula study site between 1987 and 2000.

149

average cumulative total monthly snowfall was above the regional average, the snowfall
was within 30cm above the regional average; however, in 1996 and 1997, the snowfall
exceeded the regional average by almost 100cm (Figure 2.10).

The more northern counties in this study site, Baraga and Marquette, had the

highest amount of snowfall (f = 297.8cm and f = 187.34cm, respectively), whereas the

 

more southern counties in this study site, Iron and Dickinson Counties, had the lowest

amount of snowfall (f = 94.69cm and f = 99.10cm, respectively) (Figure 2.11). The

 

snowfall in Iron and Dickinson Counties, the 2 counties with the lowest overall amount
of snowfall, was similar. However, the snowfall in Marquette and Baraga Counties, the 2
counties with the highest overall amount of snowfall, was not; the snowfall in Baraga
County exceeded the snowfall in Marquette County by over IIOcm (Table 2.3; Figure
2.11).

The overall corrected average WSI for the UP study site was 82.69 and the yearly
average WSI ranged between 55.33 in 1997 and 116.35 in 1995 (Table 2.4). For the
majority of years in this study, the corrected average WSI was below the long-term
regional average (Figure 2.12). Of the years with a below average WSI, most had a WSI
that was approximately 15 WSI points below the regional average. The exceptions were
in 1986 and 1997 in which the WSI was over 20 points below the long-term average for
the study site (Table 2.4; Figure 2.12). Of the years with an above average WSI, most
had a WSI that was within approximately 20 points of the long-term average WSI (Figure
2.12). In 1995 and 1996, however, the WSI exceeded the regional WSI by 24 (Table 2.4;

Figure 2.12).

150

 

 

Cumulative Total Monthly Snowfall

[:j Iro n (94.69cm)

ch kin s o n (99.10cm)
- Marque tte (187.34cm)
- Baraga (297.80cm)

 

Figure 2.11. Cumulative total monthly snowfall (summed from January through April) in
cm by county in the Upper Peninsula study site from 1987 through 2000.

151

 

 

Table 2.4. The mean corrected Winter Severity Index for each year between 1986 and
1999 in the Upper Peninsula study site.

 

 

 

Year Mean WSI
1986 59.85
1987 96.47
1988 100.08
1989 89.14
1990 71.57
1991 76.79
1992 77.68
1993 86.92
1994 68.62
1995 116.35
1996 104.45
1997 55.33
1998 73.48

1999 73.84

 

152

 

+Yearly Average " " ‘Long-Term Average
120.00

 

l 10.00
100.00
90.00

80.00 a

70.00

Corrected WSI

60.00

50.00

40.00

 

30.00

 

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Year

Figure 2.12. The temporal trend in the corrected WSI for the Upper Peninsula study site
between 1987 and 2000.

153

 

 

The results of the ANOVA model run to examine the relationship between male
yearling average beam diameter and each underlying mechanism that has the potential to
inﬂuence this index as a measure of herd quality indicated that in the UP study site both
the year (F13, 23 = 2.17; P = 0.05) and snowfall (F1, 23 = 4.85; P = 0.04) variables were
statistically signiﬁcant; there were no statistically signiﬁcant interactions. When this
ANOVA model was run by county in the UP study site it was clear that weather was
having an impact on male yearling average beam diameter. In each of Dickinson, Iron,
and Marquette Counties, both the temperature and WSI variables were statistically
signiﬁcant (Table 2.5). In addition, the snowfall variable was statistically signiﬁcant in
Iron County (Table 2.5).

The results of the ANOVA analysis to determine which underlying mechanisms
have the potential to inﬂuence yearling point count as a measure of deer herd quality in
the UP study site were statistically signiﬁcant and indicated that the year (F3 3, 23 = 3.29; P
= 0.006) and snowfall (F 1, 23 = 3.09; P-value = 0.09) variables were signiﬁcant; however,
there were no statistically signiﬁcant interactions. When the ANOVA model was run by
county, it was apparent that a mixture of weather and density was impacting male
yearling point count as an index of herd quality. In Dickinson and Iron Counties, deer
density was a statistically signiﬁcant variable and in Dickinson, Iron, and Marquette
Counties, the temperature variable was statistically signiﬁcant (Table 2.6). Snowfall and

WSI were also statistically signiﬁcant variables in Baraga County (Table 2.6).

NLP
Based on the sex-age-kill estimate method, the number of deer in the NLP study

site ranged between 19,897 in 1999 and 48,561 in 1991; while the long-term regional

154

 

 

Table 2.5. Results of the ANOVA model run by county in the Upper Peninsula study
site to determine which underlying mechanisms that have the potential to inﬂuence male
yearling average beam diameter as a measure of deer herd quality in Michigan were
signiﬁcant at an alpha level of 0.10.

 

 

Mechanism Baraga Dickinson Iron Marquette
Density X X X X
Temperature X 0.03 0.07 0.001
Snowfall X X 0.02 X
WSI X 0.08 0.02 0.01

 

155

 

Table 2.6. Results of the ANOVA model run by county in the Upper Peninsula study
site to determine which underlying mechanisms that have the potential to inﬂuence male

yearling point count as a measure of deer herd quality in Michigan were signiﬁcant at an
alpha level of0. 10.

 

 

Mechanism Baraga Dickinson Iron Marquette
Density X 0.04 0.03 X
Temperature X 0.07 0.03 0.03
Snowfall 0.07 X X X
WSI 0.03 X X X

 

156

 

average number of deer in the NLP study site was estimated to be 30,843 (Table 2.7).
The temporal trend in the average number of deer in the NLP study site illustrated that in
9 out of the 14 years of this study, the number of deer was at or below the long-term
average for the study site (Figure 2.13). For most of these years, the number of deer was
within 5,000 deer of the average, although, the number of deer never exceeded 10,000
deer below the average for the study site (Table 2.7; Figure 2.13). In the years that the
number of deer was above the long-term average for the NLP study site, it was within
5,000 deer for each of those years except in 1991 when the average was exceeded by
approximately 20,000 deer (Table 2.7; Figure 2.13).

The counties along the coast, Alcona, Alpena, and Presque Isle Counties, had the
highest deer numbers (J? = 39,278, E = 33,869, and 36 = 29,701, respectively), whereas
inland counties, Montmorency and Oscoda Counties, had lower deer numbers (2? =
26,539 and f = 25,034, respectively) (Figure 2.14). Alcona and Alpena Counties had the
highest number of deer, as well as the highest deer densities in the NLP study site at 22
deer/km2 (58 deer/miz) and 23 deer/km2 (59 deer/miz), respectively. Montmorency
County, however, had a lower number of deer, but a higher deer density at 19 deer/km2
(49 deer/miz). Oscoda County had the lowest number of deer and the lowest deer density
in the NLP study site at 17 deer/km2 (44 deer/miz); the deer density in Presque Isle
County was similar to that in Oscoda County at 17 deer/km2 (45 deer/miz). Although
were differences in deer densities among the counties in the NLP study site, the
difference between the county with the lowest deer density and the county with the

highest deer density was only approximately 6 deer/km2 (15 deer/miz).

157

 

 

Table 2.7 . The total number of deer, as estimated from the sex-age-kill estimate method,
in each county in the Northern Lower Peninsula study site between 1987 and 2000; the
yearly average total number of deer and the standard deviation for the entire study site,
and the average total number of deer for each county.

 

 

 

 

Alcona Alpena Montmorency Oscoda Presque Isle Yearly Average StDev
1987 29579 46888 21212 15628 36607 29983 12378
1988 38996 28581 27046 21752 27311 28737 6304
1989 53534 35584 30022 24348 29042 34506 11362
1990 52123 38918 31512 36102 20237 35779 11585
1991 61735 44072 49950 44422 42625 48561 7874
1992 29248 54854 25696 14383 55317 35900 18355
1993 43809 22593 30134 24368 32954 30772 8412
1994 26182 24432 rVa 27902 16151 23667 5207
1995 46763 41768 13768 35386 22981 32133 13588
1996 36841 31939 24683 20602 39610 30735 8013
1997 41754 34917 23665 26752 27318 30881 7354
1998 39274 27953 22959 23619 26284 28018 6608
1999 19994 20653 21136 15133 22572 19897 2827
2000 30055 21008 23226 20075 16803 22233 4945
lknal 39278 33869 26539 25034 29701 30843 5785

 

 

 

 

 

 

 

 

 

*n/a denotes that the data were unavailable

158

 

_‘

 

+Yearly Average ' ' 'Long-Term Average

 

65000
60000 *
55000
50000'
45000
40000
35000
30000

25000

Average Total Number of Deer

 

20000

 

15000

 

 

 

10000 , r r 4
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.13. The temporal trend in the average total number of deer, as estimated by the
sex-age-kill estimate method, for the Northern Lower Peninsula study site between 1987
and 2000.

159

 

 

 

Total Deer Numbers by County

E Oscoda (25,034)
Mmtmorency(26,539)
- Presque kle (29,701)
- Alpena (33,869)
- Alcona(39,278)

 

Figure 2.14. The average number of deer by county in the Northern Lower Peninsula
study site from 1987 through 2000.

160

 

The overall cumulative mean monthly temperature in the NLP study site was
208.72 °C and ranged between 169.88 °C in 1988 and 232.17 °C in 1998 (Table 2.8). In
the majority of years in this study, the yearly average cumulative mean monthly
temperature was at or below the long-term average for the NLP study site (Figure 2.15).
For most of these years, the average cumulative mean monthly temperature was within 5
°C of the long-term regional average, except in 1987 and 1988 when it exceeded the
study site average by 30 °C to 40 °C (Table 2.8; Figure 2.15). Of those years in which
the average cumulative mean monthly temperature was above the long-term regional
average, it was still within 25 °C (Table 2.8; Figure 2.15).

There was no distinct spatial pattern in the distribution of the cumulative mean
monthly temperature by county in the NLP study site (Figure 2.16). Oscoda and Presque
Isle Counties had the highest temperature readings (f = 216.50 °C and f = 223.75 °C,
respectively), whereas Montmorency and Alpena Counties had the lowest temperature
readings (f = 180.45 °C and 56 = 214.17 °C, respectively) (Figure 2.16). Even though
Alpena, Oscoda, and Presque Isle Counties had comparable cumulative mean monthly
temperatures, Alpena and Oscoda Counties had the most similar temperature readings of
the counties in the NLP study site (Table 2.8; Figure 2.16).

The cumulative total monthly snowfall in the NLP study site ranged between
68.26cm in 2000 and 236.98cm in 1997 while the overall cumulative total monthly
snowfall for the NLP study site was 110.67cm (Table 2.9). In almost every year, the
snowfall was within i 20cm of the long-term NLP study site average, except in 1987,

1999, and 2000 when the cumulative total monthly snowfall exceeded 25cm below the

161

Table 2.8. The cumulative mean monthly temperature (summed from January through
September) in °C for each county in the Northern Lower Peninsula study site from 1987
to 2000; the yearly average cumulative mean monthly temperature and the standard
deviation (StDev) for the entire study site, and the average cumulative mean monthly

temperature for each county.

 

 

 

 

 

 

 

 

 

 

 

Alcona Alpena Montmorency Oscoda Presque Isle Yearly Average StDev
1987 n/a 226.44 44.50 231.67 237.33 184.98 93.76
1988 n/a 216.50 18.72 218.78 225.50 169.88 100.84
1989 n/a 209.89 187.61 210.39 220.72 207.15 13.95
1990 n/a 219.11 200.06 222.28 230.50 217.99 12.88
1991 n/a 224.94 220.33 226.39 233.50 226.29 5.46
1992 n/a 205.11 203.67 194.06 214.33 204.29 8.30
1993 n/a 206.78 202.28 209.89 216.78 208.93 6.09
1994 n/a 204.00 202.39 204.94 208.61 204.99 2.64
1995 n/a 216.50 183.22 218.94 222.67 210.33 18.25
1996 n/a 201.94 194.72 204.33 210.28 202.82 6.44
1997 n/a 204.1 1 201.83 207.61 212.83 206.60 4.79
1998 n/a 227.72 227.44 235.78 237.72 232.17 5.35
1999 n/a 218.89 224.22 223.94 230.94 224.50 4.95
2000 n/a 216.44 215.33 221.94 230.78 221.12 7.05
Total n/a 214. 17 180.45 216.50 223.75 208.72 19.28

 

*n/a denotes that the data were not available

162

 

+Yearly Average ' ' 'Long-Term Average
270.00' 777* 77.7-. ...-,.,

250.00 '

230.00

210.00

190.00 -

Average Cumulative Temperature in °C

 

170.00 '

150.00 ' - T T . 2 f

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 I999 2000

Year

Figure 2.15. The temporal trend in the average cumulative temperature (summed from

January through September) in °C for the Northern Lower Peninsula study site between
1987 and 2000.

163

 

 

Cumulative Mean Monthly Temperature

C] Alcona (n/a)

Mon tm ore ncy (180.45 °C)

.5: j; Alp e n a (214.17°C)

- Oscoda (216.50 °C)
Presque Isle (223.75 °C)

 

Figure 2.16. Cumulative mean monthly temperature (summed from January through
September) in °C by county in the Northern Lower Peninsula study site from 1987
through 2000.

 

 

Table 2.9. The cumulative total monthly snowfall (summed from January through April)
in cm for each county in the Northern Lower Peninsula study site from 1987 to 2000; the
yearly average cumulative total monthly snowfall and the standard deviation (StDev) for
the entire study site, and the average cumulative total monthly snowfall for each county.

 

 

 

 

 

 

 

 

 

 

 

 

 

Alcona Alpena Montmorency Oscoda Presque Isle Yearly Average_ StDev
1987 n/a 103.89 35.56 68.58 119.63 81.92 37.56
1988 n/a 160.02 49.53 63.50 207.77 120.21 76.30
1989 n/a 171.45 48.26 91.44 207.77 129.73 72.88
1990 n/a 138.43 139.70 5.08 185.67 117.22 77.92
1991 n/a 96.77 90.17 46.48 158.75 98.04 46.21
1992 n/a 141.73 161.29 2.54 162.56 117.03 76.92
1993 n/a 142.24 148.59 0.00 85.09 93.98 68.85
1994 n/a 199.39 153.67 0.00 115.57 117.16 85.29
1995 n/a 122.43 84.58 5.08 164.34 94.11 67.70
1996 n/a 134.87 163.07 1.78 76.71 94.11 71.29
1997 n/a n/a 540.24 0.00 170.69 236.98 276.16
1998 n/a n/a 173.99 0.00 152.40 108.80 94.84
1999 n/a n/a 133.10 0.00 82.30 71.80 67.17
2000 n/a 93.98 105.41 0.00 73.66 68.26 47.36
Total n/a 136.84 144.80 20.32 140.21 1 10.67 60.24

 

 

*n/a denotes that the data were not available

165

long-term average and in 1997 when the cumulative total monthly snowfall exceeded
120cm above the long-term average (Table 2.9; Figure 2.17).

The northern counties in the NLP study site had a higher amount of snowfall
(Figure 2.18). Alpena, Presque Isle, and Montmorency Counties had the highest
cumulative total monthly snowfall (56 = 136.84cm, J? = 140.210m, and f = 144.80cm,
respectively), whereas Oscoda County had the lowest cumulative total monthly snowfall
(f = 20.32cm) (Table 2.9; Figure 2.18). In addition, the difference among the snowfall in
Alpena, Presque Isle, and Montmorency Counties was very slight; the amount of snowfall
among the 3 counties was within 8cm (Table 2.9; Figure 2.18). Conversely, the amount
of snowfall in Oscoda County was very different than the snowfall in each of the other 3
counties in the NLP study site for which temperature data were available (Table 2.9;
Figure 2.18).

The overall corrected average WSI for the NLP study site was 56.85 and ranged
between 35.20 in 1997 and 78.41 in 1993 (Table 2.10). For the majority of years in this
study, the yearly average corrected WSI was below the NLP study site average (Figure
2.19). In the years that the corrected average WSI was below the long-term regional
average, most had a corrected average WSI that was approximately 5 to 10 points below
the long-term average; however, in 1986 and 1997 the corrected average WSI was over
20 points below the long-term regional average (Table 2.10; Figure 2.19). Conversely, in
the years in which the corrected average WSI was above the long-term average, most had
a corrected average WSI that was approximately 15 points above the long-term average,
except in 1993 when the corrected average WSI exceeded 20 point above the study site

average (Table 2.10; Figure 2.19).

166

 

 

290.00

270.00 ,

250.00
230.00
210.00
190.00
170.00
150.00
130.00

Average Cumulative Snowfall (cm)

90.00

70.00 e

50.00

+Yearly Average " " ‘Long-Term Average

 

110.00 *

 

 

 

f f r l ' 7

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.17. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Northern Lower Peninsula study site between 1987

and 2000.

167

 

 

 

Cumulative Total Monthly Snowfall

[j Ale 0 n in (n/a)

m Os c 0 da (20.32cm)

Alp e n a (136.84cm)
Pre s q ue Isle (140.21cm)
Mo n tm 0 re n c y (144.80cm)

Figure 2.18. Cumulative total monthly snowfall (summed from January through April) in
cm by county in the Northern Lower Peninsula study site from 1987 through 2000.

 

 

168

Table 2.10. The mean corrected Winter Severity Index for each year between 1986 and
1999 in the Northern Lower Peninsula study site.

 

Year Mean WSI

 

 

1986 39.20
1987 49.18
1988 63.51
1989 60.45
1990 54.99
1991 53.03
1992 55.28
1993 78.41
1994 51.94
1995 69.52
1996 67.21
1997 35.20
1998 51.19
1999 42.73

 

169

+ Yearly Average " ' 'Long-Term Average
120.00

 

l 10.00

100.00

90.00

80.00

70.00

Corrected WSI

60.00

 

50.00

40.00

 

 

 

30.00 2
1986 1987 1988 1989 1990 1991 1992 1993 l994 1995 1996 1997 I998 I999

 

Year

Figure 2.19. The temporal trend in the corrected WSI for the Northern Lower Peninsula
study site between 1987 and 2000.

170

 

 

According to Felix (unpublished data) the habitat potential for the fall and winter
food requirement was similar among the 5 counties in the NLP study site (Table 2.11).
While there was a small range in the fall and winter food potential for the counties in the
NLP study site, in general, the southem-most counties, Oscoda and Alcona Counties, and
the westem-most county, Montmorency, had the highest fall and winter food potential,
whereas Alpena and Presque Isle Counties had the lowest (Table 2.11). The habitat
potential for the spring and summer food requirement was also similar among the 5
counties in the NLP study site (Table 2.11). There, however, did not appear to be a
distinct spatial pattern to the distribution of this habitat potential. Alcona County had the
highest spring and summer food potential, followed by Presque Isle, Montmorency,
Alpena, and Oscoda Counties, respectively (Table 2.11). There was a greater range in the
thermal cover potential among the 5 counties in the NLP study site and, for the most part,
the counties along the coast had a higher thermal cover potential (Table 2.11). Alpena
and Presque Isle Counties had the highest thermal cover potential and Oscoda County
had the lowest, while the thermal cover potential in Alcona and Montmorency Counties
fell in between (Table 2.11).

The results from the ANOVA model run to examine the relationship between
male yearling average beam diameter and each underlying mechanism that has the
potential to inﬂuence this index as a measure of herd quality in the NLP study site
showed that year (F 1 3, 34 = 2.90; P = 0.006) and county (F4, 24 = 6.22; P = 0.0007) were
statistically signiﬁcant; there were no statistically signiﬁcant interactions. When the
ANOVA model was run by county, it showed that both density and weather were

inﬂuencing male yearling average beam diameter as a measure of herd quality. In

171

 

 

Table 2.11. The potential of the fall and winter food, spring and summer food, and
thermal cover to provide suitable habitat for white-tailed deer in each county in the
Northern Lower Peninsula study site (0 is the lowest suitability and 1 is the highest

suitability) according to Felix (unpublished data).

 

 

F&W S&S Therm
Alcona 0.659 0.628 0.182
Alpena 0.596 0.536 0.551
Montmorency 0.629 0.558 0.17
Oscoda 0.676 0.512 0.033
Presque Isle 0.581 0.566 0.39

 

172

Alpena and Presque Isle Counties, both the temperature and WSI variables were
statistically signiﬁcant, and in Oscoda County the density variable was statistically
signiﬁcant (Table 2.12).

The results from the regression analyses run with the habitat potential value for
each of the 3 habitat requirements illustrated that the habitat potential for fall and winter
food, spring and summer food, and thermal cover had different affects on male yearling
average beam diameter. The fall and winter food potential was signiﬁcantly negatively
correlated with male yearling average beam diameter, whereas the thermal cover
potential was signiﬁcantly positively correlated with male yearling average beam
diameter (Table 2.13). There was, however, no relationship between spring and summer
food potential and male yearling average beam diameter (Table 2.13).

The results of the ANOVA analysis to determine the relationship between each
underlying mechanism that has the potential to inﬂuence male yearling point count and
this index as a measure of herd quality in the NLP study site showed that both the year
(F1134 = 1.84; P = 0.08) and county (F4, 34 = 6.51; P = 0.0005) variables were statistically
signiﬁcant; there were no statistically signiﬁcant interactions. When the ANOVA model
was run by county, it was apparent that density and weather were inﬂuencing male
yearling point count as a measure of herd quality. In Montmorency and Presque Isle
Counties, the density variable was statistically signiﬁcant, and in Alpena County, both
the temperature and WSI variable were statistically signiﬁcant (Table 2.14).

The results from the regression analyses run with the habitat potential value for
each of the 3 habitat requirements illustrated that the habitat potential for fall and winter

food, spring and summer food, and thermal cover had different affects on male yearling

173

 

Table 2.12. Results of the ANOVA model run by county in the Northern Lower
Peninsula study site to determine which underlying mechanisms that have the potential to
inﬂuence male yearling average beam diameter as a measure of deer herd quality in
Michigan were signiﬁcant at an alpha level of 0. 10.

 

 

Mechanism Alcona Alpena Montmorency Oscoda Presque Isle
Density X X X 0.04 X
Temperature X 0. l X X 0.03
Snowfall X X X X X
WSI X 0.04 X X 0.08

 

174

 

Table 2.13. Results of the regression analyses conducted to determine the relationship
between each habitat requirement and male yearling average beam diameter in the
Northern Lower Peninsula study site at an alpha level of 0. 10.

 

 

Habitat Potential [10 B, P-Value R2

Fall & Winter 24.49 -11.48 0.05 0.79
Spring & Summer 16.17 1.98 0.79 0.03
Thermal Cover 16.62 2.51 0.003 0.96

 

175

 

 

 

Table 2.14. Results of the ANOVA model run by county in the Northern Lower
Peninsula study site to determine which underlying mechanisms that have the potential to
inﬂuence male yearling point count as a measure of deer herd quality in Michigan were
signiﬁcant at an alpha level of0. 10.

 

 

Mechanism Alcona Alpena Montmorency Oscoda Presque Isle
Density X X 0.04 X 0.03
Temperature X 0.09 X X X
Snowfall X X X X X

WSI X 0.04 X X X

 

 

176

 

3';

point count. The fall and winter food potential as well as the spring and summer food
potential were not correlated with male yearling point count, whereas the thermal cover
potential was signiﬁcantly positively correlated with male yearling point count (Table

2.15).

SLP
Based on the sex-age-kill estimate method, the overall number of deer for the SLP

study site was 25,745 and the yearly average number of deer for this study site ranged
between a minimum of 16,080 in 1990 and a maximum of 41,424 in 1998 (Table 2.16).
The temporal trend in the yearly average number of deer in the SLP study site showed
that, in general, the average number of deer for each year between 1987 and 1997 was at
or below the average number of deer for the study site, except 1988 in which the average
number of deer was slightly above the long-term study site average (Table 2.16; Figure
2.20). In 1998 there was a sharp increase in the average number of deer for the SLP
study site and, although, there was a slight decline in the average number of deer for 1999
and 2000, the density of deer was still at least 10,000 deer above the long-term average
(Table 2.16; Figure 2.20).

Calhoun County, the southem-most county in the study site, and Barry County,
the westem-most county in the study site, had the highest overall numbers of deer (f =
27,970 and 55 = 25,392, respectively) (Figure 2.21). Eaton County, the eastem-most
county in the study site, had the lowest overall number of deer of the 3 counties in the
SLP study site (i = 23,978) (Figure 2.21). Calhoun County, which had the highest
number of deer in the SLP study site, had the lowest density of deer at 15 deer/km2 (40

deer/miz). Eaton County, however, had the smallest number of deer in the SLP study

177

Table 2.15. Results of the regression analyses conducted to determine the relationship
between each habitat requirement and male yearling point count in the Northern Lower
Peninsula study site at an alpha level of 0. 10.

 

 

Habitat Potential Po [3, P-Value R2

Fall & Winter 5.29 -3.41 0.17 0.52
Spring & Summer 1.99 2.07 0.42 0.22
Thermal Cover 2.94 0.80 0.06 0.74

 

 

178

 

Table 2.16. The total number of deer, as estimated from the sex-age-kill estimate
method, in each county in the Southern Lower Peninsula study site between 1987 and
2000; the yearly average total number of deer and the standard deviation for the entire
study site, and the average total number of deer for each county.

 

 

 

 

 

 

 

 

 

 

Barry Calhoun Eaton Avera e StDev
1987 28592 18563 16876 21344 6334
1988 24997 21453 34839 27096 6936
1989 29803 20484 14325 21538 7792
1990 17948 20030 10262 16080 5145
1991 15693 27449 18210 20451 6190
1992 22554 22433 rva 22494 86
1993 14256 38015 13693 21988 13883
1994 18894 22848 20462 20735 1991
1995 33700 24907 16431 25013 8635
1996 23033 26237 24110 24460 1631
1997 22362 31134 24618 26038 4555
1998 32862 35260 56150 41424 12809
1999 37461 41145 29921 36176 5722
2000 33335 41617 31815 35589 5276
Tknal 25392 27970 23978 25745 2024

 

179

*n/a denotes that the data were unavailable

 

 

+Yearly Average " " 'Long-Term Average

 

 

65000
60000
55000 *
50000 *
45000

40000 -
35000 4

30000 4

Average Total Number of Deer

25000

20000 r

 

15000

 

 

 

10000 . . , ,
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.20. The temporal trend in the average total number of deer, as estimated by the
sex-age-kill estimate method, for the Southern Lower Peninsula study site between 1987
and 2000.

180

 

Total Deer Numbers by County
E] Ea to n (23,978)

m Barry (25,392)
- Calhoun (27,970)

 

Figure 2.21. The average total number of deer by county in the Southern Lower
Peninsula study site from 1987 through 2000.

181

site, but had a slightly higher deer density than Calhoun County at 16 deer/km2 (42
deer/miz). Barry County had the highest deer density of the 3 counties in this study site at
18 deer/km2 (46 deer/miz). While there were differences in deer densities among the 3
counties in the SLP study site, the difference, however, was only approximately 3
deer/km2 (6 deer/miz).

The cumulative mean monthly temperature in the SLP study site ranged between
a minimum of 204.26 °C in 1999 and a maximum of 252.43 °C in 1998, with an overall
cumulative mean monthly temperature of 23 l .54 °C (Table 2.17). In general, the
temporal trend in the yearly average cumulative mean monthly temperature oscillated
around the long-term average cumulative mean monthly temperature for the study site
(Figure 2.22). In 9 out of the 14 years of this study, the average cumulative mean
monthly temperature was below the long-term study site average, but was within 5 °C,
with the exception of 1999 in which the average cumulative mean monthly temperature
was 25 °C below the long-term average (Table 2.17; Figure 2.22). For those years in
which the average cumulative mean monthly temperature was above the long-term
average, it was within approximately 20 °C (Table 2.17; Figure 2.22).

The westem-most county, Barry County, and the southem-most county, Calhoun
County, had the highest temperature readings (f = 235.69 °C and f = 232.66 °C,
respectively) (Figure 2.23). Eaton County, the eastem-most county in the SLP study site,
had the lowest cumulative mean monthly temperature of the 3 counties in the study site
(i = 226.26 °C) (Figure 2.23).

The overall cumulative total monthly snowfall for the SLP study site was

81.17cm; the minimum was 52.83cm in 1991 and the maximum wa5128.10cm in 1999

182

 

 

 

Table 2.17. The cumulative mean monthly temperature (summed from January through
September in °C for each county in the Southern Lower Peninsula study site from 1987 to
2000; the yearly average cumulative mean monthly temperature and the standard
deviation (StDev) for the entire study site, and the average cumulative mean monthly
temperature for each county.

 

 

 

 

 

Barry Calhoun Eaton Average StDev
1987 249.89 255.00 241.61 248.83 6.76
1988 238.89 239.83 200.94 226.56 22.18
1989 228.44 232.78 225.28 228.83 3.77
1990 238.50 242.61 235.56 238.89 3.54
1991 244.67 245.1 1 240.94 243.57 2.29
1992 227.22 231.06 224.06 227.44 3.51
1993 227.22 228.50 223.11 226.28 2.82
1994 224.22 232.11 220.56 225.63 5.90
1995 235.67 243.22 231.44 236.78 5.97
1996 224.44 231.67 220.61 225.57 5.61
1997 226.44 230.00 222.67 226.37 3.67
1998 251.83 256.89 248.56 252.43 4.20
1999 241.11 175.56 196.11 204.26 33.53
2000 241.17 212.89 236.22 230.09 15.10
Total 235.69 232.66 226.26 231.54 4.82

 

 

 

 

 

 

183

 

+Yearly Average ' " 'Long-Term Average
270.00

 

250.00

230.00

210.00

 

190.00

Average Cumulative Temperature in °C

 

170.00

 

 

150.00 T -
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Year

Figure 2.22. The temporal trend in the average cumulative temperature (summed from

January through September) in °C for the Southern Lower Peninsula study site between
1987 and 2000.

184

 

 

    
    
  

   
   

ll " -
III'I‘
ESE

 

‘I

   
 

.!
I}

If
I}:
I
I
.ll

  

l

l!
=Il
.-
ll:

     
 

 
 

I
I
E
ll“

   

 

Cumulative Mean Monthly Temperature
C Ea to n (226.26 °C)

- Calhoun (232.66 °C)

- Barry (235.69 °q

 

Figure 2.23. Cumulative mean monthly temperature (summed from January through
September) in °C by county in the Southern Lower Peninsula study site from 1987

through 2000.

185

 

(Table 2.18). From 1987 through 1992, the yearly average cumulative total monthly
snowfall was approximately 5cm to 20cm below the long-term average for the study site
(Figure 2.24). After that time period, the temporal trend in the yearly average cumulative
total monthly snowfall more or less oscillated around the long-term average (Figure
2.24). There were peaks in snowfall in 1993, 1997, and 1999, some of which were more
than 50cm above the study site average, whereas the lows in the yearly average
cumulative total monthly snowfall never exceeded a 30cm difference (Table 2.18; Figure
2.24).

The western and southem-most counties, Barry and Calhoun Counties, had the
highest overall amount of snowfall (f = 99.7lcm and f = 77.34cm, respectively) (Figure
2.25). Conversely, the eastem-most county in the SLP study site, Eaton County, had the
lowest overall cumulative total monthly snowfall (f = 66.44cm) (Figure 2.25). The
difference in the cumulative total monthly snowfall between the county with the largest
amount of snowfall and the county with the least amount of snowfall was approximately
33cm. However, the overall cumulative total monthly snowfall in Eaton and Calhoun
Counties was within approximately 11cm (Table 2.18; Figure 2.25).

The corrected average WSI for the SLP study site ranged between a minimum of
33.05 in 1997 and a maximum of 69.49 in 1993, with an the overall corrected average
WSI for this study site of 47.03 (Table 2.19). The temporal trend in the corrected
average WSI showed that, in general, the yearly corrected average WSI was within i 10
points of the long-term study site average, except in 1993, 1995 and 1997 (Figure 2.26).
In those years in which the yearly corrected average WSI exceeded 10 points of the long-

tenn, it was within i 25 points (Table 2.19; Figure 2.26).

186

 

 

Table 2.18. The cumulative total monthly snowfall (summed from January through
April) in cm for each county in the Southern Lower Peninsula study site from 1987 to
2000; the yearly average cumulative total monthly snowfall and the standard deviation
(StDev) for the entire study site, and the average cumulative total monthly snowfall for
each county.

 

 

 

 

Barry Calhoun Eaton Yearly Average StDev
1987 33.53 75.44 54.61 54.53 20.96
1988 73.91 74.17 73.66 73.91 0.25
1989 88.65 78.99 57.15 74.93 16.14
1990 93.47 55.88 74.17 74.51 18.80
1991 71.63 49.28 37.59 52.83 17.29
1992 85.09 68.33 53.34 68.92 15.88
1993 119.63 120.14 82.55 107.44 21.56
1994 121.67 97.79 61.47 93.64 30.31
1995 123.19 64.26 59.69 82.38 35.42
1996 84.33 62.48 53.34 66.72 15.92
1997 179.32 97.79 94.49 123.87 48.06
1998 78.23 38.61 50.80 55.88 20.29
1999 155.19 107.70 121.41 128.10 24.45
2000 88.14 91.95 55.88 78.66 19.82
Total 99.71 77.34 66.44 81.17 16.96

 

 

 

 

 

 

187

 

+ Yearly Average " " 'Long-Term Average

 

 

290.00
270.00
250.00
230.00
210.00 ~«
190.00 .
170.00

150.00 . __5
130.00 . 1
110.00 +
90.00
70.00
50.00

Average Cumulative Snowfall (cm)

 

 

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.24. The temporal trend in the average cumulative snowfall (summed from
January through April) in cm for the Southern Lower Peninsula study site between 1987
and 2000.

188

 

Cumulative Total Monthly Snowfall
E Ea to n (66.44cm)

- Calhoun (77.34cm)
- Barry (99.7lcm)

 

Figure 2.25. Cumulative total monthly snowfall (summed from January through April) in
cm by county in the Southern Lower Peninsula study site from 1987 through 2000.

189

Table 2.19. The mean corrected Winter Severity Index for each year between 1986 and
1999 in the Southern Lower Peninsula study site.

 

Year Mean WSI

 

1986 n/a

1987 n/a

1988 42.16
1989 49.17
1990 41.68
1991 41.37
1992 49.91
1993 69.49
1994 45.88
1995 57.71
1996 n/a

1997 33.05
1998 47.41
1999 45.18

 

*n/a denotes that the data were not available

190

80 00 + Yearly Average " ' 'Long-Teml Average

 

70.00

60.00

50.00 *

 

40.00

Corrected WSI

30.00

20.00

 

10.00 -

 

0.00

 

 

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Year

Figure 2.26. The temporal trend in the corrected WSI for the Southern Lower Peninsula
study site between 1987 and 2000.

191

 

The results of the ANOVA analysis to examine the relationship between the
underlying mechanisms that have the potential to inﬂuence male yearling average beam
diameter and this index as a measure of herd quality in the SLP study site revealed that
the county (F2, 22 = 1.75; P = 0.06) and density (F 1, 22 = 4.05; P = 0.06) variables were
statistically signiﬁcant; there were no signiﬁcant interactions. When the ANOVA model
was run by county, it was apparent that there was a mixture of factors inﬂuencing male
yearling average beam diameter; in each county a different factor was statistically
signiﬁcant. In Barry County, the snowfall variable was statistically signiﬁcant, whereas
in Calhoun County the WSI variable was statistically signiﬁcant (Table 2.20). The
density variable was statistically signiﬁcant in Baton County (Table 2.20).

The results of the ANOVA analysis to determine the relationship between each
underlying mechanism that has the potential to inﬂuence male yearling point count and
this index as a measure of herd quality in the SLP study site showed that the year (F 13, 22
= 3.19; P = 0.008), county (F2, 22 = 4.46; P = 0.02), and temperature (F1, 22 = 3.31; P =
0.08) variables were statistically signiﬁcant; there were no statistically signiﬁcant
interactions. When the ANOVA model was run by county, however, no variables were

detected as being statistically signiﬁcant (Table 2.21).

Comparisons Among Regions
Deer Numbers

In each year from 1987 through 1996, the number of deer in the UP study site was
consistently the highest out the 3 regional study sites, while the number of deer in the

SLP study site was consistently the lowest (Figure 2.27). Furthermore, during the time

192

Table 2.20. Results of the ANOVA model run by county in the Southern Lower
Peninsula study site to determine which underlying mechanisms that have the potential to
inﬂuence male yearling average beam diameter as a measure of deer herd quality in
Michigan were signiﬁcant at an alpha level of 0.10.

 

 

Mechanism Barry Calhoun Eaton
Density X X 0.06
Temperature X X X
Snowfall 0.05 X X
WSI X 0.08 X

 

193

 

Table 2.21. Results of the AN OVA model run by county in the Southern Lower
Peninsula study site to determine which underlying mechanisms that have the potential to
inﬂuence male yearling point count as a measure of deer herd quality in Michigan were
signiﬁcant at an alpha level of 0.10.

 

 

Mechanism Barry Calhoun Eaton
Density X X X
Temperature X X X
Snowfall X X X
WSI X X X

 

194

e UP +NLP +SLP

 

65000
60000
55000 . ‘3 . a.
50000 « ”L
45000.
40000 1
35000 «
30000

25000 .

Average Total Number of Deer

20000 +

 

15000

 

l 0000

 

 

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.27 . The comparison of yearly trends in the average total number of deer, as
estimated by the sex-age-kill estimate method, for each of the 3 regional study sites
between 1987 and 2000.

195

period from 1987 through 1996, the deer density among the 3 regional study sites
generally appeared to follow similar patterns, especially in the UP and NLP (Figure
2.27). Speciﬁcally, if the number of deer in 1 study site increased or decreased, in
general, so did the number of deer in at least 1 of the other study sites. In 1997, however,
there was a sharp decline in the number of deer in the UP study site and in 1998 there
was a sharp increase in the number of deer in the SLP study site (Figure 2.27). After
1997, deer numbers in the UP study site increased, but were still lower than in the
previous years, whereas deer numbers in the SLP study site decreased slightly in 1999
and 2000, but were still higher than in the previous years (Figure 2.27). Unlike the trends
in deer numbers in the UP and SLP study sites, deer numbers in the NLP study site did
not undergo any sharp increases or decreases; instead, they steadily declined throughout
the 14 years of the study (Figure 2.27).

Although deer numbers followed a north to south regional gradient in Michigan,
the actual density of deer did not follow the same trend. The density of deer per regional
study site was lowest in the UP study site at 13 deer/km2 (34 deer/mil) and highest in the
NLP study site at 20 deer/km2 (51 deer/miz), while density of deer in the SLP study site
fell in between the deer density for the other 2 study sites at 16 deer/km2 (42 deer/miz)
(Figure 2.28). Based on the Tukey-Kramer multiple comparison test, the density of deer
in the UP and NLP study sites was statistically different, whereas the deer density in the
UP and SLP study sites was not statistically different (Table 2.22; Figure 2.28). In
addition, the deer density in the NLP study site and the SLP study site was not

statistically different (Table 2.22).

196

n
I ‘-
. ’ " .9

l3 deer/km2
(34 deer/miz)

 

 

 

 

 

 

20 deer/1cm2
(51 deer/miz)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

16 deer/km2
(42 deer/mi’)

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.28. The regional comparison of the density of deer present in each of the 3
regional study sites averaged across years from 1987 to 2000.

197

Table 2.22. The mean comparison of (a) deer population density, (b) temperature, (c)
snowfall, and (d) the winter severity index among the 3 regional study sites at an alpha

 

 

 

 

 

 

 

 

 

 

 

 

 

 

level of 0.10.

(a) Density

UP NLP SLP
UP X 0.04 0.54
NLP 0.04 X 0.26
SLP 0.54 0.26 X
(b) Temperature

UP NLP SLP
UP X 0.15 <0.0001
NLP 0.15 X <0.0001
SLP <0.0001 <0.0001 X

(c) Snowfall

UP NLP SLP
UP X <0.0001 <0.0001
NLP <0.0001 X 0.22
SLP <0.0001 0.22 X

((1) WSI

UP NLP SLP
UP X <0.0001 <0.0001
NLP <0.0001 X 0.36
SLP <0.0001 0.36 X

 

 

198

Temperature

The trend in the yearly average cumulative mean monthly temperature among the
3 regional study sites was basically what was expected. In almost every year of this
study, the UP study site had the overall coolest temperature readings and the SLP study
site had the overall warmest temperature readings, with the temperature readings for the
NLP study site falling in between the temperature readings for the other 2 study sites
(Figure 2.29). There were, however, a few exceptions such as in 1987 and 1988 when the
average cumulative mean monthly temperature in the NLP study site was warmer than it
was in the UP study site, and in 1999 when there was a sharp decrease in the average
cumulative mean monthly temperature in the SLP study site so that it was warmer in both
the UP and NLP study sites (Figure 2.29). Overall, though, the yearly average
cumulative mean monthly temperature among the 3 regional study sites appeared to
follow similar patterns. For example, if the yearly average cumulative mean monthly
temperature in 1 study site increased or decreased, in general, so did the temperature in at
least 1 of the other study sites (Figure 2.29).

In general, as you move from north to south in Michigan, the temperature
increased. The overall temperature reading for the UP study site was the coolest of the 3
regional study sites at 199 °C and the overall temperature reading for the SLP study site
was the warmest at 232 °C while the temperature reading for the NLP study site was in
the middle of the temperature readings for the other 2 regional study sites at 209 °C
(Figure 2.30). The cumulative mean monthly temperature for the UP study site and the
NLP study site was more similar to each other than the temperature for either of these

study sites was to the cumulative mean monthly temperature in the SLP. Based on the

199

UP “P NLP +SLP
270.00

 

250.00 .

230.00

210.00

190.00

 

Average Cumulative Temperature in °C

170.00

 

150.00 . 4
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

 

Year

Figure 2.29. The comparison of yearly trends in the average cumulative temperature

(summed from January through September) for each of the 3 regional study sites between
1987 and 2000.

200

 

 

 

 

 

209 °C

 

 

 

 

 

 

 

 

232 °C

 

 

 

Figure 2.30. The regional comparison of the cumulative temperature (summed from
January through September) in °C averaged across years from 1987 to 2000 among each
of the 3 regional study sites.

201

Tukey-Kramer multiple comparison test, the cumulative mean monthly temperature was
statistically different between the UP and SLP study sites as well as between the NLP and
SLP study sites (Table 2.22). There was, however, no statistical difference between the

cumulative mean monthly temperature for the UP and NLP study sites (Table 2.22).

Snowfall

The trend in the yearly average cumulative total monthly snowfall among the 3
regional study sites was also basically what was expected. Of the 3 regional study sites,
the average cumulative total monthly snowfall for each year was highest in the UP study
site and lowest in the SLP study site while the snowfall each year in the NLP study site
fell in middle of the other 2 regional study sites (Figure 2.31). Also, the yearly average
cumulative total monthly snowfall among the 3 regional study sites appeared to follow
similar patterns. For instance, if the yearly average cumulative total monthly snowfall in
1 study site increased or decreased, in general, so did the snowfall in at least 1 of the
other study sites (Figure 2.31). There were, however, exceptions in 1993, 1999, and 2000
when the average cumulative total monthly snowfall in the SLP study site exceeded that
in the NLP study site (Figure 2.31).

The trend in cumulative total monthly snowfall decreased as you move along a
north to south regional gradient in Michigan. The UP study site had the highest amount
of cumulative total monthly snowfall at 170cm and the SLP study site had the least at
810m. At 111cm, the cumulative total monthly snowfall for the NLP study site was in
between the amount of snowfall for the other 2 regional study sites (Figure 2.32). The

cumulative total monthly snowfall was more similar between the NLP and SLP study

202

UP —I—~NLP +SLP
290.00
270.00
250.00 -;
230.00 4
210.00 x
190.00
170.00
150.00 2
130.00 « 2,, at:
110.00 3.

Average Cumulative Snowfall (cm)

9000
70.00

 

50.00 ‘ _"T A” T r v 7 ~ 7 . ﬁ—i" - _ , r 7 ,
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Figure 2.31. The comparison of yearly trends in the average cumulative snowfall
(summed from January through April) for each of the 3 regional study sites between 1987
and 2000.

203

 

 

 

 

Figure 2.32. The regional comparison of the cumulative snowfall (summed from

 

 

 

 

111cm

 

 

 

81cm

 

January through April) in cm averaged across years from 1987 to 2000 among the 3

regional study sites.

204

 

 

 

sites than the amount of snowfall in either of those 2 study sites was to the amount of
snowfall in the UP study site (Figure 2.32). According to the Tukey-Kramer multiple
comparison test, the cumulative total monthly snowfall was statistically different between
the UP and NLP study sites as well as between the UP and SLP study sites (Table 2.22).
There was, however, no statistical difference in the cumulative total monthly snowfall

between the NLP and SLP study sites (Table 2.22).

Winter Severity Index

The trend in the corrected WSI among the 3 regional study sites was exactly what
was expected. In each year of the study, the UP study site had the highest corrected
average WSI, whereas the SLP study site, overall, had the lowest (Figure 2.33). The
corrected average WSI for the NLP study site fell in between the corrected average WSI
for the other 2 regional study sites (Figure 2.33). The only exception was in 1999 when
the corrected average WSI in the SLP exceeded the corrected average WSI in the NLP,
but only slightly (Figure 2.33). Therefore, the yearly average corrected average WSI
among the 3 regional study sites tended to follow similar patterns. For instance, if the
yearly corrected average WSI in 1 study site increased or decreased, in general, so did the
snowfall in at least 1 of the other study sites (Figure 2.33).

The trend in corrected average WSI decreased as you move along a north to south
regional gradient in Michigan. The UP study site had the highest corrected average WSI
at 82.69 and the SLP study site had the least at 47.03. At 56.85, the corrected average
WSI for the NLP study site was in between the corrected WSI for the other 2 regional

study sites (Figure 2.34). The corrected average WSI was more similar between the NLP

205

0 UP +NLP +SLP
120.00

 

a»
"it.

1 10.00

100.00 '

a

90.00

/ a3!

80.00

70.00

Corrected WSI

60.00

50.00

40.00

 

 

 

 

 

30.00 . . v . 1 w . - w
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Year

Figure 2.33. The comparison of yearly trends in the corrected WSI for each of the 3
regional study sites between 1987 and 2000.

206

 

 

 

56.85

 

 

 

 

 

 

 

 

47.03

 

 

 

 

Figure 2.34. The regional comparison of the corrected WSI averaged across years from
1987 to 2000 among the 3 regional study sites.

207

and SLP study sites than the corrected average WSI in either of those 2 study sites was to
the corrected average WSI in the UP study site (Figure 2.34). According to the Tukey-
Kramer multiple comparison test, the corrected average WSI was statistically different
between the UP and NLP study sites as well as between the UP and SLP study sites
(Table 2.22). There was, however, no statistical difference in the corrected average WSI

between the NLP and SLP study sites (Table 2.22).

208

DISCUSSION

As you move along a north to south regional gradient in Michigan, some of the
underlying mechanisms that have the potential to inﬂuence white-tailed deer herd quality
increased while others decreased. The mean monthly temperature was highest in the
southem-most region of the state, and of the 3 regional study sites, the SLP study site was
the only site that showed a consistent increase in the number of deer over the course of
this study. Conversely, the total monthly snowfall and the winter severity index were
lowest in the southem-most study site, and in the UP and NLP study sites, the number of
deer, in general, decreased over the course of this study.

Clearly there were temporal and spatial differences in each of the underlying
mechanisms, but which of these factors can account for the trends that exist in both male
yearling average beam diameter and point count as measures of herd quality in
Michigan? For male yearling average beam diameter, the results suggest that in each
regional study site, multiple factors were having an inﬂuence. In the UP study site,
temperature, snowfall, and winter severity were the mechanisms that had the most
inﬂuence on male yearling average beam diameter. Whereas, the underlying mechanisms
that had the most inﬂuence in the NLP study site were temperature, winter severity, and
density. In the SLP study site, density had the most inﬂuence on the male yearling
average beam diameter.

The underlying mechanisms that had an inﬂuence on male yearling point count in
each regional study site were similar to those that had an inﬂuence on male yearling
average beam diameter. Weather was the dominant factor that inﬂuenced male yearling

point count in the UP study site. Snowfall, temperature, winter severity, and density were

209

 

all factors that affected male yearling point count in this study site. Temperature, winter
severity, and density inﬂuenced the male yearling point count in the NLP study site.
These underlying mechanisms were the same as those that had an inﬂuence on male
yearling average beam diameter in the same study site. In the SLP study site, there were
no underlying mechanisms detected that had an inﬂuence on male yearling point count.

These results suggest that, in general, as you move from north to south in
Michigan, the dominant inﬂuence on the male yearling antler measurements in the
northern part of the state was weather, which gave way to a mixture of weather and
density in the mid-part of the state, and in the southem—most part of the state, density had
the most impact. The results also suggest that since there were counties in each of the 3
regional study sites in which temperature and/or snowfall were inﬂuencing the yearling
antler measurements as a measure of herd quality but winter severity was not having an
inﬂuence and vice versa. This implies that the winter severity index did not encompass
all of the impacts of snow and temperature. In addition, habitat potential for the 3 habitat
requirements had an inﬂuence on the male yearling antler measurements as a measure of
herd quality in the NLP study site. Thermal cover had the most inﬂuence on both male
yearling average beam diameter and point count; as thermal cover potential increased, the
antler measurements also increased. Fall and winter food potential had an inﬂuence on
male yearling average beam diameter; although, there was a negative relationship

between the habitat potential and the index of herd quality.

210

Data Limitations
Incomplete Data

Broad temporal and spatial trends exist in the underlying mechanisms that may
inﬂuence yearling male antler measurements as indices of herd condition and, in general,
there was a mixture of factors affecting these indices in each of the 3 regional study sites.
There were, however, some potential limitations to this study. One limitation was
incomplete, or missing, data. This is problematic because having incomplete data might
mean that certain analyses can not be conducted or that the results may not be
comprehensive. For example, in the SLP study site, there were several years in which a
corrected average WSI could not be calculated because of missing data. Thus, there were
fewer years in which analyses could be conducted with the corrected average WSI in
relation to the other underlying mechanisms or the antler measurements as indices of herd
condition.

Incomplete data were also a problem in that the habitat potential data were only
available for the NLP study site. Therefore, comparisons of fall and winter food
potential, spring and summer food potential, and thermal cover potential cannot be made
among the 3 regional study sites in Michigan. Also, to date, there is only 1 habitat
potential for each of the 3 habitat requirements in each county in the NLP study site for
the entire course of this study; they do not change on any temporal basis. Thus, no yearly
comparisons can be made among fall and winter food potential, spring and summer food
potential, and thermal cover potential in the NLP study site. Since there can be no

regional or yearly comparisons of the habitat potential for each of the 3 habitat

211

 

requirements, at this point and time, we can only draw limited conclusions as to how

habitat quality inﬂuences male yearling antler measurements as indices of herd condition.

Data Aggregation

Aggregating data was also a possible limitation to this study. The temperature
data were aggregated into a 9—month time period, and the snowfall data were aggregated
into a 4-month time period so that analyses could be conducted. By aggregating these
data, however, some of the detail that existed at a ﬁner level (e. g., months) was
overlooked in the analyses and, hence, affected the ﬁnal outcomes. The habitat potential
for fall and winter food, spring and summer food, and thermal cover was aggregated from
a ﬁne scale (i.e., habitat types) to a broad scale (i.e., counties). Thus, how habitat quality
may inﬂuence male yearling antler measurements as indices of herd condition at ﬁner
scales was completely overlooked and some of the ﬁner nuances that could more

accurately explain this relationship were lost.

Biased Data

Some of the data used in the study was extracted from the biodata. There are,
however, various biases associated with this data set. This includes compositional,
seasonal, and geographical biases which may have impacted the results and,
subsequently, the conclusions drawn from these results. The biases in the biodata and

ways in which these biases could be improved are discussed at length in Chapter 1.

212

Assumptions

To conduct analyses, assumptions about the data were necessary because of the
data limitations. Several assumptions were necessary for the SAK estimate method to
accurately estimate deer density. These assumptions include: 1) the population is at
stable age distribution; 2) the sample sexed and aged at the check stations is
representative of the population; 3) fawn production is accurately measured; 4)
nonharvest mortality, or the survival rate, is known and constant throughout the year; 5)
buck harvest pressure is uniform from year to year (Eberhardt 1960, Creed et a1. 1984,
Mattsen-Hanson 1998). For this study, these assumptions were not tested and it was
assumed they were met.

There were also assumptions associated with the temperature, snowfall, and
winter severity data. It was assumed that all of these data were appropriately and
accurately collected and recorded. In addition, even though these data were obtained
from only 1 weather station from within each county in the 3 regional study sites, it was
assumed that these data, more or less, accurately represented the weather conditions in
each study site.

The ecoregion, land-type association, soil association, and presettlement
vegetation layers used to develop the habitat types for the habitat potential models were
assumed to be accurate and that the resulting habitat types are in fact valid. It was also
assumed that the habitat models used to predict the habitat potential for fall and winter
food, spring and summer food, and thermal cover in the NLP study site do accurately

reﬂect the habitat potential of the 3 habitat requirements in that area of the state.

213

 

Additional Approaches

In addition to the underlying mechanisms we examined in relation to the herd
quality in Michigan, there are other possible inﬂuences on herd quality that could be
examined. For example, the amount of rainfall may indirectly inﬂuence herd condition by
altering the growth and distribution of forage and browse for deer. Deer that have little
access to nutritious forage during a harsh winter may experience detrimental effects such
as loss of body weight, low productivity, and undersized antler grth (Cheatum and
Severinghaus 1950, Severinghaus et al. 1950). However, if deer can obtain adequate
nutrition during the spring and summer, often they can recover from the nutritional stress
brought on by a harsh winter and will not experience such detrimental effects (Verme and
Ullrey 1984). During this period of time, 1 of the limiting factors for adequate forage
grth becomes rainfall, especially in dryer regions. The amount of rainfall can affect
both the quantity and quality of forage and/or browse available to deer (Shea and
Osborne 1995) and long periods of drought can negatively affect deer habitat and, hence,
deer condition (Teer 1984). Although rainfall is generally not a limiting factor in
Michigan, it is another possible underlying mechanism that has the potential to inﬂuence
herd condition (Kammermeyer and Thackston 1995). Therefore, it may be beneﬁcial to
examine rainfall in relation to male yearling antler measurements as indices of herd
quality, especially in areas of Michigan that have experienced drought, to gain more
insight into which underlying mechanisms that are driving deer herd condition.

The lag time between changes in deer density and range condition in relation to
herd quality in Michigan is another aspect that could be examined. Some studies suggest

that there may be as much as a 2-year lag between the time that range conditions change

214

and quality indices such as antler measurements show a signiﬁcant response (Jacobson
and Guynn 1995). For example, bucks that are undernourished or stressed due to
overcrowding may have undersized antler measurements the subsequent year (Ozoga et
al. 1995). This may be partially due to the fact that antler growth is inﬂuenced by the
quality and quantity of forage deer are able to access during the previous winter, and that
changes in deer density do not affect changes in habitat quantity or quality for almost a
year (Severinghaus et al. 1950, Jacobson and Guynn 1995). Thus, conducting analyses
that stagger deer density, habitat potential for each of the 3 habitat requirements, and both
male yearling average beam diameter and point count on a yearly basis may actually
provide a more accurate picture of the interactions between the underlying mechanisms

and herd quality.

Conclusion

Understanding the underlying mechanisms that drive white-tailed deer herd
condition in Michigan will enable mangers to develop scientiﬁcally sound management
strategies that incorporate not only population numbers but also those factors that
inﬂuence population characteristics. And although data limitations exist, the data used in
this study were the most readily available, and the basic trends in the data supports the
overall pattern in both the antler measurements as indices of herd condition and the

underlying mechanisms than inﬂuence the characteristics of populations.

215

LITERATURE CITED

R. I. Blouch. 1984. Northern Great Lakes states and Ontario forests. Pages 391-409 in
L. K. Halls, ed. White-tailed Deer: Ecology and Management. Wildlife
Management Institute, Harrisburg, PA.

Cheatum, E. L. and C. W. Severinghaus. 1950. Variations in fertility of white-tailed
deer related to range conditions. Proceedings of the 15th North American Wildlife
Conferece.

Cook, S. L. and S. R. Winterstein. 2000. The evaluation of the MDNR’s white-tailed
deer lactation data.

Cook, S. L., B. D. Hughey, and S. R. Winterstein. 2001. The evaluation of the MDNR’s
Winter Severity Index.

Cowan, R. L. and T. A. Long. 1962. Studies on antler grth and nutrition of white-
tailed deer. Proceedings of the ﬁrst National White-Tailed Deer Disease
Symposium.

Creed, W. A., F. Haberland, B. E. Kohn, and K. R. McCaffery. 1984. Harvest
management: The Wisconsin experience. Pages 243-260 in L. K. Halls ed.
White-tailed Deer: Ecology and Management. Wildlife Management Institute,
Harrisburg, PA.

Eberhardt, L. 1960. Estimation of vital characteristics of Michigan deer herds.
Michigan Department of Conservation Game Division Report No. 2282.

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

Ford, W. M., A. S. Johnson, and P. E. Hale. 1997. Inﬂuences of forest type, stand age,
and weather on deer weights and antler size in the Southern Appalachians.
Journal Society of American Foresters 21:11-18.

French, C. E., L. C. McEwen, N. D. Magruder, R. H. Ingram, and R. W. Swift. 1956.
Nutrient requirements for grth and antler development in the white-tailed deer.
Journal of Wildlife Management 20:221-232.

Harlow, R. F . 1984. Habitat evaluation. Pages 601-628 in L. K. Halls ed. White-tailed
Deer: Ecology and Management. Wildlife Management Institute, Harrisburg, PA.

216

 

Hill, H. R., J. Meister, and J. Pohl. 1981. Deer checking station data. Michigan
Department of Natural Resources, Wildlife Division Report 2924.

Jacobson, H. A. and D. C. Guynn Jr. 1995. A primer. Pages 81-102 in K. V. Miller and
R. L. Marchington, eds. Quality Whitetails: The How of Quality Deer
Management. Stackpole Books, Mechanicsburg, PA.

Johnson, F. W. 1937. Deer weights and antler measurements in relation to population
density and hunting effort. Transactions of the North American Wildlife
Conference 22446-457.

Kammerrneyer, K. E. and R. Thackston. 1995. Habitat management and supplemental
feeding. Pages 155-168 in K. V. Miller and R. L. Marchington, eds. Quality
Whitetails: The How of Quality Deer Management. Stackpole Books,
Mechanicsburg, PA.

Kams, P. D. 1980. Winter-the grim reaper. Pages 47-51 in R. L. Hine and S. Nehls, eds.
White-tailed deer population management in the north central states. The
Wildlife Society, Eau Claire, WI.

Kie, J. G., M. White, and D. L. Drawe. 1983. Condition parameters of white-tailed deer
in Texas. Journal of Wildlife Management 47(3):583-594.

Leberg, P. L. and M. H. Smith. 1993. Inﬂuence of density on grth of white-tailed
deer. Journal of Marnmalogy 74(3):723-731.

Leopold, A. 1933. Game Management. C. Scribner’s Sons, New York, NY.

Matschke, G. H., K. A. Fagerstone, F. A. Hayes, W. Parker, D. 0. Trainer, R. F. Harlow,
and V. F. Nettles. 1984. Population inﬂuences. Pages 169-188 in L. K. Halls ed.
White-tailed Deer: Ecology and Management. Wildlife Management Institute,
Harrisburg, PA.

Mattson-Hansen, K. M. 1998. Integration of archery white-tailed deer (Odocoileus
virginianus) harvest data into a sex-age-kill population model. MS Thesis.
Michigan State University. 95 pages.

Marchington, R. L. and D. H. Hirth. 1984. Behavior. Pages 129-168 in L. K. Halls, ed.
White-tailed Deer: Ecology and Management. Wildlife Management Institute,
Harrisburg, PA.

McCullough, D. R. 1982. Antler characteristics of George Reserve white-tailed deer.
Journal of Wildlife Mangement 46:821-826.

217

 

Miller, K. V., R. L. Marchington, and J. J. Ozoga. 1995. Deer Sociobiology. Pages 118-
128 in K. V. Miller and R. L. Marchington, eds. Quality Whitetails: The How of
Quality Deer Management. Stackpole Books, Mechanicsburg, PA.

National Climatic Data Center, National Oceanic and Atmospheric Administration.
2002. http://www.ncdc.noaa.gov/ol/climate/stationlocator.html.

Ozoga, J. J. 1968. Variations in microclimate in a conifer swamp deeryard in northern
Michigan. Journal of Wildlife Management 32(3):574-585.

Ozoga, J. J. and L. W. Gysel. 1972. Response of white-tailed deer to winter weather.
Journal of Wildlife Management 36(3):892-896.

Ozoga, J. J ., E. E. Langenau Jr., and R. V. Doepker. 1995. The north-central states.
Pages 210-237 in K. V. Miller and R. L. Marchington, eds. Quality Whitetails:
The How of Quality Deer Management. Stackpole Books, Mechanicsburg, PA.

Payne, R. L. 1970. White-tailed deer physiological indices. Pages 29-35 in Deer
population dynamics and census methods: a review. Deer Population Dynamics
Subcommittee, Forest Game Committee, Southeastern Association of Game and
Fish Commissioners.

Rasmussen, G. P. 1985. Antler measurements as an index to physical condition and
range quality with respect to white-tailed deer. New York Fish and Game Journal
32297-113.

Sams, M. G., R. L. Lochmiller, C. W. Qualls, Jr., and D. M. Leslie, Jr. 1998. Sensitivity
of condition indices to changing density in a white-tailed deer population. Journal
of Wildlife Diseases 34(1)]10-125.

Severinghaus, C. W. 1947. Relationship of weather to winter mortality and population
levels among deer in the Adirondack Region of New York. Proceeding of the 12th
North American Wildlife Conference.

Severinghaus, C. W. 1955. Deer weights as an index of range conditions on two
wilderness areas in the Adirondack Region. New York Fish and Game Journal
2: 154-160.

Severinghaus, C. W. and H. F. Maguire. 1955. Use of age composition data for
determining sex ratios among adult deer. New York Fish and Game Journal
2(2):242-246.

Severinghaus, C. W., H. F. Maguire, R. A. Cookingham, and J. E. Tanck. 1950.
Variations by age class in the antler beam diameters of white-tailed deer related to
range conditions. Transactions of the North American Wildlife Conference
15:551-5 70.

218

 

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(1):30-38.

Shea, S. M. and J. S. Osborne. 1995. Poor-quality habitats. Pages 193-209 in K. V.
Miller and R. L. Marchington, eds. Quality Whitetails: The How of Quality Deer
Management. Stackpole Books, Mechanicsburg, PA.

Teer, J. G. 1984. Lessons from the Llano Basin, Texas. Pages 261-292 in L. K. Halls,
ed. White-tailed Deer: Ecology and Management. Wildlife Management
Institute, Harrisburg, PA.

Ullrey, D. E. 1983. Nutrition and antler development in white-tailed deer. Pages 49-59
in R. D. Brown, editor. Antler development in cervidae. Caesar Kelberg Wildlife
Research Institute, Kingsville, TX.

Verme, L. J. 1965. Swamp conifer deeryards in northern Michigan: their ecology and
management. Journal of Forestry 63(7):523-529.

. 1968. An index of winter weather severity for northern deer. Journal of
Wildlife Management 32(3):566-574.

Verme, L. J. and D. E. Ullrey. Physiology and nutrition. Pages 91-118 in L. K. Halls,
ed. White-tailed Deer: Ecology and Management. Wildlife Management
Institute, Harrisburg, PA.

Weeks, H. P., Jr. 1995. Mineral supplementation for antler production. Pages 155-168 in

K. V. Miller and R. L. Marchington, eds. Quality Whitetails: The How of Quality
Deer Management. Stackpole Books, Mechanicsburg, PA.

219

CONCLUSIONS AND MANAGEMENT IMPLICATIONS

In 1996, the Michigan voters passed Proposal G, a ballot initiative, to ensure that
wildlife would be managed based on scientiﬁcally sound information and procedures.
The processes used, and conclusions reached in this study, may aid managers in
developing better scientiﬁcally sound management strategies that incorporate deer
population numbers and the condition of deer. Along with social and ecological
considerations, herd condition can be incorporated into a more holistic deer management
strategy that addresses deer management at a landscape-level, rather than simply
manipulating local deer numbers. Also, better understanding the relationship between
underlying mechanisms such as deer density, winter severity, and habitat quality that
drive changes in deer population-level quality indices and herd condition could provide
additional information for managers to incorporate into a deer management strategy.

These results may also be beneﬁcial to managers for communicating with the
public as to how and why management decisions are made as well as how different
management strategies will sustain a healthy deer herd. Having a deﬁnition of herd
condition that uses quality indices that the public understands and can visually identify
enables managers to more easily communicate and maybe gain credibility with the
public. In addition, understanding the underlying mechanisms that are inﬂuencing herd
quality may enable managers to explain more details to the public as to why deer are in a
certain physical condition and how different management strategies will alter that
physical condition of deer in a speciﬁc area.

In Chapter 1, it was determined that population-level quality indices such as

yearling average beam diameter, point count, and lactation status can reﬂect the relative

220

condition of the deer herd in the 3 distinct regions of Michigan. Furthermore, it was
determined that the differences in each of the quality indices among the 3 regional study
sites were detectable. In Chapter 2, it was determined that not only did the inﬂuences of
herd condition differ among the distinct 3 regions in Michigan but also that these
inﬂuences were likely to be predictable.

The results of this study indicate that there are differences in the quality indices
among the 3 distinct regions of Michigan. This suggests that deer management strategies
should, at the largest spatial scale, be tailored to each distinct region. In reality, though,
there are differences in the quality indices even within regions. Therefore, managers
need to consider the appropriate scale at which to collect meaningful data and manipulate
deer populations. Deer management decisions should be based on a combination of
population dynamics, herd condition, life requisites, and public desires at the appropriate
scale in which management makes the most sense.

There is some question, though, as to how well the results of this study represent
the condition of the deer herd across all of Michigan. The basic trend in each of the
quality indices does support the overall pattern in which the percent of yearling does
lactating, the male yearling average beam diameter and point count increased as one
moves from north to south. This is also the case for the underlying mechanisms; for the
most part, temperature, and deer density increased as you move from north to south,
whereas snowfall and winter severity decreased. These broad trends appear logical, but
the basic trends for both the quality indices and the factors that inﬂuence herd condition
were based solely on data gathered within each of the 3 regional study sites. Therefore,

when managers are targeting a speciﬁc area for management more information may need

221

 

 

 

to be gathered to develop a deer management strategy that is appropriate for that area.
For example, if managers are targeting an area for management that falls outside of the 3
study sites used in this study, managers may want to gather speciﬁc information on the
herd condition, the factors that inﬂuence herd condition (e.g., density, weather, habitat),
and the public attitudes that exist in that area.

If the biodata were collected at a ﬁner scale than county level at either political
boundaries (e.g., township or section) or ecological boundaries (e.g., ecoregions) over a
signiﬁcant period of time, those data could be used to predict the overall quality of the
deer herd at speciﬁc spatial scales. After quantifying herd condition by examining
quality indices such as lactation status and antler measurements, these indices could be
put into a model to quantify the overall condition of the deer herd in different
management areas on an annual basis. The underlying mechanism data, collected at an
appropriate scale, could also be used in this capacity. Starting with a population number
in a speciﬁc area, the underlying mechanisms that inﬂuence herd condition could be
manipulated through a predictive model to determine how they would affect the deer herd
over the long-term. This may be useful for managers who are trying to pinpoint
management in areas that require special consideration. For example, knowing the
quality of the deer herd may enable managers to pinpoint current areas where there is the
potential for disease (e. g., Bovine TB or Chronic Wasting Disease) to be introduced or to
be spread. Being able to predict where deer will be in poor condition in the future may
help managers take preventative actions to diminish the risk of disease in that area.

Even though this study may only provide guidelines to managers on ways to

approach a more holistic deer management strategy, the processes and parameters used in

222

 

this study may be applied as a template for managers to develop management strategies
for other species by examining population trends and other factors that inﬂuence

populations at different scales.

Recommendations

By examining the physical condition of deer as well as those underlying
mechanisms that inﬂuence herd quality and incorporating these factors into a deer
management strategy, managers can devise deer management strategies that move
beyond simply manipulating the number of deer in a speciﬁc area. This, however, is only
1 aspect of deer management; to truly make it a holistic approach to deer management,
managers need to consider the ecological and social aspects of deer management. When
the 3 sections of the “umbrella” project are complete, they will be integrated into a model
to quantify deer management decisions based on herd condition, factors that inﬂuence
herd condition, habitat quality, and public desires. This will allow managers to tailor
deer management because knowing how these different aspects of deer management
interact with one another will enable managers to set population goals, assign hunting
pressures, initiate habitat manipulations, and develop public outreach in different areas of
the state.

To determine the target population number for deer in a speciﬁc area, managers
should take into consideration the current condition of the herd, habitat quality, and
public tolerance as well as the desired future status of the population. Using all of this
information, managers could then determine what kind of actions need to be taken to

ensure that the desired status of the deer herd is achieved. For example, if the

223

management goal is to improve the condition of a deer herd that is in poor condition
because the population is so large that resources have become scarce, and the public in
that area tends to support that goal, then a logical approach would be to reduce the
number of deer to ensure that the deer have adequate resources. Hunting pressures could
then be assigned accordingly to ensure that the appropriate number of deer is maintained
to sustain a healthy population. Promoting habitat improvements in conjunction with
using appropriate hunting pressures to control deer numbers, could increase quantity and
quality of resources; subsequently allowing for an improvement in the physical condition
of the deer herd in that area. This scenario may change if the public were not supportive
of actions needed to improve the condition of the deer herd. In this case, it may be
appropriate to ﬁrst address the public and decide what kind of educational or hands-on
programs would be effective in trying to educate the public on how and why these actions

should be taken to ensure the existence of a healthy deer herd.

224

 

llllll‘lllll‘l‘llllllll1|lllll'lllllLll ll

93 02327 0758