hm”

 

 

 

 

 

 

rut—T"

v

_.:.‘..'-pv

 

I! . . 1.01.7 :1.
. . ex..7..5!l.\

3.13. (mm..&

:r .v.2vn\h..v.......
1:13.. ~ ‘5..." P
.Jiﬂr) .: 115%.

4.3.
l. ﬁrdnuuzht
‘9. if

[.5 . q
‘ I. a: agilh
vﬂitnnuur. Hus . V I.Wl(..\sald
1:53.. I .

3-1.1.3.. u
.....l.s.|. 3,! z I
5.3.51.1: .7... .1!
351': ..r 1.23:5...
ﬁtnnﬂnz, .1 .wd).}..~!’)
.1 n. 5..

{.5 UV- . twat. 4
I. 3 v .71.
.1

1:4. .ufl.‘ 4..

.29! (.2563 wk"?

ﬁuﬁﬁﬂ

I HIGAN ST IVER Y
w

lull!“ lzllullllmﬁ

 

 

 

l Illllllllllll

3 1 93 01572 5538

This is to certify that the

thesis entitled
'Seasonal Movements, Habitat Use Patterns, and Population
Dynamics of White—tailed Deer (Odocoileus virginianus)
in an Agricultural Region of Northern Lower
Michigan

 

presented by

Kristie L. Sitar

has been accepted towards fulfillment
of the requirements for -

 

 

Master of Science degree inFish. & Wildl.

srﬂ‘wgma:

Major professor

Date May 21, 1996

 

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

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE II RETURN BOX to romovo this chockout "om your rocord.
TO AVOID FINES rotum on or botoro duo duo.

DATE DUE DATE DUE DATE DUE
aﬂlﬂ5%3 '

'~...L-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU is An Affirm-tin ActiorVEquol Opportunity Institution
Wanna-m

 

SEASONAL MOVEMENTS, HABITAT USE PATTERNS, AND POPULATION
DYNAMICS OF WHITE-TAILED DEER (Odocoileus virginianus) IN AN
AGRICULTURAL REGION OF NORTHERN LOWER MICHIGAN
By

Kristie L. Sitar

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

MASTER OF SCIENCE

Department of Fisheries and Wildlife

1 996

ABSTRACT
SEASONAL MOVEMENTS, HABITAT USE PATTERNS, AND POPULATION
DYNAMICS OF WHITE-TAILED DEER (Odocoileus virginianus) IN AN
AGRICULTURAL REGION OF NORTHERN LOWER MICHIGAN
By

Kristie L. Sitar

In recent years, damage to agricultural crops by white-tailed deer (Odocoileus
virginicmus) has been a growing concern of farmers and wildlife biologists in Michigan’s
northeastern lower peninsula. Biologists issue crop damage control permits and vary
harvest quotas to manage localized, high density populations. The distance, direction
and initiation dates of spring and fall migration of deer greatly inﬂuence the degree of
crop damage. Seasonal habitat use patterns and home range sizes of deer also potentially
impact damage levels by dictating how much time deer forage in and around crop ﬁelds.

I trapped deer during the winters of 1994 and 1995 and marked deer with radio
collars or ear-tags. Radio collared deer were tracked during all seasons and at various
times of the day to determine the magnitude and direction of migrations, home range
sizes and habitat use patterns. Home range size was compared among years, seasons,
migratory groups, migratory destinations, and sexes using AN OVA models and deer

were evaluated for traditional use of home ranges between years. Habitat use was

evaluated for individual deer and analyses were made by period of the day, year, season,
migratory groups, migratory destination, and sex. Overwinter deer densities were
estimated for the study area using pellet group counts performed each spring. Deer
counts ﬁ'om monthly road surveys were used to compare yearly trends and estimate fawn
to doe productivity ratios, fall-recruitment ratios, and fall buck to doe ratios. Annual and
period survival of adults and yearlings were estimated ﬁ'om collared and tagged deer.
Additional productivity ratios were estimated from observations of collared does and
incidental sightings of does and fawns during summer. Fall recruitment ratios were
estimated from the fawn to doe ratios of the captured deer. The sexes and ages of the
trapped deer were also used to estimate buck to doe ratios and age structure.

In the two years, 73 deer were radio collared and 4,172 deer locations were
obtained. Approximately half of the collared deer in each year were migratory. Spring
migrations of collared deer began ﬁ'om mid-to late March and fall migrations began in
October or November and were mostly completed by the end of ﬁrearm season. Deer
migrated an average of 10 km with males tending to travel farther than females although
diﬁ‘erences were not signiﬁcant. Home range sizes were not diﬁ'erent by year, season,
sex, migratory category or migratory destination, although pairwise comparisons of
seasonal ranges indicated smaller winter than summer ranges in a heavy snow winter and
larger winter than summer ranges in a light snow winter. Summer ranges were used with
greater ﬁdelity than winter ranges, probably due to the relative mildness of winters.
Habitat use diﬁ‘ered between years, migratory groups during 1994, migratory

destinations, and between sexes. No overall seasonal diﬁ‘erences were detected.

However summer habitat use differed by sex, migratory group and migratory destination.
Wooded habitats were often the most ﬁ'equently used habitats by all deer groups. All
groups showed signiﬁcant habitat selection or avoidance patterns, with agricultural
selection indicated by varying degrees in all deer groups except for deer migrating to
non-agricultural summering areas. Overwinter deer densities ranged from 8.94 - 34.86
deer km2 across the study area. Most deer were counted in March and April monthly
surveys, corresponding with snow melt in each year, and counts increased ﬁ'om May
through early fall and decreased in late fall. Annual and period survival rates for adults
and yearlings were not signiﬁcantly diﬁ‘erent although adult survival was higher than
yearling survival both annually and during all periods. Productivity estimates ranged
from 0.49 - 1.36 fawns per doc and fall-recruitment ratios ranged ﬁom 0.67 - 2.28 fawns
per doe. Sex ratios determined from observed counts approximated 1 buck for every 4 -
5 does. The age structure of the captured deer was skewed toward younger age classes
with males more highly skewed than females.

The timing of deer arrival on winter ranges is crucial to determining the
vulnerability of deer to block permits. Adjustments to crop damage control permits or
Deer Management Unit boundaries might increase the probability of harvesting deer
responsible for damage. High value crops should not be planted adjacent to wooded
habitats. However, if necessary, habitat management practices could be employed to

increase the quality and quantity of forest openings in non-adjacent wooded habitats.

ACIGNIOWLEDGMENTS

I greatly appreciate the support of the agencies that ﬁmded this research; the
Michigan Department of Natural Resources Wildlife Division, the Michigan Agricultural
Experiment Station, and the Michigan State University Cooperative Extension Service.
Their insights made this research possible.

I am especially grateﬁrl to my major advisor, Dr. Scott Wmterstein, for all of his
guidance, patience, advice and support of me. I also owe a great deal of thanks to my
committee members, Drs. Rique Campa and Jim Sikarski, for their valuable insights and
suggestions throughout my research and in the preparation of this manuscript.

The biologists in the study area, Torn Carlson, Elaine Carlson, Glen Matthews
and Bob Odum deserve a special thanks for all of their support and help throughout the
project. Their assistance and advice in selecting landowners, establishing a check station,
locating missing deer, collecting information on marked deer, trapping, and instruction
on pellet courses was invaluable and greatly improved the research. I also thank Paul
Friedrich of the Rose Lake Wildlife Research Station for his expertise in aging a very

cumbersome set of deer teeth.

The cooperation of many landowners made this research possible and I am
grateﬁrl for their support and enthusiasm throughout this long process and appreciate
their patience in waiting for results.

Katherine Braun’s detailed ﬁeld notes enabled me to select successﬁrl trapping
areas with minimal effort. I thank Katherine for her help in the ﬁeld, her insights into the
project, and her ﬁiendship. I also thank my fellow graduate students, Allison Gormley,
Teresa Mackey, Mark Moore, Wendy Sangster, Delia Raymer, Ed Roseman, and Tim
Van Deelan for their technical support, trapping assistance, and ﬁiendship.

My ﬁeld assistants, interns, and undergraduate assistants -- Colleen Trese, Becky
Fedewa, Bryan Knowles, Julie Car, Dan Kennedy, Cal Steinorth and Jason Hullman --
were invaluable to me and I appreciate all of their eﬂ‘orts, ideas, hard work, tolerance
(for my sometimes demanding nature), and ﬁiendship.

I thank Drs. Paul Haeﬁrer Jr., M. Joseph Klingensmith, and Franz Seischab of the
Department of Biology at Rochester Institute of Technology who guided me early on
and taught me what qualiﬁed as good science.

I thank my entire family for being as understanding as they could be when I
missed many special occasions and holidays and for doing all of the traveling in the past
few years. I am grateful for their support and enjoyed all of their help in the ﬁeld despite
allergies and bad backs.

Lastly, I thank my husband, Shawn, for his love and support and for his extreme

dedication in working alongside me almost every weekend.

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................... ix
LIST OF FIGURES .................................................................................................... xii
INTRODUCTION ........................................................................................................ 1
Migration ................................................................................................................... 4
Home range and habitat use ....................................................................................... 7
OBJECTIVES ............................................................................................................. 11
STUDY AREA ........................................................................................................... 12
METHODS ................................................................................................................. 15
Capture and handling ............................................................................................... 15
Radio telemetry ........................................................................................................ l8
Migration, home range, and habitat use .................................................................... 19
Population parameters .............................................................................................. 23
Abundance ............................................................................................................ 23
Survival ................................................................................................................ 26
Productivity .......................................................................................................... 28
Sex ratio and age structure .................................................................................... 29
RESULTS ................................................................................................................... 31
Migration ................................................................................................................. 34
Home range ............................................................................................................. 41
Habitat use ............................................................................................................... 50
Population parameters .............................................................................................. 61
Abundance ............................................................................................................ 61

VII

Survival ................................................................................................................ 65

Productivity .......................................................................................................... 72
Sex ratio and age structure .................................................................................... 74
DISCUSSION ............................................................................................................. 78
Migration ................................................................................................................. 78
Home range ............................................................................................................. 83
Habitat use ............................................................................................................... 87
Population parameters .............................................................................................. 92
Abundance ............................................................................................................ 92
Survival ................................................................................................................ 94
Productivity .......................................................................................................... 95
Sex ratio and age structure .................................................................................... 97
MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS .......................... 99
APPENDIX .............................................................................................................. 105
LITERATURE CITED ............................................................................................. 120

Table 1.

Table 2.

Table 3.

Table 4.
Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

LIST OF TABLES

Age and sex of deer marked in Presque Isle (PI), Montrnorency (M), and
Alpena (A) counties in 1994 and 1995. Table entries represent number of
collared deer and numbers in parenthesis represent number of ear-tagged

deer. ........................................................................................................... 33

Migratory status of collared deer by sex and age category for spring 1994 and
1995. Numbers in parenthesis indicate deer collared in 1994 ....................... 35

Average migration distances (km) of collared and tagged deer by sex in 1994
and 1995. .................................................................................................... 40

Sex and age classiﬁcation of fall migrating collared deer in 1994 and 1995. .. 45
Average home range size (ha) of collared deer in 1994 and 1995 .................. 46

Analysis of variance results of main eﬁ'ects and interactions of home range size
by year, migratory status ‘ , and sex. ............................................................ 47

Analysis of variance results of main effects and interactions of home range size
by season and migratory destination ' of deer. ............................................. 49

Percent of deer locations within each habitat type by sampling period ‘, season,
migratory status, migratory destination b, and sex in 1994 and 1995 ............. 51

Results of Chi square analysis of habitat use of radio collared deer. Asterisks
indicate signiﬁcance. ................................................................................... 53

Table 10. Results of combined probability analysis of habitat use vs. availability

comparisons by deer group .......................................................................... 57

Table 11. Percent of deer with non-random habitat use patterns that signiﬁcantly

selected or avoided various habitats within their home ranges. ..................... 59

Table 12. Estimated overwinter deer densities and standard errors (SE) for Presque Isle,

Montrnorency and Alpena counties in 1994 and 1995. ................................. 63

ix

Table 13.

Table 14.
Table 15.

Table 16.

Table 17.

Table 18.

Mayﬁeld survival estimates (5' ) for collared and tagged deer in 1994 and
1995. .......................................................................................................... 69

Results of survival analysis of collared and tagged deer in 1994 and 1995.... 70
Yearly causes of mortality for radio collared and ear-tagged deer. ............... 71

Average fawn to doe productivity, fall-recruitment and post-hunting season
recruitment ratios for 1994 and 1995 and statistical signiﬁcance of yearly
diﬁ‘erences ................................................................................................... 73

Age structure of deer determined by the age composition of trapped deer,
deer aged at the study area check station, and road killed deer in 1994 and
1995. Table values are percentages of deer classiﬁed in each age category. 75

Age structure of male and female deer determined by the age composition of
trapped deer and deer aged at the study area check station in 1994 and 1995.
Table values are percentages of deer classiﬁed in each age category. ........... 77

Appendix Table 1. Trap sites across the study area where deer were successfully

marked in 1994 and / or 1995. .................................................... 106

Appendix Table 2. Trap night description and trapping success rate for 1994 and

1995. ......................................................................................... 107

Appendix Table 3. Number of deer locations within each time period in 1994 and

1995. ......................................................................................... 108

Appendix Table 4. Number of deer locations by season and year for 1994 and 1995. 109

Appendix Table 5. Observed (Obs), expected (Exp), and partial Chi square values for

deer locations ' in eight habitat types across three sampling periods
(years combined). ....................................................................... 110

Appendix Table 6. Sex, age, capture date, migratory status, and fate of radio collared

deer as of March 15, 1996 .......................................................... 111

Appendix Table 7. Sex, age, capture date, migratory status, and fate of ear-tagged

deer. .......................................................................................... l 16

Appendix Table 8. Doc ages and number of fawns seen with collared does in 1994 and

1995. Dashes represent deer that were not observed due to age or
sex during a given year. .............................................................. 117

Appendix Table 9. Deer counts ﬁom 1994 and 1995 road surveys and productivity

ratios from actual and adjusted ' counts. ..................................... 118
x

Appendix Table 10. Deer counts from 1994 and 1995 road surveys and buck-doe ratios
ﬁ'om actual and adjusted ' counts. .............................................. 119

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

LIST OF FIGURES

Study area comprised of Presque Isle, Montrnorency, and Alpena counties in
the northeastern lower peninsula of Michigan. ............................................ 13

Distribution of traps across study area with numbers indicating individual trap
location. Trap designations refer to numbers listed in Appendix Table 1. 32

Initiation of spring migration for collared deer in 1994 and 1995 ................. 36

Distance and direction of spring migrations of collared deer in 1994. Circles
represent trapping location and arrows represent movements for one or more
deer. .......................................................................................................... 37

Distance and direction of spring migrations of collared deer in 1995. Circles
represent trapping location and arrows represent movements for one or more
deer. .......................................................................................................... 38

Average spring migration distance of collared and tagged deer in 1994 and
1995. Error bars represent one standard error. .......................................... 39

Destination of spring migrating collared and tagged deer in 1994 and 1995. 42

Average distances of spring deer migration (error bars represent one standard
error) to agricultural and non-agricultural land in 1994 and 1995. Asterisks
indicate signiﬁcant diﬂ‘erences in destination migration distance within

years. ......................................................................................................... 43

Initiation of fall migration for collared deer in 1994 and 1995. .................... 44

Total number of deer counted in monthly road surveys (June, July and
August are average counts for two surveys) in 1994 and 1995. .................. 64

Total counts of female and male deer from June through November 1994 (A)
and June through October 1995 (B) road surveys. ...................................... 66

xii

Figure 12. Total counts of adults and sub-adults ﬁ'om February through December 1994
(A) and ﬁ'om January through October 1995 (B) road surveys. .................. 67

Appendix Figure 1. Monthly road survey route within the study area in 1994 and 1995.
Dashed lines depict the driven route and asterisks represent
northern and southern starting points. ....................................... 105

xiii

INTRODUCTION

White-tailed deer (Odocoileus virginianus) have been a valuable natural resource
in Michigan since pre-settlement times. Historically, deer were utilized principally for
subsistence purposes. The introduction of market hunting and widespread logging of
forested lands decreased deer numbers in the late 1800’s (McCabe and McCabe 1984,
Langenau 1994). Regulated hunting and habitat regeneration allowed deer populations
to expand until the early 1950’s when heavy browsing and succession depleted deer
habitat, again causing deer numbers to decline. Following this period, a growing timber
industry and habitat improvement programs enabled deer numbers to peak in 1989 at
approximately two million individuals (Langenau 1994, Ozoga et al. 1994). The 1993
population estimate was approximately 1.5 million individuals (W interstein et al. 1995).

Today, deer are still valued by both consumptive and non-consumptive users.
Bow, riﬂe, and muzzleloader hunters spend millions of days aﬁeld and generate in excess
of $300 million per year for Michigan’s economy (Dudderar et al. 1989, Reis 1990).
Non-consumptive uses consist primarily of recreational observation of deer or other
related activities (Langenau 1979). The current management goal of the Michigan
Department of Natural Resources (NHDNR) is for a statewide deer herd of 1.3 million

pre-harvest individuals (Langenau 1994).

2

High deer densities often lead to high rates of deer-vehicle accidents (Blouch
1984, Bashore et al. 1985) and complaints of crop damage (Langenau 1994, Vecellio et
al. 1994). As deer densities increase, populations expand into new habitats. Deer often
forage in agricultural croplands within their home ranges (Larson et al. 1978, Dusek et
al. 1989, Vecellio et al. 1994). Inevitably, these agricultural lands become an integral
and traditional part of deer ranges (Flyger and Thoerig 1962, Dusek et al. 1989). The
establishment of deer in habitats adjacent to agricultural areas leads to crop damage both
from foraging (Larson et al. 1978) and through the trampling of crops (Putnam 1986,
Vecellio et al. 1994).

Agricultural deer damage is a growing problem in the United States. Deer have
been reported to damage grains (winter wheat, buckwheat, millet, and rye), cash crops
(soybeans, sunﬂowers, beans, tobacco, and corn), truck crops (tomatoes, watermelon,
sweet potatoes, sugar beets, peas, and squash), and numerous other crops including ﬁ'uit
orchards, Christmas trees, and alfalfa (Hammerstrom and Blake 193 9, Swift 1948,
Moore and Folk 1977, Dusek et al. 1989). Deer damage complaints were ﬁrst recorded
in Michigan in the late 1930s (Langenau 1994). Complaints reappeared again in 1971
(Hansen 1978) and have persisted to current times.

Beginning in 1976, special kill permits were issued to landowners across the state
to help control damage (Dudderar et al. 1989). Two types of permits are commonly
issued; summer shooting permits and block permits. Summer shooting or “kill” permits
are intended for use on antlerless deer during the summer when crop damage is

occurring, and block permits are intended for the harvest of antlerless deer during the

3
hunting season. Despite the permit system, crop damage complaints reached an all time
high in Michigan’s northeastern lower peninsula in 1988 and 1989 (T. Carlson, MDNR,
pers. commun.), corresponding to the period when deer numbers peaked in Michigan.

Crop depredation leads to varying degrees of economic loss incurred by farmers
(Conover 1994). The timing of damage in relation to crop grth stage or
characteristics affects the recovery potential of the crop and extent of economic loss
(Flyger and Thoerig 1962, Putnam 1986). The crop damage issue is further complicated
by the variation farmers exhibit toward tolerance of damage, even at severe damage
levels (Brown et al. 197 8, Decker and Brown 1982, Conover 1994). Furthermore,
management of the issue requires not only controlling deer numbers and damage levels
but maintaining suﬁcient deer numbers to provide quality hunting opportunities.

It is often assumed that deer-caused crop damage increases with deer density
(Hansen 1978, Alverson et al. 1988, Vecellio et al. 1994). While the relationship
between damage and deer density is likely to be important, additional factors including
habitat interspersion, habitat type and quality surrounding croplands, ﬁeld size and shape,
crop type, deer movement patterns, seasonal habitat use patterns, and home range size of
deer should be examined to determine their inﬂuence on the severity of crop damage.
These elements must be incorporated into the evaluation of agricultural damage to
provide a better understanding of the nature of the crop damage problem. In addition,
consideration of these factors will aid in determining an eﬁ‘ective means of controlling or
reducing depredation. Deer damage to crops has been examined and quantiﬁed in

numerous studies (Palmer et al. 1982, Garrison and Lewis 1987, Austin and Urness

4
1993, Vecellio et al. 1994), but minimal eﬂ‘ort has been made to examine the underlying
factors that contribute to agricultural damage.

The ﬁrst attempts to examine some of these underlying factors and their
contributions to damage incorporated habitat area, animal density, and the area and value
of damage into a mathematical model (Gorynska 1981). The author reported positive
relationships despite large variability in the input data. Vecellio et al. (1994) found that
damage to corn and wheat was related to deer density and possibly to distance ﬁ'om
woodlands. While these studies are important ﬁrst steps toward understanding the
factors inﬂuencing damage, they do not address the relationship of deer behavioral

patterns to crop damage levels.

Migration

Movement and habitat use patterns of white-tailed deer have been studied since
the 1930’s (Bartlett 1932). Movement and habitat use vary regionally and are
characterized or inﬂuenced by behavioral traits, climatic conditions, deer densities
(Sanderson 1966), landscape level attributes, land use practices, and hunting (Kufeld et
al. 1988, Root et al. 1988).

In the northern parts of their range, where weather extremes are pronounced,
white-tailed deer are usually migratory, spending the winters congregated in yarding
areas that act to eﬁ'ectively buﬂ‘er adverse climatic conditions (Ozoga 1968, Marchington
and Hirth 1984, Nixon et al. 1988). The timing of migration is inﬂuenced by winter

severity (V errne 1968), often with a sharp decline in temperature or heavy snowfall

5
acting as the stimulus for deer migrations to winter yards (Ozoga 1968, Hoskinson and
Mech 1976). Deer yards are often comprised of dense coniferous swamps primarily
dominated by northern white cedar (Thuja occidentalis), ﬁr (A Ibies spp.) or spruce
(Picea spp.) (Habeck 1960, Ozoga 1968, Verrne 1973, Moen 1976, Marchington and
Hirth 1984). Yards provide thermal cover and browse (Hammerstrom and Blake 193 9,
Verrne 1973, Moen 1976) for wintering deer as well as serving to reduce the risk of
predation by decreasing the predator-prey ratio and facilitating escape through increased
runway densities (Hoskinson and Mech 1976, Messier and Barrette 1985, Beier and
McCullough 1990, Ozoga et al. 1994).

In the winter, deer conserve energy by reducing metabolism and activity levels in
response to reduced food intake experienced ﬁom fall to winter (Silver et al. 1969,
Moen 1976, 1978, Mautz 1978, Marchington and Hirth 1984). Reduced winter activity
is thought to inﬂuence range size. For instance, several studies report winter ranges of
migratory deer, in deep snow areas, to be considerably smaller than summer ranges
(Moen 1978, Tierson et al. 1985, Mooty et al. 1987). Others have reported larger
winter ranges than summer ranges due to reduced snow levels and shorter distances
between seasonal forages and protective cover in summer (Dusek et al. 1988, 1989,
Beier and McCullough 1990).

At the onset of spring, when the snow pack and weather conditions permit free
movement, deer leave protective yards in search of high quality spring foods (Sparrowe
and Springer 1970, Verme 1973, Hoskinson and Mech 1976, Tierson et al. 1985).

Forest openings are thought to beneﬁt deer by providing the ﬁrst spring foods; high

6
quality grasses and forbs (McCaﬁ‘ery and Creed 1969, Gladfelter 1984, Lenarz 1987).
Likewise, forested edges supply an array of high quality foods. Deer are considered an
edge species due to the opportunistic, preferential foraging behavior they exhibit along
habitat edges (\Vrlliamson and Hirth 1985).

Deer migrations tend to be traditional (V erme 197 3, Drolet 1976, Marchington
and Hirth 1984, Nelson and Mech 1984) and are passed along within a family group
from does to fawns. Several studies have suggested that migration distance is a ﬁrnction
of food availability; deer density; habitat type, interspersion, or quality; or a combination
of all of these (Dahlberg and Guettinger 1956, Verrne 1968, Rongstad and Tester 1969).
Verrne (1973) reported that deer in the western upper peninsula of Michigan were able
to meet their summer food requirements closer to winter yards due to greater habitat
interspersion. Eastern upper peninsula deer migrated an average of three kilometers
farther between seasonal ranges than did western deer. This difference was attributed to
diminished habitat interspersion resulting ﬁ'om large tracts of monotypic forest type in
the eastern portion of the upper peninsula. In addition, several studies report younger
male deer migrating greater distances between seasonal ranges (Carlsen and Farmes
1957, Marchington and Hirth 1984, Dusek et al. 1989).

Deer tend to be resident, or non-migratory, in regions where seasonal weather
extremes are minimal (Dahlberg and Guettinger 1956, Sparrowe and Springer 1970,
Larson et al. 1978, Dusek et al. 1989, Brown 1992). Several studies have reported a
portion of a deer herd to be migratory and a portion which remains sedentary

(Hoskinson and Mech 1976, Nixon et al. 1991, Brown 1992, Kufeld and Bowden 1995).

7
Springer and Sparrowe (1970) and Larson et al. (1978) found that most deer in South
Dakota and “frsconsin respectively, were non-migratory and did not develop completely
distinct winter and summer ranges. Instead, seasonal ranges of individual deer often
overlapped with seasonal shifts in primary use areas (Dahlberg and Guettinger 1956,

Larson et al. 1978, Dusek et al. 1989, Brown 1992).

Home range and habitat use

Home range is deﬁned as the area traveled on a seasonal or annual basis by an
individual in its normal activities of foraging, caring for young, and mating (Burt 1943).
Home ranges are large enough in size to provide speciﬁc yearly or seasonal requirements
and are comprised of many different vegetation types (Sanderson 1966, Marchington and
Hirth 1984). Several studies have suggested that home ranges are traditional through the
observation that two-year—old deer often establish home ranges that are adjacent or
identical to their mother’s (Hammerstrom and Blake 193 9, Marchington and Hirth 1984,
Nelson and Mech 1984, Tierson et al. 1985). In addition, strong ﬁdelity to winter (Beier
and McCullough 1990) and summer ranges (Tierson et al. 1985, Brown 1992) was
indicated by seasonal ranges having approximately equal centers of activity in successive
years.

Home range size is greatly inﬂuenced by habitat quality, interspersion, and food
and cover availability (Sanderson 1966). In regions where habitat interspersion is high,
deer home range sizes are smaller (Beier and McCullough 1990). However, regardless

of the degree of interspersion, deer oﬁen display non-uniform use of the area and

8
habitats within their home range (I-Ieezen and Tester 1967, Rongstad and Tester 1969,
Kufeld and Bowden 1995).

It is assumed that deer select habitats that optimize survival and ﬁtness
(McCullough et al. 1989, Hobbs and Hanley 1990). As a result, diﬂ‘erent habitat types
are utilized in diﬁ’erent seasons (Bender and Hauﬂer 1987). In addition, habitat use by
deer varies by sex (Dusek et al. 1989, McCullough et al. 1989, Beier and McCullough
1990), age (Dusek et al. 1989), and time of day (Kohn and Mooty 1971, Drolet 1976,
Murphy et al. 1985, Dusek et al. 1989, Kufeld and Bowden 1995).

Loss of forested habitat in the midwest region of the U. S. has resulted in
increased foraging on agricultural ﬁelds by deer (Gladfelter 1984). Agricultural openings
in forested landscapes provide both edge and high quality food. Agricultural crops are
commonly thought to comprise a large part of a deer’s year-round diet in Missouri
(Korschgen 1962) and parts of Montana (Dusek et al. 1989).

As with other habitats, use of agricultural lands by deer has been reported to vary
seasonally. One study (Austin and Umess 1993) reported the heaviest use of croplands
in early spring while others (Gladfelter 1984, Putnam 1986) have indicated heavy use of
grain residues throughout the fall and winter. Moreover, deer have been reported to
select crop types in different seasons in relation to various food and cover attributes that
each crop provides (Dusek et al. 1989). For instance, Murphy et al. (1985) found that
hay ﬁelds were used heavily during the fawning period and following mowing. Other

studies (Kohn and Mooty 1971, Putnam 1986, Dusek et al. 1988, 1989, Kufeld and

9
Bowden 1995) reported an increased use of agricultural ﬁelds from late spring-early
summer to fall.

Disproportional use of habitats in excess of availability deﬁnes selection or
preference and use below expectation (in relation to availability) denotes avoidance
(Kohn and Mooty 1971, Putnam 1986, Mooty et al. 1987, Dusek et al. 1989, Beier and
McCullough 1990). Some studies indicate that deer use agricultural lands in proportion
to their availability (Murphy et al. 1985, Putnam 1986). However, Dusek et al. (1989)
recorded cropland selection by deer despite preferred cover being greater than 1 km
away.

In some regions, differences in habitat use by male and female deer have been
observed (Dusek et al. 1989, McCullough et al. 1989, Beier and McCullough 1990).
Females are thought to forage in higher quality habitats due to the high energetic costs
associated with ovulation and lactation (Bowyer 1984, Beier 1987, Dusek et al. 1989,
Beier and McCullough 1990). In addition, fawn survival has been shown to be related to
the quality of diets of adult does in the winter and spring (Murphy and Coates 1976,
Langenau and Lerg 1976).

While research on deer ecology in agricultural landscapes has been conducted
(Hammerstrom and Blake 1939, Kohn and Mooty 1971, Larson et al. 1978, Murphy et
al. 1985, Nixon et al. 1988, Dusek et al. 1989, Nixon et al. 1991, Kufeld and Bowden
1995), the degree of potential impact that deer have on crop damage is still poorly
understood. Michigan State University began a comprehensive three part deer damage

research project in 1992 to examine the ecological, sociological and economic factors

1 0
inﬂuencing crop depredation in a region of Michigan with substantial reports of
agricultural crop damage. The ﬁrst project (Braun 1996) quantiﬁed damage levels to
three agricultural crops and examined landscape and habitat characteristics and land use
practices as possible factors impacting damage levels. The second project is the subject
of this thesis. The last project addressed the economic costs associated with crop
damage and examined the attitudes and perceptions of farmers and hunters regarding
crop damage.

The main purpose of this research was to examine seasonal movement and
habitat use patterns and home range characteristics of deer in a region of Nﬁchigan with
historical complaints of agricultural damage. The timing, distance and direction of deer
movements, the extent and timing of agricultural and non-agricultural habitat use and the
home range sizes of deer in agricultural landscapes were described in relation to their
potential inﬂuences on crop damage.

Population characteristics such as survival, productivity, age structure and sex
ratio dictate how a population changes over time (Caughley 1977). Estimation and
prediction of population size is an essential tool for managers in the manipulation of
wildlife populations. The control of crop damage requires an understanding of all
aspects that characterize deer biology and effect crop depredation. Population
characteristics of the deer herd were studied and the potential effects on current or future
crop damage are discussed. Lastly, recommendations regarding the future management
of crop damage as it relates to deer biology and behavior will be made based on

considerations of the above analyses.

OBJECTIVES

The speciﬁc objectives of this study were to:

1. Determine seasonal movement and habitat use patterns of white-tailed deer in an

agricultural landscape and evaluate potential inﬂuences on crop damage.

2. Quantify deer densities within the study area by mark-recapture techniques and

population indices.

3. Estimate survival rates, productivity, recruitment, sex ratio and age distribution

of the deer population in the northeastern lower peninsula of Michigan.

4. Make management recommendations to reduce deer damage to agricultural crops

based on deer biology and population dynamics.

11

STUDY AREA

The study was conducted during 1994 and 1995 in the northeastern lower
peninsula of Michigan. Presque Isle, Montrnorency, and Alpena counties (Figure 1)
were selected due to the high deer densities of approximately 4 to 19 deer per square
kilometer (T. Carlson, WNR, pers. commun.) associated with these counties. In
addition, these three counties received medium to high numbers of kill permits in

response to damage complaints submitted to the MDNR (Dudderar et al. 1989).

The climate of the northern lower peninsula of Michigan is more variable than
other regions of the state. Northern Michigan experiences decreased precipitation during
the growing season, cooler temperatures and greater winter snowfalls than southern
Michigan as a result of its proximity to the Great Lakes. The lacustrine climate
moderates temperatures and produces relatively long growing seasons due to slow spring
warming and slow fall cooling. The average growing season is 122 days long with a
range of 80 to 150 days. The average maximum summer temperature is moderate, 24.8
°C, and usually occurs in July, while the average daily winter minimum, - 10.8 °C,
normally occurs in February (Knapp 1988).

Average annual snowfall for the region is approximately 175 cm and average

12

13

Presque Isle

 

 

Montrnoren Alpena

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1. Study area comprised of Presque Isle, Montrnorency, and Alpena counties in

the northeastern lower peninsula of Michigan.

14
rainfall is 72.5 cm per year (Eichenlaub et al. 1990). Presque Isle and Montrnorency
counties receive slightly more rainfall than Alpena county with a mean annual
precipitation level of 77 cm. The average yearly temperature for the region is
approximately 6.2 °C, and the average daily temperature during the growing season
(May through September) was reported to be 15.9 °C (Albert et al. 1986).

Elevation in the study area is relatively consistent with a range of only 17 7 m to
427 m above sea level. The soil types are mostly composed of sand, loam, or glacial
deposit and are poorly to moderately well drained (Knapp 1988).

The majority of the land (65%) is forested, and approximately 20% of the region
is maintained in agricultural production (Knapp 1988). Conifer species such as northern
white cedar (77mja occidentalis), balsam ﬁr (A bies balsamea),white spruce (Picea
glauca), and jack pine (Picea banksiana) and deciduous species including quaking aspen
(Popular: tremuloides), maples (Acer spp.), basswood (T ilia americana), and oaks
(Quercus spp.) are common throughout (Albert et al. 1986).

The study area has several large regions that are comprised of more than 75%
agricultural lands. Forested lands almost completely surround these agricultural areas
creating agricultural “islands” in a forested landscape (T. Carlson, MDNR, pers.
commun.).

A variety of crop types are cultivated in the three counties. Low to medium
production of beans, corn, and hay in all three counties have been recorded in the study
area. High potato and low fruit tree production are reported for Presque Isle county

(Dudderar et al. 1989).

METHODS

Capture and handling

Deer were trapped in all three counties during the winter months of 1994 and
1995. The trapping period began in mid-to late January and ended in early April during
both years. Single-gate, collapsible Clover traps (Clover 1954, 1956, McCullough 1974)
were selected for trapping because of their mobility, ability to be used in remote areas
(Rongstad and McCabe 1984), and minimal personnel requirements for operation.
Additional minor modiﬁcations, including the incorporation of a rabbit bar (Roper et al.
1971) along the back of selected traps, were made to increase trapping eﬂiciency.
Shelled corn was the most frequently used bait, but cullings of farm crops, hay and
apples were occasionally used to increase capture rates. Trap sites were baited for two
to four days before setting the traps.

Summer trapping of deer was attempted at three locations within Montmorency
and Alpena counties from July 25 to August 16, 1994. Methods followed winter
trapping with the exception of using salt as the primary bait.

Traps were placed in wooded areas immediately adjacent to agricultural ﬁelds

with reported crop damage whenever possible. In addition, two hunting clubs that

15

16
bordered agricultural lands, were centrally located within the study area, and had large,
well established, supplemental deer feeding programs, were chosen as trap sites.

Traps were placed along deer travel routes in cedar swamps, to provide thermal
cover and protection from wind (Hammerstrom and Blake 193 9, Venue 1973). Trap
disturbance and subsequent deer stress were minimized by placing traps away ﬁom
coyote (Cam's Iatrans) travel routes or snowmobile trails. New trap sites were selected
when the incidence of successful trapping diminished to less than one deer captured in a
two to three week period.

Deer were manually restrained upon capture to collect sex and age information.
Handling procedures followed the methods of McCullough (1974). Deer were aged
based on the characteristic wear and replacement patterns of the molariform teeth on the
lower jaw (Severinghaus 1949, Ryel et al. 1961). Tooth eruption and replacement
patterns were used to classify deer as fawns, yearlings, or adults. In addition, adult deer
were classiﬁed as two, three, four, ﬁve, six, or older than six years, based on the wear
patterns of the lower pre-molars and molars. Inspection of adult deer jaws upon
mortality allowed for corrections of age. Following mortality, the central lower incisor
was removed in deer determined to be older than three years of age for cementum
analysis (Gilbert 1966, Ransom 1966, DeYoung 1989). Cementum analysis was
performed by Matson’s Laboratory (Milltown, Montana) for deer in 1994 and Rose
Lake Wildlife Research Station (East Lansing, Michigan) for deer mortalities in 1995.

Supplemental age information was gathered following these aging methods for all

road-killed deer encountered within the study area. Additional data were collected ﬁ'om

17
a subsample of harvested deer within the study area in 1994 and 1995. A centrally
located check station was established (for the ﬁrst ﬁve days of ﬁrearm season) where
hunters could voluntarily bring their kill to be aged. These data were used as
supplemental age data due to the biases associated with their collection (Coe et a1.
1980)

All captured deer were marked using one or two colored, serially numbered ear-
tags (National Band and Tag Co. Newport, Ky.). Distinct combinations of color and tag
placement (left or right ear) allowed for identiﬁcation of individual deer trapped in 1994.
Diﬂ‘erent colored tags were used to mark diﬂ‘erent sex and age classes of deer captured
in 1995. Date, time, location, and habitat type were recorded from sightings of ear-
tagged deer.

Newborn fawns that were encountered early in the fawning period each year
were captured by quick, noisy approaches that elicited the “drop” or “ﬁ'eeze” response
(Downing and McGinnes 1969, Nelson and Woolf 1987, Ozoga et al. 1994). Fawns
were ear-tagged according to the above protocol, sexed and released. Researchers
minimized handling time and immediately left the area to lessen the chances of
abandonment.

Radio collars (Lotek Inc. Ontario, Canada) were issued to a subsample of the
trapped deer in each year to determine movement, habitat use and home range. Collars
were equipped with 7-hour motion-sensitive mortality sensors, had a nrinirnum battery
life of three years, and weighed less than 0.453 kilograms. Fawn radio collars were

identical to adult collars except for the attachment of a bio-degrading foam lining on the

18
inside to prevent the collar from slipping before the fawn reached adult size. All collars
were marked with an appropriate address and phone number to facilitate return of the
collars.

Radio collars were distributed among all age and sex classes of captured deer and
collars were stratiﬁed across all three counties. Deer selected to be radio collared
appeared physically healthy and were not experiencing obvious severe nutritional
deprivation. Radio collars were ﬁtted so that the thickness of one hand (at the palm)
could rest between the deer’s neck and the collar. Collars recovered ﬁom deer
mortalities or slippage in 1994 were re-issued during the 1995 trapping season. All
trapping, handling, and marking procedures were reviewed and approved by the All-

University Committee on Animal Use and Care (AUF # 01/94-024-01).

Radio telemetry

Radio collared deer were located using a hand-held receiver (Lotek Inc. Ontario,
Canada) and a two (Telonics Inc. Mesa, Ariz.) or three (Advanced Telemetry Systems
Isanti, Minn.) element Yagi antenna. Collared deer were located at least twice a week
from the time of capture until collar failure or death. The location period began in
February, 1994 and continued through mid-December, 1996. Locations were stratiﬁed
across three 8-hour sampling periods (0800-1559, 1600-2359, 0000-0759 hours)
because deer activity and habitat use patterns vary throughout the day (Montgomery

1963, Larson et al. 1978, Beier and McCullough 1990).

19

Location data were gathered by recording two or more azimuths for each deer
from known map locations. Whenever possible, azimuths considered to be erroneous
due to signal bounce were corrected in the ﬁeld. An azimuth standard deviation was
determined for all personnel performing radio telemetry. Locations were plotted using
LOCATE H (Pacer, Truro, Nova Scotia). Conﬁdence ellipses were estimated for
locations with two or more azimuths (N ams 1990) using the maximum likelihood
estimator procedures of LOCATE II and the mean azimuth standard deviation.
Locations with 95% conﬁdence ellipses greater than 200 hectares (ha), were determined
to be unreliable or incorrect and were eliminated from data analysis (White and Garrott

1990)

Migration, home range, and habitat use

The initiation date, direction, destination, and distance of seasonal migrations
were analyzed for radio collared deer. In addition, reliable movement information on
three ear-tagged deer was known and included in analyses of collared deer migration.
Deer were considered migratory if their winter and summer ranges did not overlap and
were greater than 1 km apart. Migration distance for each deer was measured as the
distance between centers of activity of seasonal ranges. Destination of seasonal
migrations was categorized as agricultural land or non-agricultural, forested land. The
start of migration for an individual was the day it leﬁ its current seasonal range and
began movement toward its reciprocal seasonal range. If a deer was not located on the

day it began migrating, the day halfway between its last known location on its current

20
range, and its ﬁrst location on its reciprocal range was assigned. Transitional migratory
locations were not included in either seasonal home range. The time period when most
deer migrations occurred was considered the peak period. Seasonal migration dates
were compared between years using the Kolmogorov-Smimov test (Sokal and Rohlf
1995). Migration distances of deer were compared by sex and destination category
using the Wilcoxin-Mann-Whitney test (Siege! and Castellan 1988). Signiﬁcance for
these and all other statistical tests was set at 0.05.

Home ranges of collared deer were determined using the harmonic mean method
of Dixon and Chapman (1980) with 20 grids speciﬁed. The 95% contour was used to
eliminate outliers in calculating size and harmonic mean center of activity for each deer
using Telem88 software (Department of Wildlife Ecology, University of Wisconsin).
The minimum number of locations (5) required to estimate home range size by Telem88
was not thought to adequately represent deer home ranges, therefore home ranges were
only calculated for deer having 10 or more radio-locations. Traditional use, or ﬁdelity to
seasonal ranges, was analyzed for deer that were radio-tracked for longer than one year.
Range use was considered traditional if an individual’s seasonal ranges overlapped in
successive years. Home range size was compared between years, seasons, migratory
groups, migratory destinations, and sexes using analysis of variance models (SAS
Institute 1985). Normality was tested using the (studentized range) w/s test (Kanji
1993) and Bartlett’s test for homogeneity of variances (Sokal and Rohlf1995) was used
to test for equal variances. Home range size is not dependent upon year (correlation

coeﬂicient = 0.031, P > 0.50) therefore, yearly samples were considered independent.

2 1
Pairwise comparisons of seasonal ranges of individual migratory deer were made using
the ercoxin signed-ranks test (Sokal and Rohlf1995).

Habitat use was derived ﬁ'om deer locations. Deer locations were converted to
point coverages using the geographic information system PC ARC/INFO (ESRI
Redlands, Calif). Base land coverages of PC ARC/INFO were digitized by MDNR
personnel from 1978 aerial photographs and designated as Michigan Inventory Resource
System (MIRIS) categories. Deer locations were assigned a MIRIS code upon
intersection of point coverages with base coverages within PC ARC/INFO. The MIRIS
code designation was also used to record cover type from visual observations of tagged
or radio-collared deer.

Due to the age of PC ARC/INFO base land coverages, habitat type of all deer
locations was ground truthed in the summer of 1995. Aerial and ground surveys were
conducted to identify changes in habitat occurring since 197 8. Boundary coordinates of
altered habitats were delineated and used to ground truth future locations.

Determination of habitat use can be biased by large telemetry conﬁdence ellipses
(White and Garrott 1986, Nams 1989). Therefore, potential bias was estimated by
comparing mean telemetry error, determined from deer locations, with mean habitat
patch size, computed ﬁ'om sampling all habitat patches within three randomly selected
townships where radio collared deer were located.

Habitat selection was determined using deer home ranges and PC ARC/INFO.
Deer home ranges were positioned on habitat base maps using centers of activity and

95% contours, determined ﬁ'om Telem88. The areas of various habitat types composing

22
an individual’s seasonal range were calculated using PC ARC/INFO. The null
hypothesis of random use of habitats was tested for each deer using a Chi square
goodness of ﬁt test comparing an individual’s proportional use of habitats (as determined
from locations) with the proportional availability of those habitats within its home range.
Random habitat use was also tested within an extended home range (an area 1.5 times as
large as the individual’s home range) to compensate for potential underestimation of deer
home range sizes. Comparisons were made on individual deer since deer had different
habitats available to them. General trends in habitat use patterns were determined by
analyzing selection (use greater than expected) or avoidance (use less than expected) of
habitat types among deer groups. Overall statistical signiﬁcance of habitat use patterns
for each deer group was tested by combining probabilities ﬁ'om independent tests of ‘
signiﬁcance (Sokal and Rohlf 1995).

Habitats were grouped and classiﬁed into eight categories; cropland, northern
and central hardwoods, lowland hardwoods, aspen and white birch, lowland conifers
(primarily northern white cedar, black spruce and tamarack), pine, non-forested
openings, and other. The ‘other’ category is comprised of habitat types that commonly
made up 1% - 5% of deer locations (scrub shrub wetland, pasture land, and old
meadow). Habitat use was compared among the three daily time periods, and between
seasons, years, sexes, migrating and non-migrating deer groups, and deer migrating to
agricultural or non-agricultural areas. Signiﬁcance was tested using the Chi square test
for two independent samples (Siegel and Castellan 1988) which assumes independent

observations, no expected values equal to zero, and less than 20% of categories with

23
expected values less than ﬁve. The experimentwise error rate for each type of
comparison was set using the Bonferroni method (Sokal and Rohlf1995) where the

critical value was determined by or divided by the number of comparisons per category.

Population parameters
Abundance

Population size was estimated from mark-recapture data using the CAPTURE
program (Department of Fishery and Wildlife Biology, Colorado State University).
CAPTURE determined model selection and trapping occasions were deﬁned as three day
intervals. The main assumption of this closed population model was that population size
remained constant throughout the duration of the study. This assumption was relaxed so
that the population was unchanging and no unknown losses occurred during yearly
trapping periods (Otis et al. 1978, White et al. 1982). Furthermore, it assumes that no
animal marks were lost or unnoticed during the experiment, all animals have constant
and equal probability of capture on each capture occasion (Otis et al. 197 8), and
mortality is equal for marked and unmarked individuals.

Population indices were used to estimate relative diﬂ‘erences in population
abundance between years (Lancia et al. 1994). Pellet groups counts were used to
compare abundance between years and standardized roadside observations were used to

compare trends within years.

24

Overwinter estimates of population size were determined from pellet counts
(Bennett et al. 1940, Eberhardt and Van Etten 1956, Neﬁ‘ 1968, Ryel 1971) conducted
aﬁer snow melt in the spring of each year. In 1994, eighty random points (plots),
stratiﬁed by habitat type (hardwoods, pine, cedar, open, and aspen), were established in
the study area. Plots established in 1994 were used in 1995. At each point, a 1m x 25m
transect was sampled for pellet groups in each of the cardinal directions. A count of the
number of pellet groups deposited since autumn leaf fall along the four transects at each
point was summed. An average number of pellet groups was obtained for each of the
ﬁve strata.

An estimate of population size (N) is determined for each strata by the following

formula (Mooty 1980):

117 _ ng * (plot size) “ * size of the study area
deposition period * defecation rate * number of plots

 

where

ng = average number of pellet groups per plot.

The deposition period represents the average number of days between leaf drop
in the previous fall, and sampling date for plots. Defecation rate varies with diet, sex and
age (Neﬁ‘ 1968, Ryel 1971, Rogers 1987), but winter, ﬂee-ranging deer defecate
approximately 13 to 14 groups per day (Neﬂ‘ 1968). A rate of 13.47 groups per day is

used by the MDNR for deer in Michigan (H. Hill, MDNR, pers. commun.) and is the

25
value used here. An overall density per county was obtained by weighting deer densities
by the percent of each strata per county.

Assumptions for a pellet group based census include: 1) defecation rate is
constant and known, 2) pellet groups exist long enough to be counted and are detected
and counted accurately, and 3) a deposition period can be delineated and groups can be
aged relative to the deposition period (Ryel 1971, Mooty 1980). Differences in pellet-
group densities between years were tested using the Wilcoxin-Mann-Whitney test.

Roadside observations of deer were recorded along a 180 km standardized route
established in the study area (Appendix Figure 1). The route was designed in a ﬁgure
eight pattern that encompassed both agricultural and non-agricultural areas throughout
the study area. The route was driven beginning two hours before dusk on two
consecutive nights from February to December in 1994, and approximately one hour
before dusk on three consecutive nights from January to October in 1995. Road surveys
were conducted twice a month during summer months (June, July, and August). The
direction the route was driven (north to south, or south to north) was alternated on a
monthly basis. Routes were not driven on evenings when poor weather such as rain,
blizzards, thunderstorms, or high winds were likely due to possible effects on deer
behavior (Zagata and Haugen 1974, Gladfelter 1980). Total number of deer seen was
recorded and deer were classiﬁed into sex and age categories (adult and sub-adult)
during months when they were discernible (Downing et al. 1977). Deer of unknown sex

or age were also recorded and included in total counts.

26
Roadside observations were originally intended to be an index of abundance,
however, the driving procedure was altered in the second year to facilitate obtaining
more accurate age and sex ratios. Therefore, monthly counts were used to indicate

trends in population ﬂuctuation within years.

Survival

Radio telemetry allows for direct estimation of survival patterns by providing
information on the timing of mortality. The Mayﬁeld survival estimator (Mayﬁeld 1961,
1975) was used to estimate survival patterns of radio collared deer. A staggered entry
approach whereby exposure days accumulate as new individuals are added (similar to the
Kaplan-Meier product limit estimator staggered entry design discussed by Pollock et al.
1989) was used. The estimator requires that newly added animals have the same survival
probability up to entering the study as previously added animals. In addition, it assumes
that animals were sampled randomly, survival events are independent and constant across
animals and periods, capturing and radio collaring have no impact on future survival of
the animal, and censoring of animals (individuals with unknown fates) is independent of
deaths.

When a radio signal became lost, ground searches were made in increasingly
larger concentric circles around the last known location in attempts to locate it.
Following this, a ﬁxed wing aircraft was employed to search for the lost signal on several

occasions before the individual was classiﬁed as censored.

27

Annual and period survival rates, with 95% conﬁdence intervals, were
determined for collared deer in three categories; adults, yearlings, (deer born in that
calendar year) and both age classes combined, due to differential survival of yearlings
and adults reported by Nelson and Mech (1986) and Fuller (1990). Three periods were
established in which survival of deer was considered to be constant (Fuller 1990); the
winter or post-harvest interval (ﬁ'om January 1 through April 15) encompasses the
period of high overwinter mortality due to starvation (Fuller 1990) and predation
(Nelson and Mech 1986), the summer or pre-harvest interval (April 16 through
September 30), and harvest period (October 1 through December 31) which includes all
hunting seasons in Michigan. Additional survival information was gathered from the
observed mortality of six ear-tagged deer. Survival rates were compared between years,
age classes, and periods. Chi square 2 X 2 contingency tables and the Bonferroni
method (described above in the habitat use section) were used to determine signiﬁcant
diﬁ’erences.

Causes of mortality for radio collared deer that died during the course of the
study were determined whenever possible. Carcasses were located and closely inspected
upon detection of mortality. Sources of mortalities were classiﬁed into seven categories;
natural mortality (including predation, disease and drowning), legal and illegal harvest,
road kill, starvation, crop permit harvest, and unknown causes.

Predation was considered the cause of mortality when there were signs of a chase
(blood and hair trail) and tooth punctures on the neck and throat (Ozoga and Harger

1966, Messier and Barrette 1985). Coyote and domestic dog (Canisfamiliaris)

28
predation were combined since they were not reliably distinguishable despite differences
in hunting behaviors (Lowry and McArthur 1978). Condition of the femur marrow, in
conjunction with amount of fat on the mesentery and other connective tissues, were used
as indicators of malnutrition (Venue and Ullrey 1984, B. Odum, WNR, pers.
commun.). Mortality of collared deer found in the winter or spring that showed signs of
predation and malnutrition were classiﬁed as unknown to avoid erroneous inﬂation of

either category.

Productivity

Adult radio-collared does were repeatedly observed throughout the fawning
period (late May - end of June) in each year to estimate productivity. Does that
produced fawns were observed again in late August to estimate fall-recnritment fawn to
doe ratios (Dusek et al. 1989). Based on the results of a previous study (Hamlin et al.
1982), it was assumed that productivity of does was not affected by trapping and
handling.

Estimates of productivity were also calculated from deer counts recorded along
monthly road surveys (see the above section on abundance). Deer were classiﬁed into
age categories (adult and sub-adult) during months when fawns and does were easily
discernible to determine fawn to doe ratios (Downing et al. 197 7). Counts during late
summer and early fall surveys were used to estimate productivity and late fall surveys

were used to estimate fall-recruitment fawn to doe ratios. In addition, doe counts were

29
decreased by 10% (to allow for possible inﬂation of counts from small-antlered bucks)
and “adjusted” ratios were estimated.

Incidental observations of does with fawns recorded during the summer months
of each year were used to supplement productivity estimates. Fawn to doe ratios of the
captured deer were used to supplement fall recruitment ratios determined ﬁom road
surveys. The Wilcoxin-Mann-Whitney test was used to test for signiﬁcant diﬂ‘erences in

ratios between years.

Sex ratio and age structure

Sex ratios were estimated from the observed counts of male and female deer
recorded on monthly road surveys during months when both sexes were considered
equally observable (Downing et al. 1977). Adjusted ratios (see above) were also
calculated from deer counts. The Wilcoxin-Mann-Whitney test was used to test for
signiﬁcant differences in ratios between years. The ratios of the captured deer were used
to supplement sex ratio data.

Age structure was determined from the age composition of the trapped deer.
Captured deer were classiﬁed into ﬁve age categories for age structure calculations;
fawns (less than one year old), yearlings (one year old), 2 - 3 year-olds, 4 - 5 years-olds,
and older than six years. Age classes were combined to reduce the impact of incorrect
age determination on age structure estimates. Age or sex biases inherent in single animal
capture traps (Garrott and White 1982) may introduce bias into sampled populations,

therefore, analyses using these data were considered cautiously. Age composition data

30
gathered from harvested deer at a deer check station (centrally located within the study
area) and road killed deer encountered in the study area during the study were compared
to the age composition of trapped deer. No statistical comparisons of age structure data

were made because of the biases inherent in their collection.

RESULTS

Over the two years of the study, twenty-three winter traps sites were established
(Figure 2). Deer were ear-tagged or radio collared at 14 out of 16 traps in 1994 and 11
out of 17 traps in 1995 (Appendix Table 1). Captures totaled 190 deer during the two
years; 117 different individuals, and 73 recaptures. Ten deer were not marked due to
mortalities, or potential injuries. Thus, 107 of the trapped deer were marked with radio
collars or ear-tags. Radio collars were distributed to 73 deer and 34 additional deer
were ear-tagged during the two years of the study (Table 1). In addition, eight newborn

fawns (four males, four females) were ear-tagged during the two years.

Winter trapping success increased from approximately 33% in 1994 to
approximately 47% in 1995 (Appendix Table 2). Trapping mortality was minor during
both 1994 (3.7%, N = 4) and 1995 (1.2%, N = 1). Summer trapping resulted in no
captures in 1994 and was not attempted in 1995.

During the study period, 4,172 locations were collected for radio collared deer.
Most locations (72.9%) were gathered during the ﬁrst time period (0800 - 1600 hours)
of the day (Appendix Table 3) and deer were located approximately twice as often

during summer months (Appendix Table 4). The average number of locations obtained

31

 

 

32

 

 

 

 

 

9
60 Presque Isle
45
r* 7580
6B
90 3A 65
Montrnorency Alpena

 

Figure 2. Distribution of traps across study area with numbers indicating individual trap

location. Trap designations refer to numbers listed inAppendix Table 1.

33

Table 1. Age and sex of deer marked in Presque Isle (PI), Montrnorency (M), and
Alpena (A) counties in 1994 and 1995. Table entries represent number of collared deer

and numbers in parenthesis represent number of ear-tagged deer.

 

Female ' Male

 

Year County Adult Yearling Fawn Adult Yearling Fawn Total

 

1994
PI 15 1 11 (6) o 1 11 (4) 39 (10)
M o 1 2 o o 1 4
A o 1 o o o o 1
1995
PI 2 (5) o 4 (6) o 2 5 (4) 13 (15)
M 4 (1) o 1 (2) o 0 3 8 (3)
A 2 0(1) 1 (3) o o 5 (2) 8 (6)
Total

23(6) 3(1) 19(17) 0 3 25(10) 73(34)

 

34
for deer surviving longer than 30 days was 75 with a standard error (SE) of 0.81 (N =
54). The mean 95% conﬁdence ellipse area of triangulated deer locations was 37.2 ha

(SE = 0.02 ha). The mean azimuth standard deviation for triangulation was 9.4 °.

Migration

Most of the collared deer (67.5%) in 1994, and approximately half(48.8%) of
the collared deer in 1995 made spring migrations (Table 2). In addition, three ear-tagged
deer were sighted or harvested ﬁom 2 to 7 km from where they were trapped in 1995
and were considered migratory. Spring migrations for collared deer began
approximately two weeks later in 1994 than in 1995 (Figure 3). The Kolmogorov-
Smirnov test determined that the distribution of spring migration dates differed between
the two years (P < 0.01). The median date for the last day on winter ranges was April 8
(N = 27) in 1994 and March 27 (N = 22) in 1995. Most migratory collared deer
(>80%) left winter ranges before May lst in both years. Generally, spring migrations
were either in a northwestern or southwestern direction during both years (Figure 4,
Figure 5).

The average migration distance of collared or tagged deer in 1994 was not
signiﬁcantly diﬂ‘erent (\Vrlcoxin-Mann-Whitney 2 value (2 mm) = - 0.424, P = 0.337)
from the 1995 distance (Figure 6). Males tended to travel greater distances between
winter and summer ranges than females, however, distances were not signiﬁcantly
diﬁ‘erent in either year (Table 3).

More than halfof spring migrating collared deer migrated to non-agricultural,

35

Table 2. Migratory status of collared deer by sex and age category for spring 1994 and

1995. Numbers in parenthesis indicate deer collared in 1994.

 

Male Female

 

Year Classiﬁcation Adult Yearling Fawn Adult Yearling Fawn Total

1994 Migratory 0 1 10 9 1 6 27
Non-migratory 0 0 1 5 2 5 13
1995 Migratory 0 (2) 4 2 (7) (2) 3 20
Non-migratory (1) 2 5 4 (5) (2) 2 21

Total 1 S 20 32 7 16 81

 

36

 

 

 

 

 

 

 

 

 

 

 

 

 

10 -
8 _
I-
6 _
“g El 1995
“a I 1994
r.
.8
E
z 4 —
2 i F _
0 I I I I I II I
15- 29- 8- 22- 1- 15- 29- 8-
Mar Mar Apr Apr May May May Jun
Date

Figure 3. Initiation of spring migration for collared deer in 1994 and 1995.

37

 

 

 

 

 

 

 

Figure 4. Distance and direction of spring migrations of collared deer in 1994. Circles

represent trapping location and arrows represent movements for one or more deer.

38

 

 

 

r.—

 

 

 

Figure 5. Distance and direction of spring migrations of collared deer in 1995. Circles

represent trapping location and arrows represent movements for one or more deer.

39

 

 

 

 

 

 

 

 

 

 

14 —
12 — l
10 ~
8
.2
'6
Q)
a 6—
I-
0
>
<1
4 _
2 n
0
1994 1995
Year

Figure 6. Average spring migration distance of collared and tagged deer in 1994 and

1995. Error bars represent one standard error.

40

Table 3. Average migration distances (km) of collared and tagged deer by sex in 1994

 

 

 

 

and 1995.
Males Females P value‘
’7 SE n 8 SE n
1994 13.67 3.25 11 8.05 1.78 16 P = 0.090
19955 11.67 4.65 6 7.91 2.11 6 P = 0.999
‘ Wilcoxin-Mann-Whitney

" includesdeercollared in 1995 only

41
forested lands in 1994 and 1995. The remainder of the migratory deer established
summer ranges in neighboring agricultural areas (Figure 7). Deer that migrated to non-
agricultural forested land in the spring traveled greater average distances than deer
migrating to agricultural areas in both years (Figure 8), however differences were only
signiﬁcant in 1994 (2 W = 3.74, P < 0.0001).

The median last day on summer ranges was November 29 (N = 11) in 1994 and
November 19 (N = 12) in 1995 (Figure 9). In addition, a deer ear-tagged in 1995 was
known to return to its winter range by November 22. The Kolrnogorov-Smirnov test
determined that the distributions of 1994 and 1995 fall migration dates were not
signiﬁcantly diﬁ‘erent (P > 0.05). The sample size of fall migrating deer was small in
both years because many collared deer died during the summer and early fall. The
majority of collared deer that migrated back to winter ranges in the fall were females
(Table 4) because most collared males were harvested during the bow or ﬁrearm hunting

seasons, prior to fall migration.

Home range

Non-migratory deer established yearly home ranges in agricultural areas. The
average home range size of non-migratory deer was larger than for migratory deer in
both 1994 and 1995. Likewise, females had average home range sizes greater than male
home ranges during 1994 (Table 5). However, analysis of variance model results of
home range size indicated no signiﬁcant interactions or differences between years,

migratory status of deer, or sex (Table 6).

42

60% -

 

El Non-agriculture

50% _ I Agriculture

 

 

40% 2

30% -

Percent

20% ~

10% -

 

 

 

 

 

 

0% r
1994 1995

 

Year

Figure 7. Destination of spring migrating collared and tagged deer in 1994 and 1995.

43

 

 

 

*
16 7
14 r
12 -‘
10 r
* .1994
C1 1995

Distance (km)
00

 

 

 

 

 

 

 

 

 

6 z
4 _‘ ,.
2 _
0 _
Non-
agriculture

Destination

Figure 8. Average distances of spring deer migration (error bars represent one standard
error) to agricultural and non-agricultural land in 1994 and 1995. Asterisks indicate

signiﬁcant diﬁ’erences in destination migration distance within years.

—A

. ﬁrearm season

  

 

 

 

3.

l-

0

€21

"5 I1994
g .1995
:1“

Z

0 r .3
1-01 l-er lS-bbv 2.9-be 81h ﬂ-Ibc

1MB

Figure 9. Initiation of fall migration for collared deer in 1994 and 1995.

45

Table 4. Sex and age classiﬁcation of fall migrating collared deer in 1994 and 1995.

 

 

Year Age category Male Female Total
1994 Adult 1 7 8
Yearling 2 2 4
1995' Adult 0 10 10
Yearling o 1 1
Total 3 20 23

 

" includes deer collared in both years.

Table 5. Average home range size (ha) of collared deer in 1994 and 1995.

46

 

 

 

1994 1995
Classiﬁcation Average SE N Range Average SE N Range
(ha) (ha)

Migratory 302.8 30.7 27 524-7744 336.6 36.6 28 806-7719
Non-migratory 424.0 26.1 12 292.3-579.3 355.8 36.8 23 502-6671
Male 293.3 38.0 12 52.4-468.9 340.7 30.1 16 31.14719
Female 367.9 33.5 28 71.8-833.0 335.7 57.3 39 30.8-7649
Migratory- 201.8 64.1 4 753-3160 353.6 56.5 9 164.0-657
winter

Migratory. 336.9 32.9 22 71.8-774.4 328.6 47.8 19 806-7719
summer

M‘Eamry' 264.2 46.1 16 52.4-316.0 296.0 44.2 15 105.1-657
agriculture

Migratory- non- 354.3 30.4 12 1790-4747 383.6 59.5 13 80.6-7719

agriculture

 

47

Table 6. Analysis of variance results of main eﬂ‘ects and interactions of home range size

by year, migratory status' , and sex.

 

 

df F value P value
Main Effects
Year 1 0.00 0.9566
Migratory status 1 0.69 0.4074
Sex 1 1.37 0.8016
Interaction

Year x Migratory status 1 0.06 0.2454
Year x Sex 1 0.1 0.4772
Migramry Status x 53" 1 1.00 0.3213
Year x Migratory status 1 0.27 0.6044
x Sex

 

‘ migratory or non-migratory

48

Migratory deer tended to have smaller winter ranges than summer ranges in
1994. In 1995, the average size of migratory deer winter ranges was larger than summer
ranges (Table 5). Furthermore, deer migrating to non-agricultural, forested regions
established larger home ranges than deer migrating to agricultural areas during both
years. However, these diﬁ‘erences were not signiﬁcant and no signiﬁcant interactions
were detected for home range sizes of migratory deer between seasons or migratory
destinations (Table 7). In addition, there were no signiﬁcant interactions between year
and season (df= 1, F = 2.34, P = 0.1322). However, pairwise comparisons of seasonal
ranges of individual deer in 1994 resulted in 72.7% (N = 11) of evaluated deer having
smaller winter than summer ranges. The Wilcoxin signed-ranks test determined 1994
seasonal ranges to be signiﬁcantly different from each other (P = 0.0332). However,
while 72.7% (N = 11) of individuals in 1995 had larger winter ranges, seasonal ranges

were not signiﬁcantly diﬂ‘erent (P = 0.2060).

Approximately 49% (N=18) of migratory radio collared deer survived to be
evaluated for traditional use of seasonal ranges. Winter ranges in successive years
overlapped for 72.2% (N=l3) of the evaluated deer, indicating traditional use. In
addition, one ear-tagged deer, known to be migratory, used the same winter range in
successive years. Four collared yearlings (three males and one female) did not return to
winter ranges in 1995; two remained on summer ranges established as fawns and two
migrated part of the return distance. Furthermore, one adult female collared in 1994,

wintered 19

49

Table 7. Analysis of variance results of main eﬁ‘ects and interactions of home range size

by season and migratory destination' of deer.

 

 

Category df F value P value
Main Eﬂ‘ect
Season 1 0.85 0.3614
Migratory destination 1 0.38 0.5840
Interaction
Season x Migratory 1 2.25 0.1394
destination

 

‘ agricultural or non-agricultural (forested)

50
km beyond her original winter range in the following year. Thirteen collared deer were
evaluated for traditional summer range use. Traditional summer range use was higher
(92.3%) than for winter ranges with all but one adult male returning to the same summer

range in successive years.
Habitat use

Habitat use was signiﬁcantly different across sampling periods of the day (df = 14
x2 = 89.7, P < 0.002). Unequal sampling intensity among periods resulted in under-

estimation of cropland use during the night period (0000 - 0759 hours) (Appendix Table
5). Therefore, a random sub-sample of 400 deer locations taken within each time period
was used to make habitat use comparisons. Habitat categories in several of the
individual comparisons had to be combined to meet the assumptions of the Chi square
test, thus reducing the degrees of freedom. Most of these combinations involved pooling
cedar and pine categories or ‘other’ with lowland hardwoods or cropland categories.
Table 8 lists the percent of deer locations in each habitat type by sampling period of the
day, season, migratory status and destination, and sex.

The habitat use patterns of all deer pooled were signiﬁcantly different between
years. Seasonal habitat use was not signiﬁcantly different at a = 0.05 however,
differences were signiﬁcant at or = 0.10 (Table 9). Diﬂ‘erences in use patterns were
signiﬁcant between years for non-migratory deer, and between migratory and non-
migratory deer in 1994 (Table 9). While there were differences in habitat use patterns of

migratory and non-migratory deer in 1995, they were not statistically signiﬁcant (P =

51

Table 8. Percent of deer locations within each habitat type by sampling period ', season,

migratory status, migratory destination", and sex in 1994 and 1995.

 

 

 

Habitattypes
Yr Category Agri. Open Upl. Asp/Bir Lowl. Pine Cedar Other
Hdw Hdwd
94 Sampling Day 14.4 5.0 15.0 32.8 9.4 2.8 13.3 7.2
period
Evening 13.7 6.1 21.7 33.5 6.1 3.8 9.9 5.2
Night 24.3 8.4 18.2 29.0 2.8 3.7 9.8 3.7
Season
Winter 20.0 7.2 12.8 28.8 10.4 0.8 16.0 4.0
Summer 17.0 6.4 20.0 32.4 4.8 4.2 9.6 5.6
Migr. status
Migratory 16.6 5.2 18.4 35.8 6.2 5.2 9.8 2.8
Non-migrat 19.5 9.1 18.6 24.5 5.5 0.5 12.7 9.5
Destination
Non-agric. 0.9 2.4 18.0 55.3 4.4 7.8 10.2 0.9
Agric. 34.4 8.3 18.9 13.3 8.3 2.2 9.4 5.0
Sex
Male 5.4 4.3 18.5 42.4 4.3 8.7 15.2 1.1
Female 23.0 7.6 18.5 27.0 6.6 1.2 9.0 7.1
95 Sampling Day 8.2 7.7 15.5 30.0 11.4 4.1 19.5 3.6
period
Evening 10.6 4.3 21.8 33.0 4.8 1.6 20.7 3.2
Night 17.9 7.1 21.2 26.6 10.9 1.6 12.5 2.2
Season
Winter 12.2 6.1 18.9 34.4 5.6 2.2 2.8 17.8

Summer 12.3 6.8 20.7 29.0 11.1 2.8 14.1 3.8

52

Table 8 (cont’d).

 

 

Habitat types
Yr Category Agri. Open Upl. Asp/Bir Lowl. Pine Cedar Other
Hdwd Hdwd

 

Migr.status
Migratory 10.8 6.4 22.4 28.8 9.2 3.7 14.2 4.4

Non-migrat 13.0 6.4 16.1 30.8 9.0 1.3 21.1 2.3

Destination
Non-agric. 1.9 2.6 20.9 37.9 7.2 5.9 19.0 4.6
Agric. 22.1 11.5 19.8 19.1 12.2 0.8 9.9 4.6
Sex
Male 4.3 5.6 14.3 33.5 9.9 4.3 27.3 0.6
Female 14.8 6.7 21.0 28.4 8.8 18 14.1 4.4

 

‘ day = 0800-1559 hours, evening = 1600-2359 hours, night = 0000-0759 hours
" winter and summer locations combined

53

Table 9. Results of Chi square analysis of habitat use of radio collared deer. Asterisks

indicate signiﬁcance.

 

 

Chi
Type‘ Comparison df Square P value
value
Year 1994 vs. 1995 7 24.11 0.0022*
Season 1994: winter vs. summer 7 16.50 0.0441
1995: winter vs. summer 7 16.06 0.0505
Migratory 1994: migratory vs. non-migratory 7 31.64 <0.0001*
status
1995: migratory vs. non-migratory 7 13.05 0.1496
1994 winter: migratory vs. non-migratory 5 10.91 0.1089
1994 summer: migratory vs. non-migratory 7 35.58 <0.0001*
1995 winter: migratory vs. non-migratory 5 6.34 0.6736
1995 summer: migratory vs. non-migratory 7 18.97 0.0173
Migratory: 1994 vs. 1995” 7 19.02 0.0170
Non-migratory: 1994 vs. 1995b 7 35.50 <0.0001*
Migratory 1994: agriculture vs. non-agriculture 7 132.50 <0.0001*
destination
1995: agriculture vs. non-agriculture 7 56.35 <0.0001*
1995 winter: agriculture vs. non-agriculture 3 6.02 0.2476
Agriculture: 1994 vs. 1995” 3 7.40 0.1265

Non-agriculture: 1994 vs. 1995b 5 14.18 0.0290

 

54

Table 9 (cont’d).

 

 

Type‘ Comparison df Sign P value
value
Sex 1994: male vs. female 7 69.28 <0.0001*
1995: male vs. female 7 34.39 <0.0001*
1994 summer: male vs. female 7 68.19 <0.0001*
1995 winter: male vs. female 5 12.21 0.0678
1995 summer: male vs. female 6 20.37 0.0047 *
Female: 1994 vs. 1995 7 18.14 0.0241
Male: 1994 vs. 1995 7 15.69 0.0570

 

' Bonferroni level of signiﬁcance determined by or / number of comparisons per type (or= 0.05)

b includes deer collared in 1995 only

55
0.1496). Furthermore, no signiﬁcant diﬁ‘erences were detected between seasonal habitat
use patterns of migratory and non-migratory deer in either year except for summer 1994
(Table 9).

Deer migrating to agricultural and non-agricultural areas used habitats differently
within years but no signiﬁcant diﬁ‘erences in habitat use patterns were detected for either
deer group between years (Table 9). Likewise, seasonal habitat use patterns of
agricultural and non-agricultural migrating deer were not detected to diﬂ'er for winter
1995 (sample size for 1994 was insufﬁcient to statistically test). However, summer
habitat use patterns of deer migrating to agricultural or non-agricultural areas were
inherently different because non-agricultural migrating deer were never located in
croplands and deer with home ranges established in agricultural lands were frequently
located within crop ﬁelds.

Diﬂ’erences in habitat use patterns of male and female collared deer within years
were statistically signiﬁcant, however, habitat use patterns between years for both sexes
were not signiﬁcantly different (Table 9). Seasonally, male and female deer used summer
habitats diﬂ‘erently in both years, however, 1995 winter habitat use patterns were not
signiﬁcantly diﬂ‘erent (1994 winter sample sizes were too small to test statistically)
(Table 9).

Individual deer habitat use patterns were compared with availability of habitats
for 88 yearly or seasonal home ranges during the study. Deer used habitats as expected
in 39.77% (N = 35) of comparisons whereas habitat use diﬂ’ered signiﬁcantly from

random selection of available habitats for 60.23% (N = 53) of all comparisons. The test

56

for overall signiﬁcance of non-random habitat use by all deer was highly signiﬁcant at P
< 0.0001 (Table 10). Within years, 52.63% of comparisons in 1994 (N = 38), and
66.00% of comparisons in 1995 (N = 50), rejected the null hypothesis of random habitat
use. Combined probability tests indicated overall statistical signiﬁcance of non-random
habitat use patterns of deer in both years (Table 10). Furthermore, 89.77% (N = 79) of
all comparisons between deer use patterns and availability of habitats within the extended
home range area concurred with comparison results using the original home range.
Results during individual years were similar with 86.84% of extended home range
comparisons corresponding with original home range comparisons in 1994 and 92.00%
agreeing in 1995. Therefore, only habitat use vs. availability results from original home
range comparisons will be presented due to the strong concurrence between extended
and original home range comparison results.

Deer home ranges included an average of 6.88 habitat types (SE = 0.11, Range =
4 - 8). Most non-migratory deer (81.3%) used habitats signiﬁcantly different ﬁom their
availability while only 48.2% of migratory deer had habitat use patterns signiﬁcantly
different ﬁ'om availability. Furthermore, deer migrating to non-agricultural, forested
areas used habitats as expected more frequently (59.3%) than deer migrating to
agricultural areas (44.8%). Likewise, deer on winter ranges used habitats randomly
more ﬁequently (61.5%) than did deer on summer ranges (48.8%) and female deer
tended to select or avoid habitats more regularly (63.5%) than male deer (52.0%).

Statistical tests of overall signiﬁcance concluded that deer used habitats signiﬁcantly

57

Table 10. Results of combined probability analysis of habitat use vs. availability

comparisons by deer group.

 

 

Deer group Chi square value (if P < value"
All deer 2307.32 174 0.0001
1994: all deer 703.69 74 0.0001
1995: all deer 1603.60 50 0.0001
Non-migratory 1097.32 64 0.0001
Migratory 1210.00 1 10 0.0001
Migratory - agricultural 739.67 58 0.0001
Migratory - non-agricultural 470.33 54 0.0001
Winter 292.35 26 0.0001
Summer 954.89 84 0.0001
Male 467.75 50 0.0001
Female 1839.57 124 0.0001

 

' degrees of freedom equals 2(# of comparisons)

" Chi square combined probability test

58
different from their availability within each of the above deer groups (P < 0.0001 for all
groups) (Table 10).

Table 11 lists the percentages of deer, exhibiting non-random habitat use
patterns, that selected or avoided various habitat types. All percentages discussed
below, in this section, refer to percentages of deer with non-random habitat selection
patterns. ‘Other’ and aspen/birch habitat types were the most frequently selected (35.8%
and 26.4% respectively) by all deer. Agricultural land was selected by 18.9% of deer
and avoided by 24.5% of deer and cedar was selected or avoided by 35.9% of deer. In
addition, aspen/birch types were avoided by 20.8% of deer (Table 11).

Non-migratory and migratory deer both selected ‘other’ habitat types most
ﬁequently. More than halfof non-migratory deer (53.9%) selected or avoided
aspen/birch and large percentages of non-migratory deer selected or avoided cedar
(34.6%), and upland hardwoods (30.8%) (Table 11). In addition, agricultural areas were
selected by 18.9% and avoided by 24.5% of non-migratory deer. Migratory deer also
selected or avoided aspen/birch (40.7%) and cedar (37.0%) types, however, 44.4% of
migratory deer selected openings or lowland hardwoods compared to 26.9% of non-
migratory deer. Migratory deer selected agricultural areas 14.8% of the time and
avoided agricultural ﬁelds 29.6% of the time.

Habitat selection and avoidance patterns were very different between deer
migrating to agricultural or non-agricultural habitats. Deer with summer ranges within
agricultural areas selected agriculture (50.0%) and ‘other’ (50.0%) habitat types most

oﬁen and aspen/birch types with low frequency (12.5%). Deer with summer ranges

59

 

 

 

o go o o o 88 o 88 8:882
R u 7c
98 o o o o o Q8 3:. 882% uses»
3: o 88 o 88 o o o 88265
an u 7c 3538.8
o Q8 Q8 e8 98 Q8 0.8 o 882$ .8: - bong:
o ”2 o o n.«_ me o as: 8:863
A2 u E Eagﬁe
c8 ”.8 m.«_ 98 m.«_ n: 9.8 98 eouoaom - .5333
2 a: I. o 2: 2 o 38 8:865
A8 n 7c
e8 «.8 a: «.8 «.8 ”.3 «.8 a: dozen—om 5.33
an «.2 o o _.m« 4.2 we «.3 8.563.
G« u E
m.«e 4.2 «.2 E ”.8 4.2 «.2 _.m« c883 b98582
3 q: «.m o 88 we 3 08 88265
an u E
«.8 d2 qt 2d 4.8 a2 «.8 3: does—om 6% .2
on: A8838
.650 58 on: 33 .33 .3365. 33 .3: =25 23.884. a 25 he a 92» coon
one .58:
.momaﬁ 082—

:2: 553, 3833 «35> 3296 8 3828 language 35 3.8an 8: “332 8038-5: £5, .53 mo «868m .2 053.

60

.3838 20 can 65: 833a .3303 38.585: .8520? @388.“ ELF—um mug—ea .

 

on n.«« e o 88 8: n.« e8 8:862
on. q: n: on o8 3. n: m.«« 8828 canon
E e 4.2 o 2.8 E e :8 8.56%
E ”.8 4.2 4.2 ”.8 :8 ”.8 _.m« 8828 on:
3 3. 3 e as e e 3. 8:865
«.8 3 m «.2 «.8 m8 «.«« «.«« 2 m c6828 58:3
8: Aggie—8
Lose 58 can can. .33 6589.. 33 .3: code 83653.. a 25 do a use» can

 

25 $8:

.3283 : as;

61
established in non-agricultural, forested areas selected aspen/birch habitat types (40.0%)
most often and never selected ‘other’ or agricultural areas (Table 11). In addition, non-
agricultural migrating deer selected or avoided pine stands 40.0% of the time while
agricultural migrating deer only selected pine 12.5% of the time.

‘Other’ (60.0%), agriculture (40.0%), and openings (20.0%) were the only
habitat types selected by deer in the winter and cedar (60.0%), agriculture (20.0%), and
upland hardwoods (20.0%) were the only habitats under utilized during winter (Table
11). However, all habitats types were selected by at least 18.0% of deer in the summer
with agricultural and cedar habitat types selected most frequently. Aspen/birch habitat
types were selected or avoided by 54.6% of the deer in the summer whereas they were
used according to availability during the winter (Table 11).

Male deer selected open, aspen/birch, and cedar habitat types most often and
female deer selected ‘other’ and aspen/birch types most frequently. Agricultural land
was selected by 23.1% of males and 22.5% of females and avoided by 23.1% of males
and 25.0% of females. Aspen/birch habitat types, in addition to being commonly

selected, were avoided by both sexes approximately equally (Table 11).

Population parameters

Abundance
The CAPTURE program selected the null model (M,,) as the best estimator of

deer abundance using the mark-recapture data. This model assumes equal probability of

62
capture on each trapping occasion for all individuals in the population. The 1994 null
model abundance estimate for the trapping region was 77 individuals (SE = 8.14) and the
1995 estimate was 95 individuals (SE = 13.65). A t-test did not detect any signiﬁcant
difference between the yearly estimates (t = -1. 132, P = 0.2682). Since winter daily deer
movements were seldom greater than 0.8 km, traps were thought to attract deer from an
area as large as 2.01 kmz. Mark-recapture data originated ﬁ'om 13 traps in 1994 and 11
traps in 1995 (Appendix Table 1). Therefore, deer densities in trapping areas were
estimated to be 2.95 deer per 1cm2 in 1994 and 4.30 deer per kmzin 1995.

A second model, which assumes that capture probabilities vary by animal M),
was also used to estimate abundance because the social dominance structure of deer may
have resulted in varying capture probabilities by age and sex. Abundance estimates from
the M1. model were 104 individuals (SE = 17.81, density = 3.98 deer / kmz) in 1994 and
144 individuals (SE = 26.16, density = 6.51 deer / kmz) in 1995. No signiﬁcant
diﬁ‘erence was detected between 1994 and 1995 M1, abundance estimates (t = 1.264, P =
0.2082).

Overwinter deer densities estimated from pellet-group counts ranged ﬁ'om 8.94 -
34.86 deer per km2 across the study area and were highest in Presque Isle county in 1994
and highest in Montmorency county in 1995 (Table 12). There were no signiﬁcant
diﬁ'erences in deer densities between years for any county.

The number of deer counted during road surveys peaked in March and April
during both years (Figure 10). Similar patterns in monthly total counts occurred during

the two years with the number of deer counted increasing from May through early fall

63

Table 12. Estimated overwinter deer densities and standard errors (SE) for Presque Isle,

Montrnorency and Alpena counties in 1994 and 1995.

 

 

 

1994 1995
County Deer / km2 SE Deer / km2 SE P value‘ Average
deer/1cm2
Presquelsle 3.54 7.94 1.33 2.04 0.9999 2.44
Montmorencyb 1.53 2.00 5.20 4.14 0.2222 3.37
Alpena" 2.99 0.94 1.60 0.99 0.9999 2.30

 

a Wilcoxin-Mann-Whitney test
b includes the northern halfofthe county
c includes four northwestern townships of the county

 

1400- ,2

1200-

10004 .
1994 3

80m .1995 1

 

 

 

Number of deer

=§§§

 

L

 

Figure 10. Total number of deer counted in monthly road surveys (June, July and

August are average counts for two surveys) in 1994 and 1995.

65
and decreasing during the fall and early winter. More deer were counted along monthly
surveys in 1995 than in 1994 except during the month of March. Male and female deer
were distinguishable during summer and fall monthly surveys. Counts of females were
greater than counts of males during each month and both sexes followed the same
general trends as total deer counts by increasing over summer months and decreasing in
the fall (Figure 11). Adult deer were distinguishable from sub-adult deer (< one year
old) during winter and early spring (although deer of unknown status were still
encountered) and from newborn fawns during summer and fall. Counts of sub-adults
tended to peak in late spring and early fall of the year and were usually less than adult
counts (Figure 12). The largest differences in the number of adults and sub-adults
counted occurred during the summer months of each year and the smallest diﬂ‘erences in

observed counts occurred during late fall and winter.

Survival
Annual survival estimates (1 95% conﬁdence intervals) for non-migratory deer in

1994 (0.814 : 0.23) were not signiﬁcantly different (a = 0.05) from annual estimates for
migratory deer (0.500 i 0.18) although diﬁ‘erences were signiﬁcant at or = 0.10 (P =
0.0948). Annual survival estimates of non-migratory (0.583 : 0.19) and migratory
(0.636 : 0.19) deer in 1995 were not statistically different (P = 0.9999). Furthermore,
while summer and fall period survival of non-migratory deer (summer = 0.91 i 0.17, fall

= 0.90 i 0.19) was greater than survival of migratory deer (summer = 0.73 p: 0.17, fall =

120 5

§

804

407
20‘

Number of deer
8

 

 

 

 

 

140 4
120 -
100 4
80 3
6O 1
40 —
20 1

Number of deer

 

 

 

 

 

 

 

 

 

66

 

 

El Females
I Males

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L

J

 

L

A

Month

 

 

El Females
I Males

 

 

 

Figure 11. Total counts of female and male deer from June through November 1994 (A)

and June through October 1995 (B) road surveys.

67

140 -
120 -

é

 

80 ~ l- DAdults

I Sub-adults

 

 

 

 

40-

204

O‘lﬂ LIT-IT

JFMAMJJASOND
Month

Number of deer
8

 

 

 

 

 

 

 

 

300 -
250 7

200*

1 50 d D Adults
I Sub-adults

 

 

 

 

100 -
50 ~

0‘ 1ft
JFMAMJJASO

Month

Number of deer

 

 

 

 

 

 

 

Figure 12. Total counts of adults and sub-adults from February through December 1994

(A) and from January through October 1995 (B) road surveys.

68
0.69 i 0.20) no signiﬁcant diﬁ‘erences were detected (Pm = 0.9999, Pf,“ = 0.9999).
Annual and period survival rates for each age category and year are given in
Table 13. Differences in annual and period survival rates between years for pooled age
categories of deer were not signiﬁcant (Table 14). Likewise, 1994 annual adult survival
was not different from 1995 annual survival (P = 0.263 6) of adults. In addition, annual

survival of yearlings was not statistically different between years (P = 0.9999) (Table

“‘~*WW

14).
Adult deer had higher survival rates than yearlings during all periods in both J

years. However, diﬁ‘erences in annual or period survival rates of adults and yearlings ‘

were not signiﬁcant in either year (Table 14). Adult and yearling survival differed most

during the 1994 fall period (adults = 0.93, yearlings = 0.55, P = 0.0367) and annually in

1994 (adults = 0.71, yearlings = 0.36, P = 0.0417).
During the two years, 45 radio collared deer died, four were censored and 24

were alive at the end of the study (Appendix Table 6). In addition, the fate of six ear-

tagged deer was known (Appendix Table 7) and their causes of death were pooled with

collared deer. Legal harvest (14 collared, 5 ear-tagged) and natural mortality (12 deaths)

accounted for most deaths over the two years (Table 15). Six deer (4 males, 2 females)

were illegally harvested, or poached and ﬁve deer (4 collared, 1 ear-tagged) were hit by

vehicles. One fawn died of starvation during the ﬁrst winter and one adult doe was

harvested with a crop damage permit. Unknown causes of mortality were attributed to

seven of the collared deer. Males made up 84% of the legal harvest mortalities and 67%

69

Table 13. Mayﬁeld survival estimates (5' ) for collared and tagged deer in 1994 and

 

 

 

 

1995.
1994 1995
Classiﬁcation 95% 95%
conﬁdence conﬁdence
5' SD N interval 5' SD N interval

Age classes pooled: 0.498 0.076 45 0.498 i 0.15 0.484 0.065 56 0.48 i 0.13
annual‘
Age classes pooled: 0.852 0.068 45 0.85 i 0.13 0.819 0.052 57 0.82 i 0.10
wint
Age classes pooled: 0.772 0.067 40 0.77 i 0.13 0.859 0.054 42 0.86 i 0.11
summer°
Age classes pooled: 0.737 0.079 31 0.74 i 0.16 0.646 0.078 37 0.65 i 0.15
fall"
Adults: annual 0.711 0.108 19 0.71 i 0.21 0.531 0.082 35 0.53 i 0.16
Adult: winter 0.913 0.083 19 0.91 i 0.16 0.849 0.057 36 0.85 i 0.11
Adults: summer 0.834 0.088 18 0.83 i 0.17 0.932 0.046 29 0.93 i 0.09
Adults: fall 0.933 0.064 15 0.93 ﬁg 0.13 0.693 0.090 26 0.65 i 0.18
Yearlings": annual 0.355 0.092 26 0.36 i 0.18 0.289 0.100 19 0.29 i 0.19
Yearlings: winter 0.808 0.099 26 0.81 i 0.19 0.744 0.110 20 0.74 i 0.22
Yearlings: summer 0.719 0.097 22 0.72 i 0.19 0.769 0.128 13 0.77 i 0.25
Yearlings: fall 0.551 0.124 16 0.55 i 0.24 0.534 0.150 11 0.53 f; 0.29
' January 1 -December 31
b January 1 - April 15
° April 16 - September 30
d October 1 - December 31

deerbominJuneofthe

year

70

Table 14. Results of survival analysis of collared and tagged deer in 1994 and 1995.

 

 

Chi P
Type Survival comparison df Square value'
value

Ages pooled Annual: 1994 vs. 1995 1 0.262 0.9999
Winter": 1994 vs. 1995 1 1.591 0.5963

Summer°z 1994 vs. 1995 1 0.925 0.8330

Fall“: 1994 vs. 1995 1 0.688 0.9172

Adults Annual: 1994 vs. 1995 1 2.527 0.2636
Winter: 1994 vs. 1995 1 1.456 0.6442

Summer: 1994 vs. 1995 1 1.115 0.7654

Fall: 1994 vs. 1995 1 3.225 0.1450

Yearlingsc Annual: 1994 vs. 1995 1 0.227 0.9999
Winter: 1994 vs. 1995 1 0.627 0.9389

Summer: 1994 vs. 1995 1 0.049 0.9999

Fall: 1994 vs. 1995 1 0.008 0.9999

1994 Annual: adult vs. yearling 1 5.472 0.0386
Winter: adult vs. yearling 1 0.534 0.9719

Summer: adult vs. yearling 1 0.639 0.9346

Fall: adult vs. yearling 1 5.560 0.0367

1995 Annual: adult vs. yearling 1 2.889 0.1839
Winter: adult vs. yearling 1 0.097 0.9999

Summer: adult vs. yearling 1 4.178 0.0857

Fall: adult vs. yearling 1 0.731 0.9019

 

' Bonferroni level of signiﬁcance (or = 0.05) = 0.012, (or = 0.1) = 0.025

b Winter = January 1- April 15

° Summer = April 16 - September 30
d Fall=October 1 -December3l

° deer born in June ofthe year

71

Table 15. Yearly causes of mortality for radio collared and ear-tagged deer.

 

 

Category 1994 1995 Total
Legal harvest 6 13 19
Illegal harvest 4 2 6
Natural‘l 4 8 12
Road killed 2 3 5
Starvation 1 0 1
Crop damage permit 1 0 1
Unknown 3 4 7
Total 21 30 51

 

' includes predator and drowning.

72
of the illegal harvest mortalities. Natural mortality resulted from 11 predator kills

(coyote or dog) and one drowning.

Productivity

The technique of observing collared does with their fawns was not perfected until
the second year of the study. Only 63% of collared does were observed on single
occasions in 1994, whereas 100% of collared does were observed repeatedly in 1995.
Therefore, only productivity data ﬁom 1995 observations will be discussed. Collared
does observed in 1995 ranged from 2 to 7+ years of age (Appendix Table 8). Does
having fawns (N = 13, 59.1%) were observed an average of 2.3 times before sighting the
fawn and does without fawns (N = 9, 40.9%) were sighted an average of 4.2 times. The
1995 average productivity ratio estimated from collared does was 1.08 fawns per doe
(SE = 0.08).

Observed and adjusted fawn to doe ratios estimated from monthly road surveys
are listed in Appendix Table 9. The 1994 average adjusted survey productivity ratio
(0.49 fawns per doe) was not signiﬁcantly diﬁ‘erent ﬁom the 1995 average adjusted
survey ratio (0.51 fawns per doe) (Table 16). Productivity ratios ﬁom incidental
observations were higher than other estimates with an average of 1.36 fawns per doe (N
= 115 observations) in 1994 and 1.30 fawns per doe (N = 98 observations) in 1995
(Table 16). A t-test did not detect any statistically signiﬁcant differences between years

(t = 0.924, P = 0.3628).

73

Table 16. Average fawn to doe productivity, fall-recruitment and post-hunting season

recruitment ratios for 1994 and 1995 and statistical signiﬁcance of yearly diﬁ‘erences.

 

 

Estimate/method 1994 SE 1995 SE P
value'
Productivity
Collared does - - 1.08 : 1 0.08 -
Road surv s 0.49: 1 0.10 0.51 : 1 0.08 0.945
(adjusted)

Incidentalsightings 1.36:1 0.04 1.30:1 0.05 0.363

Recruitment - fall
Collared does - - 1.08 : 1 0.08 -

Road surveys" 0.68:1 0.08 0.67:1 0.09 0.999

Recruitment-post season
Trapped deer 2.28 : 1 - 2.06: 1 -

 

' Wilcoxin-Mann-Whitney test
" doe counts decreased 10%, July - November surveys
° August - November surveys

74
Fall-recruitment ratios averaged 1.08 fawns per doe from collared does, 0.68
fawns per doe ﬁ'om 1994 monthly fall surveys and 0.67 fawns per doe ﬁ'om 1995
monthly surveys. The post-hunting season recruitment ratios from the captured deer

were 2.28 fawns per doe in 1994 and 2.06 fawns per doe in 1995 (Table 16).

Sex ratio and age structure

Road survey deer counts during July - October resulted in an average buck-doe
ratio of0.256 : 1 (l buck per 3.91 does) in 1994 (SE = 0.015, N = 7) and 0.214 : 1 (1
buck per 4.67 does) in 1995 (SE = 0.018, N = 7) (Appendix Table 10). Diﬁ'erences
between the two years were not signiﬁcant (W ilcoxin-Mann-Whitney statistic (WMW)=
55, P = 0.8048). Sex ratios estimated from the captured deer resulted in 0.46 bucks per
doe (1 buck per 2.16 does) in 1994 and 0.63 bucks per doc (I buck per 1.59 does) in
1995.

The winter age compositions of 1994 and 1995 trapped deer were similar with
greater than 60% of the captured deer classiﬁed as fawns in each year. No other age
category constituted more than 18% of the trapped deer in either year (Table 17). The
age composition of captured females was similar to the overall age structure of the
trapped deer with 53.5% of females aged as fawns in 1994 and 48.6% aged as fawns in
1995. However, the 2 - 3 year old age category was larger, encompassing 25.6% of the
trapped females in 1994 and 20.0% in 1995. No males older than yearlings were trapped

in either year, therefore the age composition of trapped males was highly skewed

75

Table 17. Age structure of deer determined by the age composition of trapped deer,

deer aged at the study area check station, and road killed deer in 1994 and 1995. Table

values are percentages of deer classiﬁed in each age category.

 

 

 

Trapped deer Check station Road killed Combined

Age category 94 95 94 95 94 95 94 95
Fawn 66.13 67.27 3.13 7.07 30.00 11.11 24.30 27.61
Yearling 6.45 9.09 52.34 43.43 20.00 44.45 35.72 31.90
23 year old 17.74 9.09 37.50 40.40 40.00 0 31.91 27.61
4-5 year old 4.84 3.63 4.69 5.05 10.00 33.33 5.24 6.13
6 years and older 4.84 10.92 2.34 4.05 0 11.11 2.86 6.75

Total 100 100 100 100 100 100 100 100

N 62 55 128 99 20 9 210 163

 

76
towards younger age classes in both years (Table 18).

Most deer aged at the study area check station in 1994 and 1995 were in the
yearling and 2 - 3 year old age categories (Table 17). Males made up 93.8% of the deer
aged at the check station in 1994 and 85.9% of the deer aged at the station in 1995.
Therefore, the age composition of male deer followed the overall age structure of check

station deer closely (Table 18).

Fawn and yearling age categories made up 50.0% of the road killed deer in 1994
and 55.6% of road kills in 1995. In addition, 2 - 3 year old deer constituted 40.0% of
1994 road kills and 4 - 5 year old deer comprised 33.3% of 1995 road kills (Table 17).
Samples of road killed deer were small and insuﬁicient to determine age composition by

sex.

The combination of age structure data from all 3 sources resulted in a more even
distribution of deer across the ﬁrst three age classes in both years. The 4-5 year old and

the older than 6 year old age classes were also well represented (Table 17).

77

Table 18. Age structure of male and female deer determined by the age composition of
trapped deer and deer aged at the study area check station in 1994 and 1995. Table

values are percentages of deer classiﬁed in each age category.

 

 

 

 

Trapped deer Check station
1994 1995 1994 1995
Age category M F M F M F M F
Fawn 94.74 53.49 90.91 48.57 1.67 25.00 4.71 21.43
Yearling 5.26 6.98 9.09 8.57 53.33 37.50 48.23 14.28
2-3 year old 0 25.58 0 20 37.50 37.50 43.53 21.43
4-5 year old 0 6.98 0 5.72 5.00 0 2.35 21.43
6 years and older 0 6.98 0 17.14 2.50 0 1.18 21.43
Total 100 100 100 100 100 100 100 100

N 19 43 22 35 120 8 85 14

 

DISCUSSION

The average radio telemetry bearing standard deviation in this study (9.4") was
similar to the 10° bearing error reported by Mooty et al. (1987) and the 7° bearing error
reported by Van Deelan (1995). The mean 95% conﬁdence ellipse area for deer
locations in this study (37 .2 ha) was larger than the 0 - 2 ha conﬁdence area reported in
one Minnesota study (Nelson and Mech 1984) and the 10.5 ha ellipse area reported in
one Michigan study (Van Deelan 1995). However, the mean habitat size, 23.8 ha (SE =
22.2 ha), was 64% of the average 95% conﬁdence ellipse area, indicating little potential

bias in habitat use detection (Nams 1989).

Migration

Approximately half of the radio collared deer were migratory during each year
(Table 2). Other midwestern studies have reported similar ﬁndings of migratory
behavior from at least part of a deer herd (Hoskinson and Mech 1976, Larson et a1.
1978, Nixon et al. 1991, Van Deelan 1995), however, the percent of migratory deer
ranged ﬁ'om as high as 88.2% (Hoskinson and Mech 1976) to as low as 33.0% (Kufeld
and Bowden 1995). Similar migratory patterns were reported for mule deer with 74.0%
of mule deer migrating between seasonal ranges in Idaho (Brown 1992) and only 27.0%

78

79
of Colorado mule deer exhibiting migratory behavior (Kufeld and Bowden 1995). Our
results agreed closely with the ﬁndings of one study in Michigan’s upper peninsula (Van
Deelan 1995) where approximately halfof the deer herd was migratory. Some studies
have indicated that migratory behavior of deer in the fall is lower in years with mild
winters (Nixon et al. 1991, Nelson 1995), increasing the relative presence of non-
migratory deer on winter ranges.

The winter range departure dates reported in this study (March 15 - March 30)
agreed with results of other midwestern studies in Wisconsin (Hammerstrom and Blake
1939), Minnesota (Rongstad and Tester 1969, Hoskinson and Mech 1976), South
Dakota (Sparrowe and Springer 1970), Illinois (Nixon et al. 1991), and Michigan (Van
Deelan 1995) where initiation of spring migration began ﬁ'om early March to late April.
Loss of snow cover (Hoskinson and Mech 1976, Tierson et al. 1985) or warming
temperatures (Ozoga 1968, Drolet 1976, Beier and McCullough 1990) are thought to
trigger spring migrations of deer (Nelson 1995). Initiation of spring migration in this
study seemed to be related to reduced snow cover since spring migrations began later in
1994 (when snow continued to cover cropﬁelds until late March) than in 1995 (when
snow had melted ﬁom ﬁelds by the second week in March). Van Deelan (1995) also
reported that snow depth inﬂuenced the initiation of spring migration in Michigan’s
upper peninsula with deer migrating earliest in the year with the least snow cover in
March, despite it being intermediate to other years in temperature and overall winter

severity. This study, as with others (Rongstad and Tester 1969, Sparrowe and Springer

80
1970, Hoskinson and Mech 1976, Van Deelan 1995) reported the majority of spring deer
migrations to have been completed by early May (Figure 3).

Spring migrations of deer in this study were mostly in a northwestern to
southwestern direction during each year. Other studies have also reported directional
migrations of deer ﬁom winter yards in South Dakota (Sparrowe and Springer 1970),
Minnesota (Hoskinson and Mech 1976), and Michigan (Venue 1973, Van Deelan 1995).
These directional migrations of deer are often associated with stream drainages or other
landscape features (Tierson et el. 1985, Van Deelan 1995), however, the direction of
most deer migrations in this study tended to be toward heavily forested areas and away
from open, agricultural land.

The range of migration distances of deer reported in this study (1.6 km - 37.0
km) was consistent with the results of other studies in Minnesota (Carlson and Frames
1957, Rongstad and Tester 1969, Hoskinson and Mech 1976, Nelson and Mech 1984),
New York (Tierson et a1. 1985), South Dakota (Sparrowe and Springer 1970), and
Michigan (V erme 1973, Van Deelan 1995). Average migration distances of Michigan
deer were 13.6 km in one study (Verme 1973), and 5.45 km in another (Van Deelan
1995), while deer in Illinois averaged 13 km (Nixon et al. 1991), and deer in Minnesota
averaged 17 km migrations in one study (Nelson and Mech 1984), and ranged ﬁom 10 -
38 km in a second study (Hoskinson and Mech 1976). The average migration distance
of deer in this study, 9.62 km, is consistent with these studies.

Male deer tended to travel greater distances between seasonal ranges than female

deer in this study, however, differences were not signiﬁcant. Other studies (Carlsen and

8 1
Farmes 1957, Verrne 1973) also recorded greater movements of male deer than for
female deer with differences only approaching statistical signiﬁcance. The lack of
signiﬁcant diﬁ‘erence is not unexpected since migrations are thought to be passed along
from doe to fawn (Marchington and Hirth 1984, Nelson and Mech 1984) and since the
doe-fawn bond is still strong at the time of migration (Ozoga 1972), male fawns probably
migrate to summer ranges with their mother (V erme 1973, Nelson 1994), reducing
diﬂ‘erences in migration distance by sex.

More than half of the migratory deer traveled to non-agricultural, forested areas
to establish summer ranges, often migrating 3.5 times ﬁlrther than deer migrating to
neighboring agricultural areas (Figure 7, Figure 8). These forested areas tend to be less
populated, have lower road densities than agricultural regions, and are comprised largely
of aspen/birch, lowland and central hardwoods, cedar, and pine habitats.

The shorter distance of deer migrations between agricultural areas can be
partially explained by the composition of the study area. There are several large regions
comprised of mostly agriculture that are almost entirely surrounded by wooded habitats
(primarily cedar and upland hardwoods). These agricultural “islands”, although having
less diversiﬁed habitats, would be highly conducive to providing both winter thermal
requirements and nutritious summer forage in close proximity. It is not unexpected then
that some deer migrate only short distances to nearby abundant, nutritious, summer
forage.

Verrne (197 3) reported that deer migrations were greater in the eastern regions

of Michigan’s upper peninsula where habitat interspersion was low, compared to the

82
western region, where habitats were more diversiﬁed. Longer distance migrations may
therefore, be an attempt, among other things, to reach areas of greater habitat
interspersion.

The average fall migration departure dates of migratory deer in this study (mid-to
late November) are consistent with dates reported in other studies (Hammerstrom and
Blake 1939, Rongstad and Tester 1969, Sparrowe and Springer 1970, Hoskinson and
Mech 1976, Nixon et al. 1991, Van Deelan 1995) although only Nixon et al. (1991) and
Nelson (1995) reported initiation of migration beginning in early October, as reported in
this study (Figure 9).

An important ﬁnding in this study is the relationship of the timing of deer arrival
on winter ranges with the ability of block permits to successfully target deer responsible
for crop damage. Most block permits are used during the ﬁrearm deer season (T.
Carlson, MDNR, pers. commun.) which occurs ﬁ'om November 15 - November 30.
Non-migratory deer and deer migrating between agricultural areas are susceptible to
harvest by crop damage permits since their yearly or seasonal home ranges are located in
agricultural areas. However, deer with summer ranges in non-agricultural areas, are only
vulnerable to harvest by crop damage permit if they arrive on winter ranges (agricultural
areas in this study) before the end of ﬁrearm season. In 1994, 50.0% (N = 6) of fall
migrating deer with non-agricultural summer ranges returned to winter ranges before the
end of the ﬁrearm season. In 1995, 85.7% (N = 7) of migrating deer with non-
agricultural summer ranges arrived on winter ranges before the end of ﬁrearm season.

Overall, 54.5% of 1994 fall migration and 66.7% of 1995 fall migration was completed

83
before the end of ﬁrearm season. These high percentages suggest that many deer that do
not forage on agricultural crops during the summer are susceptible to harvest by fall
block permits.

Small sample sizes of fall migrating deer may seemingly limit our ability to
describe the overall probability of block permits to successfully target depredating deer,
however, the range of migration dates was extensive and although a more concentrated
period of fall migration may be detected with larger sample sizes, the range would
probably increase. Although detection of a concentrated period of fall migration would
enable block permit use to be adjusted accordingly, the probability of harvesting non-

depredating deer will likely still exist.

Horne range

Home range size estimates of non-migratory deer in this study were larger than
for migratory deer, although differences were not signiﬁcant. Kufeld and Bowden
(1995) also reported similar home range sizes of migratory and non-migratory deer in the
plains river bottoms of Colorado. Comparisons of home range size estimates with other
studies is difﬁcult because of the number of diﬂ‘erent estimators used in the literature. A
second confounding factor is how several estimators vary with sampling intensity.
Furthermore, many estimators have various calculation options that generate diﬂ‘erent
estimates. In this study, home ranges were calculated for deer having 10 or more
seasonal locations. Beier and McCullough (1990) required 20 locations for home range

determination and Nixon et a1. (1991) required 30 seasonal locations for home range

84
calculation, however, both of these studies used the minimum convex polygon estimator
(MCP) which is more severely aﬁ‘ected by sampling intensity (White and Garrott 1990).

Average winter range estimates (201.8 - 353.6 ha) in this study were comparable
to the ﬁndings of other studies. Winter ranges of deer in Minnesota averaged 264.6 -
353.2 ha using grid cell counts, (Rongstad and Tester 1969), 178 ha for Msconsin deer
(MCP, Larson et al. 1978), 341 ha in an average winter in New Brunswick (MCP,
Drolet 1976) and 368.7 ha for Illinois deer (MCP, Nixon et al. 1991). Winter range
estimates in this study were larger than winter estimates of 43 ha (MCP) in one
Minnesota study (Mooty et al. 1987), estimates of 132-150 ha (MCP) for deer in New
York (Tierson et al. 1985), and smaller than estimates of winter ranges (774 ha, MCP) in
Colorado (Kufeld and Bowden 1995) and South Dakota (698 ha, estimator unknown)
(Sparrowe and Springer 1970).

Drolet (1976) and Rongstad and Tester (1969) recorded an increase in the winter
range sizes of deer in snow-free winters compared to range sizes in deep snow winters.
The ﬁndings of this study supported this tendency with smaller winter ranges (201.8 ha)
in the heavy snow winter than in the light snow winter (3 53.6 ha), although diﬁ‘erences
between years were not signiﬁcant (P = 0.1322).

Summer ranges recorded in this study (328.6 - 336.9 ha) were larger than
estimates of 83 - 319 ha (MCP) for deer in northeastern Minnesota (Nelson and Mech
1984), 219.4 ha (MCP) for deer in north-central Minnesota (Kohn and Mooty 1971),
221 -233 ha (MCP) for deer in the Adirondacks of New York (Tierson et al. 1985), 258

ha (estimator unknown) for deer in South Dakota (Sparrowe and Springer 1970), and

85
266 ha (MCP) in New Brunswick (Drolet 1976). Areas with high habitat interspersion
support smaller seasonal ranges (Sanderson 1966, Dusek et al. 1988, Beier and
McCullough 1990) whereas home ranges in areas of low habitat interspersion or
diversity tend to be larger. Habitat interspersion in our study area may, therefore, be low
as suggested by the large summer range size of radio collared deer in this study.

In general, winter ranges of deer in northern areas tend to be smaller than
summer ranges (Hoskinson and Mech 1976, Tierson et al. 1985, Mooty et al. 1987, Van
Deelan 1995) due to reduced activity levels associated with wintering deer (Silver et al.
1969, Moen 1976, 1978, Marchington and Hirth 1984). However, one southern i
Michigan study (Beier and McCullough 1990) and one northeastern Montana study
(Dusek et al. 1988) recorded larger winter ranges than summer ranges for non-migratory
deer. These diﬁ‘erences were attributed to a high degree of habitat interspersion which
allowed for small summer ranges because of short distances between forage and
protective cover (Dusek et a1. 1988, 1989, Beier and McCullough 1990).

Although no diﬂ‘erences in seasonal range sizes were detected in this study (P =
0.3614), summer ranges of migratory deer were larger than winter ranges during the
heavy snow winter but smaller than winter ranges during the light snow winter. In
addition, pairwise comparisons of an individual’s seasonal ranges in the heavy snow
winter year (1994) indicated signiﬁcantly smaller winter ranges (P = 0.0332), while
comparisons in the light snow winter year (1995) were not signiﬁcant despite 72.7% of
the individuals having larger winter ranges than summer ranges. Since summer range

sizes were very similar between years (336.9 ha vs. 328.6 ha), this suggests that while

 

86
many factors inﬂuence home range size, the severity of winter (however indicated) is
important in determining which seasonal range is larger.

In addition, seasonal ranges of deer migrating between agricultural areas were
smaller than seasonal ranges of deer migrating to non-agricultural regions. This result is
not unexpected since deer in agricultural areas are likely to have abundant forage in close
proximity to security cover whereas deer in forested regions are less likely to have
plentiful nutritious forage within security cover areas.

Many studies report traditional use or ﬁdelity to seasonal ranges (Rongstad ad
Tester 1969, Verrne 1973, Nelson and Mech 1984, Tierson et al. 1985, Nixon et al.
1988, Beier and McCullough 1990, Van Deelan 1995). Furthermore, many studies
indicate greater traditional use of summer ranges due to variation in winter severity
(Drolet 1976, Tierson et al. 1985, Dusek et al. 1989, Beier and McCullough 1990).
Findings in this study agreed with traditional use of winter ranges by 72.7% of deer and
ﬁdelity to summer ranges by 92.3% of deer. Decreased traditional use of winter ranges
was due to two deer that remained on summer ranges during the light snow winter and
two deer that returned to areas adjacent to winter ranges for approximately one week at
the end of the light snow winter.

Nelson and Mech (1981) also reported traditional use of intermediate (spring or
fall) ranges. In this study, six deer established distinct spring home ranges; four located
between winter and summer ranges and two appearing to be distinct fawning ranges.

However, only one deer survived to be evaluated for traditional use of intermediate

87

ranges. This adult doe traveled approximately 4 km to an area where she had her fawn
and then returned to her previous winter range during each year.

Studies in Minnesota (Nelson and Mech 1984), Michigan (Beier and McCullough
1990, Van Deelan 1995), New York (Tierson et al. 1985), and Montana (Dusek et al.
1989) reported a tendency of males to have larger home ranges sizes than those of
females. However, Rongstad and Tester (1969) reported females to have larger home
ranges than those of males. The results of this study indicate no signiﬁcant differences

between home range sizes of diﬂ‘erent sexes.

Habitat use

Habitat use in this study diﬂ‘ered throughout a 24-hour period with use of open
habitat types increasing and use of forested habitat types decreasing from day to night
(Table 8). Fluctuations in diel patterns of habitat use by deer have also been reported in
other studies (Rongstad and Tester 1969, Kohn and Mooty 1971, Drolet 1976, Larson et
al. 1978, Murphy et al. 1985, Dusek et al. 1989, Kufeld and Bowden 1995) with some
studies also reporting greater use of forested types during the day and greater use of
open habitat types (including agriculture) at night (Montgomery 1963, Larson et al.
1978, Suring and Vohs 1979, Beier and McCullough 1990, Dusek et al. 1989). This
ﬂuctuation in diel use is possibly due to deer use of concealment habitats (closed forests)
during the day which decreases visibility to predators (Beier and McCullough 1990).

This ﬁnding is also likely related to the crepuscular foraging activity of deer

88
(Montgomery 1963, Beier and McCullough 1990) since both dawn and dusk hours fall in
the evening (1600 - 2359 hours) and night (0000 - 0759 hours) periods.

Few studies report random use of habitats by deer (Kohn and Mooty 1971, Beier
and McCullough 1990). More commonly habitats are utilized in excess of availability or
less than availability either seasonally (McCaﬂ‘ery and Creed 1969, Drolet 1976, Murphy
et al. 1985, Mooty et al. 1987, Dusek et al. 1988, 1989, Nixon et al. 1988, 1991) or by
diﬁ‘erent sexes (Beier 1987, Dusek et al. 1989, McCullough et al. 1989).

Wooded habitat types are more commonly used by deer during winter than
summer months (Rongstad and Tester 1969, Larson et al. 1978, Murphy et al. 1985,
Dusek et al. 1989, Nixon et al. 1991, Kufeld and Bowden 1995). Furthermore, deer in
Wisconsin (Murphy et al. 1985), Montana (Dusek et al. 1988, 1989), and Minnesota
(Mooty et al. 1987) tended to use wooded habitats more than expected during winter
and only Drolet (1976) reported winter use of forested swamps in New Brunswick to be
less than expected. In this study, 60.0% of deer with non-random winter habitat use
patterns selected forested wetlands, however, no other forested types were selected.
Furthermore, cedar was used less than expected by 60.0% of wintering deer with non-
random habitat use. This result is consistent with the ﬁndings of Drolet (1976) and is
probably due to the proximity of agricultural ﬁelds to cedar swamps. The decreased
energy expenditure associated with reduced travel distances between food and cover
would facilitate more ﬁequent feeding and consequently, under utilization of wooded

COVCI'.

89

Furthermore, agricultural and open habitat types are more heavily utilized during
spring and summer months than during winter months (Kohn and Mooty 1971, Dusek et
al. 1989, McCullough et al. 1989, Nixon et al. 1991, Austin and Umess 1993, Kufeld
and Bowden 1995). Studies in Illinois (Nixon et al. 1991), Wisconsin (Murphy et a1.
1985), and Montana (Dusek et al. 1988) report avoidance of agricultural and open range
habitat types by deer in the winter. The ﬁndings of this study are not wholly consistent
with these results. Winter use of openings and agricultural areas by all deer (27.2%) in
1994 was greater than summer use of these habitats (23.4%), and although differences
were not signiﬁcant at or = 0.05, there were signiﬁcant seasonal differences at or = 0.10.
In addition, 40.0% of deer with non-random winter habitat use patterns selected
agricultural ﬁelds and 20.0% selected openings while only 20% avoided agricultural
areas during the winter. This selection can be explained by the relative proximity of
agricultural ﬁelds and openings to winter yarding areas which would result in more
opportunistic foraging. Suring and Vohs (1979) reported use of openings to be greatest
when they were close to habitats providing cover. In addition, Moen (1968) noted that
winter cover is less critical if high quality forage is available and Kohn and Mooty (1971)
argued that available forage was one of the most important factors determining habitat
use. Sparrowe and Springer (1970) also reported frequent use of agricultural areas by
deer in the winter as a source of food (waste or residue crops).

Our ﬁndings may also reﬂect the relative mildness of northern Michigan winters.
Van Deelan (1995) recorded below average winter severity for three recent winters

(including the most severe winter during this study) in Michigan’s upper peninsula.

90

Other studies in Michigan (McCullough et al. 1989, Beier and McCullough 1990) and
Minnesota (Rongstad and Tester 1969) reported increased use of swamps during more
severe winters and greater use of openings during less severe winters.

During spring and summer months, deer tend to increase use of agriculture,
(Kohn and Mooty 1971, Dusek et al. 1988, 1989, Kufeld and Bowden 1995) openings,
(McCaﬁ‘ery and Creed 1969, Suring and Vohs 1979) and grasslands (McCullough et al.
1989), however, despite increased use, these habitats tended to be used less than
expected for deer in Montana (Dusek et al. 1988,1989). In this study, only 4.5% of deer
with signiﬁcantly different summer habitat use patterns avoided agricultural areas while
31.8% selected agriculture and 22.7% selected openings. This ﬁnding may reﬂect the
preferential use of nutritious food sources that are close to cover providing habitats.

Furthermore, deer tend to heavily use forested areas during summer (Kohn and
Mooty 1971, Murphy et al. 1985, Mooty et al. 1987, Dusek et al. 1989, Nixon et al.
1991) and often use these areas in greater proportion than their availability (Mooty et a1.
1987, Dusek et al. 1988, 1989). The results of this study were consistent, with greater
than 70% of deer locations occurring in forested habitat types during summer months.
In addition, all wooded habitat types were selected by at least 18.2% of deer with non-
random summer habitat use patterns. Cedar and aspen/birch habitat types were selected
by 31.8% and 27.3% of deer respectively, which agrees with summer selection of
aspen/birch and lowland conifers by deer in Minnesota (Mooty et al. 1987). Most

wooded habitat types provide some natural forage and supply adequate to excellent

91
security cover. Habitats that provide both cover and forage are beneﬁcial to meeting
energy demands.

Migratory and non-migratory deer diﬁ‘ered in their summer habitat use patterns in
this study (Table 9). Migratory deer tended to select lowland hardwood and cedar
habitat types more and aspen/birch and ‘other’ types less ﬁ'equently than non-migratory
deer (Table 11). In addition, non-migratory deer selected agriculture more ﬁ'equently
and avoided it less than migratory deer did. Non-migratory deer, being year-round
residents of an agricultural area, may possess an increased knowledge of their
surroundings and escape routes that might allow them to forage in higher risk habitats
(such as agricultural ﬁelds) more ﬁ'equently than migratory deer.

Deer migrating between agricultural areas and deer migrating to non-agricultural
areas also diﬁ‘ered signiﬁcantly in overall habitat use patterns (Table 9). Winter habitat
use was not signiﬁcantly diﬂ‘erent (P = 0.247 6), however, diﬂ‘erences in summer habitat
use were signiﬁcant because deer with summer ranges in non-agricultural areas do not
use agricultural habitats at all and deer with summer ranges in agricultural areas
ﬁ'equently utilize croplands. Furthermore, non-agricultural migrating deer selected
wooded habitats more and ‘other’ habitats less than agricultural migrating deer (Table
11). Agriculture was selected by 50.0% of deer migrating between agricultural areas
with non-random habitat use patterns. The implications of such large percentages of
deer selecting agriculture (in addition to non-migratory deer selection of agricultural

ﬁelds) regarding potential crop damage is readily apparent.

92

Diﬂ‘erential habitat use by sexes was reported for deer in Michigan (Beier 1987,
Beier and McCullough 1990, McCullough et al. 1989) and Montana (Dusek et al. 1989),
however, male and female deer in Washington (Suring and Vohs 1979), Illinois (Nixon et
al. 1991), and Colorado (Kufeld and Bowden 1995) did not differ in habitat use patterns.
Habitat use patterns of male and female deer were signiﬁcantly different in this study (P
< 0.0001), and although both sexes used winter habitats similarly, summer habitat use
differed between sexes (P < 0.0001). Female deer in Michigan (Beier 1987, Beier and
McCullough 1990, McCullough et al. 1989) have been reported to use grasslands more
and forested habitat types less than males. The results of this study were similar, with
females averaging more locations than males (26.1% vs. 9.8%) in agricultural ﬁelds and
openings in the summer. Likewise, females were located less often than males (73.9%

vs. 89.2%) in wooded cover types.

Population parameters

Abundance

Estimates of deer density determined ﬁ'om pellet group counts were lower than
MDNR pellet group density estimates of 7.08 deer / km2 in 1994 and 8.48 deer / km2 in
1995 (H. Hill, MDNR, pers. commun.). This diﬁ’erence maybe attributed to diﬂ‘erences
in design and calculations used by the MDNR The MDNR stratiﬁcation method is

based on assumed deer density which may introduce some bias into their estimates. In

93
addition, their estimates are based on district-wide data vs. on a per county basis. Thus,
high densities in some counties may erroneously increase district-wide estimates.

Our average winter deer densities of 2.30 - 3.37 deer / km2 are smaller than
winter deer yard densities of 16 - 39 deer / km2 reported in Minnesota (Nelson and Mech
1981), 30 - 40 deer / 1cm2 reported in Wisconsin (Larson et al. 1987), 21.3 - 30.0 deer/
1cm2 in Washington (Gavin et al. 1984), 17.5 deer / 1cm2 in Quebec (Messier and Barrette
1985), 24.6 - 65.2 deer / km2 reported for two historic deer yards in Michigan’s upper
peninsula (Van Deelan 1995) and similar to 2.3 - 10.4 deer / km2 in New York (Tierson
et al. 1985).

Deer counts on road surveys peaked during March and April of each year
(Figure 3). This was consistent with the time that snow began to melt from cropﬁelds.
The March 1994 road survey, conducted on March 30 - 31, coincided with the ﬁrst day
of spring migration in 1994. Likewise, the 1995 March road survey was conducted from
March 14 - 16 and the ﬁrst day of spring migrations of collared deer in 1995 was March
1 5th.

Deer counts increased over summer months and decreased during the fall.
Similarly, deer use of agricultural ﬁelds and openings was high (21.5%) during the
summer in this study and use of openings was reported to increase from early to late
summer and decrease during the fall in other studies (Kohn and Mooty 1971, Dusek et
al. 1989, Kufeld and Bowden 1995). The decrease in deer counts ﬁ'om summer to fall,
however, is confounded since corn and other crops were maturing during this time and

were fully mature and able to hide deer, reducing observability, in the fall.

313mg” 1.37

94
Diﬁ‘erences in feeding and bedding related behaviors of diﬂ‘erent sex or age deer
are likely to inﬂuence total deer counts and sex and age ratios of observed deer. The
change in the driven route from two nights in 1994 to three nights in 1995 would
alleviate some behavioral bias since most of the three night survey was driven within the

hour around sunset when observability of deer is highest (Zagata and Haugen 1974).

Survival

No differences in survival of migratory and non-migratory deer were detected in
this study. Van Deelan (1995) reported a similar ﬁnding for deer in Michigan’s upper
peninsula. While diﬂ‘erences in adult and yearling survival were not signiﬁcant, adult
deer had higher annual and period survival than yearling deer during all periods. Results
of survival analysis for deer in this study agree with ﬁndings reported in other studies
(Nelson and Mech 1986, White et a1. 1987, Dusek et al. 1989, Fuller 1990, Nixon et al.
1991). Winter survival was probably high due to the relative mildness of winters and
nearby agricultural ﬁelds providing winter forage.

Sources of mortality for deer reported in this study were similar to other studies
(Ozoga 1972, Gavin et al. 1984, Fuller 1990, Van Deelan 1995). Legal harvest was the
greatest source of mortality in this study and in other hunted populations of deer (Dusek
et al. 1989, Nixon et al. 1991, Van Deelan 1995). Predator kills were also a large source
of deer mortality in this study while overwinter starvation was minor. Starvation was

reported to be a minor source of mortality in other studies of northern deer in Minnesota

95
(Nelson and Mech 1986) and Montana (Dusek et al. 1989) although it has been
frequently documented as a greater source of mortality in Michigan (Ozoga and Harger
1966, Ozoga 1972), Washington (Gavin et al. 1984), and \Vrsconsin (Dahlberg and

Guettinger 1956).

Productivig

In this study, 59.1% of radio collared does produced fawns with an average ratio
of 1.08 fawns per doe. Productivity estimates from observed counts of fawns and does
in late summer and fall road surveys were lower and estimates ﬁom incidental sightings
were higher than collared doe observations (Table 16). Productivity estimates from
collared doe observations in this study agree with the productivity estimates of other
studies (Ransom 1967, Coe et al. 1980, Gavin et al. 1984, Dusek et al. 1989), however,
estimates were lower than productivity estimates of 1.88 fawns per doe in southern
Illinois (Roseberry and Klimstra 1970), 1.6 fawns per doe in Wisconsin (Dahlberg and
Guettinger 1956), 1.63 fawns per doe in Iowa (Haugen 1975), and 1.82 - 2.1 fawns per
doe in an extensively farmed region of Illinois (Nixon et al. 1991). Fall-recruitment fawn
to doe ratios of deer in this study were similar to results reported in other studies (Dusek
et al. 1989, Nixon et al. 1991).

Although collared doe observations seem to be a good method for determining
productivity estimates, precautions must be taken to observe does on numerous

occasions and at various times of the day. Approximately 60% of 1995 collared does in

. .. 'T‘v‘ll‘ilég Li”

96
this study produced fawns as compared to much higher percentages in other studies
(Roseberry and Klimstra 1970, Dusek et al. 1989). Our estimates of productivity may
therefore, be conservative, especially if collared does lost their fawns before we were
able to observe them together.
Downing et al. (1977) reported that fawns were not as visible as adult deer

during July and August when they were most easily discernible. This accounts for the

- arraigtuy

low productivity and fall-recruitment fawn to doe ratios estimated ﬁ'om road surveys.
Nixon et al. (1991) also reported fawn to doe ratios ﬁom observed counts approximately
16% lower than those estimated from observations of marked deer.

Productivity estimates from incidental sightings agree with estimates from
collared doe observations. It is logical that recordings of incidental sightings would be a
favorable method to determine fall-recruitment fawn to doe ratios since behavioral
differences in feeding and bedding behaviors between fawns and does (which directly
inﬂuence observability) are less pronounced in the fall.

Fall-recruitment ratios estimated from trap-related data overestimated actual
fawn to doe ratios since they were higher than all productivity estimates for this study.
In addition, reports of sex and age trapping bias in a mule deer population (Garrott and
White 1982) and sex bias towards females during trapping in New York (Mattfeld et al.
1974) indicate that fawn to doe ratios or sex ratios determined from capture data would

be misleading.

97

Sex ratio and age structure

Buck to doe ratios estimated ﬁom road surveys were comparable to those
reported by Gavin et al. (1984), Dahlberg and Guettinger (1956), and Dusek et al.
(1989) although our estimates used July - October road survey counts whereas Gavin et
al. (1984) utilized October - December counts and Dusek et al. (1989) used March and
April counts. Male and female deer were reported to be equally visible during August
and November (Downing et al. 1977) although Zagata and Haugen (1974) reported male
deer to be the most difﬁcult to observe at all times. Male deer in our study used wooded
habitats more than females, decreasing their observability and causing an underestimation
of buck to doe ratios. Ratios estimated from the captured deer most likely overestimate
adult buck to doe ratios due to the numerous captures of male fawns that will probably
not survive to adult status.

The age structure of the trapped deer was skewed toward younger age classes
(fawn and yearling) in both years with an average of 74.5% of trapped deer in these
segments of the population. Males were more highly skewed with 100% of deer in the
two youngest age classes while only 60.5% of females were in these age classes. The
winter age structure of a hunted population in Michigan (Van Deelan 1995) and the fall
age structure of a hunted population in Montana (Dusek et al. 1989) closely resembled

the results of this study.

98

The age structure of male and female deer voluntarily brought to the study area
check station were skewed more towards middle age classes (Table 18). Coe et al.
(1980) indicated that hunters tended to shoot antlered or large deer when groups of deer
were encountered. In addition, hunters often passed up shots to get a bigger or antlered
deer later. These hunter biases create skewed sex or age structures typical of hunted
populations.

The age structure of road killed deer is probably the least age biased source,
however, small sample sizes in this study limited its use. Combining age structure data
ﬁ'om all three methods could provide the most representative sample because the age

biases associated with capture and hunting potentially cancel one another.

MANAGEMENT IMPLICATIONS AND RECONﬂVlENDATIONS

Crop damage control permits (summer shooting and fall issued block permits)
attempt to control damage on a localized level, however, their effectiveness is greatly
inﬂuenced by seasonal deer movements. The migratory nature of deer as well as the
distance, destination, and timing of deer migrations greatly inﬂuence potential crop
damage to individual ﬁelds. Deer movements in and around agricultural areas and
seasonal habitat use patterns also impact potential crop damage levels.

Non-migratory deer reside in the habitats surrounding agricultural lands and
ﬁequently utilize crop ﬁelds during the summer growing season. These deer are most
vulnerable to harvest by crop damage control permits since they establish yearly home
ranges in agricultural areas. Migratory deer, however, fall into two categories; those
susceptible to harvest by both types of damage control permits because they migrate
between winter and summer ranges that are both within agricultural regions, and those
migratory deer that are not vulnerable to harvest with summer shooting permits because
their summer ranges are in non-agricultural, forested regions. The latter group of deer
is, however, vulnerable to harvest by block permits if they return to winter ranges in the

fall during any period in which block permits are used.

99

100

The timing of fall deer migration in relation to the use of block permits becomes
crucial to the effectiveness of block permits to successfully target the deer responsible
for damage. Most fall deer migrations were completed by the end of ﬁrearm season.
Therefore, to be most successﬁll at harvesting depredating deer, block permits should be
ﬁlled before the start of ﬁrearm season. Otherwise, the probability of block permits
harvesting deer that summer away ﬁom agricultural areas increases and the likelihood of
harvesting depredating deer that migrate ﬁ'om one agricultural area to another decreases.
Block permits could be issued with the stipulation that most be ﬁlled prior to the start of
ﬁrearm season. The initiation of fall migration however, is inﬂuenced by falling
temperatures and increasing snow fall. Therefore, the eﬁ‘ect of adjusting the time period
for block permit use will likely be inconsistent ﬁ'om year to year. Most block permits are
currently ﬁlled during ﬁrearm season, therefore convincing landowners to alter their
hunting habits would be diﬂicult at best.

The MDNR also utilizes variable antlerless harvest quotas within Deer
Management Units (DMUs) to reduce deer numbers in areas with high densities. Over
the years, the tendency has been to decrease the size of DMUs to manage populations on
a localized scale (T. Carlson, MDNR, pers. commun.) The alteration of DMU
boundaries to reﬂect and incorporate the direction and magnitude of deer migrations may
aid in resolving problems of deer migrating between seasonal agricultural areas. Smaller
DMUs, which enable managers to react to local “hot spots”, could still be enclosed
within these larger DMUs. Harvest quotas could be adjusted accordingly so that the

harvest regime of one management unit would have less impact on neighboring DMUs.

101
Expanding DMUs may also create easier consecutive management of summer densities
with desired winter yard densities.

Scare tactics are aimed at reducing deer use of crop ﬁelds and some landowners
suggest using them earlier in the spring before damage occurs, to try to get deer to
migrate earlier or entice resident deer to migrate. The traditional nature of migratory
behavior and its initiation which is closely associated with distinct weather patterns
suggests that this type of action will not be effective.

Summer shooting permits are potentially the most eﬁ‘ective at reducing deer that
forage on crops because they are used during the growing season when foraging is most
detrimental to crop production and when deer are established in summer ranges. The
efﬁcacy of summer shooting permits should be maximized to provide the greatest
damage control. Currently, summer shooting permits are issued for use on a single crop
ﬁeld, although a landowner may have permits for multiple ﬁelds. Block permits are more
commonly issued for the harvest of antlerless deer on all of the landowner’s property and
occasionally may be extended to the adjoining private property. Summer home range
sizes of migratory deer and yearly home range sizes of non-migratory deer indicate that
deer of either migratory nature probably move between numerous crop ﬁelds within their
home ranges. Summer shooting permits may be more eﬂ‘ective overall if they were
adjusted to include all of the landowner’s property thereby encompassing more of the
deer’s home range. With this system, problems associated with changes in deer foraging
behaviors or adjustments in diel habitat use patterns as a result of shooting on

agricultural ﬁelds could be reduced.

102

Another reﬁnement in the issuance of summer shooting permits that may aid in
reducing ﬁlture damage lies in the timing of permit use. Harvesting antlerless deer
earlier in the summer would ensure that fewer newborn fawns would survive being
orphaned, whereas fawns orphaned later in the summer have a higher chance of survival
(Woodson et al. 1980). Higher fawn survival, as result of orphaning, may lead to a
greater incidence of non-migratory deer because of the matriarchal inheritance patterns
associated with deer migration. Currently, summer permits are issued once damage
becomes noticeable or is considered inevitable. The nature of this adjustment however,
would pose diﬁculties for both landowners and managers.

Other recommendations regarding summer shooting permits involve increasing
their effectiveness through increased use. Summer permits are established such that the
landowner is not allowed to keep the harvested deer, whereas deer harvested using block
permits can be utilized by the landowner. Adjustments in the permit system that allow
the landowner to keep a portion of the summer harvested deer may increase the
landowner’s willingness to harvest deer in the summer. In addition, summer shooting
permits are required to have been issued for a landowner to be eligible for block permits
in the fall. This requirement could be strengthened by requiring that a certain number of
summer shooting permits be ﬁlled before becoming eligible for fall block permits. One
important consideration regarding increased summer deer harvest involves the potentially
high levels of toxins existing within carcasses due to recent pesticide treatment of crops.
Summer harvested deer, which are often times donated to charitable organizations,

should therefore, be cautiously considered before ingesting.

103

Greater than 75% of summer deer locations were in wooded habitats. In
addition, deer are thought to forage in habitats that are close to wooded cover. This
suggests that planting crops in areas with non-wooded adjacent edges would decrease
foraging by deer. Often times this may not be possible due to landownerships. Under
these circumstances, landowners may be able to reduce economic losses by rotating
crops such that high value crops are located farthest from wooded edges.

Various reductions in the quality of adjacent wooded habitats may alleviate deer
foraging in agricultural ﬁelds. Decreasing the habitat quality of preferred adjacent
wooded habitats may lessen time spent there by deer and consequently, decrease the
amount of foraging in a particular ﬁeld. Decreasing habitat quality may be as simple as
not creating additional quality habitat adjacent to agricultural ﬁelds. The consequences
to other wildlife species should be carefully evaluated before changes of habitat quality
or quantity are made.

In addition, increasing habitat interspersion in wooded areas away from
agricultural ﬁelds may result in smaller deer home ranges and decreased foraging in
agricultural areas. Speciﬁcally, creating forest openings that would provide spring and
summer forages (McCaﬂ‘ery and Creed 1969) might alleviate some agricultural foraging.
This alteration could serve to increase deer densities by increasing available forage,
therefore, its design requires careful planning and consideration and would require that
farmers cultivate high value crops away from wooded edges whenever possible.

Lastly, additional research indicating the presence of deer that summer in

agricultural areas and winter elsewhere, as well as the direction, magnitude, and timing

J7

711-11513-

l 04
of their movements would allow managers to adjust DMU harvest quotas and increase

hunting pressure on this deer group.

APPENDIX

105

 

 

 

Montrnorency Alpena

 

 

 

Appendix Figure 1. Monthly road survey route within the study area in 1994 and 1995.
Dashed lines depict the driven route and asterisks represent northern and southern

starting points.

106

Appendix Table 1. Trap sites across the study area where deer were successfully

marked in 1994 and/or1995.

 

 

 

Year
County Trap number 1994 1995
Presque Isle 9 M‘ Nc
Presque Isle 10A M N
Presque Isle IOB NM" N
Presque Isle 10C M NM
Presque Isle 12 M NM
Presque Isle 19B M N
Presque Isle 30 M M
Presque Isle 3 5 M M
Presque Isle 40 NM N
Presque Isle 45 M M
Presque Isle 55 M M
Presque Isle 60 M M
Presque Isle 70 N NM
Presque Isle 95 N M
Presque Isle 100 N M
Montrnorency 85 N M
Montmorency 90 N M
Montrnorency 3A M NM
Montrnorency 50 M NM
Alpena 65 NM NM
Alpena 6B M N
Alpena 75 N M
Alpena 80 N M

 

' M=deersuccessfullymarked
" NM=nodeermarked
° N=trapsitenotused

107

Appendix Table 2. Trap night description and trapping success rate for 1994 and

 

 

1995.

Trap night description 1994 1995
Number of trap nights 781 585
Number of inoperable trap nights‘ 426 228
Number of operable trap nightsb 355 357
Number of captures 103 83
Number of escapes° 38 25
Overall success rate‘l 33.10% 47.35%

 

' traps that were tripped (closed) by birds, small mammals, 0r weather conditions.
b trap nights - inoperable trap nights.
°successﬁlllytrappeddeerthatescapedbefore handling.

‘ (total captures + total escapes) / total operable trap nights.

108

Appendix Table 3. Number of deer locations within each time period in 1994 and 1995.

 

 

 

Time period’
1 2 3 Total
1994 Number of locations 1930 477 302 2709
% of 1994 total 71.24% 17.61% 11.15% 100%
1995 Number of locations 1110 203 150 1463
% of 1995 total 75.87% 13.88% 10.25% 100%
Total Number of locations 3040 680 452 4172
% of total 72.87% 16.30% 10.83% 100%

 

' 1 = 0800-1559 hours, 2 = 1600-2359 hours, 3 = 0000-0759 hours.

w

' 1mm”. . V

109

Appendix Table 4. Number of deer locations by season and year for 1994 and 1995.

 

 

Season 1994 1995 Total
\Vlnter 53 1 817 1348
Summer 1368 1456 2824

Total 1899 2273 4172

 

110

Appendix Table 5. Observed (Obs.), expected (Exp), and partial Chi square values for

deer locations' in eight habitat types across three sampling periods (years combined).

 

 

 

 

 

Sampling period"
1 2 3
Obs. Exp. gm. Obs. Exp. 32?: Obs. Exp. m
Cropland 275 332.9 10.08 91 76.2 2.86 92 48.8 38.10
Openings 181 181.7 < 0.01 35 41.6 1.05 34 26.7 2.05
Upland hardwoods 472 508.8 2.67 142 116.5 5.58 86 74.7 1.72
Aspen/birch 994 954.4 1.64 204 218.5 0.96 115 140.1 4.49

Lowland hardwoods 279 256.6 1.96 43 58.7 4.22 31 37.7 1.18

Pine 108 102.6 0.30 22 23.5 0.09 11 15.0 1.09
Cedar 531 503.0 1.56 111 115.2 0.15 50 73.8 7.69
Other 165 165.0 < 0.01 40 37.8 0.13 22 24.2 0.20
Subtotal 3005 3005 18.20 688 688 15.06 441 441 56.47

 

‘ includes winter and summer locations only

b l= 0800-1559 hours, 2 = 1600—2359 hours, 3 = 0000-0759 hours

111

Appendix Table 6. Sex, age, capture date, migratory status, and fate of radio collared

deer as of March 15, 1996.

 

 

Sex Age (months) Capture Migratory Fate
at capture date status‘

F 7 01/25/94 N Censored 06/26/94

F 19 01/25/94 N Alive

M 7 01/26/94 N Legal harvest 11/15/94
F 7 01/26/94 U Coyote kill 03/09/94

F 175 01/29/94 N Crop permit harvest 08/06/94
M 7 01/31/94 Y Legal harvest 11/15/94
F 32 02/01/94 N Road killed 09/21/95

F 20 02/01/94 N Alive

F 8 02/01/94 U Coyote kill 02/17/94

M 8 02/02/94 U Starvation 03/22/94

F 8 02/04/94 Y Illegal harvest 10/06/94
M 8 02/06/94 Y Legal harvest 11/15/95

F 8 02/06/94 Y Alive

F 32 02/12/94 N Coyote/Dog kill 01/12/95
M 20 02/17/94 Y Legal harvest 11/15/95
M 8 02/18/94 Y Legal harvest 11/15/94
M 8 02/19/94 Y Legal harvest 11/15/95

 

112

Appendix Table 6 (cont’d).

 

 

Sex Age (months) Capture Migratory Fate
at capture date status'
F 32 02/20/94 N Dead unknown 10/23/95
F 32 02/22/94 Y Alive
M 8 02/25/94 Y Legal harvest 11/15/95
M 8 02/26/94 Y Legal harvest 11/15/94
F 20 02/27/94 Y Legal harvest 11/16/94
F 32 02/28/94 Y Alive
M 9 03/02/94 U Legal harvest 11/15/94
F 9 03/07/94 Y Dead unknown 04/18/94
F 9 03/07/94 U Coyote kill 04/19/94
F 33 03/09/94 N Alive
F 33 03/09/94 Y Alive
F 9 03/10/94 N Legal harvest 11/15/95
F 105 03/10/94 Y Coyote/Dog kill 05/20/94
F 33 03/10/94 Y Alive
F 9 03/11/94 N Censored 06/27/95
F 33 03/12/94 Y Alive
F 9 03/12/94 Y Dead unknown 05/14/95

113

Appendix Table 6 (cont’d).

 

 

Sex Age (months) Capture Migratory Fate
at capture date status‘
M 9 03/12/94 Y Illegal harvest 07/04/94
F 33 03/15/94 Y Alive
F 33 03/16/94 Y Alive
F 9 03/16/94 Y Alive
M 9 03/16/94 Y Coyote/Dog kill 04/09/94
F 9 03/19/94 N Road killed 03/15/95
M 9 03/20/94 Y Illegal harvest 10/21/94
M 9 03/22/94 Y Road killed 04/26/94
M 9 03/27/94 Y Legal harvest 11/25/95
F 46 04/04/94 U Dead unknown 04/07/94
F 9 04/09/94 Y Road killed 09/19/94
F 140 04/12/94 Y Illegal harvest 08/15/94
M 7 1/25/95 N Alive
M 19 1/25/95 U Coyote kill 04/03/95
M 7 1/25/95 Y Illegal harvest 05/24/95
F 7 1/26/95 N Alive
F 7 1/26/95 Y Alive

114

Appendix Table 6 (cont’d).

 

 

Sex Age (months) Capture Migratory Fate
at capture date status‘
M 7 1/27/95 Y Coyote/Dog kill 04/09/95
F 7 1/27/95 Y Alive
F 7 1/28/95 Y Dead unknown 05/05/95
M 7 1/28/95 N Legal harvest 11/20/95
M 7 1/29/95 N Legal harvest 10/07/95
F 7 1/29/95 U Coyote kill 02/16/95
F 79+ 1/30/95 U Coyote kill 02/22/95
M 7 1/30/95 U Coyote kill 02/09/95
F 31 1/31/95 N Alive
M 8 2/1/95 U Coyote kill 02/13/95
F 68 2/2/95 U Dead unknown 02/13/95
M 20 2/6/95 N Legal harvest 11/05/95
F 68 2/10/95 Y Alive
M 8 2/10/95 Y Legal harvest 10/14/95
F 44 2/13/95 N Alive
M 8 2/16/95 N Alive
M 8 2/22/95 Y Censored 07/25/95

115

 

 

Appendix Table 6 (cont’d).
Sex Age (months) Capture Migratory Fate
at capture date status‘
F 80 2/24/95 N Alive
M 8 2/28/95 U Censored 03/13/95
F 81+ 3/3/95 N Alive
M 9 3/9/95 U Drowned 03/12/95
F 9 3/19/95 N Illegal harvest 08/08/95
F 81+ 3/24/95 Y Alive
M 10 4/8/95 N Alive

 

‘ Y = migratory, N = Non-migratory, U = migratory status not known.

116

Appendix Table 7. Sex, age, capture date, migratory status, and fate of ear-tagged

 

 

deer.
Sex Age (months) Capture Migratory Fate
at capture date status‘

M 8 02/06/94 Y Legal harvest 11/15/95
M 9 03/02/94 U Legal harvest 11/15/94
F 7 1/31/95 U Legal harvest 12/05/95
F 32 2/5/95 Y Legal harvest 11/22/95
F 33 3/2/95 U Road killed 03/05/95
M 9 3/7/95 Y Legal harvest 11/15/95

 

‘ Y = migratory, U = migratory status not known.

1 17
Appendix Table 8. Doe ages and number of fawns seen with collared does in 1994 and

1995. Dashes represent deer that were not observed due to age or sex during a given

 

 

 

 

year.
1994‘ 1995
Frequency Doe age # fawns Doe age # fawns
(yearS) (yearS)
151.396” 2 0 - -
151.405 3 l 4 1
151.415 3 1 4 2
151.424 3 0 4 0
151.436 3 0 4 0
151.676" 15 0 4 1
151.686 3 0 4 1
151.716 3 0 4 1
151.726" 12 0 7+ 0
151.735 2 2 3 1
151.915 3 not seen 4 1
151.895 - - 3+ 0
151.205 - - 2 0
151.774 2+ not seen 3+ 0
151.926 - - 2 1
151.745 3+ not seen 4+ 0
151.235 - - 7 1
151.496 5 not seen 6 1
151.275 - - 2 0
151.785" 4 not seen 6 1
151.176 - - 2 1
151.444 3 not seen 4 1
151.226 - - 6 0

 

b different individuals between years

‘ individuals never sighted more than once

Appendix Table 9. Deer counts ﬁ'om 1994 and 1995 road surveys and productivity

118

ratios from actual and adjusted‘ counts.

 

 

 

Actual Adjusted
# does # fawns ratio # does # fawns ratio
1994
13 July 103 15 0.146 92.7 15 0.162
25 July 60 8 0.133 54 8 0.148
5 August 98 36 0.367 88.2 36 0.408
15 August 11 1 47 0.423 99.9 47 0.470
15 September 36 26 0.722 32.4 26 0.802
6 October 73 54 0.740 65.7 54 0.822
16 November 1 1 6 0.545 9.9 6 0.606
1995
83 13 0.157 74.7 13 0.174
25 July 110 29 0.264 99 29 0.293
8 August 73 39 0.534 65.7 39 0.594
21 August 118 53 0.449 106.2 53 0.499
15 September 133 96 0.722 119.7 96 0.802
21 October 85 55 0.647 76.5 55 0.719

 

‘ #ofdoesdecreasedby 10%

t- c... o’rw

119

Appendix Table 10. Deer counts from 1994 and 1995 road surveys and buck-doe ratios

 

 

 

from actual and adjusted‘ counts.
Actual Adjusted

# bucks # does ratio # bucks # does ratio

1994
2 July 17 71 0.239 18.7 63.9 0.293
13 July 18 103 0.175 19.8 92.7 0.214
25 July 18 60 0.300 19.8 54 0.367
5 August 32 98 0.327 35.2 88.2 0.399
15 August 21 111 0.189 23.1 99.9 0.231
15 September 4 36 0.111 4.4 32.4 0.136
6 October 9 73 0.123 9.9 65.7 0.151

1995
21 June 16 77 0.208 17.6 69.3 0.254
12 July 28 83 0.337 30.8 74.7 0.412
25 July 21 110 0.191 23.1 99 0.233
8 August 3 73 0.041 3.3 65.7 0.050
21 August 25 118 0.211 27.5 106.2 0.259
15 September 24 133 0.180 26.4 119.7 0.221
21 October 5 85 0.058 5.5 76.5 0.072

 

‘ #ofbucksincreasedby 10%,#ofdoesdecreasedby10%

LITERATURE CITED

LITERATURE CITED

Albert, D. A., S. R. Benton, and B. V. Barnes. 1986. Regional landscape ecosystems of
Michigan. School of Natural Resources, University of Michigan. 32pp.

Alverson, W. S., D. M. Waller, and S. L. Solheirn. 1988. Forests too deer: edge eﬁ‘ects
in northern Wisconsin. Conserv. Bio. 22348-3 58.

Austin, D. D., and P. J. Umess. 1993. Evaluating production losses from mule deer
depredation in alfalfa ﬁelds. Wildl. Soc. Bull. 21:397-401.

Bartlett, 1. H. 1932. Live-trapping Michigan Whitetail deer. Mich. Acad. of Arts, Sci.,
and Lett. 17:487-492.

Bashore, T. L., W. M. Tzilkowski, and E. D. Bellis. 1985. Analysis of deer-vehicle
collision sites in Pennsylvania. J. Wildl. Manage. 49:769-774.

Beier, P. 1987. Sex differences in quality of white-tailed deer diets. J. Mammal 68:323-
329.

Beier, P., and D. R. McCullough. 1990. Factors inﬂuencing white-tailed deer activity
patterns and habitat use. Wildl. Monogr. 109. 5 1pp.

Bender, L. C., and J. B. Hauﬂer. 1987. A white-tailed deer HSI for the upper great
lakes region. Michigan State University, Dept. of Fish. and Wildl., unpublished.

Bennett, L. J ., P. F. English, and R. McCain. 1940. A study of deer populations by use
of pellet-group counts. J. Wildl. Manage. 41398-403.

Blouch, R. I. 1984. Northern great lakes states and Ontario forests. Pages 391-410 iir
L. K. Halls, ed. White-tailed deer ecology and management. Wildl. Manage.
Inst. Stackpole Books. Harrisburg, Pa.

Bowyer, R. T. 1984. Sexual segregation in southern mule deer. J. Mammal 65:410-
417.

120

121
Braun, K. F. 1996. Ecological factors inﬂuencing white-tailed deer damage to
agricultural crops in northern lower Michigan. M. S. Thesis. Michigan State Univ.
East Lansing, Mich.

Brown, C. G. 1992. Movement and migration patterns of mule deer in southeastern
Idaho. J. Wildl. Manage. 56:246-253.

Brown, T. L., D. J. Decker, and C. P. Dawson. 1978. Willingness of New York farmers
to incur white-tailed deer damage. Wildl. Soc. Bull. 6:23 5-239.

Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals. J.
Mammal 24:346-352.

Carlsen, J. C., and R E. Farmes. 1957. Movements of white-tailed deer tagged in
Minnesota. J. Wildl. Manage. 21 :397-401.

Caughley, G. 1977 . Analysis of vertebrate populations. John Wiley and Sons Inc. New
York, NY. 234pp.

Clover, M. R 1954. A portable deer trap and catch-net. Calif. Fish and Game. 40:376-
373.

. 1956. Single-gate deer trap. Calif. Fish and Game. 42: 199-201.

Coe, R J ., R. L. Downing, and B. S. McGinnes. 1980. Sex and age bias in hunter-killed
white-tailed deer. J. Wildl. Manage. 44:245-249.

Conover, M. R. 1994. Perceptions of grass-roots leaders of the agricultural community
about wildlife damage on their farms and ranches. Wildl. Soc. Bull. 22:94-100.

Dalhberg, B. L., and R. C. Guettinger. 1956. The white-tailed deer in Vlfrsconsin.
\Vlsconsin Conserv. Dept. Tech. Wildl. Bull. 14. 282pp.

Decker, D. J ., and T. L. Brown. 1982. Fruit growers vs. other farmers’ attitudes
toward deer in New York. Wildl. Soc. Bull. 10:150-155.

DeYoung, C. A 1989. Aging live white-tailed deer on southern ranges. J. Wildl.
Manage. 53:519-523.

Dixon, K. R, and J. A. Chapman. 1980. Harmonic mean measure of animal activity
areas. Ecology 61:1040-1044.

Downing, R. L., and B. S. McGinnes. 1969. Capturing and marking white-tailed deer
fawns. J. Wildl. Manage. 33:711-714.

122

, E. D. Michael, and R J. Poux. 1977. Accuracy of sex and age ratio counts of
white-tailed deer. J. Wildl. Manage. 41:709-714.

Drolet, C. A. 197 6. Distribution and movements of white-tailed deer in southern New
Brunswick in relation to environmental factors. Can. Field Nat. 90: 123-136.

Dudderar, G., J. Hansen, J. Hauﬂer, B. Peyton, H. Prince, and S. Winterstein. 1989.
Michigan’s deer damage problems: an analysis of the problems with
recommendations for ﬁlture research and communication. The Deer Damage
Committee, Dept. Fisheries and Wildl., MI State Univ., East Lansing Mich.

Dusek, G. L., A. K. Wood, and R. J. Mackie. 1988. Habitat use by white-tailed deer in
prairie-agricultural habitat in Montana. Prairie Nat. 20:135-142.

, R. J. Mackie, J. D. Herriges Jr., and B. B. Compton. 1989. Population ecology
of white-tailed deer along the lower Yellowstone river. Wildl. Monogr. 104.

68pp.

Eberhardt, L. L., and R. C. Van Etten. 1956. Evaluation of the pellet-group count as a
deer census method. J. Wildl. Manage. 20:70-74.

Eichenlaub, V. L., J. R. Harman, F. V. Numberger, and H. J. Stolle. 1990. The climatic
atlas of Michigan. The University of Notre Dame Press. Notre Dame, Ind.
165pp.

Flyger, V., and T. Thoerig. 1962. Crop damage caused by Maryland deer. Proc. Annu.
Conf. SE Assoc. Fish and Game Comm. 16:45-52.

Fuller, T. K. 1990. Dynamics of a declining white-tailed deer population in north-
central Minnesota. Wild]. Monogr. 110. 37pp.

Garrison, R L., and J. C. Lewis. 1987 . Effects of browsing by white-tailed deer on
yields of soybeans. Wildl. Soc. Bull. 15:555-559.

Garrott, R. A, and G. C. White. 1982. Age and sex selectivity in trapping mule deer. J.
Wildl. Manage. 46:1083-1086.

Gavin, T. A., L. H. Suring, P. A. Vohs, and E. C. Meslow. 1984. Population
characteristics, spatial organization, and natural mortality in the Columbian
white-tailed deer. Wildl. Monogr. 91. 41pp.

Gilbert, F. F. 1966. Aging white-tailed deer by annuli in the cementum if the ﬁrst
incisor. J. Wildl. Manage. 30:200-203.

123

Gladfelter, H. L. 1980. Deer population estimators in the midwest farmland. Pages 5-
11 ”m R. L. Hines and S. Nehls eds. White-tailed deer population management in
the northcentral states. Proc. 1980 Symp. North Cent. Sect. Wildl. Soc. 116pp.

Gladfelter, H. L. 1984. Midwest agricultural region. Pages 427-440 E L. K. Halls, ed.
White-tailed deer ecology and management. Wildl. Manage. Inst. Stackpole
Books. Hanisburg, Pa.

Gorynska, W. 1981. Method of determining relations between the extent of damage in
farm crops, big game numbers, and environmental conditions. Acta Theriologoca
26: 469-481.

Habeck, J. R. 1960. Wmter deer activity in the white cedar swamps of northern
Wisconsin. Ecology. 41:327-333.

Hamlin, K. L., R. J. Mackie, J. G., and D. F. Pac. 1982. Effect of capture and marking
on fawn production in deer. J. Wildl. Manage. 46: 1086-1089.

Hammerstrom, F. N. Jr., and J. Blake. 1939. Winter movements and winter foods of
white-tailed deer in central Wisconsin. J. Mammal. 20:206-215.

Hansen, C. 1978. Social costs of Michigan’s deer habitat improvement program.
Michigan Dept. of Nat. Resour. Wildl. Div. Rep. 2808.

Haugen, A. O. 1975. Reproductive performance of white-tailed deer in Iowa. J.
Mammal. 56:151-159.

Heezen, K. L., and J. R. Tester. 1967. Evaluation of radio-tracking by triangulation
with special reference to deer movements. J. Wildl. Manage. 31: 124-141.

Hobbs, N. T., and T. A. Hanley. 1990. Habitat evaluation: do use / availability data
reﬂect carrying capacity? J. Wildl. Manage. 54:515-522.

Hoskinson, R L., and L. D. Mech. 1976. White-tailed deer migration and its role in
wolf predation. J. Wildl. Manage. 40:429-441.

Kanji, G. K. 1993. 100 statistical tests. SAGE publications Newbury Park, Calif
216pp.

Knapp, B. D. 1988. Soil survey of Presque Isle county, Michigan. U.S.D.A. Soil
Conservation Service. 252pp.

124

Kohn, B. E., and J. J. Mooty. 1971. Summer habitat of white-tailed deer in north-
central Minnesota. J. Wildl. Manage. 35:476-487.

Korschgen, L. J. 1962. Foods of Missouri deer, with some management implications.
J. erdl. Manage. 26:164-172.

Kufeld, R. C., and D. C. Bowden. 1995. Mule deer and white-tailed deer inhabiting
eastern Colorado plains river bottoms. Colorado Div. of Wildl. Tech. Publ. 41.

58pp.

, , and D. L. Schrupp. 1988. Inﬂuence of hunting on movements of female
mule deer. J. Range. Manage. 41:70-72.

Lancia, R A, J. D. Nichols, and K. H. Pollock. 1994. Estimating the number of
animals in wildlife populations. Pages 215-253 'm T. A. Bookout, ed. Research
and management techniques for wildlife and habitats. Fifth ed. The Wildlife
Society, Bethesda, Md.

Langenau, E. E. 1979. Nonconsumptive uses of the Michigan deer herd. J. Wildl.
Manage. 43:620-625.

. 1994. Commemorating 100 years of licensed deer hunting in Michigan.
Michigan Dept. of Nat. Resour. Wildl. Div. Rep. 3213.

, and J. M. Lerg. 1976. The eﬁ’ects of winter nutritional stress on maternal and
neonatal behavior in penned white-tailed deer. Appl. Anim. Ethol. 21207-223.

Larson, T. J ., O. J. Rongstad, and F. W. Terbilcox. 1978. Movement and habitat use of
white-tailed deer in south-central Wisconsin. J. Wildl. Manage. 42:113-117.

Lenarz, M. S. 1987. Economics of forest openings for white-tailed deer. Wildl. Soc.
Bull. 15:568-573.

Lowry, D. A, and K. L. McArthur. 1978. Domestic dogs as predators on deer. Wildl.
Soc. Bull. 6:38-39.

Marchington, R L., and D. H. Hirth 1984. Behavior. Pages 129-168 in L. K. Halls, ed.
White-tailed deer ecology and management. Wildl. Manage. Inst. Stackpole
Books. Harrisburg, Pa.

Mattfeld, G. F., R W. Sage, J. E. Wiley, and D. F. Behrend. 1974. Seasonal difference
in sex ratio of box trapped deer. J. Wildl. Manage. 38:563-565.

125

Mautz, W. W. 1978. Sledding on a bushy hillside: the fat cycle in deer. Wildl. Soc.
Bull. 6:88-90.

Mayﬁeld, H. 1961. Nesting success calculated ﬁ'om exposure. Wilson Bulletin. 73 :25 5-
261.

. 1975. Suggestions for calculating nest success. Wilson Bulletin. 87:456-466.

McCabe, R E., and T. R. McCabe 1984. Of slings and arrows: a historical
retrospection. Pages 19-72 i_n_ L. K. Halls, ed. White-tailed deer ecology and
management. Wildl. Manage. Inst. Stackpole Books. Harrisburg, Pa.

McCaffery, K. R., and W. A. Creed. 1969. Signiﬁcance of forest openings to deer in
northern \Vrsconsin. Wisconsin Dept. of Nat. Resour. Tech. Bull. 44. 104pp.

McCullough, D. R. 1974. Modiﬁcation of the Clover deer trap. Calif. Fish and Game.
61:242-244.

, D. H. Hirth, and S. J. Newhouse. 1989. Resource partitioning between sexes in
white-tailed deer. J. Wildl. Manage. 53:277-283.

Messier, F., and C. Barrette. 1985. The efﬁciency of yarding behavior by white-tailed
deer as an antipredator strategy. Can. J. Zool. 63:783-789.

Moen, A N. 1968. Energy exchange of white-tailed deer, western Minnesota.
Ecology. 49:676-682.

. 1976. Energy conservation by white-tailed deer in the winter. Ecology. 57:192-
198.

. 197 8. Seasonal changes in heart rates, activity, metabolism, and forage intake of
white-tailed deer. J. Wildl. Manage. 42:715-738.

Montgomery, G. G. 1963. Nocturnal movements and activity rhythms of white-tailed
deer. J. Wildl. Manage. 27:422-427.

Moore, W. G., and R. H. Folk. 1977 . Crop damage by white-tailed deer in the
southeast. Proc. Annu. Conf. SE Assoc. Fish and Game Comm. 32:263-268.

Mooty, J. J. 1980. Monitoring deer populations in the northern forested areas of the
midwest. pp. 13-22 i_n_ R. L. Hines and S. Nehls eds. White-tailed deer
population management in the northcentral states. Proc. 1980 Symp. North
Cent. Sect. Wildl. Soc. 116pp.

126

, P. D. Karns, and T. K. Fuller. 1987. Habitat use and seasonal range size of
white-tailed deer in northcentral Minnesota. J. Wildl. Manage. 51:644-648.

Murphy, D. A., and J. A. Coates. 1976. Effects of dietary protein on deer. Trans.
North. Am. Wildl. Nat. Resour. Conf. 31:129-139.

Murphy, R K., N. F. Payne, and R. K. Anderson. 1985. White-tailed deer use of an
irrigated agriculture-grassland complex in central Wisconsin. J. Wildl. Manage.
49: 125-128.

Nams, V. O. 1989. Eﬂ‘ects of radiotelemetry error on sample size and bias when testing
for habitat selection. Can. J. 2001. 67:1631-1636.

. 1990. Locate II user’s guide. Pacer. Truro, Nova Scotia. 84pp.

Neﬁ‘, D. J. 1968. The pellet-group count technique for big game trend, census, and
distribution: a review. J. Wildl. Manage. 32:597-614.

Nelson, M. E. 1994. Migration bearing and distance memory by translocated white-
tailed deer in northeastern Minnesota. Can. Field Nat. 108:74-76.

. 1995. Winter range arrival and departure of white-tailed deer in northeastern
Minnesota. Can. J. Zool. 73: 1069-1076.

, and L. D. Mech. 1981. Deer social organization and wolf predation in
northeastern Minnesota. Wildl. Monogr. 77. 53pp.

, . 1984. Home-range formation and dispersal of deer in northeastern
Minnesota. J. Mammal. 65:567-575.

 

, . 1986. Mortality of white-tailed deer in northeastern Minnesota. J.
Wildl. Manage. 50:691-698.

 

, and A. Woolf. 1987 . Mortality of white-tailed deer fawns in southern Illinois. J.
Wildl. Manage. 51:326-329.

Nixon, C. M., L. P. Hansen, and P. A. Brewer. 1988. Characteristics of winter habitats
used by deer in Illinois. J. Wildl. Manage. 52:552-555.

, ,and J. E. Chelsvig. 1991 Ecology of the white-tailed deer m an
intensively farmed region of Illinois. Wlldl.Monogr.118.77pp.

 

127

Otis, D. L., K. P. Bumharn, G. C. White, and D. R. Anderson. 1978. Statistical
inference from capture data on closed animal populations. Wildl. Monogr. 62.
135pp.

Ozoga, J. J ., and E. M. Harger. 1966. Winter activities and feeding habits of northern
Michigan coyotes. J. Wildl. Manage. 30:809-818.

. 1968. Variations in microclimate in a conifer swamp deeryard in northern
Michigan. J. Wildl. Manage. 32:574-585.

. 1972. Aggressive behavior of white-tailed deer at winter cuttings. J. Wildl.
Manage. 36:861-868.

, R V. Doepker, and M. S. Sargent. 1994. Ecology and management of the
white-tailed deer in Michigan. Michigan Dept. of Nat. Resour. Wildl. Div. Rep.
3209.

Palmer, W. L., G. M. Kelly, and J. L. George. 1982. Alfalfa losses to white-tailed deer.
Wildl. Soc. Bull. 10:259-261.

Pollock, K. H., S. R. Winterstein, C. M. Bunck, and P. D. Curtis. 1989. Survival
analysis in telemetry studies : the staggered entry design. J. Wildl. Manage. 53:7-
15.

Putnam, R. J. 1986. Foraging by roe deer in agricultural areas and impact on arable
crops. J. Appl. Ecol. 23:91-99.

Ransom. B. A. 1966. Detem1ining age of white-tailed deer from layers in cementum of
molars. J. Wildl. Manage. 30:197-199.

. 1967. Reproductive biology of white-tailed deer in Manitoba. J. Wildl.
Manage. 31:114-123.

Reis, T. F. 1990. 1990 Michigan deer seasons. Michigan Dep. of Nat. Resour. Wildl.
Div. Rep. 3147. 15pp.

Rogers, L. L. 1987. Seasonal changes in defecation rates of free-ranging white-tailed
deer. J. Wildl. Manage. 51:330-333.

Rongstad, O. J ., and R. A. McCabe. 1984. Capture techniques. Pages 655-676 in L. K.
Halls, ed. White-tailed deer ecology and management. Wildl. Manage. Inst.
Stackpole Books. Harrisburg, Pa.

128

, and J. R. Tester. 1969. Movements and habitat use of white-tailed deer in
Minnesota. J. Wildl. Manage. 33:366-379.

Root, B. G., E. K. Fritzell, and N. F. Giessman. 1988. Effects of intensive hunting on
white-tailed deer movement. Wildl. Soc. Bull. 16:145-151.

Roper, L. A, R L. Schmidt, and R. B. Gill. 1971. Techniques of trapping and handling
mule deer in northern Colorado with notes on using automatic data processing
for data analysis. Proc. West. Assoc. Game and Fish Comm. 51:471-477.

Roseberry, J. L., and W. D. Klimstra. 1970. Productivity of white-tailed deer on Crab
Orchard National Wildlife Reﬁrge. J. Wildl. Manage. 34:23-28.

Ryel, L. A 1971. Evaluation of pellet-group surveys for estimating deer populations in
Michigan. Ph.D. Diss. Michigan State Univ. East Lansing, Mich. 237pp.

, L. D. Fay, and R C. Van Etten. 1961. Validity of age determination in
Michigan deer. MI Acad. Sci., Arts, and Lett. 46:289-316.

Sanderson, G. C. 1966. The study of mammal movements - a review. J. Wildl.
Manage. 30: 215-235.

SAS Institute 1985. SAS” user’s guide: statistics, version 5 ed. SAS Institute Inc.,
Cary, NC. 956pp.

Severinghaus, C. A. 1949. Tooth development and wear as criteria of age in white-
tailed deer. J. Wildl. Manage. 13:195-216.

Siegel, S. and N. J. Castellan, Jr. 1988. Nonparametric statistics for the behavioral
sciences. Second ed. McGraw-Hill Inc. New York, NY. 399pp.

Silver, H., N. F. Colovos, J. B. Holter, and H. H. Hayes. 1969. Fasting metabolism of
white-tailed deer. J. erdl. Manage. 33:490-498.

Sokal, R. R. and F. J. Rohlf. 1995. Biometry. Third ed. W. H. Freemen and Co. New
York, NY. 887pp.

Sparrowe, R D., and P. F. Springer. 1970. Seasonal activity patterns of white-tailed
deer in eastern South Dakota. J. Wlldl. Manage. 34:420-431.

Suring, L. H., and P. A. Vohs. 1979. Habitat use by Columbian white-tailed deer. J.
Wildl. Manage. 43:610-619.

Swift, R W. 1948. Deer select most nutritious forages. J. Wildl. Manage. 12: 109-1 10.

129

Tierson, W. C., G. F. Mattfeld, R W. Sage Jr., and D. F. Behrend. 1985. Seasonal
movements and home ranges of white-tailed deer in the adirondacks. J. Wildl.
Manage. 49:760-769.

Van Deelan, T. R 1995. Seasonal migrations and mortality of white-tailed deer in
Michigan’s upper peninsula. MS. Thesis. Michigan State Univ. East Lansing,
Mich. 158pp.

Vecellio, G. M., R. H. Yahner, and G. L. Storm. 1994. Crop damage by deer at
Gettysburg Park. Wildl. Soc. Bull. 22:89-93.

Venue, L. J. 1968. An index of winter weather severity for northern deer. J. Wildl.
Manage. 32:566-574.

. 1973. Movements of white-tailed deer in upper Michigan. J. Wildl. Manage.
37 2545-551.

, and D. E. Ullrey. 1984. Physiology and nutrition. Pages 91 -118 'm L. K. Halls,
ed. White-tailed deer ecology and management. Wildl. Manage. Inst. Stackpole
Books. Harrisburg, Pa.

White, G. C., D. R. Anderson, K. P. Bumharu, and D. L. Otis. 1982. Capture-recapture
and removal methods for sampling closed populations. Los Alamos Natl. Lab.,
LA 8787-NERP. 235pp.

, and R A. Garrott. 1986. Eﬂ‘ects of biotelemetry triangulation error on detecting
habitat selection. J. Wlldl. Manage. 50:509-513.

, . 1990. Analysis of wildlife radio-tracking data. Academic Press Inc.
San Diego, Calif. 383pp.

 

, , R. M. Bartmann, L. H. Caprenter, and A W. Alldredge. 1987. Survival
of mule deer in northwestern Colorado. J. Wildl. Manage. 51 :852—859.

Williamson, S. J ., and D. H. Hirth. 1985. An evaluation of edge use by white-tailed
deer. Wlldl. Soc. Bull. 13:252-257.

Wmterstein, S. R., H. Campa HI, K. F. Millenbah. 1995. Status and Potential of
Michigan Natural Resources Special Report 75. 39pp.

Woodson, D. L., E. T. Reed, R. L. Downing, and B. S. McGinnes. 1980. Effect of fall
orphaning on white-tailed deer fawns and yearlings. J. Wildl. Manage. 44:249-
252.

130

Zagata, M. D., and A O. Haugen. 1974. Inﬂuence of light and weather on observability
of Iowa deer. J. Wildl. Manage. 38:220-228.

 

 

 

tritiu

 

.1194

(.2...

1ti..l|..t
1

(at:
I.\v$.r
I}
3&7de
.33 We.
3... . .n r 17.
. 25.4..- 1.. 1... r
. n- P.“ ”u... .Wuthxlntt
. t . t f. at!
11 .r.... . . . Lama‘sﬁaﬂ .
1 t... .r. e: 11 r
1. .

lg
Janus.
KR. n. r

1:? r
H}...
atom