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

 

DATE DUE

DATE DUE

DATE DUE

 

 

I'
(,3 ha)

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDatoDue.p65-p.15

 

POPULATION DYNAMICS OF MOOSE IN THE
WESTERN UPPER PENINSULA OF MICHIGAN, 1999-2001

By

William B. Dodge, Jr.

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

2002

ABSTRACT

POPULATION DYNAMICS OF MOOSE IN THE WESTERN UPPER
PENINSULA OF MICHIGAN, 1999-2001

By

William 8. Dodge, Jr.

Moose (Alces alces amen’cana) are native to Michigan but were possibly
extirpated from the entire state by the end of the 19th century. With the goal of
reestablishing a self sustaining population of moose in the Upper Peninsula of
Michigan, the Michigan Department of Natural Resources (MDNR) translocated
61 moose (59 of which were released) from Ontario, Canada to western
Marquette county in the winters of 1985 (n=31) and1987 (n=30). Based on
evaluations of potential habitat and optimistic population growth projections the
objective was to have a self-sustaining population of 1000 moose by the year
2000. However, population size estimates from aerial surveys conducted in the
winters of 1996 and 1997 were well below 1000. To evaluate the possible
reasons for the slower than expected population growth, 84 moose were
captured and radio-collared in the winters of 1999-2001. Survival of all age
classes has been excellent; annual survival of adults was >85% and first-year
calf survival >70%. Seventy-four percent of adult cows were pregnant which is
lower than the 84% average for moose in North America. In addition, radio-
tagged moose, primarily yearlings, have dispersed out of the study area at an
annual rate of about 6%. Preliminary indications are that low reproductive output

is the likely cause of the slower than predicted population growth.

ACKNOWLEDGMENTS

I would like to thank the Wildlife Division of the Michigan Department of
Natural Resources and Michigan State University for providing the primary
funding for this research. The Michigan Involvement Committee of Safari Club
lntemational also deserves special recognition for their continued funding support
and donations of equipment to The Michigan Moose Project since the initial
moose release in 1985.

I am grateful to my major professor, Dr. Scott Winterstein for giving me the
opportunity to work on this exciting and high profile research project. I also thank
my committee members, Dr. Rique Campa of the Fisheries and Wildlife
Department, Dr. Dean Beyer of the MDNR and Dr. Barbara Lundrigan of the
Zoology Department for their comments and suggestions that greatly improved
this manuscript. Dr. Beyer deserves additional recognition for initially proposing
the project, organizing and managing the moose captures and teaching me how
to perform a necropsy.

I thank the dedicated personnel of the MDNR Wildlife Division, Rob Aho,
Torn Bach, Terry Gouza, Jim Hammill, John Hendrickson, Brad Johnson, Terry
Minzey and Brian Roell for their efforts during the moose captures. Additionally, I
would like to thank Rob Aho for sharing his aerial radio-telemetry expertise and
giving me pointers to improve my skill at relocating moose from the air. Brian
Roell deserves a special thank you for helping me do much of the tedious,
mundane tasks associated with wildlife research (e.g., making copies of data

sheets, verifying moose relocation points, shipping samples, etc). Brian was also

always available to search for a dead moose and help with the necropsy, which
inexplicably seemed to occur late on a Friday afternoon.

I appreciate the skill, dedication and high priority given to safety of the
MDNR Forest Management Division airplane pilots, Neil Harri, Dean Minett and
Steve Adkins who spent countless hours relocating and searching for moose.
Neil deserves an additional thank you for making our moose relocation flights a
high priority of the Forest Management Division and for helping train many new
employees how to do aerial radio-telemetry. An additional thanks to pilots Pete
Petosky and Kevin Jacobs who filled in when other pilots were unavailable.

I appreciate all the efforts, suggestions, hard work, dedication, and
camaraderie of my field assistants, Rob Atkinson, Bill Martin, Vic Lane, Mark
Westbrock, Nate Seward, David Jentoft, Eric North, and Jerry Hill.

I would also like to thank Clarence and Vie Rivers for allowing me to rent
their vacant house in Negaunee on short notice the first year of the project. Paul
and Peg Gravedoni whom I also rented a house from in Negaunee were not only
wonderful landlords but also great neighbors.

Lastly, but most importantly, I thank my wife, Trish, for her love and

support and for encouraging me to follow my dreams.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

page
LIST OF TABLES ................................................................................................ vii
LIST OF FIGURES .............................................................................................. ix
INTRODUCTION .................................................................................................. 1
RESEARCH OBJECTIVES ................................................................................. 12
STUDY AREA ..................................................................................................... 13
METHODS .......................................................................................................... 17
Capturing and radio-tagging ............................................................................. 17
Radio-telemetry relocating ............................................................................... 18
Aerial relocations .......................................................................................... 18
Ground relocations ........................................................................................ 19
Survival and mortality ....................................................................................... 20
Movements ...................................................................................................... 26
Home range si_ze and dispersion ................................................................... 26
Dispersal and home range development ...................................................... 28
Reproduction ................................................................................................... 29
Pregnancy determinations ............................................................................ 29
Productivity ...................................................................................................... 31
Natalitv and Recruitment ............................................................................... 31
Population size estimation and projection ........................................................ 33
RESULTS ........................................................................................................... 35
Capturing and radio-tagging ............................................................................. 35
Radio telemetry relocations ............................................................................. 37
Aerial relocations .......................................................................................... 37
Ground relocations ........................................................................................ 39
Survival and mortality ....................................................................................... 39
Adult and yearling ......................................................................................... 39
Bull and cow .................................................................................................. 46
First-year calf ................................................................................................ 46
Calves 0—6 months of age ............................................................................. 46
Calves 6-12 months of ag_e_ ........................................................................... 48
Mortalities ..................................................................................................... 48
Movements ...................................................................................................... 50
Home range size and dispersion ................................................................... 50
Espersal and home range devel0pment ...................................................... 57
Reproduction ................................................................................................... 71
Pregnancy determination .............................................................................. 71

 

Productivity ...................................................................................................... 76

Natality and Recruitment ............................................................................... 76
Population size estimation and projection ........................................................ 78
DISCUSSION ..................................................................................................... 84
Survival and mortality ....................................................................................... 84
Movements ...................................................................................................... 87
Reproduction and productivity .......................................................................... 95
Population size estimation and projection ........................................................ 96
LIMITING FACTORS .......................................................................................... 99
MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS ....................... 103
APPENDIX ........................................................................................................ 107
LITERATURE CITED ........................................................................................ 114

vi

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 1 1.

LIST OF TABLES

page

Sex and age breakdown of moose radio-tagged in 1995, January-
February 1999, January 2000, and February 2001 in the western
Upper Peninsula of Michigan. . . . 35

Annual and seasonal survival rates of radio-tagged adult and
yearling moose studied during 1999-2001 in the western Upper
Peninsula of Michigan. . . . . 41

First-year, 0-6 months of age, and 6-12 months of age joint survival
rates of radio-tagged and un-tagged calf moose of radio-tagged cows
seen each spring during 1999-2001 in the western Upper Peninsula of
Michigan. . . . . . . . 42

Moose id, age at death, sex, femur marrow fat %, approximate date of
death and diagnosed cause of death of radio-tagged moose studied
during 1999-2001 in the western Upper Peninsula of Michigan. . 49

Percentage of overlap and distance between summer and winter
home ranges of migratory adult radio-tagged moose studied during
1999 - 2001 in the western Upper Peninsula of Michigan. . 51

Annual and seasonal home range sizes of resident and migratory
adult radio-tagged moose studied during 1999-2001 in the western
Upper Peninsula of Michigan.. . . . . . 52

Annual home range sizes of radio-tagged bull and cow moose studied
during 1999-2001 in the western Upper Peninsula of Michigan. . 55

Annual and seasonal Euclidian (or linear) distance (AD) between pairs
of relocations of adult and yearling radio-tagged moose studied during
1999-2001 in the western Upper Peninsula of Michigan. . . 56

Radio-tagged moose that dispersed out of the core study area during
1999-2001 in the western Upper Peninsula of Michigan. . . 58

Annual dispersal rates of radio-tagged adult and yearling moose
studied during 1999-2001 in the western Upper Peninsula of Michigan.
. . . . . 60

Pregnancy specific protein B (PSPB) concentration (ng/mL) in blood
serum collected in winter from radio-tagged cow moose in the western

vii

Table 12.

Table 13.

Table 14.

Table 15.

Upper Peninsula of Michigan, 1999-2001. All cows with detectable
levels of PSPB were considered to be pregnant. . . . 72

Progesterone concentration (ng/mL) in blood serum collected in winter
from radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. The 95% Upper tolerance limit (95% UTL) or
dividing serum progesterone level between pregnant and non-
pregnant cows was 1.68 ng/mL. . . . . . 73

Progesterone concentration (pg/g) in fecal material collected in winter
from radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. The 95% Upper tolerance limit (95% UTL) or
dividing fecal progesterone level between pregnant and non-pregnant
cows was 5.17 pg/g. . . . . . . . 75

Annual reproductive parameters of radio-tagged cow moose studied
during 1999-2001 in the western Upper Peninsula of Michigan. . 78

Reproductive parameters of radio-tagged cow moose and survival of
their offspring during 1985-1990,1995-1996 In the western Upper
Peninsula of Michigan. . . 80

Appendix Table 1 3'”. Records of moose sightings or evidence of moose by

county in Michigan’s Upper Peninsula, 1954-1966. . 108

Appendix Table 2 ’1’. Records of moose sightings or evidence of moose by

county in Michigan’s Upper Peninsula, 1967-1983. . 109

Appendix Table 3. Physical examination scores and description for assessing

moose condition during capture and when conducting field
necropsies. . . . . . . . 110

Appendix Table 4. Moose id, sex, age at capture, date captured, type of radio

collar, and fate as of 30 June, 2001 of radio-tagged moose
studied during 1999-2001 in the western Upper Peninsula of
Michigan. . . . . . . . 11 1

viii

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 1 1.

LIST OF FIGURES

page

Moose study area during 1999-2001, and 1985 and 1987 moose
translocation release sites in the western Upper Peninsula of
Michigan. . . . . . . . . 7

Moose population size estimates for 1985-1997 in the western Upper
Peninsula of Michigan based on a deterministic book-keeping model,
POP-ll, a population modeling computer program, and aerial surveys
conducted by the MDNR in 1991,1996, and 1997.. . . 9

Capture locations of moose radio-tagged in 1995 and during 1999-
2001 in the western Upper Peninsula of Michigan.. . . 36

Distribution of radio-tagged moose aerial relocations (A) and ground
relocations (B) by time of day in the western Upper Peninsula of
Michigan during 1999-2001. . . . . . . 38

Mean number of moose observations by month during aerial
relocation flights when moose were seen in the western Upper
Peninsula of Michigan during February 1999-June 2001. . 40

Number of radio-tagged adult and yearling moose deaths by month in
the western Upper Peninsula of Michigan during 1 June 1999-31 May
2001. . . . . . . . 43

Locations of recovered mortalities of radio-tagged moose within the
core study area during 1999-2001 in the western Upper Peninsula of
Michigan. . . . . . . . . 45

Number of radio-tagged and un-tagged calf moose deaths by month
in the western Upper Peninsula of Michigan from 1 June 1999-31
May 2000. . . . . . . . . 47

Relocations and movements of six radio-tagged moose that
dispersed out of the core study area during 1999-2001 in the western
Upper Peninsula of Michigan. . . . . . 59

Relocations and movements of radio-tagged cow moose (No. 185)
during 2001 in the western Upper Peninsula of Michigan. . 61

Relocations and movements of radio-tagged cow moose (No. 102)
during 1999—2001 in the western Upper Peninsula of Michigan. 62

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Relocations and movements of radio-tagged yearling male moose
(No. 104) during 1999-2000 In the western Upper Peninsula of
Michigan. . . . . 64

Relocations and movements of radio-tagged yearling female moose
(No. 105) during 1999- 2000 In the western Upper Peninsula of
Michigan. . . . . . 66

Relocations and movements of radio-tagged yearling female moose
(No. 134) during 1999-2000 In the western Upper Peninsula of
Michigan. . . . . 67

Relocations and movements of radio-tagged yearling male moose
(No. 106) during 1999- 2000 In the western Upper Peninsula of
Michigan. . . . . . 68

Relocations and movements of radio-tagged yearling male moose
(No. 131) during 1999-2000 In the western Upper Peninsula of
Michigan. . . . . . 69

Progesterone concentration in blood serum collected in winter from
radio-tagged moose in the western Upper Peninsula of Michigan,
1999-2001. Each symbol represents the serum progesterone
concentration of a single moose. The 95% upper tolerance limit

(95% UTL) or dividing serum progesterone level between non-
pregnant (0, O, A) and pregnant (0, O, A) cows was 1.68 ng/mL. 73

Progesterone concentration in fecal material collected in winter from
radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. Each symbol represents the mean fecal
progesterone concentration of a single moose. The 95% Upper
tolerance limit (95% UTL) or dividing fecal progesterone level
between non-pregnant (O, O, A) and pregnant (9, O, A)cows was
5.17 ug/g. . . . . . . . . 75

Numbers of radio-tagged cow moose, their spring calf production,
and observed spring twinning rates in the western Upper Peninsula
of Michigan, 1999-2001. . . . . . . 78

POP-ll moose population size estimates in the western Upper
Peninsula of Michigan, 1987-2001. Results reflect the use of: (A)

adult S=0.87, calf S=0.83, calf:cow=0.84, (8) adult S=0.88, calf
S=0.71, calf:cow=0.81, and (C) adult S=0.88, calf S=0.71,
calf:cow=0.81, and yearling D =0.06. . . . . 81

Figure 21.

Figure 22.

Figure 23.

POP-ll moose population size estimates in the western Upper
Peninsula of Michigan, 1987-2001. Results reflect the use of adult

S=0.88, calf:cow=0.84, and (A) calf s=o.71 and (B) calf S=0.83. 82

POP-ll moose population size estimates in the western Upper
Peninsula of Michigan, 1987-2001. Results reflect the use of adult

S=0. 88, calf S=0. 77, calf: cow=0. 84 and a yearling dispersal rate of
(A) 0%, (B) 6% and (C) 12%. . . . 83

Estimated relocation points and movements of radio-tagged yearling
female moose (No. 65) during March 1989-December 1990, and
yearling male moose (No. 74) during March-October 1994 in the
western Upper Peninsula of Michigan. . . . . 94

xi

INTRODUCTION

The Eastern sub-species or Taiga moose (Alces alces americana) is
native to Michigan and once ranged throughout the entire state, except for the
most southwestern portion of the Lower Peninsula (Baker 1983, de Vos 1964,
Peterson 1955, Schroger 1942, Wood 1914). Moose were probably never
numerous in the Lower Peninsula because of its geographic location at the
extreme southern limit of moose range. This factor in combination with habitat
degradation and unregulated hunting resulted in the extirpation of moose in the
Lower Peninsula by the mid-18805. The last credible sighting in the Lower
Peninsula may have been John Roger’s report of a moose at Black Lake in
Presque Isle County in 1883 (Wood and Dice 1923, Baker 1983). In an effort to
protect any remaining moose in the Lower Peninsula and those in the Upper
Peninsula, the Michigan Legislature passed a law in 1889 giving moose full
protection.

The Upper Peninsula of Michigan may have supported a substantial
moose population prior to extensive human settlement after the American Civil
War (1861-1865). However, by the end of the 19th century, despite legal
protection, moose in the Upper Peninsula had declined and may have been
briefly extirpated. Dramatic changes in habitat from logging and fires, heavy
hunting pressure, wolf (Canis lupus) predation and diseases were speculated as
factors contributing to the population decline (Verme 1984). The last record in

the 19th century of a moose in the Upper Peninsula is possibly Hickie’s (1944)

report of the poaching of a yearling female in the vicinity of Breevort Lake in
Mackinac County in 1899. However, periodic sightings of moose in Chippewa,
Luce and Schoolcraft counties (Wood and Dice 1923) provides some anecdotal
evidence that a remnant population may have persisted in the eastern Upper
Peninsula.

According to Baker (1983) the 1St Biennial Report of the Michigan
Department of Conservation (MDC) for 1920-1921 estimated that a population of
>1000 moose existed in Chippewa and Luce counties. A closer reading
however, reveals that this estimate included the large moose herd on Isle Royale
(see below). The 1St Biennial Report actually states that a herd of 25 or more
moose had been located near the southern boundary ofthe Lake Superior Forest
Reserve in Luce County and that moose had also been seen in western
Chippewa county. A decade later in the 6th Biennial Report (1931-1932) the
M00 states that the population in Chippewa and Luce counties had increased,
although no numbers are given. In addition, sightings of moose had been
reported from Alger, Mackinac and Schoolcraft counties in the eastern Upper
Peninsula and Keweenaw County in the western Upper Peninsula. It is not clear
if the observations in the eastern Upper Peninsula were of moose from a small
remnant population or whether the observations were of animals that periodically
immigrated from Ontario. Also, the reported sighting in Keweenaw County is
questionable because (1) no moose had been reported in any county in the

western Upper Peninsula in Biennial reports prior to 1930 and (2) no evidence or

sightings of moose in Keweenaw County were recorded during 1954-1983
(Appendix Table 1, Appendix Table 2).

Although moose in the Upper Peninsula of Michigan were fairing poorly,
the population on Isle Royale, Michigan — a 544-km2 island in Lake Superior, 24
km from the Canadian shore - was proliferating. Dr. Adolph Murie (1934)
reported that there were at least 1000 moose (~1.84 moose/kmz) on Isle Royale
and that the vegetation was severally over-browsed. To reduce the population
on Isle Royale with the added benefit of re-establishing moose in the Upper
Peninsula, the MDC live-trapped and transported 71 moose (33 adult males, 38
adult females) to the Upper Peninsula in the winters of 1935, 1936, and 1937.
Thirty-four moose were released in the Cusino Game refuge in Schoolcraft
County, 17 in the Escanaba River tract (Marquette County), 18 on the Keweenaw
Peninsula, and two were given to the Detroit Zoological Park (Hickie 1944). In
addition, 6 moose were kept in captivity at the Cusino Game Refuge for life
history studies (Verme 1970). The 9th MDC Biennial Report (1937-1938)
recorded 108 sightings of moose following these releases. Furthermore,
observations or evidence of moose had been recorded in all counties of the
Upper Peninsula except Gogebic and Ontonagon counties by 1941. However,
by the end of World War II (1942-1946) the population had again declined
(Verme 1984).

Moose are not mentioned again until the 21‘"t (1961-1962) Biennial Report,
which notes an increase in observations, again primarily in the eastern Upper

Peninsula. Additionally, 2 small groups of moose were purported to exist in

northern Marquette and northern Iron counties in the western Upper Peninsula.
However this is not substantiated by the observations of moose documented by
Verrne (1984) who reported only 9 observations of moose between 1961-1966 in
the western Upper Peninsula (Appendix Table 1). Through the remainder of the
1960’s and into the early 1970’s Biennial Reports produced by the MDC
(renamed the Michigan Department of Natural Resources [MDNR] in 1968) are
nearly identical. Paraphrasing these reports, “Observations or evidence of
moose continue to be reported from most counties of the Upper Peninsula,
primarily in Chippewa, Luce and Mackinac counties in the eastern Upper
Peninsula”. No attempts were made to actually estimate the moose population
size during this period, however, the 23rd Biennial Report (1965-1966) states that
“...25-50 animals seems to be a reasonable figure”.

Observations of moose appeared to increase in the entire Upper
Peninsula after 1966. However it is likely that the increase in tourism following
completion of the Mackinac Bridge in 1957 was responsible, rather than an
actual increase in moose numbers. Evidence supporting this is that only 14
observations of calves, 10 of which were from the eastern Upper Peninsula, were
made during 1961-1983 (Appendix Table 1, Appendix Table 2) suggesting that
reproduction was low. From 1967 to 1983 sightings and/or evidence of moose
averaged approximately 13 per year in the eastern Upper Peninsula and 3 per
year in the western Upper Peninsula (Appendix Table 2).

Despite the failed initial attempt to re-establish moose to the Upper

Peninsula the idea was kept alive by MDNR Regional Wildlife Biologist Ralph

Bailey (deceased June 2001) (Verme 1984). However, it was not given serious
consideration until the white-tailed deer (Odocoileus virginianus) density in the
Lake Superior watershed had decreased to acceptable levels. White-tailed deer
are the normal definitive host of two parasites, the large American liver fluke
(Fascioloides magna), and meningeal or brain worm (Parelaphostrongylus
tenuis), that are potentially harmful to moose. Moose acquire these parasites by
inadvertently ingesting the intermediate host (aquatic snails in the case of liver
flukes and terrestrial gastropods in the case of brain worm) harboring the
infective form of the parasite along with their food. It is speculated that
transmission of these parasites to moose is correlated with white-tailed deer
densities. In Maine, Gilbert (1974) found a positive correlation between deer
density and prevalence of moose infected with brain worm. In addition, moose
declines in Minnesota, New Brunswick, Nova Scotia, and Maine occurred when
white-tailed deer densities exceeded 5.0/km2 (Whitlaw and Lankester 1994a,
1994b). Fewer white-tailed deer may reduce the incidence of moose contacting
and being infected by these parasites. However, liver flukes alone have not been
shown to be detrimental to moose and the relationship between white-tailed
density, the prevalence of moose infected with brain worm and the impact of
brain worm on moose populations is more subtle than previously hypothesized
(Gilbert 1992, Nudds 1990, Nudds 1992, Whitlaw and Lankester1994a, 1994b).
To assess potential moose habitat in the Upper Peninsula, H. Cummings
from the Ontario Ministry of Natural Resources (OMNR), Ontario, Canada was

invited to conduct ground and aerial surveys in the winter and fall of 1972. He

determined that available habitat in the entire Upper Peninsula could probably
support a population of approximately 1000 moose. Additionally, he reported
that available habitat in the Tracy Creek drainage (an area west of the city of
Marquette and south of State route 28) could likely support 500 moose (MDNR
1974). A decade later (October 1982) M. Wilton of the OMNR conducted a more
thorough analysis of potential moose habitat. Based on 10 criteria (e.g., aquatic
feeding sites, human activity), the Lake Michigamme area in the western Upper
Peninsula was rated the highest. This was primarily due to the greater
topographic relief than in other areas analyzed, and the good interspersion of
summer and winter habitat (Wilton 1982).

Finally, in January-February 1985, the MDNR translocated 28 adult (22.5-
years of age) and 3 yearling moose (21 females, 10 males) from Algonquin
Provincial Park (APP) in Ontario, Canada to a 1,540 km2 area north of Lake
Michigamme in the western Upper Peninsula (Figure 1). Only 29 moose were
actually released, 2 cows that developed complications related to the recycling of
carfentanil, the drug used to immobilize moose in Ontario, died 3—days after
arriving in Michigan (Schmitt and Aho 1986). One died after aspirating
regurgitated rumen content and the other could not stand, and was euthanized.
To facilitate monitoring, each moose was fitted with a motion-sensitive radio-
collar (Telonics, Mesa, Arizona) at the staging area in Ontario. At the time of
release, the white-tailed deer density in the western Upper Peninsula was
approximately 2.9 deer/km2 (Hill and Pohl 1983) and the incidence of parasite

transmission was considered to be relatively low. A second translocation of 30

.5922: Go mimcEcn. Loan: £3ch
9: 5 9.37.. 33.8 cosmooficw: $09: $9 new $2. cam 58-89. 0:56 coca >36 38.2 .F 8:9...

_ btzoo :omeoLQ _ \. _ c I\ w.

1

 

 

 

3560 $8362

 

 

 

o c \
.. E45... .. a a K a. e.
. I

 

 

 

 

 

 

 

 

 

 

 

 

.6236
an? 95:00..
5 we“ ..
occamooz . adv .. c a
CM” my.“ :29:ch “an“...v . \.\ \Il
. . 36530
K
............ .. a .a
coo
Emmofiommco &
E u v
$5.3...“ mtg? _
9. 46>:
529:3! i
.8858 .830 D
alga-932:8 i ll.
2:6... *1
r :5 382.88: O
H

 

 

 

 

 

 

adult moose (15 cows, 15 bulls) occurred in 1987, prompted by the low
pregnancy rate (50%) of radio-tagged females in 1986 (Schmitt and Aho 1986).
Two cows died within 1 week after this second release, both were heavily
parasitized with hydatid cysts (the larval stage of the tapeworm Echinococcus
grandulosus) and also suffered from chronic hepatitis, peritonitis, and metritis.

The long-term goal of the 2 translocations was to reestablish a self-
sustaining population of moose in the Upper Peninsula of Michigan. Based on
Wilton’s 1982 report and rudimentary population growth projections (Beyer 2001
personal communication), the objective was to produce a population of 1000
moose by the year 2000 (MDNR 1991 a).

For several years following the translocations it was fairly simple to
estimate the population. The observed number of calves born in the spring to
radio-tagged cows was added to the number of radio-tagged moose that were
alive, from which was subtracted the annual number of fatalities of radio-tagged
moose. However, as the proportion of radio-tagged moose in the population
decreased, a deterministic book-keeping model that incorporated the mortality
rates and intrinsic rate of increase from 1985-1990, was used to estimate the
size of the population on an annual basis for 1994-1997. Using this method, the
1997 population in the western Upper Peninsula was estimated at 494 moose.
(Figure 2). In addition, POP-ll (Fossil Creek Software, Fort Collins, CO), a
population modeling computer program was used to estimate the population
(Figure 2). POP-ll projected a population of >850 moose by the year 2000 (Aho

etaL1995)

   

600 , .7- ,- ”i 3

1—0—Book-keeping model C) a
500- ‘--I---POP-ll : 3 '

g * 9 Aerial surveys

o 400 l ’ T

o

E

O 300 -

h

'3 .

a l

g 200 .

2 107120

(30 N
o 1—
0 "' "’ ,,.. TV ""7? a; “ ‘V 'T *7” fi '1
LO (0 N w G) O 1- N (‘0 V I!) (D N
CO CD CD GD co 0) O) O) O) O) O) O) O)
eeeeeemeeeeee
Year

Figure 2. Moose population size estimates for 1985-1997 in the western Upper
Peninsula of Michigan based on a deterministic book-keeping model,
POP-ll, a population modeling computer program, and aerial surveys
conducted by the MDNR in 1991,1996, and 1997.

Because it is not feasible to maintain a statistically adequate sample of
radio-tagged animals indefinitely, other methods are used to estimate wildlife
populations. The most practical method for estimating moose numbers is to
count them on their winter range from a fixed-wing aircraft (Timmerrnann 1993).
In January-March 1991, the MDNR conducted an experimental aerial survey of
moose in the Upper Peninsula using techniques developed by the Ontario
Ministry of Natural Resources (Aho et al. 1995). Although the entire Upper
Peninsula was sampled, no moose were seen in the eastern Upper Peninsula.
Based on this survey the population was estimated at approximately 210 moose

(Figure 2, MDNR 1991b), a value higher than the estimate generated by the

book-keeping model (175 moose). To make the aerial surveys more rigorous,
and to account for unseen animals, a stratified sampling scheme and sightability
model were developed (Drummer and Aho 1998). Use of the new survey
methods produced population size estimates of 107 moose in the winter of 1996
and 120 in the winter of 1997. Values much lower than those generated by the
book-keeping model (1996, n = 452; 1997, n = 494, Figure 2). At the time there
was no evidence of a massive die-off of moose (MDNR 1997), so the differences
between the estimation methods could not be readily explained.

The stimulus for the current research grew out of the inability of the
MDNR, after extensive evaluation, to explain the discrepancies between the
population size estimates from the aerial surveys and those produced by the
book-keeping model. In addition, recent declines in moose populations in the
Great Lakes region, specifically northwest Minnesota (Cox et al. 1997) and
Ontario, Canada, have MDNR biologists concerned that factors effecting these
populations are, or will eventually, impact the moose population in Michigan.
Only through a better understanding of the population dynamics of moose in the
Upper Peninsula can these issues be addressed. To accomplish this, data must
be collected on the three factors that control population growth, (1) births, (2)
deaths, and (3) emigration/immigration. The use of radio-tagged animals is a
widely accepted practice for gathering this information. Movement of animals
into the population (immigration) is assumed to be negligible and will not be
determined. This study of moose population dynamics will produce quantitative

measures of birth, death, and dispersal rates, and will provide the basis for

10

developing a scientifically sound management objective for the moose population

in the western Upper Peninsula of Michigan.

11

RESEARCH OBJECTIVES

The data obtained from this study will help the MDNR Wildlife Division
(MDNR-W0) develop a more scientifically rigorous moose population model and
provide a better understanding of the spatial movements of moose. This
information will help guide future management strategies and set realistic
objectives for the moose population in the western Upper Peninsula of Michigan.
The main objective of this study is to quantify the population dynamics of moose

in the western Upper Peninsula by:

1) Estimating pregnancy rates and natality of cow moose,
2) Estimating sex- and age-specific survival rates of moose,
3) Estimating dispersal rates of moose, and

4) Identifying potential limiting factors of moose population growth.

12

STUDY AREA

The research was conducted in the western Upper Peninsula of Michigan
between January 1999 and June 2001. The core study site (defined as the area
within which >95% of all radiotelemtry relocations were obtained) included
portions of Baraga, Iron, and Marquette counties and covered an area of
approximately 3000 km2 (Figure 1). This area was chosen because both the
1985 and 1987 release sites are within the area and it has the greatest known
density of moose in the Upper Peninsula.

The western Upper Peninsula is bounded to the north by Lake Superior
and to the south and west by the state of Wisconsin. Approximately 90% of the
western Upper Peninsula is wooded, primarily in secondary growth forests.
During the mid- to late 1800’s iron mining and logging were the primary
enterprises. Today, recreation and timber production are the most important land
uses.

The Upper Peninsula is part of the deciduous/coniferous ecotone or
“northem-hardwood boreal forest” (Theberge and Theberge 1998), an area of
transition between the northern boreal forests and the more temperate deciduous
woodlands to the south. This is reflected in the heterogeneous composition of
forests throughout the region.

In the western portion of the Upper Peninsula, northern hardwoods forests
lacking American beech (Fagus grandifolia, except along the Lake Superior

shoreline) dominate upland areas. Exposed rocky ridges support scattered white

13

pine (Pinus strobus), red pine (Pinus resinosa), and trembling aspen (Populus
tremuloides) which are often disturbed by windthrow and lightning strike caused
fires. On moderately to poorly drained sites balsam fir (Abies balsamea), black
ash (Fraxinus nigra), eastern hemlock (Tsuga canadensis), northern white cedar
(Thuja occidentalis), red maple (Acer rubrum), trembling aspen, white spruce
(Picea glauca), and yellow birch (Betula alleghanr’ensis) may be found either as
mixed conifer-hardwoods or pure conifer or hardwoods stands. A variety of
wetlands including conifer bogs, hardwood swamps, conifer swamps, and
speckled “tag” alder (Alnus incana) thickets occur where bedrock is at or near the
surface. Large contiguous stands of willow (Salix spp.) - often an important
moose food item - found along river and stream banks are conspicuously
absent, although small scattered patches do occur.

Two distinct physiographic regions (Albert 1995) — the Michigamme
Highlands (Subsection lX.2) to the north and northeast and the Upper
Wlsconsin/Michigan Moraines (Subsection lX.3) to the south and southwest -
divide the study area.

Granite bedrock at or near the ground surface dominates much of the
Michigamme Highlands Subsection. Outwash plains and steep sandy plains are
present where gaps in the bedrock exist. The terrain consists of a mosaic of low
(< 15.0 m) rocky knolls in the southern parts of the Subsection and higher (160-
250 m) exposed ridges in the Huron Mountains to the north. Numerous lakes
and swamps occur in glacially formed depressions in the bedrock. Elevation

ranges from 184 to 604 m.

14

The Upper Wisconsin/Michigan Moraines Subsection consists primarily of
course-textured (sand and sandy loam) end moraines. Steep, well-drained
ridges; deep, steep-sided kettle lakes, and poorly drained depressions
characterize the terrain. Elevation ranges from 183 to 595 m.

Because of the absence of a large body of water to its south, the climate
of the western Upper Peninsula is more continental than that in the eastern
Upper Peninsula. The lack of lacustrine moderation results in a greater variation
in seasonal temperatures. In addition, average interior winter temperatures are
often 9-12 °C colder than those along the Lake Superior shoreline. Thunder-
storms are also more likely in summer. Average annual rainfall at Champion,
Michigan from 1951 to 1980 was 85 cm; sixty-three percent (54 cm) of which fell
during the April to September growing season (Bemdt 1988). The following
temperature and precipitation readings were recorded at Van Riper State Park in
Champion, Michigan from 1999 to 2001. In January, the average daily minimum
temperature was -20.2 °C, average daily maximum temperature was -4.3 °C.
Average daily minimum and maximum temperatures in July were 8.4 and 26.5
°C, respectively. In comparison, the average daily minimum temperature for
1951-1980 at Champion, Michigan in January was -17.5 °C and the average daily
maximum temperature was -6.0 °C. Also, for the same period, the average daily
minimum temperature in July was 10.0 °C and the average daily maximum
temperature was 25.8 °C (Berndt 1988). Average seasonal snowfall for 1999-
2001, all of which occurred between November-April, was 211.5 cm. Mean

annual snow-fall at Champion, Michigan during 1951-1980 was 350.5 cm. Mean

15

daily snow depth on the ground at Van Riper State Park during January-February
was 57 cm in 1999, 24 cm in 2000 and 54 cm in 2001.

A number of animal species have the potential to positively or negatively
impact moose. Predators of moose, primarily calves, are black bears (Ursus
amen’cana) and wolves. At present the black bear population in the Upper
Peninsula is estimated at >12,000 and the increasing wolf population is
estimated at 200-250. Ponds created from beaver (Castor canadensis) dammed
rivers and streams allow for escape from summer heat and insects as well as
providing aquatic vegetation, an important summer food source. White-tailed
deer may have the greatest potential to negatively impact moose because they
are the normal definitive host of the meningeal or brain worm; 50 to 90 percent
may be infected, but signs of the disease are rare (Lankester and Samuel 1998).
In moose, however, it can cause “moose sickness” (Parelaphostrongylosis or
Cerbrospinal Nematodiasis) which manifests itself in severe neurological

disorders (Anderson 1964) and often results in death.

16

METHODS

Capturing and radio-tagging

Hawkins & Powers Aviation, Greybull, Wyoming captured moose via net-
gunning from a helicopter (Hughes 500 or Bell Long Ranger L-3). Captures were
conducted in January-February 1999, January 2000, and February 2001. MDNR
Forest Division (MDNR-FD), MDNR-WD, and Michigan State University (MSU)
personnel provided air and ground support. For at least a week prior to each
year’s capture effort the study area was surveyed from fixed-wing aircraft
(Cessna 172, 182 or 201) flown by MDNR-FD pilots to find and record locations
of moose. Upon arrival of the capture crew these locations were re-flown to look
for potential moose targets. When a moose was spotted its coordinates
(Latitude/Longitude in decimal degrees) were stored (Le, a waypoint was
recorded) in a Precedus Global Positioning System (GPS) unit (UPS Aviation
Technologies, Salem, Oregon). These coordinates were then radioed to the
helicopter pilot who attempted to drive the animal to a suitable netting location
(e.g., frozen swamp, lake, or pond). For the safety of moose and netting crews
steep hills, cliffs, and open rivers and streams were generally avoided.

Exclusively MDNR and Hawkins & Powers Aviation personnel did all
handling of captured moose. Moose were fitted with radio-collars (transmitters)
equipped with 4-hour motion sensitive switches (Telonics, lnc., Mesa, AZ) and
ear tags were attached to both ears. Blood was drawn from the jugular
(preferred) or femoral artery of all moose to assess winter health condition and to

determine the pregnancy status of cows. Health condition was evaluated

17

through measures of hemoglobin (Hb) and packed cell volume (PCV) In whole
blood and calcium (Ca), copper (Cu), iron (Fe), magnesium (Mg), phosphorous
(P), potassium (K), sodium (Na), selenium (Se), vitamin A, and zinc (Zn) in blood
serum. Pregnancy status was determined through assays of blood serum for
pregnancy specific protein B (PSPB) and progesterone concentrations. Fecal
samples were collected to measure glucocorticoid concentrations as an indicator
of physiological stress (Millspaugh et al. 2001) and to test cows for pregnancy
through assays of fecal progesterone concentrations. A 2-cm diameter patch of
hair (including shafts and follicles) was plucked from the shoulder hump of each
captured animal. Hair samples were archived for future determination of
elemental deficiencies and accumulation of toxins (Flynn et al. 1975), and
possible genetic profiling. The following data were recorded for each captured
moose: (1) coordinates of capture location, (2) estimated age, (3) sex, and (4)
physical examination score (Appendix Table 3). Because the capture process
can be highly stressful, each animal’s body temperature (normal = 38.4-38.8 °C),
pulse rate (normal = 70-91 beets/min), and respiration rate (normal = 13-40
breaths/min) were closely monitored. If excessive stress or overheating was
detected, moose were quickly radio-tagged, released, and closely observed over
the next several days for continuing problems.
Radio-telemetry relocating
Aerial relocations

Following release, all radio-tagged moose were relocated at least weekly

throughout the year from a fixed-wing aircraft (Cessna 172, 182 or 201) equipped

18

with radio-telemetry tracking equipment (side-facing, 2-element yagi antennas
mounted to each wing strut, connected by coaxial cable to a switchbox in the
cockpit). Transmitters were located and point locations of detected moose were
recorded as waypoints (map datum: WGS-84, Latitude/Longitude: decimal
minutes) in a Precedus GPS. If radio-tagged moose were relocated near open
areas (e.g., open water, wetlands or logging clear-cuts) visual confirmations were
attempted. In addition, if untagged moose were seen, their location, age and sex
(if possible) were recorded. Vegetation cover type and moose activity (bedded,
feeding, or moving) were also noted.

Ground relocations

Radio-tagged moose were also approached on the ground to collect fecal
samples, and to determine calving status and calf survival (see below). Before
going into the field the most recent waypoint recorded for the targeted radio-
tagged moose was converted to Universal Transverse Mercator (UTM) grid
coordinates and plotted on the appropriate 1:25000 United States Geological
Survey (USGS) quadrangle map.

In the field, radio-telemetry equipment was used to take at least three
directional bearings, which were then drawn on the USGS quad map to establish
a more precise location. To determine the approximate distance to the targeted
moose the UTM coordinates of this location were recorded in a Garrnin GPS lll+
unit (Garrnin, lntemational, lnc., Olathe, Kansas).

After moving to within 0.15-0.30 km of the targeted moose, it was

approached more closely by directionally homing-in on the transmitter signal.

19

Transmitter signal strength, change in angle, approximate distance to the
targeted moose, and auditory cues (e.g., breaking of saplings, moose
vocalizations) were used to determine proximity to the animal. Visual sightings
were always attempted to verify the identity of the radio-tagged moose and to
make health condition and behavioral observations. When collecting fecal
samples, after a sighting was made or evidence (e.g., tracks, beds) of the
targeted moose was found, observers walked 100-150 meters in each of the four
cardinal directions (N, S, E, W) to determine if other moose or their sign were
present in the area. Fecal pellets were collected only if we were certain of the
identity of the moose that had deposited them. Fecal pellet size (Cox et al.
1997), track size, stride length, ventral surface drag marks, and bed dimensions
were used to differentiate between adults and calves. The following information
was recorded at the collection site: GPS coordinates, estimated position error
(EPE), vegetation cover type and dominate plant species, snow depth (cm), bed
dimensions (cm), and elevation (m).
Survival and mortality

Survival monitoring of radio-tagged moose (i.e., listening for a slow pulse
rate [alive=~60 beats/sec] or a rapid pulse rate [dead=~120 beats/sec] from
each transmitter) was generally conducted once or twice a week from a fixed-
wing aircraft. If inclement weather or other priorities (e.g., fire patrols, aircraft
maintenance) prevented us from flying, ground monitoring of transmitters was

conducted if >1 week had passed since the previous survival check. Survival

20

monitoring of calves seen with radio-tagged cows during spring calf checks was
done at monthly intervals as described above (see Ground relocations).
Whenever a mortality pulse signal was heard during an aerial relocation
flight the signal was homed-in on, a GPS waypoint was taken, and a visual of the
animal was attempted. As soon as possible the transmitter emitting the signal
was approached on the ground. While in the field, a transmitter emitting a
mortality pulse signal was approached immediately by homing-in on the signal.
Upon verifying that the moose was dead, the surrounding area was
examined for signs of predators (Roffe et al. 1994), evidence of humans,
evidence of behavior associated with P. tenuis (Anderson and Lankester 1974),
or other possible causes of death. Photographs of the surrounding area were
also taken prior to approaching the carcass. If possible the whole carcass was
transported to the MDNR Rose Lake Diagnostics Lab (RLDL) in East Lansing,
Michigan. In most instances this was not feasible and a gross field necropsy was
performed. When conducting necropsies we followed the guidelines of Nettles
(1981 ). Depending on the degree of decomposition and amount of scavenging
the following items were examined and collected: the entire head, 13t incisors,
femur, liver, heart, lungs, kidneys, reproductive tract of females, and samples of
hide, feces, and blood. In addition, photographs of abnormal-appearing tissues,
organs, or other abnormalities (Roffe et al. 1994) were taken. Additional
information recorded included date, time of day, GPS location, elevation,
vegetation cover type, dominant plant species, weather conditions, snow depth

(when applicable), presence or evidence of insects or parasites (e.g., ticks

21

(Dermacentor spp., liver flukes), physical condition score (Appendix Table 3),
and preliminary cause of death.

All samples were sent to the RLDL for examination and analysis. First
incisors were sectioned and cementum annuli counted to estimate ages of dead
moose (Sargeant and Pimlott 1959, Gilbert 1966). The meninges or membranes
covering the brain were inspected for presence of meningeal worms, and moose
livers were sectioned and inspected for large American liver flukes. Assessment
of subcutaneous and visceral (e.g., kidney) fat deposits, physical condition score
(Appendix Table 3) and chemical assays of femur bone marrow were used to
evaluate the nutritional condition of moose near time of death (Neiland 1970,
Mech and DelGiudice 1985). Causes of death were classified as accidents,
unknown, disease or predation.

Survival rates of moose with 95% confidence intervals (95% Cl) were
calculated using MICROMORT, a DOS-based microcomputer program
developed by Heisey and Fuller (1985) which incorporates the Mayfield survival
estimator (Mayfield 1961, 1975). MICROMORT uses the Taylor series method
(which is based on symmetric intervals about the log-transformed survival
function) to calculate 95% Cl’s, therefore, the 95% Cl’s were truncated to lie

between 0 and 1. The Mayfield estimator calculates the interval survival rate

(Si ) for the period of interest (L) as:

 

Q. = 1_ number of deaths in the period
' total exposure intervals

22

where the total exposure intervals (e.g., months) is equal to the sum of all
intervals each animal is at risk of death (i.e., the number of intervals from which

an individual enters the study, to, until it is last observed t.-). The period survival

rate (Sp) is then Sp= SiL. When used with radiotelemetry data the Mayfield

model assumes that the survival rate remains constant over the period of interest
(Trent and Rongstad 1974). To reduce the bias associated with relocation
probability (Bunck et al. 1995), we attempted to check (listen for dead or alive
signal) each individual frequency at least weekly (see above). However, if flights
were shortened because of scheduling constraints or inclement weather we tried
to check at least 90% of the radio-tagged animals (Kenward 2001) during the
flight with the remainder being checked as soon as possible. In addition to
constant period survival, all survival models require these additional
assumptions: (1) radio-collars do not impact an animals’ fate, (2) the exact time
of death of each animal is known, (3) individual animals are independent, (4) the
survival function of newly radio-tagged animals is equal to that of previously
radio-tagged animals, (5) radio-tagged animals are randomly sampled, and (6)
censoring of radio-tagged animals is random (Winterstein et al. 2001). Finally,
because the Mayfield estimator uses the total period (e.g., animal months) an
individual is at risk of dying, it permits staggered entry (i.e., different start times or
to) (see Pollock et al. 1989) of newly radio-tagged animals.

Annual, summer and winter survival rates of radio-tagged adult (2 2-years
of age) and yearling (13-24 months of age) moose were calculated. Because

adult and yearling moose were checked weekly I initially calculated a weekly-

23

interval survival rate. This weekly rate was then used to calculate a monthly (the
period of interest) rate, over which survival was assumed to be constant. Annual

(12 month) and seasonal (6 month) survival rates were then equal to the product

of the rate for each month (e.g., SAnnuafi S June* S My * . . . *SMay ). Annual

and seasonal adult moose survival rates were also determined for bulls and cows
separately. Annual survival rates were calculated for the biological year from 1
June-31 May. Winter was defined as 1 November-30 April and summer as 1
May-31 October. The season starting and ending dates roughly correspond to
the first freezing temperature in the fall and the last freezing temperature in the
spnng.

Survival rates were calculated jointly for radio-tagged and un-tagged
calves (<12 months of age) of radio-tagged cows seen each spring. So as not to
positively bias calf survival rates, calves radio-tagged in the winter but not
previously seen in the spring of that year were not included in survival
calculations. Calf survival was calculated for the first-year of life and for two
periods, 1 June-30 November (birth to 6 months of age) and 1 December-31 May
(6-12 months of age). These periods correspond with times when calves are
most vulnerable to black bear and wolf predation in the first period and wolf
predation in the second (Van Ballenberghe and Ballard 1998). Although, survival
of radio-tagged calves was checked at weekly intervals, survival of un-tagged
calves was checked only monthly. It was therefore necessary to assume that
calf survival was constant within the 0-6 month and 6-12 month periods. First-

year calf survival was then equal to the product of the rate from both periods.

24

Moose radio-tagged after the start of the biological year (1 June) or
seasonal period (winter = 1 November, summer = 1 May) entered the study on
the day they were radio-tagged. Except for moose that were radio-tagged in
January-February 1999 which entered the study on 1 June 1999. The date of
death of radio-tagged moose was estimated at halfway between the last
recorded live signal and the date that the moose was first known to be dead
(Mayfield 1961, 1975). Determining dates of death for un-tagged calves was not
as straight forward. If the radio-tagged cow of an un-tagged calf died during the
calf’s first 6 months of life, it was assumed that the calf did not survive and it was
assigned the same date of death as the cow. If a radio-tagged cow previously
attended by an un-tagged calf was found alone for two consecutive months the
calf was considered to have died unless it had attained an age of at least nine
months. In this case the un-tagged calf’s date of death was estimated at halfway
between the last date it was seen with its cow and the date that the cow was first
seen alone. Un-tagged calves not seen with their radio-tagged dams after
attaining the age of 9 months (the earliest time of cow-calf separation) were
assumed to have survived the entire year.

Survival rates of adults were compared between years, seasons, and
sexes. Yearling survival rates were compared between years and seasons and
also to adult survival rates. F irst-year and interval (birth to 6 months of age, 6—12
months of age) survival rates of calves were compared within and between

years. In instances where sample sizes were adequate (>30), comparisons

25

between survival estimates where made with a simple approximately normal test
statistic based on the equation:
2 = é'101) - é2(t1)
JVar $161) + Var é2(t1)
where S(t1) is the estimate of the first survival curve at time t1 and S2 (t1) is the
estimate of the second survival curve at time t1 (Pollock et al.1989). Statistical
significance was accepted at P s 0.05.

Movements

Home range size and dispersion

 

Relocations (radio-fixes) obtained during weekly flights, winter fecal
sample collection, spring natality determinations and calf survival monitoring
were used to (1) estimate annual and seasonal home range size, (2) monitor
home range development and (3) differentiate exploratory movements from
permanent dispersal. To standardize radio-fixes collected in different coordinate
systems (e.g., UTM, Latitude/Longitude) all were converted to the Michigan
GEOREF coordinate system.

Annual (1 May-30 April) and seasonal (summer: ~1 May-31 October,
winter: ~1 November-30 April) home range sizes were estimated using the
minimum convex polygon (MCP; Mohr 1947) and fixed kernel (FK; Worton 1989)
approaches. The Animal Movement Analyst Extension (AMAE; Hooge et al.
1999) to ArcVIew® (Environmental Systems Research, Inc., Redlands,
California) GIS was used to analyze these data. The FK smoothing factor was

calculated via least squares cross validation (LSCV). I used the FK-95%

26

utilization distribution (UD) to estimate home range sizes and the FK-25% UD to
estimate the center of seasonal home ranges, from which distances between
seasonal ranges could be determined. If the FK-25% UD resulted in more than
one home range polygon, the distance between the centers of the multiple home
ranges were averaged. Annual and seasonal home ranges were estimated only
when >19 and >9 relocations per moose respectively, were available. When
determining seasonal home ranges, relocations that deviated from grouped
relocations were considered transitory points and were excluded from seasonal
home range estimates. Moose were considered to be migratory if there was
<25% overlap of their seasonal home ranges and 2 2 km between the centers of
their seasonal home ranges. The Mann-Whitney U test was used to make
comparisons among FK-95% UD home ranges. The 100%-MCP method was
used to permit comparisons to other studies.

To determine if movements of radio-tagged moose were more or less
spatially dispersed at certain times of the year, I calculated the average Euclidian
(or linear) distance (AD) between all pairs of relocations for individual adult and
yearling moose. AD was calculated annually (May-April) and for 3 periods:
summer (May-August), fall (September-December) and winter (January-April).
Conner and Leopold (2001) demonstrated that AD was more precise and less
biased than the kernel home range method, especially with sparse data. AD was
calculated only when >19 relocations annually and >9 per season, per moose
were available. Comparisons between annual ADs were made with the Mann-

Whitney U test. Differences among seasonal ADs were determined using the

27

Kruskal-Wallis test. Unless otherwise noted, P s 0.05 was required for statistical
significance in all home range comparisons.
Dispersal and home range develcmt-Lnt

A moose was considered to have dispersed if it moved beyond the
boundary of the core study area and remained there until it established a new
home range, died, dropped its radio-collar or the study ended. Annual (1 June-
31 May) dispersal rates (with 95% Cl’s) of radio-tagged moose were estimated
with the Mayfield estimator (Mayfield 1961, 1975). The Mayfield is a time-to-
event model generally used to estimate a survival probability, where death is the
event of interest. However, the rate of any event such as dispersal, that can be
situated in time can be calculated with the Mayfield estimator. Because moose

that dispersed were relocated at approximately monthly intervals it was assumed
that dispersal was constant among months. The monthly dispersal rate ([5m ) for

the period of interest (L) is then estimated as:

[5 = number of dispersers in the period
m total exposure months

 

and the period dispersal rate (DD) is then Dp= [1—(1-Dm)L]. All other

assumptions of the Mayfield estimator were adhered to when estimating
dispersal rates. Dispersal distance was approximated by measuring between the
centers of the FK-25% UD of pre- and post-dispersal home ranges. In cases
where dispersing moose did not establish a new home range, dispersal distance
was measured between relocations obtained prior to, and after detection of

dispersal.

28

Home range development of young-of-the-year was quantified by
determining the linear distance (km) between and the percentage of overlap of
their cow’s home range(s) and the areas they occupied after separation from
their cow. Moose were classified as yearlings when they became independent of
their cow or they attained their first birthday (7 June, the median birthing date).
Reproduction
Pregnancy determinations

The pregnancy status of recently radio-tagged cows was determined
through assays of blood serum for pregnancy-specific protein B (PSPB; Haigh et
al. 1993, Stephenson et al. 1995) and progesterone levels (Haigh et al. 1982,
Stewart et al. 1985) and assays of fecal material for fecal progesterone levels
(Monfort et al. 1993, Schwartz et al. 1995).

PSPB is secreted by the placenta (Reimers et al. 1985) and has been
shown to reliably detect pregnancy 40 days after conception in moose. It may
also be possible to detect the number of fetuses present using assays of PSPB
(Huang et al. 2000). Stephenson et al. (1995) was 100% accurate at diagnosing
pregnancy in ten captive moose at the Kenai Moose Research Center (Kenai
Peninsula, Alaska) in March using assays of PSPB. In addition, Huang et al.
(2000) was 84% correct in categorizing single fetuses and 97% correct in
categorizing twin fetuses, using PSPB. Again, at the Kenai Moose Research
Center (Kenai Peninsula, Alaska) captive cows carrying twin fetuses had serum

PSPB concentrations >365 ng/mL, whereas those with singletons had

29

concentrations <365 ng/mL. Because of its accuracy serum PSPB was used as
the primary indicator of pregnancy when available.

Assays of the hormone progesterone are less reliable than PSPB at
detecting pregnancy because it is not pregnancy-specific (Huang et al. 2000)
Progesterone is elevated during gestation but is also present at lower levels in
blood serum of males and non-pregnant females. Blood serum progesterone
concentrations of pregnant moose are generally >3.0 ng/mL, while levels for non-
pregnant moose are <0.5 ng/mL (Schwartz 1997).

Assays of fecal progesterone are less reliable at detecting pregnancy than
assays of PSPB or serum progesterone. However, the method is noninvasive
and it is relatively easy to collect fecal samples in the winter. Assays of fecal
progesterone can be used to reliably detect pregnancy by the eighth week of
gestation (Schwartz et al. 1995). Monfort et al. (1993) was 85% accurate in
diagnosing pregnancy in captive moose using fecal progesterone levels.
Schwartz (1997) and Monfort et al. (1993) reported that pregnant moose
generally have fecal progesterone concentrations >2.0 pg/g, whereas levels in
non-pregnant moose are <0.5 pg/g.

Assays of fecal material alone were used to determine the pregnancy
status of previously radio-tagged cows. Fecal samples were collected at monthly
intervals following the first significant snowfall and continued until snowmelt each
season. To reduce the chance of a false-positive or false-negative result we
attempted to collect at least two fecal samples from each cow. Fifteen to twenty

fecal pellets were collected into a zip-lock plastic bag and kept frozen until

30

analysis. Moose stalking procedures and sampling protocols are outlined above
(see Ground Locations).

PSPB assay results, when available, and a cow’s calving status were
used to determine if a radio-tagged cow was actually pregnant or not. In
addition, if a radio-tagged cow was not seen with a calf during spring calf checks
or its calving status was unknown it was considered to be pregnant if (1) it had a
positive PSPB result and serum or fecal progesterone levels indicative of
pregnancy, or (2) it had fecal progesterone levels indicating pregnancy from at
least two independent fecal samples. When fecal material was used for
pregnancy determinations, measures of fecal progesterone concentrations of

non-pregnant cows were used to calculate the 95% upper tolerance limit (95%
UTL) of pregnancy (i.e., )7 +t-value x SD; Messier et al. 1990). The 95% UTL for

serum progesterone was also calculated from serum progesterone levels of non-
pregnant cows. The Mann-Whitney U-test was used to compare progesterone
concentrations between pregnant and non-pregnant cows.

The Conservation and Research Center (CRC, Front Royal, Virginia) of
the Smithsonian Institute performed the analyses of fecal samples and blood
semm for progesterone levels. BioTracking (Moscow, Idaho) conducted
assaying of blood sera for PSPB levels.

Productivity
NataJity andLRecrulment
To determine annual reproduction and roughly estimate calf birthing dates,

radio-tagged cows were observed at least once each year between 15 May-30

31

June, the approximate moose calving period in Michigan (Verme 1970).
Although attempts were made to see all radio-tagged cows, priority was given to
those that had positive pregnancy test results (see above). Each radio-tagged
cow was approached on the ground until a visual observation of the cow and calf
or calves was made or strong evidence (e.g., tracks, beds, fecal material) of a
calf or calves was found. Observations of radio-tagged cows and their offspring
were also attempted during radiotelemetry aerial re-location flights, but due to
heavy vegetative cover were very difficult to make. Pregnant cows that were
observed alone at the beginning of the calving period were approached multiple
times until their calving status was determined.

Determining the annual recruitment of calves into the adult population was
accomplished in the same manner as for calving. We attempted to visually re-
locate radio-tagged cows and their offspring from the ground and/or air at
monthly intervals beginning 1 July and continuing until calves became
independent of their cows, the next seasons calving period began, the radio-
tagged cow died, or it was believed that the calf had died. Natality, measured as
the number of calves per adult female (calf:cow) was determined each spring
after calving (post-calving) and prior to calving the following year (pre-calving) as
an index of calf recruitment. The frequency of twinning, which has been shown
to be an important indicator of habitat quality and female health condition
(Franzmann and Schwartz 1985) was calculated each spring as the number of

twins produced divided by the number of radio-tagged cows giving birth.

32

Population size estimation and projection

Difficulties replicating the book-keeping model developed by the MDNR
precluded using it to make comparisons between the population size estimates
obtained using survival and reproductive values determined for 1985-1996 and
those determined during the current study. Instead, POP-ll (WindowsTM Version
1, Fossil Creek Software, Fort Collins, Colorado), a population modeling
computer program was used to make these comparisons. POP-ll is a
deterministic, density-independent model that uses age-specific mortality rates
(i.e., 1 - survival rates) and age-specific reproductive rates (i.e., the number of
offspring per reproducing females or calf:cow ratio) to simulate the dynamics of
geographically closed animal populations.

Previous estimates of annual survival of adult radio-tagged moose and

survival of calves of radio-tagged cows were derived by the Apparent Percent

Success (APS) or Simplistic estimator. Using this approach the survival rate (S)
for a selected period of time is estimated as

é = 1_ number of deaths in the period
number at risk in the period

 

where the number at risk is the initial number of animals at the start of the period
(Winterstein et al. 2001). Calf production (and thus the calf:cow ratio) was
estimated by relocating radio-tagged cows on the ground and observing calves
or finding evidence of calves during May-July each year. Calf survival was
determined by relocating radio-tagged cows from the air and looking for their
calves during December-January 1985-1995 and April 1985-1988 (Aho et al.

1995). Because reproductive determinations and calf monitoring were

33

temporarily suspended during 1991-1993 and the sample of radio-tagged cows in
1994, 1997 and 1998 was small (<10), estimates of the calf:cow ratio and calf
survival rates for 1985-1990, 1995-1996 were used in POP-ll. Also, estimates of

adult survival were only available for 1985-1989.

Additionally, yearling dispersal (D ) was incorporated in the POP-ll
population size estimates by increasing the average yearling mortality rate.
Comparisons were made between population size estimates using the mean
survival and reproductive values collected from 1985-1990, 1995-1996 (MDNR
[1 985-1996]) to average estimates of survivorship and fecundity obtained during
the current study (MSU [1999-2001]). Additional comparisons looked at how
changes in calf survival and yearling dispersal affected moose population size
estimates. Estimates of the population size were made with the following
assumptions (1) a 1:1 sex ratio at birth, (2) only cows 2 2-years of age
reproduce, (3) all cows of reproductive age reproduce (this is necessary because
POP-II uses a per cow productivity value), (3) only yearling moose disperse and

(4) dispersal is not biased toward either sex.

RESULTS

Capturing and radio-tagging

Seventy-four moose were captured and fitted with 4-hour motion sensitive
VHF radio-collars in January-February 1999 (n=26), January 2000 (n=26), and
February 2001 (n=22). In addition, GPS collars were placed on four moose (2
bulls, 2 cows) in 2000 and six moose (3 bulls, 3 cows) in 2001. Twelve moose (6
bulls, 6 cows) radio-tagged in 1995 were also part of the initial sample population
(Table 1). No moose died or were injured during capture, and no signs of
capture myopathy (e.g., muscle stiffness, lethargy) were seen following capture
in any year.

Table 1. Sex and age breakdown of moose radio-tagged in 1995, January-
February 1999, January 2000, and February 2001 in the western

 

 

 

 

Upper Peninsula of Michigan.

Year Adults Yearlings Calves

Bulls Cows Males Females Males Females Total
1995 6 6 0 0 0 0 12
1999 1 12 0 1 8 4 26
2000a 3 14 0 1 7 5 30
2001b 2 1 1 1 2 8 4 28
Totals 12 44 1 4 23 13 96

 

aIncludes 2 bulls and 2 cows outfitted with GPS radio-collars.
bIncludes 3 bulls and 3 cows outfitted with GPS radio-collars.

Capture locations of all moose radio-tagged in 1995, and between 1999-
2001 appear in Figure 3. All radio-tagged moose that were alive as of 30 June

2001 (the end of this study) are currently being monitored.

35

.5822: a
835ch Logo: Emma; 8.: E 58-me 05:6 cam mam? 5 838.068 3008 Co 85:80. oSEmO .m 8:9...

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

_ 3:200 28:20.5 F \. ...
I}
a
32:00 ca 2532 I
more: . 8 K .
me. a.
«5%
one .6208 4 J irrhl / Jr
0032.50.— 68 D a. 4
...... 4 one o I we
I? w 3 o N
é ?
accommoz . a 1
. . mummy. Q g .
u S .E 83620 . ..... o
_ o... I I 88 I I
‘ II o P5922: Jam 3.... ‘
x 4.. mIF.118 E as 82
some 380.... I O 3 4 .
.m .28 4% rag. meuuw a
. C
J: ....
ox .. .02 D
4
x 5.0 mum:
C 0 mm \ .§ I
4 08~ 4
82 C _.ll.
32 D
r A 2883 2.88
H
....ouoEQO. or 0 OF \ k5?)—

 

 

 

 

 

36

Radio telemetry reloca tions
Aeria_l relocam

Twelve moose radio-tagged in 1995 and twenty-five moose radio-tagged
in 1999 were relocated on 1,573 occasions (7:43 relocations per moose,
range=12-59) from February 1999 through May 2001. The mean number of
relocations for males (n=14) was 41 and for females (n=23) was 44 during this
period. Twenty-nine moose radio-tagged in 2000 were relocated on 668
occasions (7:23 relocations per moose, range=5-37) from January 2000- May
2001. During this span, the mean number of relocations per males (n=9) was 22,
whereas for females (n=20) it was 23. Twenty-eight moose radio-tagged in 2001
were relocated on 143 occasions (i=5 relocations per moose, range=2-7) from
February 2001 through May 2001. The mean number of relocations for males
(n=11) and females (n=17) was approximately 5 for the period.

All aerial relocations were obtained during daylight between the hours of
0900 and 1800 (Figure 4). From February 1999 through May 2001, 768 visual
observations of moose (586 radio-tagged, 182 un-tagged) were made during 99
of 195 relocation flights. Eighty percent (615 of 768) of all the observations
occurred during 57 relocation flights (33:11 observations per flight, range=1-41)
conducted during the winter. The mean number of observations per flight during
42 summer flights was approximately 4 and ranged from 1 to 13. During a 19-
day period in late January-early February 2001, when snow and weather
conditions were apparently excellent, we observed 234 moose (149 radio-

tagged, 85 un-tagged) during 10 relocation flights. Aerial observations of moose

37

700

600
500 i
400 -
300

200 1

Number of observatlons

 

Number of observatlons

100

152
32
8’78 7’8 o
(‘0 In I‘ 8
6
1
. - I
8 8 8 8
(‘0 I!) f‘ O)

 

583
393
351
299299
158
84
31
O O 0 TO
0 O O O
:2 .‘1’ :9 it
Time of day
59
53 55
l
39
34
23
3 |
I .
8 8 8 8
S ‘2 :9 It
Time of day

40

2
‘F’T‘Y‘TT‘I *fi—i
O O O
O O O
O) v- (0
‘— N N
16
6
12 2 1
.. ILL-H
O O
a e .8,
s- N N

Figure 4. Distribution of radio-tagged moose aerial relocations (A) and ground
relocations (B) by time of day in the western Upper Peninsula of
Michigan during 1999-2001.

38

began to increase in the late autumn when leaves were gone and snow cover
was present (Figure 5). They peaked during the winter, and the number of
observations was greatly influenced by weather conditions. A recent snowfall
event and 100% cloud cover at an altitude of approximately 1,000 m. appears to
be ideal. The number of moose observed decreased rapidly after snow-melt and
sightings continued to decrease as leaves appeared and were at their lowest
throughout the summer.
Ground relocations

During the study we approached radio-tagged moose on 340 occasions.
We observed 293 individual moose during 190 (59%) of these approaches. Each
season as we gained more experience our success rate increased. It was 40%
in 1999, 55% in 2000, and 74% in 2001. Ninety-two percent of the animals we
approached were females, our primary targets. Males were only sporadically
approached and only male calves of radio-tagged cows were approached with
any consistency. Moose were approached between the hours of 0900 and 2300,
although 87% of approaches occurred after midday (Figure 4).
Survival and mortality

Annual, summer, and winter survival rates for each year for adult and
yearling moose are given in Table 2. First-year and interval survival rates for
each year for calves appear in Table 3.
Adult and yearling

Between 1 June 1999-31 May 2001, 10 adult (2 bulls, 8 cows) and 5

yearling (3 male, 2 female) radio-tagged moose died. At least one radio-tagged

39

.58 822.68.. beacon. mezzo coming—2 Co m_:m:_con_ Loan: Emacs 8:
5 58 8o; 88:. con; 8:9: 5:828 .mtom octao EcoE E 826288 089.: Go LonEzc coo—2 .m 8:9“.

  

 

 

 

 

 

 

 

 

 

5.32
m .... m. ...... w w m m. m ......
ril \\ ii lLll l i Ll - liil ri IT o
. W s
m
m
\ T m
\
w m
W n
\ l m
\ . 3 w
w . w
ll ; m l . m
m Sewn: \ . m. n4u
“cecal m l W 3 W
88-5 m m m
m m m.
\ I . ow s
w _ w
a . a
m a.
k . N
.. mm

.- on

 

co; “m 3.855 $3.25 Logo: 8
.63 F. 59 Eogmzv 5.88:8 82>an Boga—2 8.: ago: 8283.8 828 $225 a

 

 

 

..oooé opoo m was om Noo.o Zoo ..oooé mono N 50 Nm wood :oo 8500
..ooo... mono _. Ron 3. ooo.o No.0 ..oooé oooé 0 EN 0 ooo.o ooo._. 2.3m
aoooé :No N :m 3. wood Noo.o .oooé ohno _. ooN E. ooo.o o..o.o mag—50>
ooo.o ooo.o v No: on _.oo.o mroo ..oooé ooo.o N 34o mv Noo.o ooo.o 3:66.
COHE>>
..oooé Noo.o N on» on Noo.o Noo.o ..oooé ooo.o F mom or Noo.o ooo.o m>>oO
..ooo... ooo.o v oNo n.. ooo.o Pmoo ..ooo... ooo._. o For n ooo.o ooo... m__:m
«ooo._. 55o F ooN or god VNoo ..oooé oouo _. MEN 3 ooo.o goo mocEmm>
.oooé ooo.o m tom mv ..ooo ooo.o ..ooo... Noo.o _. moo oN Food ooo.o m=an<
EOE—tam
ooo.o oomo v N53 5‘ Food Noo.o ooo.o nouo v on: on Noo.o ogo m>>oo
.oooé vooo N moo or wood Noo.o ..ooo._. ooo... o No o ooo.o oooé m=3m
.ooo... MNoo v woo or god ooo.o ..ooo... ooo.o N tum NF Vooo ovoo moEtmm>
«moo vono o moNN om Pooo mmoo ooo.o otho v one No Food ooo.o 3:64.
_m3cc<
.03 .04 acumen .925 3008 mm m .03 .01. mcummc .33 3008 mm am $80
Axcmo a\omo .02 .02 .oz o\omo a\omo .02 .oz .02
rooNoooN oooN-ooo _.

 

.5922: Go 8855a Loon: E933 on“ E Foomooow
octao 863m 809: 05:88» new :39... 888.068 Co 88 E223» .8538 new .macc< .N aim...

41

.8; am 850ch E295 Loon: ..
.638. Bow Boga—2V LoomEzmo E228 29.382 9: 9.3: “02830.00 v.98 .0225 a

..oooé 3.0.0 m ..or or ooo.o ooo.o ..ooo._. nmoo N on or N_.o.o ooo.o .moE N_. I o
..oooé ooo.o F Now oN ooo.o mvoo ooo.o ooo.o v no 5. Zoo «ammo .moE o I o
ooo.o NNoo v moN ON Noo.o nono m..o.o omvo o 59. t o..o.o ooo.o 50>..um2E

 

 

 

.03 _O._ 2:000 .on $200 mm m _03 .04 05020 .moE @0200 mm mm P53005880
Rho abmo .02 .02 .02 a\omo a\..@@ .02 .02 .02
PooN-oooN ooON-ooo _.

 

.cmofozz 20 88:20.“.

Loan: 50885 9: c. Foowoomr 92% 928 20mm :08 8500 889.068 00 $00.: :00 ooooflé:
ocm 889-068 00 $28 3228 220.. com Co 20:06 N ...o ucm .oom ..0 mchE o-o 203-28.”. .m 030...

42

.58 52 5-82 9:; r 95% 5952.).
D6 0.35000 .0003 E00002, 05 5 £00.: >0 05000 0000:. mc__._mm> 0:0 0.000 0089.060. 00 000052 .m 059“.

5:05.
W m m

mm mam

P
m w
, .J mltfi O
0
w
m
I I I Mr F
ii}: : I. w
,mmc__._mw> D,
£32. _

 

 

 

 

 

 

 

 

 

 

p

 

 

“#0- T___ 0
N

43

adult or yearling moose died in every month except February (Figure 6).
Estimated ages of dead radio-tagged moose ranged from 1.0 to 13.0 for females
(Y=4.75 :l: 1.24, n=10) and 1.5 to 7.5 for males ()7=3.0 :l: 1.12, n=5). Forty-eight
adult (12 bulls, 36 cows) and twelve yearling radio-tagged moose were alive at
the end of the study (Appendix Table 4). The fates of three radio-tagged moose
(1 bull, 1 cow and 1 yearling female) that managed to slip their radio collars were
unknown. In addition, the fate of three adult moose (1 bull, 2 cows) outfitted with
GPS radio-collars were also unknown because the collars were removed to
collect the data each contained. Of the remaining 7 moose radio-tagged with
GPS collars, 1 bull had his radio-collar replaced with a VHF transmitter and 6
were still wearing their radio-collars when this study ended. The final ground
location of each adult or yearling moose that died is shown in Figure 7.

Annual adult survival was 0.88 in 1999 and 0.87 in 2000. Annual and
winter survival rates for adults between years (annual: Z=0.089, P=0.4645;
winter: Z=0.345, P=0.3677) were not significantly different. Although more adult
moose died in the winter (1 Nov-30 Apr, n=7) than in the summer (1 May-31 Oct,
n=4) (Figure 6) no difference was detected between winter and summer adult
survival rates in either year (1999: Z=-0.554, P=0.3010; 2000: Z=-O.357,
P=0.3639). Annual survival of yearling moose (1999 = 0.84, 2000 = 0.80) was
less than the annual survival of adult moose in both 1999 and 2000. Winter
survival of yearlings in 1999 (0.92) was 6% higher than in winter 2000 (0.87) and

in both years, summer survival rates of yearling moose were >0.91.

00022.2 00 0.305000 .0003 E00003 05 5
58-003 0550 000 >030 0.00 05 £53, 0000.: 000000060. 00 00=__0t0E 00.0500. 00 0000000.. .x. 050E

 

 

 

 

 

 

 

 

 

— c KwV
r 5:000 0000.205 _ \ _ . T.
a l
50:00 0000802 . .00 00.:
. g
0 3w . ma . \ x . K .V . a
> 2.5... 0x 0 .
0..., .../flag 0 5500 .0
. t 4 . 0
.. j .-
...... “$00 .. .. A . . . . .
‘ . .
00. 0x
5 y av t 00 0on ~ woman « 0
0830002 0 00 .. u ”4(a
. . . 48.? Q . .. ..mumn 0 B o ‘
1 . CO- E ..H.»mmmmi_.....
m_. M i . 20 . ....” yew—w ) . # \\\ll
0 20022.2 4 .... o 903%.... 86550
000 . .... a \.
a Q 0. COO .
5.030 00.0.0 .. /. 0 0..

 

I 0:22 .
‘

 

 

 

 

 

 

 

 

 

 

 

3:000 000000
4
, ,8... I
coon ‘ l
A 08. O
r 8883 0.0.8..
”MI-Ill“
o .. 2 \ - x035.

 

 

45

 

Bull and cow

In 1999, annual bull survival (1.00) was approximately 19% higher than
annual cow survival (0.84). In 2000, annual cow survival increased to 0.88 and
annual bull survival decreased to approximately 0.86. The high bull survival in
1999 is likely related to the small sample of radio-tagged bulls (n=9) that year.
Winter and summer bull survival rates in 2000 were nearly identical (winter=0.93,
summer=0.92). Annual survival rates between years (Z=-0.449, P=0.3345) and
survival rates between winter and summer in 2000 (Z=-0.398, P=0.3508) for
cows were not significantly different. Cow survival in winter 1999 (0.9107) was
neady identical to that in winter 2000 (0.9112). Summer survival rates of cows
between years were also very similar (1999=0.95, 2000=0.94).
First-year calf

Seventy percent of calf deaths (1999 n=6, 2000 n=4) occurred in the
winter (Nov-April) (Figure 8). However, this seasonally skewed calf mortality
pattern could be a result of (1 ) the frequency (monthly) at which calf survival was
checked and (2) the difficulty of detecting calf mortalities shortly after birth. It is
interesting to note that all radio-tagged (compared to un-tagged) calves that were
seen with radio-tagged cows in the spring survived at least their first year. First-
year calf survival in 1999 (0.63) was 20% lower than in 2000 (0.79).
Calves 0-6 mcfls of age

Sixty-seven percent (4 of 6) of calf mortalities in 1999 and 25% (1 of 4) in
2000 occurred within the first six months of life. 0-6 month calf survival was 20%

lower in 1999 (0.75) than in 2000 (0.94).

46

EON >05. 5 - 000? 0:2. F E0... 000.20:).
.0 0.300.000 .0003 E00003 05 0. 5:0... .5 05000 0000:. :00 00000.0: 0:0 000030.00. .0 .00502 .0 0.30.“.

0.0.0 0 m mlm rm . m . w ..m m

1L1 .lllt

 

 

 

 

 

 

 

 

. 5-000.“! .
__ 00.009 U.

 

 

831“?!)me

47

_(3_aLves 6-12 mont_h_s of age

Thirty-three percent (2 of 6) of calf mortalities in 1999 and 75% (3 of 4) in
2000 occurred between 6 and 12 months of age. 6-12 month calf survival was
1% higher in 1999 (0.84) than in 2000 (0.83).

Mortalities

Moose died of various causes (Table 4). Only one moose, a yearling male
(No. 106) was struck and killed by a motor vehicle during the study. All other
moose died of natural causes or accidents. The 2 accidents merit further
description. One yearling male (No. 114) died from stress and shock after
breaking through ice at the edge of a pond and being unable to get out of the
hole he created. Evidence at the site indicated that wolves had scavenged the
carcass, but it could not be determined if they had actually killed him. A bull (No.
74) also died from stress and shock after becoming mired in a mud hole. The
carcass of this moose was found mostly intact, claw marks on his back and rump
indicated that he was attacked by a black bear post-mortum.

Only 1 moose, a yearling female (No. 160) was confirmed as being killed
by wolves during the study. When captured she was originally identified as a calf
because of her size relative to other calves previously radio-tagged. However,
upon her death, aging by counting cementum annuli revealed that she was an
extremely small yearling perhaps stunted by her ailment (see Table 4). In
addition, capture stress may have contributed to her death.

Three moose were verified as having died from complications related to

cerebrospinal nematodiasis (brain worm). Necropsies revealed that 47% percent

48

000.000. .0: 0.00 o
.0.0>.0:0 .0225 ..00 5 0003.05 .0: ”0:20 00.3000 ..
0.0.. 0:0 020:5... 0

000.03 .00 o\: .0. 30:08 .350“. n

.._:::0 :.2:0:.00 a

 

 

:898230: 0.8.2288.888.: 58.02 F 5.5 u. 0: 8.
0.8.888: 0:08.880 88 8. 2 8.8 . 8.: 8:.
0.8.0.2008.58:58:. 0.8.888: 0:08.880 38.9.. 8 5.8 s. 8.8 8.
0.8.88: 0.8.0288..:o_.8__::.8 8.5.0 88 >22 .0 :2 u. .32 mi
88.8.:8...:28558. 0.8.0.2008. 88 a: :2 s. .00 3:.

08:8... :2 88 :8 8 :2 . 8.. m:

8.08.8...858. .08....288. 88 8: 8.0 . 8.0 8.

.:0..000:00 30:95:: 80.005000 :082. :00000000 000.. >02 3 N080 “. om.N 0:
0.8.2288..:o_._.s:_m:. 5.888 089:3: 88 80 t 2.: u. 8... 0:
.8808 088: .888... 03.88. 9:85.808 08:0 88 8.2 8 0.8. s. 0: v:
0.8.0.2008.88.80:. .:o_.80:8 00:053.. 88 80 8 «Ev . 0: :3
08:8... :2 88 .62 0. :2 s. 8.0 .8

00.818.88.058: 80. 80 2 8.8 s. 0: 08

.808 89.0 58.98 8.8 u. 8... 8.

.80 2:. 0858:: 80.88: 88 :2. 8 :2. u. .53. 8

.8808 :88: 08.8.8.8 8:08.880 08. 80 8 8.8.. U. 8.0 8

.0.00.0.o.0.000.. .0.:0E:0:0

8.28:8 .888... 03.08:. 058.....88 08:0 88 03:: 3.8 .2 8: E

2
0.00:005 0.00 £000 £005.... x0m ~00< 0000.2

 

.:00.:0..2 .0 0:05:00 .000: E0002, 0:. 5 58-003 05.00 00.030 0000:. 000090.00. .0 £000

.0 00:00 0000:00.0 0:0 £000 .0 0.00 0.0808500 6.. .0. 2.9.0:. SE0. .00 .5000 .0 000 .0. 0000.2 .0 0.00 ..

49

(8 of 17) of moose that died during the study were parasitized by the large
American liver fluke. Five moose with liver flukes were also malnourished
Movements

Home ranqe si_ze and dispersion

Twenty-two percent (4 of 18) of adult radio-tagged moose were seasonally
migratory during 1999-00. Three moose moved between distinct winter and
summer home ranges. One cow was relocated in the same general area from
winter until late-spring, then moved to a separate, distinct summer-fall home
range. During 2000-01, 38% percent (14 of 37) of adult radio-tagged moose
were seasonally migratory. Thirteen moose migrated between winter and
summer home ranges. Again, as in 1999-00, the same cow migrated between a
distinct summer-fall and winter-summer home range.

Seventy-five percent (3) of migratory moose in 1999 and 64% (9) in 2000
had seasonal home ranges that did not ovedap. The remaining 6 (1 in 1999, 5 in
2000) had adjacent seasonal home ranges that overlapped < 25% (Table 5).
Migratory moose with adjacent seasonal home ranges moved an average of 6
km (range=2-1O km) between summer and winter ranges, whereas those with
spatially separate seasonal home ranges moved an average of 14 km (range=4-
26 km).

The median size of winter, summer and annual home ranges of migratory
adult moose were 23, 39, and 53 km2 respectively (Table 6). Seasonal home
ranges of migratory adult moose did not differ between winter and summer

(Mann-Whitney U test, ZMwu=-1.5615, P=O.1184). All migratory adult moose

50

.83. 0:0..00...0.0 508...... 00002“. .08.; 0:0 .08800 .0 0.0800 c0080.. 05.80.85 0
00:0. 080: 8000.00. I < 60:0. 080: 0.0.0000 I m ...

 

 

51

878 02-8 .0... 8.0 0 u. 8.
80-0 02-8 00... 8.0 0 n. 00.
80-8 82-8 0.... 8.0 0 . .0.
8.30 82-8 8.0 8.0 < u. 8.
82-0 82-0. .0... 8.0 < n. 00.
00-0 82.8 8.0 a... < 2 8.
82-. . 82-0. 8.0. 8.0 0 u. 8.
82-. 82-0. 8.0 8.8 < u. 8.
00-8 82-8 0 ..0 0.8 < n. 0. .
3.8 8.0 0 n. 0:
870 82-0. 00.0. 8.0 0 n. 2.
80-. 82-8 00... 8.0 0 . 8.
80-. 870. 00.0 8.0 0 .2 00
00-0. :80 .00. 8.0 0 u. 8
8-88
82... 82-0. 8.0 8.0. < n. 0:
3.8 8.0 0 u. 0:
870. 8.38 00.. 8.0 0 2 00
80-8 82-0. 00.3 8.0 0 2 8
8000.
001:0. .0.:..s 001:0. .088:0 0.80.. 00t0>0 000.». 0.
:0 .02.... :0 .0>...< 00:0.0.0 00 00:0. 0801 x00 0000 .00>

 

80080.5. .0 0.00:800 .000: 80.002, 0:. :. Sow - 000. 08.00 00.030 00008 00000.
-0.00. .300 80.0.0.8 .0 000:0. 080: .08.; 0:0 .08800 :00500 00:0.0.0 0:0 0000>0 .0 000800.00 .0 0.00..

 

 

 

 

 

 

 

 

 

001. 00.. 0 0.. vol. 00.0 0. mm 0 .0 mm .0.0. .0 M
owl. 00. . n 0 SI. 00.0 N. 00 0 ..I0 om .:00.00m.
wmlm 00.0 0 0. Elm 00.... mm mm 0.10. 0 28.0.0.5.
.0.:.>>
whim 00.. 00 mm 02.4 00.0 mm 00 VNI: 00 .0.0. .0 M
more 00.. 00 cm 02.4 00.0 00 ..0 VNI. .. 3 .:00.00m.
mlm 0 . .0 cm 00 0010 00.0 00 00 VNI. . 0. 80.0.0.5.
.08800
.0Io. 00.0 .0 .0 mm ...-3 0.0 00 00 000. mm .0.0. .0 M
0910. 00.0 .0 cm 00L: 00.0 .0 0.. 000.. .0 800.001
.010. 8... 8 8 8.18 N... 00 00 8:8 0. 80.0.2
.0:::<
.085 N08... ANEM. $80.. 0:0..000. 00008 002.
00:0m mm :0.005. :005. 00:0m mm :0.005. :005. .02 .02 00:0. 080...
502.00 x0>:00
8088.8 0000.. .0800. 00x... 0000

 

80080.5. .0 0.00:800 .0003 80.003 0:. :. Sow-000. 0:..00
00.030 00008 00000.0.00. .....00 80.0.0.8 0:0 .:00.00. .0 00~.0 00:0. 080: .0:00000 0:0 .0:::< .0 0.00...

52

were found on their summer ranges by the end of June each year. The earliest
date a migratory moose was detected on its summer range was 3 May, the latest
19 June. Seventy-seven percent (17 of 22) of migratory adults had moved to
their winter range by early December each year. The earliest date a migratory
moose was found on its winter range was 27 September and the latest date 7
December. These dates do not include 3 cows that were not found on their
2000-O1 winter range until early January. However, because these animals were
not relocated in December it is possible that they had migrated earlier than was
detected.

In addition, a bull and a cow were not found on their winter range until 13
January and 25 January, respectively. The bull made sporadic, relatively long
distance movements ()7 = 5.22 km) between successive relocations during the
autumn and early winter before moving to his winter home range. Prior to
moving to her winter home range the cow established a distinct autumn range
(16 August to 19 December, 2000-01).

The lone cow with atypical migratory behavior was found on her 15-km2
summer-fall range from 13 July to 11 October in 1999-00. In 2000-01 she was
found in the same, but smaller area (10-km2) from 4 August to 9 October. In
1999-00 she was located on her winter-summer range (43 kmz) from 28 January
to 20 July and in 2000-01 from 1 November until her death on approximately 20
December.

The remaining 37 adult radio-tagged moose were classified as resident

animals. When winter and summer home ranges of resident moose were

53

delineated, they overlapped >75%. Resident adult moose had median winter,
summer, and annual home ranges of 12, 36, and 41 km2 respectively (Table 6).
Resident adult moose had significantly larger median summer home ranges than
median winter home ranges (ZMWU=-3.4255, P=0.0006). Annual median home
ranges of migratory adult moose were larger (ZMwu=2.4664, P=0.0136) than
those of residents, whereas summer (ZMwu=0.5266, P=0.5985) and winter
(ZMwu=0.7374, P=0.4609) home ranges did not differ between the groups.

Median home ranges of migratory and resident adult moose combined
were 15 km2 in winter, 37 km2 in summer and 44 km2 annually (Table 6). Median
summer home ranges were significantly larger than median winter home ranges
(ZMwu=-3.7313, P=0.0002). There was no difference between median home
ranges of cows and bulls (ZMWU=0.0102, P=0.9919) (Table 7). Adult cows
attended by calves had annual median home ranges that were 10% smaller than
those of cows without calves (44 km vs. 49 km), but the difference was not
significant (ZMWU=0.6664, P=0.5052).

Median FK-95°/o adult annual home ranges were 42% larger than MCP-
100% adult annual home ranges. Summer and winter median FK-95% home
ranges of adults were 85% and 88% larger respectively, than summer and winter
median MOP-100% home ranges of adults.

Annual median AD values (Table 8) did not differ between bulls and cows
(ZMwu=0.2166, P=0.8286), between yearling males and yearling females (ZMWU=-
0.6429, P=0.5203) or between cows with calves and cows without calves

(ZMwu=-0.1.4336, P=0.1517). However, at the 0.10 level of significance, annual

54

 

 

 

001.. .00 .0 00 00100 .00 00 .0 00:00 . . 828
.30 \>> 0250
0010. 00.0 00 00 0.70. 0.0 00 00 001.0 00 0028
.3 0.500
.010. 00.0 00 .0 00.10. 0.0 00 00 00 0260 __<
00.0. 0.0 00 .0 00:00 .00 00 .0 0010. N. 230
0.33.
ms... .05.. .05.. .00... 08082 88:.
0000a mm 00.005. 0005. 0000.“. mm 00.005. 0005. .02 .02 02.02.0006

 

00000.00 x0>000
808.0.8 0\000.

.0E0x 00.0.. 0\..00

00080.5. .0 0.000.000 .0003
80.002. 0... 0. .000000. 00000 00.030 00008 300 000 =00 00000.0.00. .0 00N.0 0000. 0800 .0:00< ... 0.00..

55

Table 8. Annual and seasonal Euclidian (or linear) distance (AD) between pairs
of relocations of adult and yearling radio-tagged moose studied during
1999-2001 in the western Upper Peninsula of Michigan.

 

Distance (km)
Classification No. No. Mean Median SE Range
moose locations

 

 

 

 

 

Adults
Bulls 11 20 - 25 4.33 4.02 0.42 2.37 - 6.69
All Cows 45 20 — 34 4.05 3.72 0.47 1.89 - 14.51
Cows w/
calves 32 20 - 34 4.75 3.43 0.58 1.89 - 14.51
Cows w/ out
calves 11 21 — 30 5.23 4.02 0.94 3.12 — 13.90
Yearlings
Males 6 21 -23 5.54 4.80 0.91 3.56 — 9.56
Females 5 20 - 24 7.74 3.95 3.91 3.05 - 23.30
Summera
Bulls 4 10 - 12 3.12 2.84 0.56 2.18 — 4.63
Cows 30 10 - 17 4.06 3.09 0.61 1.26 - 5.17
Yr. males 3 10 — 10 3.26 3.70 0.54 2.19 — 3.91
Yr. females 5 10 — 16 3.14 3.22 0.60 1.60 - 5.08
Fall”
Cows 8 10 - 14 4.60 3.05 1.21 2.20 — 12.51
Winterc
Cows 20 10 - 13 2.94 3.04 0.34 0.54 — 6.52

 

aSummer = May - August.

t’Fall = September - December.

°Winter = January - April.

median AD of yearlings was greater than that of adults (ZMWU=1.8646,
P=0.0612). Adequate numbers of relocations (>9 per moose) were available to
make statistical comparisons among seasonal AD values for radio-tagged cows
only. Mean seasonal AD values indicated that the movements of cows were less

dispersed during the winter (Y=2.94 km) than during the summer (7:4.06 km)

or fall (3?=4.60 km) (Table 8). However, the Kruskal-Wallace test detected no

56

significant difference (x§=1.243, P=0.5372) among median seasonal AD values

of cows.
Dismersal and home range development

Six radio-tagged moose permanently dispersed out of the core study area,
3 each in 1999-00 and 2000-01 (T able 9, Figure 9). Dispersers included 3
yearling males and 2 yearling females (see below), and 1 adult female (cow).
The cow, 1 yearling male and both yearling females dispersed in April-June. The
other 2 yearling males dispersed in January and September. The mean :l: SE
linear distance dispersed was 80:l: 16 km and ranged from 30 to 134 km. The
average number of days a yearling moose wandered from the time a dispersal
movement was detected until it ceased because the moose established a new

home range, it died or dropped its radio-collar or the study ended, was 164 days
(range=25-592). The estimated annual dispersal rate in 1999 (D =0.068) was

26% greater than in 2000 (1") =0.054) (Table 10).

Two moose that dispersed, a yearling male (No. 104, see details below)
and a cow (No. 185, radio-tagged 13 February 2001) established new home
ranges in northeastern Wisconsin (Figure 9). During a 16-day period (8 April-24
April 2001) the cow moved 64 km in a southwesterly direction to an area in
Forest County Wisconsin where she permanently settled (Figure 10).

In addition, a cow (No. 102) captured as a yearling on 29 January 1999
permanently shifted her range within the core study area (Figure 11). Following
capture she established a home range of 23 km2 that was 3-4 km southwest of

the town of Nestoria which she occupied through March 2000. On 28 April 2000

57

.00000 >030 .0 0.000 ..0__00.0.00. 00000.0 .0000. 080.. 300 .0 8085.80.00 ..8: 0
00000020. 0080808. 0000.00. e
.000..000.0. .00.000.0-.000 000
0.0 0003.00 .0 00000. 0800 .00.000.0..000 000 0.0 .0 an. 00..00...0.0 00..0~....0 a\a00 x“. .0 0.0800 0003.00 00:80.0..0 0
..00.000.0 .000 0000. .0... 0.00 000 0000. 0800 .00000 .0000. .0 .0.00 00 0000. .00. 0.00 0003.00 0.00 00.005. 0

 

 

.000-00..-00 00.0020. .00.. 0. .0 00 .000.>05.-0 > n. 00.
0.00000.>> 0. 0000. 080... 0 00 00 .000..0<.0. < u. 00.
0000-070. 0:8 8000.0 00 00 .0 0000..0<-00 > u. 00.
500-0070 00.0020. .00.. 0.. ..0. 0.. 00000070 > 5. .0.
0.00000.>> 0. 0000. 080... 000 000 00. 000.0000. > 5. 00.
000.0000 00.0 00 00. 00 000703.00 > 5. 00.
.050 .8...
0000.0.0 0000.0.0 000.0...0. 0.
0.0“. 00.00. 80.0005 .00.000.n. 08000.0 0.00 000. x00 00005.

 

00080.5. .0 0.000800
.0003 80.002, 00. 0. .000000. 00000 00.0 030 0.00 0... .0 .00 000.0008 .00. 00008 000030.000 .0 0.00...

58

000.025. .0 0.000.000 .000: 80.002, 00. 0. 500-000.
00000 00.0 >020 0.00 00. .0 .00 000.000.0 .00. 00008 00000.0.00. .00 .0 0.008058 000 000..000.0m. .0 0.00.”.

 

0000

 

 

 

 

S

 

 

  
  

00000.05

/

00000.05.
0:

 

 

 

 

b.

 

00.00

.000000 00.0..

 

 

 

 

 

H”
0.0.0520. 00 0 00
... z_wzoow.>>
~ c
00.. .7 3
I §
0
I r

‘l‘l

‘ 000000.00
00.0000...

.00000m 00.00

 

 

 

 

 

 

0.00000

{0351 at: \<
1.858 058 \/\
0.230 8! €30 \<

8. 4

a. o

.2 I

8. <

8. o

H

 

 

 

 

59

Table 10. Annual dispersal rates of radio-tagged adult and yearling moose
studied during 1999-2001 in the western Upper Peninsula of

 

 

Michigan.
Year Rate SE No. No. No. 95% 95%
moose months dispersing LCI UCI
1999 0.068 0.0017 42 514 3 0.000* 0.139
2000 0.054 0.0012 56 657 3 0.000* 0.112

 

‘Dispersal rates calculated using the Mayfiled estimator (Mayfield 1961,1975).
'Lower interval truncated at 0.00.

she was relocated 5 km south of Shank Lake in Iron County. Approximately 1-
month later (25 May) she was relocated 3-km northeast of Van Riper State Park,
35 km to the northeast of her previous location. Four days later (29 May) she
was found18 km further east. Thereafter, she established a new home range of
21 km2 several kilometers west of the Dead River Storage Basin by 20 June
2000. She occupied this home range until her death on 25 May 2001.
Twenty-six calves - the offspring of 23 radio-tagged cows - were radio-
tagged during captures in 1999, 2000, and 2001. The areas occupied and
movements after cow-calf separation of 11 (7 males, 4 females) of these moose
(henceforth referred to as yeadings) for which an adequate number of relocations
were available, were examined. In addition, the movements of 2 yearling males
who wandered extensively prior to permanently dispersing were investigated.
The mean:l: SE age at which a calf became independent of its cow was
1110.25 months and ranged from 7.5 to 13.0 months. Calves (n=5) of cows that
were barren the next calving season remained with their cow 11-days longer than

calves (n=20) who’s cow gave birth the following season. The median date of

60

-'

.5953 B m_:wc_con_
.03: E283 9: 5 room 9:8 33 .05 $008 Boo $89.28. So 35890:. new mcozmoemm .o_. 059“.

 

{M 220205. 2

 

\

 

 

 

 

 

H
o 0..

 
     

. «so:

 

 

 

 

 

 

 

 

/..

 

tmmmo .305 on.—

2.3

 

 

 

 

 

 

 

 

9.
9.
9.3 32.5 as
.933. oEEmmEQs.
... 9.3 neon.
%
>5 4
fishes m 358 5..
33 82*
0x3 95. I
9.3 8:25

 

 

 

 

 

 

61

:mm_:o=2 :o m_:m:_:mn_ .63:

52.0.95 9: :_ 58-89 95.5: ANS. .oz. 88:. Boo 883-069 B mEoEo>oE Em mcozmoofim ...P 939“.

 

 

 

E29530. m o m

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

 

C

 

3:300 :9.
19.3 95..

 

 

 

 

 

A

 

 

 

 

 

 

 

a .
3:80 2.30.5.2 5%me xcmzm
. .6208”. «are A:
@9803: 3.0 “we
a. ... 23 32
Mv o
s it 29... p
5e 82 52.2 an
...... 5580.. 9.2qu @
SEEN—.0 W 9.3 .
.. mm
magma e was
s i . ..E >.
m ilwr. .1 Jr
Hemaammmmmx_ av r A» 9.5302 mm
9.3 m3 #4! . as ..
a 382m ...... fiv. b A
2m :80 av a. 529:! \/\
9.9. 85.: a, b 9.3 .90 s... as. k z
5% ...... a s _,
. ._ ._ to: o .. 3:300 393 35.338 2
Emmm 9.3 525 I ۤ8E<sam >\
‘9 «8:39.: 38: Q \
N N

 

 

 

 

62

separation for the former group was 21 May, whereas for the latter group it was 1
April.

During their first one and a half years following independence, yearling
moose exhibited three movement patterns, they either (A) permanently dispersed
out of the core study area over 1 to 16 months, (B) occupied areas that partially
overlapped their cow’s summer and/or winter home range for periods of 9 to 13
months or (C) occupied areas in the summer that partially overlapped their cow’s
summer home range but moved to completely separate areas in the winter.

Two yearlings (1 male, 1 female) captured in 1999 and 2 (1 male, 1
female) captured in 2000 permanently dispersed after separating from their dams
(pattern A). Also, a yearling male (No. 106), captured alone on 29 January 1999
permanently dispersed. Three yearlings occupied temporary home ranges that
were completely separate from their cow’s home range prior to dispersing.
Yearling male, No. 104 (radio-tagged 30 January 1999) occupied a 17 km2 home
range, yearling female, No. 105 (radio-tagged 30 January 1999) a 39 km2 home
range and yeaning female, No. 134 (radio-tagged 20 January 2000) a 54 km2
home range.

No. 104 occupied his temporary home range through early September of
1999, then wandered in a westerly direction (Figure 12). He was next found near
Lake Gogebic where he remained until 6 September 2000. Then, from mid-
September 2000 until late-March 2001 he moved 82 km in a southwesterly
direction. In late-April 2001 he appeared to establish a home range northwest of

Mercer, Wisconsin. Taken in whole, over a 242 day period (7 September 2000-6

63

.:mm_:o__>_ .6 m_:m:_:on. :95: E933

9: :_ 58-89 9.5: $9 .02. 38:. 63:. m:_:mo> comm—3.0.3.. ho mEoEm>oE ccm 20.60261 .9 9:9“.

 

 

 

H
920.539. 0 P o 9.

 

 

 

...!
to;

 

 

.. $9 52.2. on
\J :883 93:00

 

 

 

 

 

 

V
2: mac...

 

Z_wZOOw_>>

 

 

 

5.595 9.3

 

 

 

 

 

 

 

room 0:2. v
:883 $3

 

 

 

...-29: at:
8:85am €80
.9938 82 .68

28309:. 38!

 

$222

 

 

 

 

64

May 2001), No. 104 covered a distance of at least 205 km.

Following separation from her cow in late February 1999, No. 105 was
found on her temporary home range through March 1999. Thereafter, over a 2-
month period she moved 48 km in an easteny direction (Figure 13).
Unfortunately, any further monitoring of this moose ended when her radio-collar
was recovered on 14 July 2000.

From the date she was first relocated alone on 25 may 2000, until her
signal was temporarily lost after 18 April 2001, No. 134 wandered throughout her
temporary home range. On 15 May 2001 we picked-up her signal and relocated
her 5-km southeast of Lake Independence, 31-km to the northeast of her prior
location (Figure 14). When last relocated (29 June 2001) she had moved 3-4 km
to the southeast.

The other 2 males (Nos. 106 and 131) that exhibited pattern A movements
wandered over large areas (>250 kmz) before eventually dispersing. After
roaming nomadically for 4 months, No. 106 was struck and killed on 19
September 1999 by a westbound vehicle on state route M-28, 0.4 km west of the
Sand River (Figure 15). Eight days after being captured on 20 January 2000,
No. 131 was found alone, >7 km from where his cow (No. 123) was relocated.
Following long movements both south (36 km) and north-northwest (60 km) of his
capture location over a 9-month period, his signal was lost in late-October 2000.
Five months later he was found 85 km southwest of his previous location. He

was last relocated on 5 June 2001 near Lake Gogebic (Figure 16).

65

.:0o_:0__2 ..0 0.:0:.:0n. .0003 £0.00?
05 :_ 000N002. 0:030 Sop .02. 0000.: 0.050.. 05:00» 00390.00: 00 $:0E0>0E 0:0 050000.01 .0? 0.50....

 

 

20.0805. 0

 

o m

 

 

 

Wrim).alrw00. 0003902

9.3 2.

      

.6208”.
0003:0020

 

 

 

 

 

.. . 40
mm

A?

0% 535.20
88 ...: 3
......m... ..N=OO 000 90 V W“
......m... mv—NJ DOI—

.0055
.0000 0v

9.0.. 255

 

Camim. QXNA hm>=w .
$7

 

.. .. . C
. a
3:300 :2.
\
\txx03 0:0m
.% _
0 0 :0
t. a. J /.M v. .. v. ...m.
6
0
.00
.\¢
a“?
1.1.1
..........mmwwwwwmama...”0....H....Mummmfi...
........ mam? E0200 om WW
. :000003 0.3000
m l.~(n.. 0:07.02 .
be... a. r
.0. G
a, 0.0 9.0.. ..0.0 A? $2913.... \/\ r
A 25 .4 ...
F9500 02< 63% 2
08308". 0082 U

 

 

 

 

 

 

 

 

    
   

 

 

 

 

 

 

 

 

 

 

66

 

.:0m_:0_.2 ..0 0:05:00 :0003 E00003
05 :_ ..ooméoow 0:030 $2 .02. 0000:: 0.09:0. 05:00.» 008000.00: ..0 0.:0E0>0E 0:0 m:0=000.0m .3. 0.39“.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/ _ RN \J \.\‘
$0592 0 o m 3:300
fl 0:030:05.
Sow B0200 om
/ . . 5:80.. 0.3000
5000 000.20
.021 0000
Sow 0:3... am
:00003 .00..
19%
......so
800030 . .
9.0..
00:00:00005
00.0..

5000
00.0.. 02.0

 

 

 

 

 

/§*

 

.0560
000:0m

 

 

 

83¢ €30
5.8500 €30
0.358 82 .630

.831 Gog

 

3222

 

 

 

67

00030.5. 00 0.30:.:0n. .0003 £0.00;
05 :_ 800-000.. 0:030 89 .02v 0000:. 0.08 05000.» 000000200: ”.0 0.:0E0>0E 0:0 200000.01 .0? 0:30.“.

.... ..W

 

 

 

  

 

 

 

 

 

 

 

 

/ . / 50:00.0 0.2:“...wa

00.0.. 0.0.

.. . 0 .. W1
I r .. :0: .

bu 0W0. .82” a

 

 

.»

 

 

 

 

 

 

      

 

 

 

 

 

9.0.. 95.
n. 5.0:.000 .
82.85 802 0«0:20.002 0
80000.. .0022 0.550.... 0x0.
1 . as”. 000.002. . a
a .
5.00“... 083000. /.
8:003 0 .
. 0 .3 a? 5.8 if... >\ n
mums—000.2 0.0.. 02.0
000.00 :2 0..
808500.330 >\
...-95815165 2
08310034 Q I
J 1 q

 

 

 

 

 

 

68

.:00_:0__2 .0 0.:m:_:0n. .0003 50.003
0:. :_ 0000-000. 0:000 30. .02. 0000:. 20:. 05:00.. 000090.00: .0 0.:0E0>0E 0:0 0:0..000_0m .0. 050.”.

 

In“).

 

 

 

 

0.30805. or o 0..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

z.wzoow_>>
d
V ... .
0. .0 .
o h C
88253
:00000..m0.
0.00000 . . VII.
. 00.0.. 050000
x
. $.00... j
\\.\ x. \
000:0m :000:0.:O if: >\
8000500330
\ COHSODOI $950008<33m &
A? 0 3.50.6 00.0. .s§%¢.§: -

 

 

 

 

 

69

 

K‘

 

Six yearlings (4 males, 2 females) exhibited pattern B movements after
separation from their cows. Four were offspring of resident cows and 2 were
offspring of seasonally migratory cows. The meaniSE and median areas
occupied by these yearlings was 57 :l: 13 and 42 km2 (range=30-108 km). There
was a 47% (range=31-72%) overlap of the area occupied by offspring of resident
cows and their cow’s annual home range. The distance between concentrated
areas of activity (i.e., the FK-25% UD) of a resident cow’s seasonal home ranges
and the area occupied by her offspring averaged 4 km in the summer and 7 km
in the winter. One yearling male (No. 107), the offspring of a cow that migrated
between summer and winter home ranges occupied an area that overlapped
27% of his cow’s summer home range. His area of concentrated activity was 11
km from that of his cows’ when she was on her summer home range and was 13
km after she had migrated to her winter home range. The remaining yearling
male (No. 114) was the offspring of the cow (No. 115) that migrated between
summer-fall and winter-summer home ranges. He occupied an area of 24 km2
between 26 April 1999-31 March 2000 that included 18% of his cow’s winter-
summer home range. The linear distance between the cow’s center of activity
and that of her yearling was 6 km when she occupied her winter-summer home
range and was 29 km when she occupied her summer-fall home range. On two
separate occasions, No. 115 made exploratory movements outside the area he
occupied. On 13 September 1999 he was located 16 km to the northeast, within

4 days he had returned. The second movement began in mid-October1999 and

70

lasted until late-January the following year. The linear round-trip distance he
moved during this period was 50 km.

Three yearling males (Nos. 138, 148 and 153) exhibited pattern C
movements after separation. Between late-May and mid-September No. 138
occupied an area of 27 kmz, No. 148 an area of 10 km2 and No. 153 an area of
15 km2. The overlap of these areas with their cow’s summer home range was
20% for No. 138, 8% for No. 148 and 10% for No. 153. The distance between
the activity centers of these areas and that of their cow’s summer home range
varied from 1 to 3 km. In late-September all 3 moved to areas that were
completely separate from the winter home range of their cow. No. 138 occupied
an area of 12 km2 that was 50 km distant, No. 148, an area of 9 km2 that was 41
km distant and No. 153 an area of 20 km2 that was 18 km distant from their cow’s
winter home range. All 3 remained on their winter range until late-April. No. 153
died on 28 April 2001, No. 138 was last relocated 16 km south of his winter range
and No. 148, 28-km southwest of his winter range.

Reproduction
Pregnancy determination

Relative to the 84% average pregnancy rate of adult moose in North
America (Boer 1992), pregnancy rates in the western Upper Peninsula of
Michigan were moderate each year: 78% (n=14) in 1999, 70% (n=19) in 2000,
and 74% (n= 31) in 2001.

Seventy percent (24 of 34) of cows from which blood samples were

collected in 1999, 2000, and 2001 had detectable PSPB levels indicating

71

pregnancy. Pooled 2000-2001 meaniSE and median PSPB values were
411.051: 69.38 and 387.30 ng/mL respectively (T ab|e1 1 ).

Table 11. Pregnancy specific protein B (PSPB) concentration (ng/mL) in blood
serum collected in winter from radio-tagged cow moose in the western
Upper Peninsula of Michigan, 2000-2001. All cows with detectable
levels of PSPB were considered to be pregnant.

 

 

Year 7 SE Median Min Max n
2000 297.01 51.38 261.55 76.70 553.70 8
2001 541.39 123.43 475.40 76.50 996.60 7

Pooled 411.05 69.38 387.30 76.50 996.60 15

Due to the small number of twin calves produced by cows for which we
had blood samples, we did not attempt to correlate PSPB to calf production.
Eighty-nine percent (8 of 9) of cows that had positive PSPB results in the winter
of 1999 were seen with calves in the spring. The lone cow not seen with a calf
was missing at the time calf checks were conducted. Three cows that had
negative PSPB results in the winter of 1999 did not reproduce calves. In 2000,
83% (5 of 6) of cows that had positive PSPB results were believed to have
reproduced. We were unable to assess the calving status of one cow until mid-
July and it is possible that she had given birth and that her calf had subsequently
died. PSPB results in winter 2001 diagnosed 70% (7 of 10) of captured cows as
being pregnant, however only 3 pregnant cows were seen with calves in the
spring. Serum and fecal progesterone levels also indicated that these animals
were pregnant. It is difficult to believe that all 3 testing methods would result in
false-positive results. Although these cows were approached multiple times

during the calving season it is possible that we missed the calf (calves), (1)

72

Table 12. Progesterone concentration (ng/mL) in blood serum collected in winter
from radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. The 95% Upper tolerance limit (95% UTL) or
dividing serum progesterone level between pregnant and non-
pregnant cows was 1.68 ng/mL.

 

 

 

Pregnant moose Non-Premant moose
Year x SE Median Min N )7 SE Median Max n
1999 5.74 0.97 4.82 2.47 9 0.99 0.45 1.26 1.60 3
2000 4.44 0.21 4.50 3.55 8 0.18 0.08 0.18 0.33 3
2001 3.16 0.20 3.47 2.23 7 0.65 0.16 0.56 0.96 3

Pooled 4.56 0.42 4.28 2.23 24 0.61 0.18 0.43 1.6 9

 

 

12.00 —,
11.00 ‘ .
10.00 1 ’
:7 .
g 9.00 .
g BOOT
o 7.00 x
C
g 6.001 .
a 5.001 *
a . 8
g 4.001 . ! ‘
5:} 3.001 ‘
= 200 '1 . ‘
L. - . W_ _-_+_¢mmi
0 2
m 1.00—; 8— 3
0.00 +~—-—————~——s>—~_._~—9»—._——a .
1999 2000 2001
Year

Figure 17. Progesterone concentration in blood serum collected in winter from
radio-tagged moose in the western Upper Peninsula of Michigan,
1999-2001. Each symbol represents the serum progesterone
concentration of a single moose. The 95% upper tolerance limit
(95% UTL) or dividing serum progesterone level between non-
pregnant (0, O, A) and pregnant (0, O, A) cows was 1.68 ng/mL.

73

because the calf (calves) died shortly after birth or, (2) the calf (calves) died
sometime in the interval between cow sightings.

Useable blood serum samples were obtained from 33 cows. Blood serum
progesterone concentrations fell into fairty distinct pregnant and non-pregnant
groups each year (Figure 17). Mean :l:SE and median values pooled across
years were 4.56 $0.42 and 4.28 ng/mL for pregnant cows and 0.61 :l:0.18 and
0.43 ng/mL for non-pregnant cows (Table 12). As expected, blood serum
progesterone concentrations of pregnant cows were significantly different from
those of non-pregnant cows (Z=-4.3454, P<0.0001) (Mann-Whitney U test). The
95% UTL of serum progesterone concentration for non-pregnant cows was 1.68
ng/mL. Using the 95% UTL, the pregnancy status of all cows would have been
correctly identified (Figure 17).

Between January 1999 and May 2001, we collected 119 fecal samples
from 36 radio-tagged cows. Multiple samples were collected from 16 cows in the
winter of 2000 (2 =2.19 :I: 0.100) and from 19 cows in the winter of 2001 ()_(=2.05
i005). The results of 3 fecal samples with very low progesterone levels (<1.4
ug/g), 1 of 2 samples collected from 3 different cows, were not included in the
analysis. They were removed because the cows were seen with calves during
spring calf checks, results from other fecal samples indicated that the cows were
pregnant, and the origin of the samples, when collected in the winter, was in
doubt. Fecal progesterone concentrations also fell into fairly distinct pregnant
and non-pregnant groups each year, although the results were not unequivocal

(Figure 18). Pooled across years, the meaniSE and median values for

74

Table 13. Progesterone concentration (pg/g) in fecal material collected in winter
from radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. The 95% Upper tolerance limit (95% UTL) or
dividing fecal progesterone level between pregnant and non-pregnant
cows was 5.17 ug/g.

 

 

 

Premant moose Non-PLegnant moose
Year 2 SE Median Min N )7 SE Median Max n
1999 6.07 1.26 5.36 1.13 9 0.97 0.35 1.12 1.50 3
2000 11.82 0.68 10.45 4.29 36 3.23 0.45 3.10 5.9 12
2001 22.30 1.40 22.89 1.35 42 1.65 0.44 1.27 4.22 11

Pooled 16.29 0.98 14.46 1.13 87 2.30 0.32 2.66 5.90 26

 

3500 -~‘
l A
30.00 »‘ g
g 25.00 A
5 ;
O l
20.00 -;
<3 i ' 2
3 i . A
in 15.00 i ,
0 . A
8’ : I
a 10.00 , ' ‘
5 ‘ 0
g 5.00 ~ f -- 8 A -
0.00 ~- ag$+wfig__fi
1999 2000 2001
Year

Figure 18. Progesterone concentration in fecal material collected in winter from
radio-tagged cow moose in the western Upper Peninsula of
Michigan, 1999-2001. Each symbol represents the mean fecal
progesterone concentration of a single moose. The 95% Upper
tolerance limit (95% UTL) or dividing fecal progesterone level
between non-pregnant (<>, 0, A) and pregnant (0, O, A) cows was
5.17 119/9.

75

pregnant cows were 16.29 :l:0.98 and 14.46 pg/g, whereas for non-pregnant
cows the values were 2.30:l:0.32 and 2.66 ug/g (Table 13). As expected, fecal
progesterone concentrations of pregnant cows were significantly different from
those of non-pregnant cows (Z=-7.1730, P<0.0001). The 95% UTL of fecal
progesterone concentration for non-pregnant cows was 5.17 ug/g. Ninety-two
percent (44 of 48) of pregnant cows and 94% (17 of 18) of non-pregnant cows
would have been correctly identified using the fecal 95% UTL. The 4 false-
positive fecal results were from single samples collected from cows during
capture in 1999, however PSPB and serum progesterone levels correctly
identified them as being pregnant.

Ten radio-tagged yearling females survived long enough to possibly
reproduce. Only one (No. 109) gave birth to a single calf in 1999. Among the
remaining 9, 3 had negative pregnancy test results, 3 were seen alone during
spring natality checks, 2 were unknown as regards to calving, and 1 dispersed
the summer of her second year of life. The maximum yearling pregnancy rate, if
we include the 2 moose with unknown calving status, then would be only 30%.
Productivity
Natalitv and Recruitment

The method in which parturition was determined did not lend itself well to
estimating the birth dates of calves. As expected, 90% of calves were seen
between 15 May and 30 June each spring. The eafliest observation of a radio-
tagged cow with calves was 21 May in 1999, 24 May in 2000, and 15 May in

2001. In spring 1999, pregnant cows (n=14, 78%) were approached on the

76

ground an average of 1.86 times before their calving status was determined.
Two pregnant cows (1 per flight) were observed with offspring during aerial
relocation flights on 10 June and 26 June 1999. In spring 2000, pregnant cows
(n=20, 71 %) were approached on the ground an average of 1.50 times before
their calving status was determined. Three pregnant cows (1 per flight) were
seen with calves during aerial relocation flights on 25 May, 27 May, and 29 June
2000. In addition, 4 radio-tagged cows classified as non-pregnant were each
seen alone in the spring of 2000 during separate aerial relocation flights. In
spring 2001, pregnant cows (n=31, 74%) were approached an average of 1.71
times before their calving status was determined. No pregnant cows attended by
calves were seen from the air in 2001.

Among cows that tested positive for pregnancy, 100% in 1999 (n=8), 94%
in 2000 (n=17), and 76% in 2001 (n=29) were observed with at least 1 calf in the
spring. Radio-tagged cows produced 19 calves in 1999, 20 calves in 2000, and
29 calves in 2001. Post-calving calf:cow ratios decreased each year: 1.06 in
1999, 0.71 in 2000, and 0.69 in 2001. Frequency of twinning differed among
years (Table 14, Figure 19). The highest rate, 36% occurring in 1999 when the
fewest number of cows (n=18) were radio-tagged. The lowest rate was observed
in 2000 when only 6% of cows that reproduced gave birth to twins. The rate
increased 71% the next year when 21% of cows that gave birth had twins.
Annually, twinning rates averaged approximately 19%. Calfzcow ratios
decreased throughout the year due to the loss of a greater number of calves than

cows. Between post-calving and pre-calving the following year, calf:cow ratios

77

Table 14. Annual reproductive parameters of radio-tagged cow moose studied
during 1999-2001 in the western Upper Peninsula of Michigan.

 

Year Total no. % cows No. calves %

Spring Fall

cows pregnant produced twins calf: cow calf: cow

 

 

1999 18 78 19 36 1.06 : 1 0.76 : 1

2000 28 71 20 6 0.71 : 1 0.62 : 1
2001 42 74 29 21 0.69 : 1

7 or total 88 74 68 19 0.77 : 1 0.72 : 1
80 r 40%

N
C)

-. ....LLE -

 
  

////

Number of cow moose
8 8 8 8

N
O

 

       

 

 

29

 

 

2000 2001
Year

«30%

j E
j 25% t p
E c [‘Cows l
i 20% E [:0alves ’
' fllwifls.’

l 15% "

 

Percent

10%
.i. 5%

l 0%

Figure 19. Numbers of radio-tagged cow moose, their spring calf production, and
observed spring twinning rates in the western Upper Peninsula of

Michigan, 1999-2001.

decreased from 1.06 to 0.76 in 1999-2000 (-28%) and from 0.71 to 0.62 (-13%)

in 2000-2001. Post-calving calf:cow ratios for 2001-2002 were unavailable

because the study ended on 30 June 2001.

Population size estimation and projection

It needs to be emphasized that, although POP-ll is generic (i.e., it can be

used to estimate the population size of any species in any geographic location),

78

the survival and reproductive values used in the model are unique to the
population of interest. Therefore, the moose population size estimates produced
by POP-ll are unique to the western Upper Peninsula and immediate region only.
Population size estimates (or simulations) were made for 15 years (1987-2001)
starting with a population of 103 moose (the estimated 1987 post-partum
population), consisting of 65 adults, 9 yearlings, and 29 calves.

Using the MDNR (1985-1996) values of adult S=0.87, calf S=0.83 and
calf:cow=0.84 (Table 15) resulted in an estimated population of 812 moose by

2001 with a finite rate of increase (1 ) of 1.158. In comparison, the population

size estimate for 2001 using the MSU (1999-2001) values of adult S=0.88, calf

A

S=0.71 and calf:cow=0.81 was 44% smaller (455 moose, ,1 =1.112). When 6%
annual yearling dispersal was added, the result was a 13% smaller population
size (395 moose, ,1 =1.101) than without dispersal (Figure 20).

l examined the influence of calf survival by comparing the population size
estimates obtained using the MSU (1999-2001) calf S=0.71 to that using the
MDNR (1985-1996) calf S=0.83. As a control, adult S and the calf:cow ratio
were averaged over both studies (adult S=0.88, calf:cow=0.84). The higher calf
8 resulted in a 49% larger population size by 2001 (739 moose, A =1.151) than
the result using the lower calf S(496 moose, ,1 =1.090) (Figure 21).

The influence of yearling dispersal (l5 ) was more closely examined by

comparing population size estimates obtained if no annual yearling dispersal

occurred and if annual yearling f) were 0.06 or 0.12. The above control

79

parameters (adult S=0.88, calf:cow=0.84) and a mean calf S= 0.77 were used.
Without dispersal the population size estimate was 608 moose (,1 =1.135) by
2001. In comparison, the estimate of the 2001 population size was 14% smaller

(523 moose, 1 =1.123) when annual yeanlng 6 =0.06 and 26% smaller (446

moose, ,1 =1.110) when annual yearling 0 =0.12 (Figure 22).

Table 15. Reproductive parameters of radio-tagged cow moose and survival of
their offspring during 1985-1990, 1995-1996 in the western Upper
Peninsula of Michigan (Source: MDNR files, Marquette, Michigan).

 

 

Year"1b No. No. Calves per (33” $6
cows calves 100 cows

1985 17 21 123 0.76
1986 14 10 71 0.90
1987 27 30 111 0.80
1988 23 27 117 0.81
1989 16 17 106 0.71
1990 16 12 75 “ND
1995 10 7 70 0.86
1996 10 5 50 1.00
All years

combined 133 129 91 0.83

 

a Spring calving checks not conducted in 1991-93.

” Small sample of cows precluded use of data in 1994, 1997-1998
c Simplistic estimator.

d ND - No data available.

80

 

900 .- , ,7, 7 :11
"-0- -- (A) NDNR [1985-1996]; 00
800 . 8 p
+(B)MSU[1999-2001] m N...
a: 700 . 8 ..
§ 600 1 --*--(C)MSU[1999-2001+ it: ...
E 6% dispersal] 8 L0 . . 8
500— , v v ,. 2 ‘r
‘5 o: ... co V
L (f) m
a, 400 ' m ‘
.n _ ‘ m
E 300 8 g
3 00
z 200 -
100
0 z 8 8
N O) ‘- (‘0 ln‘ N O) ‘-
co <13 0') o) O) O) O) o
‘02 93 93 9 93 93 93 a
Year

Figure 20. POP-ll moose population size estimates in the western Upper
Peninsula of Michigan, 1987-2001. Results reflect the use of: (A)

adult S=0.87, calf S=0.83, calf:cow=0.84, (B) adult S=0.88, calf
S=0.71, calf:cow=0.81 and (C) adult S=0.88, calf S=0.71,
calf:cow=0.81 and yearling D =0.06.

81

800

700 -

600

400
300
200
100

Number of moose

500 .

1987 '

103

    

+7 (sateen?

----... (B) Calf S=0.83 5
7’ ”7” ID .
.O'

' to
co ‘2 00
O (D ‘—
ln ('0 V ‘—
,.‘: F, .‘_ z ,3 , _ . _ , 2,—7.72,
O) ‘- 0') L0 N O) ‘—
CD 0) O) O) O) O) O
O) O) O) O) O) O) O
1— ‘— ‘- ‘— ‘— ‘— N

Year

Figure 21. POP-ll moose population size estimates in the western Upper

Peninsula of Michigan, 1987-2001. Results reflect the use of adult
S=0.88, calf:cow=0.84 and (A) calf S=0.71 and (B) calf S=0.83.

82

700

 

co
---e---(A) 0% dispersal N 8
a, 600‘ —l——(B) 6% dispersal <r 8 ...
‘8 500 I —-A—- (C)12% dispersal '33. 523
o .O’
E 400 .
I.—
O
3 300~
.o
E 200 .
g -
‘ O
1001,, a 9 :t :2
o S ‘- ‘—
o.‘-. , .2 ,a . . 72V,,". .
N CD ‘- 0') IO N O) \—
co 1!) O) O) O) O) O) O
9 e 92 92 e e 9 8
Year

Figure 22. POP- ll moose population size estimates in the western Upper
Peninsula of Michigan, 1987- 2001. Results reflect the use of adult

S =0 88, calf S=0. 77, calf: cow=0. 84 and an yearling D of (A) 0. 00, (B)
0. 06 and (C) 0.12.

83

DISCUSSION

Survival and mortality

Survival rates of moose in the western Upper Peninsula (Table 2, Table 3)
were comparable to other non-hunted moose populations where predation by
wolves and bears is minimal. ln north-central Alberta Canada, Mytton and Keith
(1981) reported mean annual survival rates of 0.86 for adults and 0.83 for
yearlings. In addition, the mean survival rate of calves s 7-months of age was
0.67, which they believed to be a good estimate of first-year survival. In a newly
established moose population in southwest Colorado, Olterman and Kenvin
(1998) reported only slightly higher mean annual bull survival (0.94, 1%>) and a
similarly lower mean annual cow survival (0.83, 2%<) than were found in this
study (bull=0.93, cow=0.85). Also, mean annual adult survival in this study
(Y=0.877, range=0.873-0.880) was nearly identical to that in Michigan from
1985-1989 (x =0.871, range=0.818-0.956).

In two separate moose populations (Agassiz National Wildlife Refuge
[ANWR] and Beltrami Island State Forest [BISF]) Cox et al. (1997) estimated
annual cow survival rates of 0.67 and 0.72, rates that were 22% and 16% lower,
respectively, than annual cow survival (0.86) in the western Upper Peninsula of
Michigan for 1999-2001. Also, average 1999-2001 winter and summer survival
rates of adult moose in the western Upper Peninsula of Michigan (summer=0.95,
winter=0.92) were similar to those reported by Cox et al. (1997) at ANWR

(summer=0.91, winter=0.89).

Cox et al. (1997) reported 0-6 month calf survival at ANWR of 0.87, which
was 3% higher than in the western Upper Peninsula (0.84). However, first-year
survival of calves (0.56) and annual survival of yearlings (0.41) at ANWR were
21 % and 50% lower than first-year calf survival (0.71) and annual yearling
survival (0.82) in this study. In every year except 1985 and 1989, estimates of
first-year calf survival in the western Upper Peninsula of Michigan for 1985-1996
(T able 15) were greater than first-year calf survival rates found in the current
study (Table 3). This is likely the result of the small sample of calves that were
monitored in 1986 (n=10), 1995 (n=7), and 1996 (n=5) and the use of the
simplistic survival estimator.

Where predation is a major source of early calf mortality, survival rates of
calves are much lower than reported here. For example, in interior south-central
Alaska during 1976-1986, Ballard et al. (1991) found that 61% of calves died
within the first 5-months of life and that brown bear (Ursus arctos) predation
accounted for 73% of all calf deaths. The survival of calves s 5-months of age
was estimated at 0.39. Also, on the Copper River delta in coastal south-central
Alaska, MacCracken et al. (1997) reported calf survival rates of only 0.03 in
1987, 0.24 in 1988, and 0.05 in 1989 and that 94% of calves had died by 30 June
each year. This low calf survival was associated with inclement spring weather
and predation by brown bears. Predation on neonates may have been higher in
this study than was detected, primarily because only un-tagged calves were
determined to have died in their first year of life, and their bodies could not be

recovered. Nevertheless, compared to the above studies and others in Canada

85

and Alaska (Franzmann, et al. 1980, Ballard et al. 1981, Larsen et al. 1989,
Gasaway et al. 1992) predation of neonates in the Upper Peninsula of Michigan
by wolves and bears is minimal.

Previous to the current study, 23% (13 of 57) of radio-tagged moose that
died were diagnosed with cerebrospinal nematodiasis (i.e., adult P. tenuis
nematodes were found on the meninges or brain); of these 13 moose, 54% (7)
also suffered from malnutrition. During 1999-2001 only 17% (3 of 17) of dead
moose were diagnosed with cerebrospinal nematodiasis and of these 3, only 1
was malnourished. Whether malnourished moose are more susceptible to
infection by meningeal worms or infection prevents moose from getting adequate
nutrition is not known. Currently it does not appear that brain worm is a major
contributing cause of moose deaths in the western Upper Peninsula of Michigan.
However, if the region experiences a period of mild winter weather and the white-
tailed deer density increases, we may see more moose dying from brain worm
related problems. In areas managed primarily for moose, Kams (1967) has
suggested that white-tailed deer densities be maintained at $4.6 deer/kmz. The
most recent estimate of white-tailed deer density in the western Upper Peninsula
of Michigan was 4 to 11 deer/km2 (Hill 2001). However, the transmission of P.
tenuis may be more complex, and its long-term effects on moose health less
severe than previously suspected (see Lankester and Samuel 1998).

Additionally, prior to the current study 12% (n=7) of radio-tagged moose
that died were found to have liver flukes or liver damage evident of them (fibrous

scar tissue and/or thick-walled capsules). In this study 44% (n=8) of radio-

86

tagged moose that died were infested with liver flukes. However, alone, the
parasite has not been clearly shown to be harmful to moose (Lankester and
Samuel 1998). A heavy infestation of liver flukes accompanied by some other
ailment, such as malnutrition may contribute to death. All dead radio-tagged
moose that had liver flukes (Table 4) were also suffering from other ailments.
Movements

Twenty to 40% of moose in the western Upper Peninsula of Michigan
annually migrated between spatially separate winter and summer home ranges
(Table 6). Addison et al. (1980), and Phillips et al. (1973) reported similar
proportions of migratory moose in northwest Ontario (37%) and northwest
Minnesota (20%), respectively.

Migration distances are mainly a function of habitat dispersion or terrain,
and therefore, vary both within and among populations (Addison et al. 1980).
Generally, moose that inhabit more level terrain migrate shorter distances
between seasonal ranges than do those in mountainous regions. Distances
between summer and winter ranges in this study (2-26 km) were similar to those
reported by Mytton and Keith (1981) in Alberta (bulls, 13 km; cows, 7 km),
MacCracken et al. (1997) in coastal south-central Alaska (maximum distances:
bulls, 15 km; cows, 25 km), Addison et al. (1980) in northwest Ontario (2-13 km),
and Phillips et al. (1973) in northwestern Minnesota (14-34 km). In contrast, the
average migration distance between seasonal ranges in interior south-central
Alaska was 48 km (range=10-68 km) (Ballard et al. 1991) and southeast of the

Brooks Range in Alaska and Canada, 123 km (range=18-196 km) (Mauer 1998).

87

Moose migrate to summer ranges to optimize nutrient intake, however in
the winter these areas may not provide adequate forage-or cover (Hundertmark
1998). In mountainous terrain, moose generally migrate from lower elevation
winter ranges to higher elevation summer ranges (Ballard et al. 1991). Although
the opposite elevation movement pattern has been observed (Gasaway et al.
1983, Mauer 1998). In the relatively non-mountainous Great Lakes region of the
coniferous-deciduous ecotone, moose often migrate in the spring to poorly
drained areas dominated by black spruce (Picea man'ana) and interspersed with
bogs and lakes (Van Ballenberghe and Peek 1971, Phillips et al. 1973, Peek et
al. 1976). In addition to the palatable, high quality aquatic vegetation (e.g.,
Equisetum spp, Nuphur spp.) that is available and consumed by moose in these
lowland areas, they may also provide escape from insects and safe calving sites
for cows.

Migration to winter ranges occurs in response to seasonal changes in
forage plant species (i.e., leaf abscission and plant dormancy) and in the
availability of these plants because of snow accumulation. Many authors, for
example Addison et al. (1980), MacCraken et al. (1997), and Sweanor et al.
(1992) found that late-autumn and winter migrations began when snow depths
approached 50-60 cm. In the early part of winter in eastern North America,
moose are often found in areas with a comparatively open canopy and well-
developed undergrowth of coniferous and deciduous regeneration. Later, when
snow depths begin to cover available forage, moose move to dense spruce-fir

forests where snow depths are less. Thereafter, if snow conditions improve,

88

moose may move into open woodlands to feed and return to dense spruce-fir
forests to bed-down (Telfer1970, Peek et al. 1976, McNicol et al. 1980). In
Michigan in early winter, Minzey and Robinson (1991) found moose in relatively
open (<50% canopy closure) hardwood stands dominated by sugar maple.
However, later in winter, moose sought out areas with a dense overstory (50-
70% canopy closure) of balsam fir and eastern hemlock. In the present study,
the proportion of seasonally migratory moose and their movement patterns were
generally typical of those reported for moose in the Great Lakes region.
Because of the numerous methods used to estimate home range size,
and the different algorithms and parameter values that software programs
incorporate to apply these methods, it is often difficult to make comparisons to
other studies (Lawson and Rogers 1997). The MCP method is the oldest (Mohr
1947) and most frequently used estimator of home range size (Harris 1990).
However, the technique is severely affected by sample size (White and Garrot
1990, Boulanger and White 1990) and the distribution of outlying points
(Kenward 2001). Ballard et al. (1991) found that 240 relocations were needed to
adequately define 75% of total home range sizes of moose using the MCP
method. Probabilistic methods, such as the FK are better estimators than the
MCP approach because (1) they are not as severely affected by sample size, (2)
home range boundaries are based on the complete distribution of relocation
points, and (3) they can better handle outliers (Kemohan et al. 2001). Kernel
methods were introduced as home range estimators by Worton (1989), but have

only recently become popular as more software has become available (Seaman

89

et al. 1998). Only 1 study in the literature was found that used the fixed kernel
method to estimate moose home ranges and only winter ranges were reported.
Therefore, unless othenlvise specifically stated, all of the studies referred to
below used the MCP home range estimator. How authors define their seasonal
time periods further complicates comparisons of seasonal ranges between
studies.

Estimates of mean 95%-PK summer home ranges in this study (4-115
kmz, Table 6) were smaller than those found by Ballard et al. (1991) in interior
south-central Alaska (23456 kmz), by MacCracken et al. (1997) in coastal south-
central Alaska (21-162 kmz), and by Stenhouse et al. (1995) in the Northwest
Territories (681 35 kmz). Smaller summer home range estimates were found in
northwest Minnesota (1 .5-24 km, Phillips 1973, home range fill method), Central
Alberta (1-34 kmz, Mytton and Keith 1981 ), and northern New York (36.3 km2,
bulls only, Garner and Porter 1990). Similarly sized estimates of summer home
ranges were noted by Hauge and Keith (1981) in northeast Alberta (18-97 kmz)
and Addison et al. (1980) in northwestern Ontario (6-90 kmz).

Average 95%-PK estimates of the size of winter home ranges in the
western Upper Peninsula of Michigan (1-64 kmz, Table 6) were smaller than
those found in interior south-central Alaska (15-375 kmz, Ballard et al. 1991),
coastal south-central Alaska (19-146 kmz, MacCracken et al. 1997), and the
Northwest Territories (57d: 58 kmz, Stenhouse et al.1995). Winter home range
estimates were larger than those found by Phillips (1973) in northwest Minnesota

(0.5-5.0 kmz, home range fill method), Addison et al. (1980) in northwest Ontario

90

(2-12 kmz), and Garner and Porter (1990) in northern New York (7.5 kmz, bulls
only). Estimates of winter home range sizes similar to the findings of this study
were noted by Lawson and Rogers (1997) in northwest Ontarion (33 kmz, Fixed
Kernel estimator), Mytton and Keith (1981) in central Alberta (2-54 km?) and
Hauge and Keith (1981) in northeast Alberta (3-111 kmz).

The winter movements and size of winter home ranges of moose are likely
influenced by snow depth. During extended periods of deep snow (>70 cm),
moose often restrict their movements to relatively small areas where snow is not
as deep, thus influencing the overall size of their winter home range. For
example, in northern Maine, winter home ranges (median=7.1 kmz) were nearIy
4.5 times larger during a winter when snow depths were low compared to when
they were >70 cm (median=1.6 kmz) (Thompson 1987). In this study no
difference (P=0.5372) was detected between the movements of cows in winter
compared to other times of the year (Table 5). Because the influence of snow
depth on winter home range size was not closely examined during this study no
strong conclusions can be made. However, 2 adult cows, for which consecutive
winter home range estimates were obtained, occupied different sized areas from
one winter to the next. But the trend was not the same for each animal. Moose,
No. 76, had a larger winter home range in 1999-00 (16 km2) than in 2000-01 (5
kmz), whereas moose No. 121, had a smaller winter home range in 1999-00 (25
kmz) than in 2000-01 (42 kmz). Although we often found snow depths >700m
during the present study, mean daily snow depth never exceeded 60 cm and no

rapid movements of moose were noticed following early winter snowstorms.

91

There is no consistent trend that moose have larger (or smaller) home
ranges in the winter than in the summer (Hundertmark 1998). In this study,
median home ranges of resident and migratory adult moose combined were
greater in the summer than in the winter. Addison et al. (1980) and Garner and
Porter (1990) also found that moose had larger summer than winter home
ranges. In contrast, Doerr (1983) found that moose had larger winter home
ranges, whereas Ballard et al. (1991), Hauge and Keith (1981) and Stenhouse et
al. (1995) found no difference between the size of seasonal home ranges. Males
typically have larger home ranges than females (Hundertmark 1998, Ballard et al.
1991) due to their greater movements in the fall during the rut (Hauge and Keith
1981). Although no studies reported the opposite, Phillips et al. (1973), Hauge
and Keith (1981), MacCracken et al. (1997), and this study found no difference
between the size of male and female home ranges. Ultimately, season, habitat
quality, interspersion of habitat components, terrain, population density, age, sex
and weather influence the size of areas occupied by moose.

All offspring eventually separate from their dams, either voluntarily or by
being driven off. Separation generally occurs as parturition approaches, but a full
and permanent parting may not occur until sometime during the offspring’s
second year of life (Hundertmark 1998). In this study all radio-tagged offspring
(n=26) had broken completely from their cows at around 1-year of age. Fourteen
radio-tagged offspring were closely followed for up to 1-year following separation
to analyze their movements. Six offspring occupied areas that spatially

overlapped their cow’s home range and were often found within close proximity

92

to her for several months after separation. Three offspring completely
abandoned the areas occupied by their cow shortly following separation, and 5
(19% of all offspring) dispersed out of the core study area. Ballard et al. (1991)
reported that 9 of 15 offspring partially (n=4) or fully dispersed (n=5) from their
parental home range, and that more male than female offspring dispersed. In
most mammals, juvenile males are more likely to disperse than other individuals,
although other age and sex classes are represented (Greenwood 1980). No sex-
biased dispersal was noticed in this study, but all moose that dispersed except 1
were yearlings. There are a limited number of reports in the literature regarding
moose dispersal and nowhere, were actual dispersal rates stated. Lynch (1976)
noted that 50% of subadults (<2-years of age) and 17% of adult moose dispersed
off his study area in central Alberta. However, because he considered moose for
which radio contact had been lost to have dispersed, these values may be high.
In north-central Alberta, Mytton and Keith (1981) documented long distance
dispersal movements of at least 50 km for 3 young moose, and 1 yearling bull
moved 250 km off the study area over a 2-year period. In the western Upper
Peninsula of Michigan, Aho et al. (1995) reported that a yearling female, over a
9-month (March 1989-December 1990) and a yearling male, over a 7-month
period (March 1989-December 1990) emigrated at least 160 km to Wisconsin
(Figure 23). These moose and their movements are notable because (1) they
were born 5-years apart to the same cow, and (2) their dispersal routes, and final
destination in Wisconsin were very similar to each other, and to 2 moose in the

current study (Figure 12, Figure 16).

93

denies. so m_:mc_con_ Loan: E26?

65 5 33 59900.5sz 9.56 A: .ozv omooE 06E mc__._mm> new .83 538830-32 scum—2
acts—o Go .ozv omooE ofiEfi 05.59» 888.059 ho macoEo>oE new $58 cosmoofie umumEzmm .mm 65?.

 

 

I.

 

 

 

 

H
22080.6. 2. c or

 

 

 

 

 

s.
w

 

Z_w200w_>> a a

 

 

 

 

 

 

 

 

 

 

B l

I, sot-undo €30

 

5558 82 .68

 

n..—082.1 z. 0.
2.03.8.0”. no 0.

   

99222

Lotcaaw 0x3

 

 

 

 

 

 

94

Although the exact mechanisms prompting dispersal are multifaceted and
complex, population density in relation to habitat productivity and social stress
appear to be two important factors (Ballard et al. 1991).

Reproduction and productivity

Moose pregnancy rates in the western Upper Peninsula of Michigan
()7 =74%) were lower than those in the Northwest Territories (96%, Stenhouse et
al. 1995), but higher than those in 2 moose populations in northwest Minnesota
(ANWR=37.5%, BISF=51%, Cox et al. 1997). In his review of 12 studies that
reported information on moose reproduction, Boer (1992) found relatively
constant adult pregnancy rates across North America (i=84%, range=77-97%).
Although pregnancy rates in this study were 7 to 15% lower than the North
American average, they were fairly consistent from year to year (Table 14). The
rates reported here were most similar to the 79% found by Boer (1987) and the
76% of Edwards and Ritcey (1958).

Yearling pregnancy rates, unlike adult pregnancy rates are less stable.
Among different moose populations, Boer (1992) found that pregnancy rates of
yearlings averaged 42% and ranged from O to 93% (CV=69%). Furthermore,
within the same moose population, Blood (1974) reported that 15 to 93%

()7 =29%, CV=89%) of yearlings were pregnant. Although the number of
potentially reproducing yearling females (n=10) in this study was small, it appears
that yearling pregnancy rates in the western Upper Peninsula (10-30%) are very

low compared to other moose populations in North America.

95

Twinning rates also vary considerably among moose populations in North
America (Boer 1992, Schwartz 1997). Values from as low as 1% in southern
Newfoundland (Albright and Keith 1987) to as high as 88% in north-central
Alberta (Mytton and Keith 1981) have been recorded. Lower rates (<5%) are
indicative of populations above carrying capacity, whereas moderate (5-25%)
and higher (25-90%) rates are associated with populations near and below
carrying capacity, respectively (Gasaway et al. 1992, Schwartz 1997). Although
the average observed twinning rate reported in this study (19%) fell in the range
of rates for populations near carrying capacity, among years, each rate fell into a
different carrying capacity category (Table 14, Figure 19). An evaluation of
available habitat in the western Upper Peninsula indicated that the moose
population is apparently below carrying capacity (Patterson et al. 1995). in
south-central Alaska, Ballard et al. (1991) reported twinning rates which ranged
from 17 to 63% that varied by year, area, and possibly collection method, and
Blood (1974) reported that they ranged from 4 to 48% at Elk Island National
Park, Edmonton, Alberta. For the period 1985-1995, MDNR wildlife biologists
reported an average twinning rate of 36% that was also highly variable
(range=24-69%, Aho et al. 1995).

Population size estimation and projection

Population models are useful tools that are often used to assist in
formulating management strategies and making decisions. The type and
complexity of the model needed is driven by the objectives of the study. For the

purposes of this study (i.e., comparisons between estimates using different input

96

variables), a deterministic model, such as POP-ll is adequate. Because the
results of deterministic models are based on fixed input variables, it is critical that
the values of these variables are accurately estimated. This requires that an
adequate and representative sample of animals is radio-tagged. For survival
analysis, Winterstein et al. (2001) suggests that >25 animals, per treatment (e.g.,
sex, age class) are radio-tagged. From 1985 to 1990 the number of radio-tagged
adult moose varied from 22 to 50, which permitted fairly accurate estimates of
adult moose survival. However, after 1990, despite the MDNRs attempts to
maintain an adequate sample size - through periodic darting and radio-collaring
of moose, and the capture and radio-collaring of 26 moose from a helicopter in
1995 - the number of radio-tagged moose at any one point in time was probably
insufficient to provide accurate results. Additionally, the Simplistic estimator used
by the MDNR to estimate survival, does not permit staggered entry of newly
radio-tagged individuals and is biased if animals are censored (see Winterstein et
al. 2001). Estimates of natality (calves per cow), prior to 1991, were also fairly
accurate, but became less accurate thereafter because (1) spring calf checks
were suspended during 1991-1993, and (2) the number of radio-tagged cows in
any one year during 1994-1998 was <10.

Inconsistencies in data collection, periodically inadequate sample sizes
and the use of a biased survival estimator call into question the accuracy of the
1985-1996 values of moose survival and reproduction and the population size
estimates derived from these values. Because a more robust survival estimator

was used and the sample size per treatment was larger, more accurate moose

97

survival rates were obtained during 1999-2001. Additionally, more intensive
monitoring of radio-tagged cows and their offspring resulted in better estimates of
reproduction and calf survival.

Using the survival and reproductive rates obtained during 1999-2001
POP-II produced 15-year moose population estimates of 395 moose (1 =1 .101)
with dispersal and 455 moose (,2. =1.112) without dispersal (Figure 20). In
addition, preliminary Mark-Resight population estimates from the 2002 winter
aerial survey extrapolated to include all high-density stratum were 448 moose
(0.28 moose/kmz) from a fixed-wing aircraft and 467 moose (0.29 moose/kmz)
from a helicopter. Based on the above estimates of moose population size,
information found in MDNR documents and published reports, conversations with
MDNR wildlife biologists and my involvement in the current study, an estimated
population of 400-500 moose in the western Upper Peninsula of Michigan is

reasonable.

98

LIMITING FACTORS

One of the objectives of this study was to identify the factors that could be
limiting the growth of the moose population in the western Upper Peninsula of
Michigan. As defined by Gasaway et al. (1992), “limiting factors retard the rate of
increase in population size or density through density-dependent and/or density-
independent processes”. Evidence indicates that the moose population in the
western Upper Peninsula is likely below carrying capacity and density-dependent
processes, such as competition for food, are not operating. Predation, disease,
snow accumulation >90 cm. and nutrition have been identified as the primary
factors that have the potential to limit moose population growth. In the following
paragraphs the impact of these factors on moose population growth in the study
area will be assessed.

Predation by wolves and bears (black and/or brown) combined, has been
shown to keep moose at low densities for extended periods of time (Gassaway et
al. 1992). Although the Upper Peninsula of Michigan supports a healthy black
bear population (n>12,000) and an increasing wolf population (n=200-250), there
seems to be little predation on moose. Only 1 confirmed case of a wolf-killed
moose was made during the current study, and of the 43 moose deaths reported
from 1985-1994, none was attributed to wolf or black bear predation (Aho et al.
1995).

No evidence was found to indicate that disease or parasites is limiting the

growth of the moose population in the western Upper Peninsula. Moose are

99

susceptible to many diseases, viruses, and parasites, but the population
regulating effects of these are largely unknown or of little consequence.
Necropsies determined that only 3 moose likely died of brain worm and that 8
were parasitized with liver flukes. However, in almost every case these animals
had additional maladies that likely contributed to their deaths (Table 4).

There was no evidence that severe winter weather or deep snow had a
large impact on moose survival and reproduction. Average snow depths were
<70 cm, the level above which adult moose may have difficulty travelling and
reaching browse, and large losses of calves has been recorded in interior Alaska
(Gasaway et al. 1983). Although snow depths >70 cm were often found where
fecal samples were collected, movements of moose did not appear to be
restricted. Furthermore, first-year calf survival was lower in 1999-00 (0.63) than
in 2000-01 (0.78) when snow depths were greater.

Nutritional stress has the potential to limit population growth irrespective of
the density of the population. Although, at high density, increased competition
for food resources make certain classes of the population (e.g., calves, pregnant
cows, rutting bulls) more vulnerable to malnutrition. Possible signs of acute
nutritional stress include, low adult survival and retarded body growth, reduced
recruitment and low calf survivorship, poor physical condition and starvation prior
to old age, and poor reproductive output (Messier and Créte 1984, Gasaway et
al. 1992). The quality, quantity, and availability of forage ultimately determine an
animal’s nutritional status. Therefore, the marginal habitats often found at the

periphery of an animal’s range likely effect its nutritional condition. No studies

100

have yet, however looked at the consequences of marginal habitat on moose
nutrition and its effect on survival and fecundity.

The manifestations of nutritional stress were not reflected in moose
survival, recruitment, physical condition, body growth or longevity. Annual adult
survival was relatively high (>85%) during both years. Calf survival and
recruitment were comparable to other moose populations where predation is low.
In addition, analysis of blood samples collected during capture, and physical
examination score (3? =8.5:l: 0.15, range=2-10) indicate that moose are in
adequate physical condition for early winter. Also, only 3 of 12 adult moose that
died, for which marrow fat analysis was done, had femur marrow fat levels <20%
indicating severe malnutrition (Peterson et al. 1984).

On the other hand, marginal nutrition, the result of less than optimal
forage, could be the reason for the below average pregnancy rates and low
twinning observed in this study. Crete and Courtois (1997) claimed that the low
fecundity of moose in east central Québec at the northern extent of moose range,
was the result of suboptimal production of summer and winter forage. They
attributed this low productivity to the mature age of the forest stands and
absence of recent disturbances, and the unsuitability of a large portion of the
plant communities on the study area. Franzmann and Schwartz (1995)
demonstrated that twinning rates might also be a good indicator of nutritional
status and indirectly of habitat quality. On the Kenai Peninsula of Alaska, they
found that 70% of cow moose living on high quality habitat, gave birth to twins,

whereas on poor quality habitat only 20% of cows produced twins. Furthermore,

101

Boer (1992) found a direct relationship between adult twinning rates and yearling
pregnancy rates. Yeariings that are well nourished grow faster, mature earlier
and frequently breed and give birth to a single calf (Schwartz and Hundertmark
1993). The relationship between forage, nutrition, and reproduction is intuitive,
but no study has yet related forage quality, quantity, and availability to moose
fecundity (Crete and Courtois 1997).

A coarse evaluation of habitat suitability in Baraga County using historic
information and remotely sensed forest cover data indicated that the available
habitat in the county could support between 1094 and 1845 moose depending on
which data-set was used (Patterson et al. 1995). Also, the relatively young age
of many forest-stands and the intensity of logging and natural disturbance in the
western Upper Peninsula would suggest that forage productivity and quality is
good. However, habitat suitability has to be verified on the ground and a more
quantitative analysis of forage productivity should be conducted before
eliminating these factors as sources of nutritional deficiencies.

Moose inhabiting the Upper Peninsula of Michigan are at the southern
limit of moose range in the non-mountainous regions of North America. Given
this, the habitat and climatic conditions that have prevented further range
expansion of moose in the region may also be limiting their population growth in

the western Upper Peninsula of Michigan.

102

MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS

The 1985 and 1987 translocations of moose from Ontario Canada to the
western Upper Peninsula of Michigan generated great excitement and high
expectations. Although some of this enthusiasm has waned, there is still great
interest and some concern, both among the citizenry and wildlife biologists of
Michigan regarding the slower than expected growth of the moose population.
However, this concern may be unwarranted. The original population goal of
1000 moose in the western Upper Peninsula by the year 2000 was overly
optimistic. It is important to inform personnel within the MDNR and the general
public that from most indications the moose herd is currently faring well. It
should also be conveyed that the number of moose in the Upper Peninsula of
Michigan might never reach 1000.

Brain worm is another issue that needs to be clarified. Popular articles
and MDNR reports inadvertently over emphasized the negative effect of P. tenuis
on the moose population. Recent experiments have shown that moose infected
with P. tenuis do not always exhibit clinical signs of the disease, nor is infection
always fatal. It appears that the consequences of infection are both dose and
age dependent (Lankester and Samuel 1998). Also, because moose have
historically coexisted and continue to coexist with white-tailed deer infected with
P. tenuis, the extent to which the disease effects moose populations needs

further investigation.

103

The current study of the population dynamics of moose in the western
Upper Peninsula of Michigan is being conducted, in part, to determine why 2
independent methods, a deterministic population model and aerial surveys
produced drastically different population size estimates. The study is divided into
2, 3-year phases. The primary focus of Phase-l, my portion of the study, was to
determine the current rates of moose survival, reproduction, and dispersal.
Phase-II, which began in the winter of 2001, will concentrate on evaluating the
aerial survey methodology and sightability model as well as estimating the “true”
size of the moose population. Although the focus has changed, intensive
monitoring of moose survival, reproduction, and movements will continue for at
least the next 3 years. Weekly monitoring of radio-collared moose for timely
recovery of mortalities is taking place. Also, collection of fecal samples for
pregnancy determinations and evaluation of stress levels is ongoing and natality
checks will be done in the spring each year. Several more seasons of collecting
this data will hopefully provide a more complete understanding of moose
population dynamics.

As with most wildlife research, the findings from this study generated
further questions, pointing to the need for more research, much of which is
currently being conducted. The study did reveal the answer as to why the moose
population in the western Upper Peninsula has not increased as quickly as
predicted. Low reproduction, exhibited in (1) below average adult reproductive
rates, (2) low adult twinning rates, and (3) extremely low yearling reproductive

rates is the reason for the slow population growth. The more important question

104

and more difficult one to answer is, Why is moose reproduction low? And,
should we be concerned? Or, is low reproduction endemic to this moose
population at the southern edge of its range?

Because nutrition plays such an important role in reproduction, an analysis
of the nutritional quality and productivity of the available forage in the western
Upper Peninsula is warranted. This information would indicate how well (or
poorly) the habitat in the western Upper Peninsula is meeting the year round
nutritional and energetic requirements of moose. As a preliminary step, I began
collecting samples of terrestrial and aquatic plants both incidentally and those
browsed by moose. Plants were analyzed for selenium (Se), a trace element
that has been linked to white muscle disease, low reproduction and reduced
fertility primarily in livestock. We were specifically interested in Se because the
mean whole blood Se levels of moose radio-tagged in 1999 was only 0.036-ppm,
a level considered deficient by livestock standards (<0.04-ppm, Oliver 1990).
Preliminary analysis found no detectable levels of Se in the plants collected.
Because forage and grain plants grown in Michigan for livestock consumption are
deficient in Se (Kubota and Allaway 1972) these results were not unexpected.
However, the Se requirements of free-ranging moose are not known and the
relationship between moose health and Se Is not as yet completely understood.

Also, although moose appear to be in adequate health condition in the
winter, their summer health condition is unknown. Extended periods of high
summer temperatures may cause heat stress and interfere with feeding, which

could ultimately effect moose body growth, fat storage, and reproduction.

105

Assays of fecal glucocorticoid steroids have been used to study physiological
stress in elk (Cervus elephus) in North Dakota (Millspaugh et al. 2001). Moose
fecal samples, collected primarily in the winter, were analyzed by Dr.
Millspaugh’s lab at the University of Missouri this past year for glucocorticoid
levels. However, base-line glucocorticoid levels for moose need to be
established before comparisons among individuals can be made. Nevertheless,
the collection of fecal samples in the summer (in addition to those collected in the
winter) and analysis of these samples for glucocorticoid levels would permit
comparisons between summer and winter stress levels of moose.

Finally, a stochastic model that allows survival, reproduction, and
dispersal rates to vary randomly and uses the statistical distribution of these
values rather than their means would probably produce more accurate estimates
of the moose population in the western Upper Peninsula of Michigan. In addition,
climatic conditions such as snow depth and a summer severity index could be

incorporated into such a model.

106

APPENDIX

107

$82.3 9.8: ..adv 338 So 8:330 ocm dam new xom Esocxc: Co $02: $3.05 0
.22an05 bemmoa 2m 2023 .5952: .cBchEm £255 cocmomom 0:630 65 .m we no 8809 co oommm a
.53 25> 3 2mm .02 toamm c2920 3:33 ”.292 Eot 6232a

 

mm

N_.

m

m
0:26:ch

£330

or

N

_.

mom—359.

9500

m

N

F
m=3m

om

V's-‘-

N.

or
0:265.ch

o

wozmo

2

n..

m

com Tvmmw

«.260

rm

NP

m
2.3m

.oom Tvmmv

.92.

tesoozom
59302.5
23305.2
omsxoms.
mos;
Bocomzox
c9.
:oEmaoI
ofiomoo
com—.220
memo
$58qu
39mm
5m?

3:300

.335ch Econ: @5955 E 3:38 3 omooE Co 8525 .o 3:335 9,608 So 388m ....1 flow... x_ucooo<

108

A326: .969“ 5.3 9.39: So 8:830 ocm .omm ccm xom 565:: So omooE 3:205 0
.20.an05 >368: 9m 22:? .:mu_:o__2 .:2o_u:_:m 60.3% 55801 87.30 05 am 0.: :o 3:82 :o oommm a
.39 .oEo> E mnmw .oz toaom 5335 3:33 «20.2 So: 6232 a

 

Fm

mN
NF

Fm
F

F
0565::

m m F mm Nm

‘_
(009N010
N

‘_
N

m

m F

m

m

32.3 9500 m__:m o:Bo:x:3
mmm Tmnm F

N NF
F N

NFNC‘OFF

N VN

v mF
v

mozmo
Nnm Thom F

F

Fm

COP-FCC

.mwm 750m F

_So._.

cm:o_oo:ow
:ommcoEO
23:ng
8:20.22
mos;
2,2339.
:o:
5396:
ofiomoo
:om:_xo_o
$.00
msoaazo
www.mm
8m?

3:300

.m_:m:_:cn_ :cao: m.:mmEo_5_ :. 3:38 3 omooE Co 8:225 :o 3:336 809: ..o mnBoom .a.mN 2an x_n:oaa<

109

Appendix Table 3. Physical examination scores and description for assessing

moose condition during capture and when conducting field
necropsiesa.

 

Score
10

O—‘Nw-fifl‘l

Description of condition

A prime fat moose with thick firm rump by sight. Well
fleshed over back and loin. Shoulders round and full.

A choice fat moose with evidence by feel of rump fat.
Fleshed over back and loin. Shoulders round and full.

A good fat moose with slight evidence by feel of rump fat.
Bony structures of back and loin not prominent. Shoulders
well fleshed.

An “average" moose with no evidence of rump fat but well
fleshed. Bony structures of back and loin evident by feel.
Shoulders with some angularity.

A moderately fleshed moose beginning to demonstrate one
of the following conditions: (a) definitions of neck and
shoulders, (b) upper foreleg musculature distinct from chest,
or (c) prominent rib cage.

When two of the characteristics in class 6 are evident.
When all three of the characteristics in class 6 are evident.
When the hide fits loosely about the neck and shoulders.
Malnutrition obvious. Outline of the scapula is evident.
Point of no return

Dead from malnutrition/starvation.

 

a Adapted from Franzmann 1998.

110

Appendix Table 4. Moose id, sex, age at capture, date captured, type of radio
collar, and fate as of 30 June, 2001 of radio-tagged moose
studied during 1999-2001 in the western Upper Peninsula of

 

 

Michigan.
Moose Sex Age at Date Last Radio collar Fate°
id capturea captured located typeb
28 F A 1 Feb 1995 7 Jun 2001 VHF Alive
70 M A 1 Feb 1995 22 Jun 2001 VHF Alive
73 F A 1 Feb 1995 29 Jun 2001 VHF Alive
74 M A 1 Feb 1995 18 Aug 2000 VHF Dead
75 F A 1 Feb 1995 7 Jun 2001 VHF Alive
76 F A 2 Feb 1995 12 Jun 2001 VHF Alive
77 M A 1 Feb 1995 27 Jun 2001 VHF Alive
78 M A 31 Jan 1995 11 Jun 2001 VHF Alive
79 M A 31 Jan 1995 22 Jun 2001 VHF Alive
80 F A 1 Jan 1995 30 Oct 1999 VHF Dead
86 F A 1 Jan 1995 26 Jun 2000 VHF Dead
92 M A 2 Feb 1999 22 Jun 2001 VHF Alive
101 M C 30 Jan 1999 2 Feb 1999 VHF Dropped
102 F Y 29 Jan 1999 9 Apr 2001 VHF Dead
103 M C 30 Jan 1999 29 Jun 2001 VHF Alive
104 M C 30 Jan 1999 4 Jun 2001 VHF Alive
105 F C 30 Jan 1999 3 Jul 2000 VHF Dropped
106 M C 29 Jan 1999 19 Sep 1999 VHF Dead
107 M C 30 Jan 1999 16 Nov 2000 VHF Dead
108 F Y 31 Jan 1999 11 Jun 2001 VHF Alive
109 F C 28 Jan 1999 7 Jun 2001 VHF Alive
110 F C 28 Jan 1999 31 Oct 2000 VHF Dead
111 F C 29 Jan 1999 16 Jun 2001 VHF Alive
112 M C 31 Jan 1999 22 Jun 2001 VHF Alive
113 M C 28 Jan 1999 14 Apr 2001 VHF Dropped
114 M C 7 Feb 1999 27 Mar 2000 VHF Dead
115 F A 7 Feb 1999 17 Dec 2000 VHF Dead
116 F A 31 Jan 1999 7 Jun 2001 VHF Alive
117 F A 31 Jan 1999 27 Jun 2001 VHF Alive
118 F A 30 Jan 1999 11 Jun 2001 VHF Alive
119 F A 30 Jan 1999 11 Nov 1999 VHF Dead
120 F A 28 Jan 1999 29 Jun 2001 VHF Alive
121 F A 2 Feb 1999 29 Jun 2001 VHF Alive
122 F A 30 Jan 1999 7 Jan 2000 VHF Dead
123 F A 29 Jan 1999 22 Jun 2001 VHF Alive
124 F A 7 Feb 1999 22 Jun 2001 VHF Alive
125 F A 29 Jan 1999 29 Jun 2001 VHF Alive
127 M A 1 Oct 1998 27 Jun 2001 VHF Alive
130 F A 19 Jan 2000 29 Jun 2001 VHF Alive

111

 

Appendix Table 4. (Continued).

131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174

ngZ‘anZZmZZZ‘nm'fl‘nm‘n'fl'flZmmmmg'nm'nz'nZZ‘n‘nZngTI-nmz

O>>O>OOOO>O>-<><O>O>>>O>>>OO>>>O<NN>>N>>>N>NN

20 Jan 2000
20 Jan 2000
20 Jan 2000
20 Jan 2000
20 Jan 2000
21 Jan 2000
22 Jan 2000
19 Jan 2000
19 Jan 2000
19 Jan 2000
19 Jan 2000
19 Jan 2000
19 Jan 2000
19 Jan 2000
17 Jan 2000
17 Jan 2000
17 Jan 2000
17 Jan 2000
17 Jan 2000
19 Jan 2000
15 Jan 2000
15 Jan 2000
15 Jan 2000
15 Jan 2000
15 Jan 2000
15 Jan 2000
16 Jan 2000
16 Jan 2000

7 Feb 2000
10 Feb 2001
10 Feb 2001
10 Feb 2001
10 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
11 Feb 2001
12 Feb 2001
12 Feb 2001

5 Jun 2001
29 Jun 2001
25 Apr 2000
29 Jun 2001
22 Jun 2001
21 Apr 2000
16 May 2000
29 Jun 2001
17 Jun 2001

7 Jun 2001
11 Jun 2001
22 Jun 2001
30 Jun 2000

3 Apr 2000

31 May 2000

7 Jun 2001
22 Jun 2001
29 Jun 2001
22 Jun 2001
27 Jun 2001
16 Jun 2001
29 Jun 2001
28 Apr 2001
22 Jun 2001
21 Jun 2001
29 Jun 2001
11 Jun 2000
29 Jun 2001
13 Feb 2000

1 Mar 2001
17 Jun 2001
17 May 2001
29 Jun 2001
11 Jun 2001
29 Jun 2001
11 Jun 2001
29 Jun 2001
29 Jun 2001
11 Jun 2001
11 Jun 2001
27 Jun 2001
29 Jun 2001
22 Jun 2001
29 Jun 2001

112

VHF
VHF
GPS
VHF
VHF
GPS
GPS
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
GPS
GPS
GPS
VHF
VHF
VHF
VHF
VHF
VHF
VHF
VHF
GPS
VHF
VHF

Alive
Alive
Off line
Alive
Alive
Off line
Off line
Alive
Alive
Alive
Alive
Alive
Dead
Dead
Dead
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Dead
Alive
Alive
Alive
Dropped
Alive
Dead
Dead
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive

Appendix Table 4. (Continued).

175
176
177
178
179
180
181
182
183
184
185
186
187

a A - adult, C - calf, Y - yearling.

mmmmmmememm
>>>>OOO>O>>O>

12 Feb 2001
12 Feb 2001
12 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
13 Feb 2001
14 Feb 2001
14 Feb 2001

22 Jun 2001
29 Jun 2001
13 Apr 2001
29 Jun 2001
22 Jun 2001
22 Jun 2001
22 Jun 2001
11 Jun 2001
22 Jun 2001
27 Jun 2001

4 Jun 2001
25 Jun 2001
18 Jun 2001

b GPS - global positioning system, VHF - very high frequency.
b Dropped - radio-collar removed by moose.

1B

VHF
VHF
GPS
VHF
VHF
VHF
VHF
VHF
VHF
VHF
GPS
VHF
VHF

Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive
Alive

LITERATURE CITED

Addison, R. B., J. C. Williamson, B. P. Saunders, and D. Fraser, 1980. Radio-
tracking of moose in the boreal forest of northwestern Ontario. Canadian
Field-Naturalist. 94: 269-276.

Aho, R. W., S. M. Schmitt, J. Hendrickson, and T. R. Minzey. 1995. Michigan’s
translocated moose population: 10 years later. Wildlife Division Report
Number 3245, Michigan Department of Natural Resources, Lansing,
Michigan. 18pp.

Albert, D. A. 1995. Regional landscape ecosystems of Michigan, Minnesota, and
Wisconsin: a working map and classification. General Technical Report. NC-
178. St. Paul, Minnesota: US. Department of Agriculture, Forest Service,
North Central Forest Experiment Station. 250 pp.

Albright, C. A., and L. B. Keith. 1987. Population dynamics of moose, Alces
alces, on the south-coast barrens of Newfoundland. Canadian Field-
Naturalist 101: 373-387.

Anderson, R. C. 1964. Neurological disease in moose infected experimentally
with Pneumostrongylus tenuis from white-tailed deer. Pathological
Veterinarian 1: 289-322.

Anderson, R. C., and M. W. Lankester. 1974. Infectious and parasitic diseases
and arthropod pests of moose in North America. Naturaliste Canada 101: 23-
50.

Baker, R. H. 1983. Michigan mammals. Michigan State University Press. East
Lansing, Michigan. 642pp.

Ballard, W. B., T. H. Speaker, and K. P. Taylor. 1981. Causes of neonatal moose
calf mortality in southcentral Alaska. Journal of Wildlife Management 45: 335-
342.

Ballard, W. B., J. S. Whitman, and D. J. Reed. 1991. Population dynamics of
moose in south-central Alaska. Wildlife Monographs No. 114.49 pp.

Bemdt, L. W. 1988. Soil survey of Baraga county area, Michigan. United States
Department of Agriculture, Soil Conservation Service. 306 pp.

Blood, D. A. 1974. Variation in reproduction and productivity of an enclosed herd

of moose (Alces alces). Transaction of the lntemational Congress of Game
Biologists 1: 59-66.

114

Boer, A. H. 1987. Reproductive productivity of moose in New Brunswick. Alces
23:49-60.

Boer, A. H. 1992. Fecundity of North American moose (Alces alces): a review.
Alces Supplement 1: 1-10.

Boulanger, J. G., and G. C. White. 1990. A comparison of home-range
estimators using Monte Cario simulation. Journal of Wildlife Management.
54(2): 310-315.

Bunck, C. M., C. L. Chen, and K. H. Pollock. 1995. Robustness of survival
estimates from radio-telemetry studies with uncertain relocation of animals.
Journal of Wildlife Management 59: 790-793.

Conner, M. J., and B. D. Leopold. 2001. A euclidian distance metric to index
dispersion from radiotelemetry data. Wildlife Society Bulletin 29(3): 783-
786.

Cox, E. W., D. L. Murray, and T. K. Fuller. 1997. Moose population dynamics in
northwestern Minnesota - annual progress report and recommendations.
Minnesota Department of Natural Resources, Wildlife Populations and
Research Unit 1997 report, Minneapolis, Minnesota, pp. 66-99.

Crete M., and R. Courtois. 1997. Limiting factors might obscure regulation of
moose (Cervidae: Alces alces) in unproductive boreal forests. Journal of
Zoology (London) 242: 765-781.

de Vos, A. A. 1964. Range changes of mammals in the Great Lakes region.
American Midland Naturalist 71(1): 210-231.

Doerr, J. G. 1983. Home range size, movements and habitat use in two moose
populations in southeastern Alaska. Canadian Field-Naturalist. 97: 79-88.

Drummer, T. D., and R. W. Aho. 1998. A sightability model for moose in upper
Michigan. Alces 34(1): 15-19.

Edwards, R. Y., and R. W. Ritcey. 1958. Reproduction in a moose population.
Journal of Wildlife Management 23: 261-268.

Flynn, A., A. W. Franzmann, and P. D. Ameson. 1975. Sequential hair shaft as
an indicator of prior mineralization in Alaskan moose. Journal of Animal
Science 41 : 906-910.

Franzmann, A. W., C. C. Schwartz, and R. 0. Peterson. 1980. Moose calf
mortality in summer on the Kenai Peninsula, Alaska. Journal of Wildlife

115

Management 44: 764-768.

Franzmann, A. W., and C. C. Schwartz. 1985. Moose twinning rates: a possible
population condition assessment. Journal of Wildlife Management 49: 394-
396.

Gamer, D. L., and W. F. Porter. 1990. Movements and seasonal home ranges of
bull moose in a pioneering Adirondock population. Alces 29: 80-85.

Gasaway, W. C., R. O. Stephenson, J. L. Davis, P. E. K. Shepherd, and O. E.
Burris. 1983. Interrelationships of wolves, prey, and man in interior Alaska.
Wildlife Monographs 84. 50 pp.

Gasaway, W. C., R. D. Boertje, D. V. Grandgard, K. G. Kelleyhouse, R. O.
Stephenson and D. G. Larsen. 1992. The role of predation in limiting moose
at low densities in Alaska and Yukon and implications for conservation.
Wildlife Monographs 120. 59 pp.

Gilbert, F. F. 1966. Aging white-tailed deer by annuli in the cementum of the first
incisors. Journal of Wildlife Management 30(1): 200-202.

Gilbert, F. F., 1974. Pare/aphostrongylus tenuis in Maine: ll-prevalence in
moose. Journal of Wildlife Management 38(1): 42-46.

Gilbert, F. F., 1992. Retroductive logic and the effects of meningeal worms: a
comment. Journal of Wildlife Management 56(3): 614-616.

Greenwood. P. J. 1980. Mating systems, philopatry and dispersal in birds and
mammals. Animal Behavior 28: 1140-1162.

Haigh, J. C., E. W. Kowal, W. Runge, and G. Wobeser. 1982. Pregnancy
diagnosis as a management tool for moose. Alces 18: 224-253.

Haigh, J. C., W. J. Dalton, C. A. Ruder, and R. G. Sasser. 1993. Diagnosis of
pregnancy in moose using a bovine assay for pregnancy-specific protein B.
Theriogenology 40: 905-911.

Hauge, T. M. and L. B. Keith. 1981. Dynamics of moose populations in
northeastern Alberta. Journal of Wildlife Management. 45: 573-597.

Harris, 8., W. J. Cresswell, P. G. Forde, W. J. Trewhella, T. Woollard, and S.
Wray. 1990. Home-range analysis using radio-tracking data - a review of
problems and techniques particularly as applied to the study of mammals.
Mammal Review 20(2/3): 97-123.

Heisey, D. M., and T. K. Fuller. 1985. Evaluation of survival and cause-specific

116

mortality rates using telemetry data. Journal of Wildlife Management 49:
688-674.

Hickie, P. F. 1944. Michigan moose. Michigan Department of Conservation,
Lansing, Michigan. 57pp.

Hill, H. R., and J. C. Pohl. 1983. The 1983 deer pellet group surveys. Michigan
Department of Natural Resources, Wildlife Division Report 2963, Lansing,
Michigan. 18 pp.

Hill, H. R. 2001. The 2001 deer pellet group surveys. Michigan Department of
Natural Resources, Wildlife Division Report 3349, Lansing, Michigan. 15 pp.

Hooge, P. N., W. M. Eichenlaub, and E. K. Solomon. 1999. Using GIS to
analyze animal movements in the marine environment. United States
Geological Survey, Alaska Biological Science Center, Glacier Bay Field
Station, Gustavus, Alaska. 20 pp.

Huang, F., D. C. Cockrell, T. R. Stephenson, J. H. Noyes, and R. G. Sasser.
2000. A serum pregnancy test with a specific radioimmunoassay for moose

and elk pregnancy-specific protein B. Journal of Wildlife Management 64(2):
492-499.

Hundertmark, K. J. 1998. Home range, dispersal and migration. Pages 303-335
in A. W. Franzmann and C. C. Schwartz, editors. Ecology and management
of the North American moose. Smithsonian Institution Press, Washington,
DC.

Kams, P. D. 1967. Pneumostrongylus tenuis in deer in Minnesota and
implications for moose. Journal of Wildlife Management 31: 299-303.

Kenward, R. E., 2001. A manual for wildlife radio tagging. Academic Press, San
Diego, California. 311 pp.

Kemohan, B. J., R. A. Gitzen, and J. J. Millspaugh. 2001. Analysis of survival
data from radiotelemetry Studies. Pages 125-166 in J. J. Millspaugh and J.
M. Marzluff editors. Radio tracking and animal populations. Academic Press,
San Diego, California.

Kubota, J., and W. H. Allaway. 1972. Geographic distribution of trace element
problems. Pages 525-554 in J. J. Mortvedt, P. M. Giordano, and W. L.
Lindsay, editors. Soil Science Society of America, Madison Wisconsin, USA.

Lankester, M. W., and W. M. Samuel. 1998. Pests, parasites and diseases.

Pages 479-518 in A. W. Franzmann and C. C. Schwartz, editors. Ecology
and management of the North American moose. Smithsonian Institution

117

Press, Washington, DC.

Larsen, D. G., D. A. Gauthier, and R. L. Markel. 1989. Cause and rate of moose
mortality in the southwest Yukon. Journal of Wildlife Management 53: 548-
557.

Lawson, E.J., and A. R. Rodgers. 1997. Differences in home-range size
computed in commonly used software programs. Wildlife Society Bulletin.
25(3): 721-729.

Lynch, G. M. 1976. Some long-range movements of radio-tagged moose in
Alberta. Proceedings of the North American Moose Conference Workshop.
12: 220-235.

MacCracken, J. G., V. Van Ballenberghe, and J. M. Peek. 1997. Habitat
relationships of moose on the Copper River Delta in coastal south-central
Alaska. Wildlife Monographs No. 136. 52 pp.

Mauer, F. J., Moose migration: northeastern Alaska to northwestern Yukon
Territory, Canada. 1998. Alces 34(1): 75-81.

Mayfield, H. 1961. Nesting success calculated from exposure. Wilson Bulletin
73: 255-261.

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

McNicol, J. G., and F. F. Gilbert. 1980. Late winter use of upland cutovers by
moose. Journal of Wildlife Management 44(2): 363-371.

Mech, L. D., and G. D. DelGiudice. 1985. Limitations of the marrow-fat
technique as an indicator of body condition. Wildlife Society Bulletin 13:
204-206.

Messier, F. D., and M. Créte. 1984. Body condition and population regulation by
food resources in moose. Oecologia 65: 44-50.

Messier, F ., D. M. Desauliners, A. K. Goff, R. Nault, R. Patenaude, and M. Crete.
1990. Caribou pregnancy diagnosis from immunoreactive progestins and
estrogens excreted in feces. Journal of Wildlife Management 54(2): 279-

283.

Michigan Department of Natural Resources. 1962. 1961-1962 Biennial Report.
Michigan Department of Natural Resources, Lansing, Michigan.

Michigan Department of Natural Resources. 1974. Abstract of Harold Cumming’s

118

report on the potential of moose range in the Upper Peninsula. 1973-1974
Biennial Report. Michigan Department of Natural Resources, Lansing,
Michigan.

Michigan Department of Natural Resources. 1991a. Draft moose management
plan. Michigan Department of Natural Resources, Wildlife Division Report.
Lansing, Michigan. 8 pp.

Michigan Department of Natural Resources. 1991b. Experimental Michigan
moose survey. Wildlife Division, Michigan Department of Natural Resources,
Lansing, Michigan. 8 pp.

Michigan Department of Natural Resources. 1997. Michigan’s annual moose
count apparently affected by harsh U.P. winter. DNR News. Lansing,
Michigan. 2 pp.

Millspaugh, J. J., R. J. Woods, K. E. Hunt, K. J. Raedeke, G. C. Brundige, B. E.
Washbum, and S. K. Wasser. 2001. Fecal glucocorticoid assays and the
physiological stress response in elk. Wildlife Society Bulletin 29(3): 899-907.

Minzey, T. R., and W. L. Robinson. 1991. Characteristics of winter bed sites of
moose in Michigan. Alces 27: 150-160.

Mohr, C. O., 1947. Table of equivalent populations of North American small
mammals. American Midland Naturalist 37:223-249.

Monfort, S. L., C. C. Schwartz, and S. K. Wasser. 1993. Monitoring reproduction
in moose using urinary and fecal steroid metabolites. Journal of Wildlife
Management 57: 40-407.

Murie, A. 1934. The moose of Isle Royale. University of Michigan Museum,
Zoological Miscellaneous Publication 25. 44 pp.

Mytton, W. R., and L. B. Keith. 1981. Dynamics of moose populations near
Rochester, Alberta, 1975-1978. The Canadian Field-Naturalist 95(1): 39-49.

Neiland, K. A. 1970. Weight of dried marrow as indicator of fat in caribou femurs.
Journal of Wildlife Management 34: 904-907.

Nettles, V. F. 1981. Necropsy procedures. Pages 6-16 in Diseases and
parasites of white-tailed deer. W. R. Davidson Editor. University of Georgia,
Athens, Georgia.

Nudds, T. D. 1990. Retroductive logic in retrospect: the ecological effects of
meningeal worms. Journal of Wildlife Management 54(3): 396-402

119

Nudds, T. D. 1992. Retroductive logic and the effects of meningeal worms: 3
reply. Journal of Wildlife Management 56(3): 617-619.

Oliver, M. N., D. A. Jessup, B. B. Norman, and C. E. Franti. 1990. Selenium
concentrations in blood of free-ranging mule deer in California. Pages 80-86
in Transactions of the western section of the Wildlife Society.

Olterman, J. H., and D. W. Kenvin. 1998. Reproduction, survival, and occupied
ranges of Shiras moose transplanted to southwestern Colorado. Alces
34(1): 41-46.

Patterson, R. L., S. L. Ockey, C. E. Olson, and A Brenner. 1995. Analysis of
population statistics (1985-1994) and habitat (1955-1991) for moose (Alces
alces) in the western Upper Peninsula of Michigan. Final research submitted
to the Wildlife Division, Michigan Department of Natural Resources. 42 pp.

Peek, J. M., D. L. Urich, and R. J. Mackie. 1976. Moose habitat selection and
relationships to forest management in northeastern Minnesota. Wildlife
Monographs No. 48. 66 pp.

Peterson, R. L. 1955. North American moose. University of Toronto Press.
Toronto, Ontario. 280pp.

Peterson, R. O., J. D. Woolongton, and T. N. Bailey. 1984. Wolves of the Kenai
Peninsula, Alaska. Wildlife Monographs No. 88. 52 pp.

Phillips, R. L., W. E. Berg, and D. B. Sniff. 1973. Moose movement patterns and
range use in northeastern Minnesota. Journal of Wildlife Management 37:
266-278.

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

Reimers, T. J., R. G. Sasser, and C. A. Ruder. 1985. Production of pregnancy-
specific protein B by bovine binucleate trophoblast cells. Biology of
Reproduction 32 (Supplement 1): 65.

Roffe, T. J., M. Friend, and L. N. Locke. 1994. Evaluation of causes of wildlife
mortality. Pages 324-348 in Research and management techniques for
wildlife and habitats. T. A. Bookhout editor. The Wildlife Society. Bethesda,
Maryland.

Sargeant, D. W., and D. H. Pimlott. 1959. Age determination in moose from
sectioned incisor teeth. Journal of Wildlife Management 23: 315-321.

120

Schmitt, S. M., and R. H. Aho. 1986. Reintroduction of moose from Ontario to
Michigan. Pages 258-274 in L. Nielsen and R. D. Brown, editors.
Translocation of wild animals. Wisconsin Humane Society, Milwaukee, and
Kleberg Wildlife Research Institute, Kingsville, Texas.

Schroger, A. W. 1942. Extinct and endangered mammals and birds of the upper
Great Lakes Region. Transactions of the Wisconsin academy of Science,
Arts, and Letters 34:23-44.

Schwartz, C. C., and K. J. Hundertmark. 1993. Reproductive characteristics of
Alaskan moose. Journal of Wildlife Management 57: 454-468.

Schwartz, C. C., S. L. Monfort, P. H. Dennis, and K. J. Hundertmark. 1995. Fecal
Progestagen concentration as an indicator of the estrous cycle and
pregnancy in moose. Journal of Wildlife Management 59:580-583.

Schwartz, C. C. 1997. Reproduction, natality, and growth. Pages 141-171 in A.
W. Franzmann and C. C. Schwartz, editors. Ecology and management of the
North American moose. Smithsonian Institution Press, Washington, DC.

Seaman, D. E., B. Griffith, and R. A. Powell. 1998. KERNELHR: a program for
estimating animal home ranges. Wildlife Society Bulletin 26(1): 95-100.

Stenhouse, G. B., P. B. Latour, L. Kutny, N. MacLean, and G. Glover 1995.
Productivity, survival, and movements of female moose in a low-density
population, Northwest Territories, Canada. Arctic 48(1):57-62.

Stephenson, T. R., J. W. Testa, G. P. Adams, R. G. Sasser, C. C. Schwartz, and
K. J. Hundertmark. 1995. Diagnosis of pregnancy and twinning in moose by
ultrasonography and serum assay. Alces 31 :167-172.

Stewart, R. R., L. M. Comishen-Stewart, and J. C. Haigh. 1985. Levels of some

reproductive hormones in relation to pregnancy in moose: Preliminary report.
Alces 21: 393-402.

Sweanor, P. Y., F. Sandgren, R. Bergstrom, and G. Cederlund. 1992. A synopsis
of moose movement studies in Furudal, Sweden. Alces Supplement 1: 115-
120.

Telfer, E. S. 1970. Winter habitat selection by moose and white-tailed deer.
Journal of Wildlife Management. 34(3): 553-559.

Theberge, J. B., with M. T. Theberge. 1998. Wolf Country, eleven years tracking
the Algonquin wolves. McClelland & Stewart Inc, Toronto, Ontario. 306 pp.

Thompson, M. E., 1987. Seasonal home range and habitat use by moose in

121

northern Maine. M. S. thesis, University of Maine, Orono. 57 pp.

Timmerrnann, H. R. 1993. Use of aerial surveys for estimating and monitoring
moose populations - a review. Alces 29:35-46.

Trent, T. T. and O. J. Rongstad. 1974. Home range and survival of cottontail
rabbits in southwestern Wisconsin. Journal of Wildlife Management 38: 459-
472.

Van Ballenberghe, V., and W. B. Ballard. 1998. Population dynamics. Pages
223-245 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and
management of the North American moose. Smithsonian Institution Press,
Washington, DC.

Van Ballenberghe, V., and J. M. Peek. 1971. Radiotelemetry studies of moose in
northeastern Minnesota. Journal of Wildlife Management 35: 63-71.

Verme, L. J. 1970. Some characteristics of captive Michigan moose. Journal of
Mammalogy 51: 403-405.

Verme, L. J. 1984. Some background on moose in Upper Michigan. Michigan
Department of Natural Resources, Wildlife Division Report 2973, Lansing,
Michigan. 6pp.

Whitlaw, H. A., and M. W. Lanksester. 1994a. A retrospective evaluation of the
effects of parelaphostrongylosis on moose populations. Canadian Journal of
Zoology 72: 1-7).

Whitlaw, H. A., and M. W. Lanksester. 1994b. The co-occurrence of moose,
white-tailed deer, and Pare/aphostrongylus tenuis in Ontario. Canadian
Journal of Zoology 72: 819-825).

White, G. A., and R. A. Garrott. 1990. Analysis of radio-tracking data. Academic
Press, Inc., Toronto, Canada. 383 pp.

Wilton, M. L. 1982. Report to the Michigan Department of Natural Resources
concerning potential moose habitat. Michigan Department of Natural
Resources. Wildlife Division, Lansing, Michigan. 25pp.

Winterstein, S. R., K. H. Pollock, and C. M. Bunck. 2001. Analysis of survival
data from radiotelemetry studies. Pages 351-380 in J. J. Millspaugh and J.
M. Marzluff editors. Radio tracking and animal populations. Academic Press,
San Diego, California.

Wood, N. A. 1914. An annotated check-list of Michigan mammals. University of
Michigan Occasional Papers No.4.13pp.

122

Wood, N. A., and L. R. Dice. 1923. Records of the distribution of Michigan
mammals. Michigan Academy of Science, Arts and Letters. 3: 425-469.

Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in
home-range studies. Ecology 70:164-168.

123