9:55?-

‘un

-115

'ygéiﬁ': s. "n; Egg-ya.

3“: L' AW

33%?

7"" -‘.!

i-o-m

.53

1‘ 5f». m

ya,

R ‘

"I‘H‘ A

I

T-

13%

 

 

 

 

 

II I IIIIIIIIIIIIIIIIIIIIIIII I III IIIIIIII

1174

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Movements, Habitat Use, and Survival of
Ruffed Grouse (Bonasa umbellus) in Northern Michigan

 

presented by

Margaret E. Clark

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Fish. & Wildl.

 

 

gm Qua-A

Major professor

 

Date November 4, 1996

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

PLACE IN RETURN BOX to remove thie checkout from your record.
TO AVOID FINES return on or More dete due.

DA'DIEGQQE DATE DUE DATE DUE

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Ie An Affinnetive Action/Equel Opportunity lnetltution
WHO-N

 

MOVEMENTS, HABITAT USE, AND SURVIVAL
OF RUFFED GROUSE (Bonasa umbellus) IN NORTHERN MICHIGAN

By

Margaret E. Clark

A THESIS

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

MASTER OF SCIENCE

Department of Fisheries and Wildlife

1 996

ABSTRACT

MOVEMENTS, HABITAT USE, AND SURVIVAL
OF RUFFED GROUSE (Bonasa umbellus) IN NORTHERN MICHIGAN

By

Margaret E. Clark

The ruﬁ‘ed grouse (Bonasa umbellus) is a valued game bird throughout the United
States, and northern Michigan is no exception. Declining grouse populations have met
with concern by hunters and wildlife biologists. This project was initiated in 1993 to
compare the movements, habitat use and survival of ruffed grouse on a site closed to
hunting with those for ruffed grouse on a site open to hunting. An analysis of the
composition and quality of habitat for ruffed grouse was also undertaken on the study
sites. Such a comprehensive study was necessary to provide insight on the characteristics
of ruﬁ‘ed grouse populations through fall and winter, perhaps the birds’ most vulnerable
times of year.

Ruﬂ‘ed grouse were radio-collared in the fall of 1993 and 1994, and locations were
recorded for each bird several times per week to determine movements and habitat use.
Dispersal movements and home ranges were similar on sites open and closed to hunting.
No consistent trends were found in the movements of ruffed grouse by site, sex, or age

class. Habitat use differed between sites, with birds in the open site ﬁ'equently choosing

older aspen. Birds in the closed site used openings, lowland hardwoods, older pine and
aspen types more frequently than expected.

Radio-collared grouse were monitored for survival through the fall and winter of
1993 and 1994. Survival between sites differed signiﬁcantly in 1993, with the closed site
birds having a higher survival probability. In 1994, survival was similar between sites.
Avian predation was the largest source of mortality in both sites in both years, and hunting
accounted for a moderate number of mortalities.

Habitat composition and quality for ruffed grouse differed between sites, with the
open site having more aspen and higher quality habitat according to the Habitat Model for
Ruﬂ‘ed Grouse in Michigan (Hammill and Moran 1986). Quality of habitat was generally
lower than expected for aspen, and other habitat types met the requirements of the model

as well as or better than aspen.

ACKNOWLEDGMENTS

Thank you to Dr. Scott Winterstein for taking on this project, and for advice,
patience, and a great sense of humor. I have so appreciated having you as a role model.
I also thank my committee members, Dr. Rique Campa and Dr. Don Beaver for their
time and effort on the part of this study.

This project could not have taken place without ﬁmding provided by the
Michigan Department of Natural Resources Wildlife Division, the Ruffed Grouse
Society, and Michigan State University. Wildlife Division has also provided us with the
help of numerous personnel, and their efforts were greatly appreciated; Tom Cooley for
innumerable grouse necropsies, Jeff Green for teaching us how to catch those pesky
birds, Bill Green for telemetry ﬂights, and John Urbain for wholeheartedly supporting
and adding his knowledge to this project.

The entire staff at the Pigeon River Headquarters has welcomed me with warmth
and friendship from the beginning of this study, and has bent over backwards to help me
make it work. Thank you to Joe, Char, Dan, Ken, Rick and Arch for grouse tips,
equipment, advice, and for using the big red truck to pull us out of the mud and ice.
Thanks also to MDNR biologists Glenn Matthews and Doug Whitcomb of Gaylord for
providing me with needed information, and for spreading the word to the public about

the project.

iv

Special thanks goes to Katherine Bieme, for a friendship that made it through
some rough holidays, and earthworms, frogs and snakes in the apartment. I’ll never
forget the message I got in Alaska: “Some girl from Virginia wants to be your
roommate.” How thankﬁrl I am that it was you. Thanks also to Matt Bieme, for all the
laughs and friendship.

Many graduate students in the department made ﬁeld work and school work
enjoyable. Delia Raymer, Kelly Millenbah, Christine Hanaburgh, Gina Karasek, Steve
Negri, Bruce Pfeffers, Mike Larson, Mark Moore, Doug Peterson, and others have made
the department a comfortable and friendly place to be. Thanks also to Allison Gormley,
my colleague on the grouse project, for her ﬁiendship and hard work.

I would like to thank all the interns and fall/winter workers who made this project
possible: Mike Parker, Mandy Dunlap, Meagan Williams, Linda Briggs, Prentice
Pylrnan, Jeremiah Breuker, Mary Beth Wirth, Brad Kassuba, Zoe Chao, Amy Hinkson,
Steve Husted, Ty Ratliff, Ann Freel, Rebecca Graving, Joe Cea, Ryan White, Kathy
Wicktor, Dan Kennedy, Dave Denomme, Todd Felkey, Grant Goltz, and last but not
least, Margi Chriscinske, Petosky Stone Queen and everlasting grouse trapper.

My sincere thanks to Jeff Breuker for being a great friend, and a terriﬁc help on
this project. Your conﬁdence, patience and love have been invaluable. I couldn’t have
done it without you.

Special thanks to my parents, my sister Pam, and my brother Ed. Throughout my
life, they have shown me more love and support than a person could ask for. For as long

as I can remember, they have instilled in me a belief that I could be anything I wanted to

be. I know they didn’t count on me to choose wildlife biology as a career, but the love
and support have been there none the less.

Lastly, I would like to thank my niece Stephanie and my nephews Christopher,
Nick, and little Eddie. All I had to do was look at their pictures and letters and even the
biggest problems melted away. Thanks to them for giving me the inspiration to continue

through the roughest times in graduate school and in life.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

LIST OF TABLES -- - - ....... - - -- ........ - -- - - -- -viii
LIST OF FIGURES - _ - - - - - -- xi
INTRODUCTION - - -- l
OBJECTIVES - -- - - _ -- - - - 4
STUDY AREA - 5
METHODS _ 8
Trapping and collan'ng ................................................................................................... 8
Vegetation sampling .................................................................................................... 10
Survival and habitat use ............................................................................................... l4
Movements ................................................................................................................. 16
Data analysis ............................................................................................................... 18
RESULTS AND DISCUSSION __ - - - 20
Trapping ..................................................................................................................... 20
Habitat ........................................................................................................................ 26
Habitat quality ............................................................................................................. 31
Movements ................................................................................................................. 46
Habitat use .................................................................................................................. 53
Survival ....................................................................................................................... 7O
Proportional hazards model ......................................................................................... 83
Conclusions ................................................................................................................. 89
Management Implications ............................................................................................ 92
APPENDICES- _ _ 95
Appendix 1. Production ﬁrnctions from the Habitat Model for Ruﬁ‘ed Grouse in
Michigan (Hammrl' l and Moran 1986). ..................................................................... 95
Appendix II. Examples of ruﬁ‘ed grouse dispersal patterns in the Pigeon River Country
State Forest, 1993. .................................................................................................. 98
Appendix III. Weekly harmonic centers of individual ruffed grouse in the open and
closed sites of the Pigeon River Country State Forest, in 1993 and 1994 ................ 101
LITERATURE CITED -- 109

 

vii

LIST OF TABLES

Table 1. Michigan Department of Natural Resources cover type components used for
each sampling category in 1994 and 1995 vegetation sampling in the PRCSF ............... 11

Table 2. Number of stands sampled in each site and plot size used for each sampling
category in 1994 and 1995 vegetation sampling in the PRCSF. .................................... 12

Table 3. Trapping results for 1993, PRCSF open and closed sites. Birds caught includes
recaptures and birds not collared. ................................................................................ 21

Table 4. Trapping results for 1994, PRCSF open and closed sites. Birds caught includes
recaptures and birds not collared. ................................................................................ 23

Table 5. Number of birds radio-collared in the PRC SF by site, year, sex and age class.
No signiﬁcant differences in age or sex distributions were detected between sites or years
(Chi-square, p > 0.10). ................................................................................................ 25

Table 6. Distribution of habitat types in the Pigeon River Country State Forest open and
closed sites. ................................................................................................................. 27

Table 7. Distribution of habitat types in the North and South areas of the Pigeon River
Country State Forest closed site. ................................................................................. 30

Table 8. Overall HSI values for the PRCSF open and closed sites. Weighted HSI is the
mean HSI for the habitat type * the percent of area in that habitat type. ....................... 32

Table 9. Overall HSI values for the North and South areas of the Pigeon River Country
State Forest closed site. Weighted HSI is the mean HSI for the habitat type * the
percent of area in that habitat type. .............................................................................. 35

Table 10. Pearson correlation coefficients (r) and associated probabilities (P) for
individual variables versus HSI values in each habitat type in the open site of the PRCSF.36

Table 11. Pearson correlation coefﬁcients (r) and associated probabilities (P) for
individual variables versus HSI values in each habitat type in the closed site of the
PRCSF. ....................................................................................................................... 36

Table 12. Mean deciduous, conifer and shrub stems/ha (SE), and equivalent stem density
for each habitat type sampled in the open site of the PRCSF. ....................................... 38

viii

Table 13. Mean deciduous, conifer and shrub stems/ha (SE), and equivalent stem density
for each habitat type sampled in the closed site of the PRC SF. ..................................... 38

Table 14. Percent of mean equivalent stem density for deciduous, conifer and shrub
stems for each habitat type sampled in the open site of the PRC SF. ............................. 40

Table 15. Percent of mean equivalent stem density for deciduous, conifer and shrub
stems for each habitat type sampled in the closed site of the PRCSF. ........................... 40

Table 16. Pearson correlation coefﬁcients (r) and associated probabilities (P) for separate
stem density components of the Equivalent Stem Density versus HSI values in each
habitat type in the open site of the PRCSF. .................................................................. 41

Table 17. Pearson correlation coefﬁcients (r) and associated probabilities (P) for separate
stem density components of the Equivalent Stem Density versus HSI values in each
habitat type in the closed site of the PRCSF ................................................................. 42

Table 18. Mean home range size (ha) for Type I and Type II non-dispersers in the
PRCSF open and closed sites, 1993. ............................................................................ 47

Table 19 . Mean dispersal distance (m) for ruffed grouse in the PRCSF open and closed
sites, 1993. .................................................................................................................. 47

Table 20. Mean home range size (ha) for Type I and Type II non-dispersers in the
PRCSF open and closed sites, 1994. ............................................................................ 49

Table 21. Mean dispersal distance (m) for ruﬂ‘ed grouse in the PRCSF open and closed
sites, 1994. .................................................................................................................. 49

Table 22. Mean home range size (ha) for Type I and Type 11 birds and dispersal
distances (m) for Type III birds in the North and South areas of the closed site of the
PRCSF, 1993. ............................................................................................................. 51

Table 23. Mean home range size (ha) for Type I and Type 11 birds and dispersal
distances (m) for Type III birds in the North and South areas closed site of the PRCSF,
1994 ............................................................................................................................ 51

Table 24. Number and percent of grouse locations in each habitat type by site and year.54

Table 25. Number and percent of grouse locations in each habitat type in the open site by
dispersal type and year. ............................................................................................... 56

Table 26. Number and percent of grouse locations in each habitat type in North area of
the closed site by dispersal type and year. .................................................................... 57

Table 27. Number and percent of grouse locations in each habitat type in the South area
of the closed site by dispersal type and year. ................................................................ S8

ix

Table 28. Number and percent of radio-collared ruﬁ‘ed grouse by status in the PRC SF
open and closed sites in 1993 and 1994, as of May 15 of the following year. This table
contains only birds which were used in the analysis of survival. .................................... 74

Table 29. Number and percent of known deaths by category in the open and closed sites
of the PRCSF in 1993 and 1994. No signiﬁcant differences were detected in non-
hunting mortalities between sites within the same year (Chi-square, p > 0.10). ............. 74

Table 30. Number and percent of known deaths by category in the North and South
areas of the closed site of the PRCSF in 1994. No signiﬁcant differences were detected
in non-hunting mortalities between areas (Chi-Square, p > 0.10) .................................. 76

Table 31. Risk ratios and associated P values for individual variables tested for
association to survival with the Cox Proportional Hazards Model (Cox 1972). ............ 84

Table 32. Risk ratios and corresponding P values for explanatory variables found in a
stepwise regression through the Cox proportional hazards model (Cox 1972), and
overall model P values for male and female ruffed grouse in the open site of the PRC SF,
1993 and 1994. ........................................................................................................... 86

Table 33. Risk ratios and corresponding P values for explanatory variables found in a
stepwise regression through the Cox proportional hazards model (Cox 1972), and
overall model P values for male and female ruffed grouse in the closed site of the
PRCSF,1993 and 1994. .............................................................................................. 87

LIST OF FIGURES

Figure 1. Location of the site open to hunting in Cheboygan County and the site closed
to hunting in Otsego County, in the northern lower peninsula of Michigan. .................... 6

Figure 2. Aspen stems per hectare and equivalent stem density in various aged aspen
stands in the open site of the PRC SF. Dashed lines indicate maximum (19700) and
minimum (7400) stems/ha called for by Gullion (1970), and optimum (12085) stems/ha
in the Michigan model (Hamrnill and Moran 1986). ..................................................... 44

Figure 3. Aspen stems per hectare and equivalent stem density in various aged aspen
stands in the closed site of the PRCSF. Dashed lines indicate maximum (19700) and
minimum (7400) stems/ha called for by Gullion (1970), and optimum (12085) stems/ha
in the Michigan model (Harnmill and Moran 1986). ..................................................... 44

Figure 4. Fall and winter habitat preference for ruffed grouse in the Pigeon River
Country State Forest in 1993 and 1994. Grouse habitat preference was not different
(Waller Duncan multiple comparison procedure, p > 0.10) between underlined habitats.
Mean HSI values for each habitat type in each site are given for reference. .................. 60

Figure 5. Mean habitat use values by site and year. Error bars are one standard error
above and below the mean. Different letters indicate signiﬁcant difference among habitat
use values (Kruskal-Wallis one-way AN OVA, Multiple Comparisons, p < 0.10). ......... 65

Figure 6. Mean habitat use values by dispersal type in the open site of the PRC SF in

1993. Error bars are one standard error above and below the mean. No signiﬁcant
diﬂ‘erences were detected (Kruskal-Wallis one-way AN OVA, p > 0.10). ..................... 66

Figure 7. Mean habitat use values by dispersal type in the open site of the PRCSF in
1994. Error bars are one standard error above and below the mean. No signiﬁcant
differences were detected (Kruskal-Wallis one-way ANOVA, p > 0.10). ..................... 66

Figure 8. Mean habitat use values by dispersal type in the South area of the closed site of
the PRCSF in 1993. Error bars are one standard error above and below the mean. No
signiﬁcant diﬁ‘erences were detected (Kruskal-Wallis one-way ANOVA, p > 0.10). ..... 67

Figure 9. Mean habitat use values by dispersal type in the North area of the closed site of
the PRCSF in 1994. Error bars are one standard error above and below the mean. No
signiﬁcant diﬂ‘erences were detected (Kmskal-Wallis one-way ANOVA, p > 0.10). ..... 68

Figure 10. Mean habitat use values by dispersal type in the South area of the closed site
of the PRCSF in 1994. Error bars are one standard error above and below the mean.
No signiﬁcant differences were detected (Kruskal-Wallis one-way ANOVA, p > 0.10).68

Figure 11. Survival probability for ruffed grouse radio-collared in the PRCSF in 1993.
Arrows approximate hunting season, September 15 through December 31, with a 2 week
break between November 15 and November 30. .......................................................... 71

Figure 12. Survival probability for ruffed grouse radio-collared in the South area of the
PRCSF closed site in 1993. ......................................................................................... 71

Figure 13. Survival probability for ruffed grouse radio-collared in the PRCSF in 1994.
Arrows approximate hunting season, September 15 through December 31, with a 2 week
break between November 15 and November 30. .......................................................... 73

Figure 14. Survival probability for ruffed grouse radio-collared in the North and South
areas of the PRCSF closed site in 1994. ....................................................................... 73

Figure 15. Survival probability for adult (AHY) and juvenile (HY) ruﬁ‘ed grouse in the
open site of the PRCSF, 1993 ...................................................................................... 78

Figure 16. Survival probability for male and female ruffed grouse in the open site of the
PRCSF, 1993. ............................................................................................................. 78

Figure 17. Survival probability for adult (AHY) and juvenile (HY) mﬁ‘ed grouse in the
closed site of the PRCSF, 1993. .................................................................................. 79

Figure 18. Survival probability for male and female ruffed grouse in the closed site of the
PRCSF, 1993. ............................................................................................................. 79

Figure 19. Survival probability for adult (AHY) and juvenile (HY) ruffed grouse in the
open site of the PRCSF, 1994 ...................................................................................... 80

Figure 20. Survival probability for male and female ruffed grouse in the open site of the
PRCSF, 1994. ............................................................................................................. 80

Figure 21. Survival probability for adult (AHY) and juvenile (HY) ruﬁ‘ed grouse in the
closed site of the PRCSF, 1994. .................................................................................. 81

Figure 22. Survival probability for male and female ruffed grouse in the closed site of the
PRCSF, 1994. ............................................................................................................. 81

Appendix I, Figure 1. Index value for equivalent stem density in the Michigan model.. 95

Appendix I, Figure 2. Index value for average height of deciduous trees in the Michigan
model. ......................................................................................................................... 95

xii

Appendix I, Figure 3. Index value for height above ground of lower coniferous branches
in the Michigan model. ................................................................................................ 96

Appendix I, Figure 4. Index value for average height of deciduous shrubs in the
Michigan model ........................................................................................................... 96

Appendix I, Figure 5. Index value for distance between life requisites in the Michigan
model. ......................................................................................................................... 97

Appendix II, Figure 1. Weekly harmonic centers for a single bird plotted spatially and
numbered by week. This bird was categorized as a Type 1 non-disperser ..................... 98

Appendix H, Figure 2. Distance moved by a single bird ﬁom the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a Type I non-disperser (same individual as in Appendix II,
Figure 1). Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at
week 16 indicates November 30, after which movements are not considered fall
dispersal. ..................................................................................................................... 98

Appendix II, Figure 3. Weekly harmonic centers for a single bird plotted spatially and
numbered by week. This bird was categorized as a Type II non-disperser. .................. 99

Appendix 11, Figure 4. Distance moved by a single bird from the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a Type II non-disperser (same individual as in Appendix 11,
Figure 3). Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at
week 16 indicates November 30, after which movements are not considered fall
dispersal. ..................................................................................................................... 99

Appendix II, Figure 5. Weekly harmonic centers for a single bird plotted spatially and
numbered by week. This bird was categorized as a disperser ..................................... 100

Appendix H, Figure 6. Distance moved by a single bird from the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a disperser (same individual as in Appendix 11, Figure 5).
Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at week 16
indicates November 30, after which movements are not considered fall dispersal. ...... 100

Appendix III, Figure 1. Weekly harmonic centers for individual male ruffed grouse
(identiﬁed by radio frequency) in the open site of the PRC SF, September - December,
1993. Map is sealed with Universal Transverse Mercator (UTM) coordinates. .......... 101

Appendix III, Figure 2. Weekly harmonic centers for individual female ruffed grouse
(identiﬁed by radio frequency) in the open site of the PRCSF, September - December,
1993. Map is scaled with Universal Transverse Mercator (UTM) coordinates. .......... 102

Appendix III, Figure 3. Weekly harmonic centers for individual male ruffed grouse
(identiﬁed by radio frequency) in the open site of the PRC SF, August - December, 1994.
Map is scaled with Universal Transverse Mercator (UTM) coordinates ...................... 103

Appendix III, Figure 4. Weekly harmonic centers for individual female and unknown sex
ruffed grouse (identiﬁed by radio frequency) in the open site of the PRC SF , August -
December, 1994. Map is sealed with Universal Transverse Mercator (UTM)
coordinates ................................................................................................................ 104

Appendix III, Figure 5. Weekly harmonic centers for individual male ruffed grouse
(identiﬁed by radio frequency) in the closed site of the PRCSF, September - December,
1993. Map is scaled with Universal Transverse Mercator (UTM) coordinates. Dashed
line indicates the approximate location of the division between North and South areas of
the closed site ............................................................................................................ 105

Appendix 111, Figure 6. Weekly harmonic centers for individual female ruffed grouse
(identiﬁed by radio frequency) in the closed site of the PRCSF, September - December,
1993. Map is sealed with Universal Transverse Mercator (UTM) coordinates. Dashed
line indicates the approximate location of the division between North and South areas of
the closed site ............................................................................................................ 106

Appendix IH, Figure 7. Weekly harmonic centers for individual male ruffed grouse
(identiﬁed by radio ﬁ'equency) in the closed site of the PRCSF, August - December,
1994. Map is scaled with Universal Transverse Mercator (UTM) coordinates. Dashed
line indicates the approximate location of the division between North and South areas of
the closed site ............................................................................................................ 107

Appendix 111, Figure 8. Weekly harmonic centers for individual female and unknown sex
ruffed grouse (identiﬁed by radio ﬁ'equency) in the closed site of the PRCSF, August -
December, 1994. Map is sealed with Universal Transverse Mercator (UTM)
coordinates. Dashed line indicates the approximate location of the division between
North and South areas of the closed site. ................................................................... 108

xiv

INTRODUCTION

Ranging from central Alaska to the southern Appalachians, the ruffed grouse
(Bonasa umbellus) has the most extensive distribution of any resident game bird in North
America (Johnsgard 1973). It is a highly valued species throughout its range for hunting
as well as nonconsumptive recreation, such as photography and bird watching. Since the
early 19003, there has been considerable attention paid to the periodic changes in the
number of grouse present in the forests of North America. Still, little is known about how
certain factors (e.g. seasonal movements, habitat use and hunting pressure) inﬂuence
overall grouse survival. It has been recognized that a high survival rate through the fall
and winter is important to future population numbers (Gullion 1970), and that several
factors must play a role in determining this rate in any given year.

In the midwest, ruﬁ‘ed grouse generally travel several kilometers during the fall
before settling in their wintering areas (Small and Rusch 1989). Fall and winter
movements of grouse are assumed to be related to food supply and cover (Chambers and
Sharp 1980). As the leaves begin to drop, the presence of dense vertical stems of
regenerating deciduous trees, shrubs, or coniferous trees becomes important for cover
(Hammill and Moran 1986). During dispersal, birds are likely to occupy less suitable
habitats, at least temporarily, as they seek what they judge to be good cover (Barber et al.

1989). According to Small, et al. (1991), an increased probability of death is thought to

be an inherent risk of this increased movement. Their study, in Waushara and Marquette
counties in Wisconsin, showed that increased mortality during transient dispersal was
negligible. However, grouse that had dispersed during the autumn season had a higher
winter mortality than those that remained on an established winter range.

Providing adequate habitat may be a key to increasing the survival of birds over
the fall and winter. Food and cover, as mentioned above, are the 2 factors which
determine habitat quality for grouse. Because of the relatively small amount of food that
is needed to sustain a grouse, it may be assumed that if a forest stand provides suﬂicient
cover, it will have a sufﬁcient food source as well (Harnmill and Moran 1986). Habitat
components which provide suitable cover from mammalian and avian predation include
high stem densities, high deciduous tree and shrub height, and low conifer branch height
(Cade and Sousa 1985, Hamrnill and Moran 1986). In Michigan, these components are
principally associated with deciduous hardwood forests, especially those with aspen as a
dominant species (Cade and Sousa 1985). Other habitat types which may provide suitable
cover are lowland areas and young conifer stands. The amount and quality of these
habitat types that is available in northern forests may help determine grouse dispersal
movements in addition to affecting survival.

Hunting is assumed to be an important factor affecting fall and winter grouse
survival, but its role is poorly understood. In the history of grouse hunting, management
attitudes have varied dramatically, with the most recent regulations being fairly liberal
(DeStefano and Rusch 1982). During the present period of increasing hunting pressure,
however, there is a question as to whether the inﬂuence of humans has an additive effect

on grouse mortality. Research has yet to demonstrate convincingly any impact of harvest

on subsequent numbers of ruffed grouse, but there is recent evidence of high harvest rates
on some public lands. Rusch et al. (1984) in Wisconsin followed the mortality of banded
juveniles through the summer, and subsequent hunting season. Chick mortality through
September was 63%, and another 58% of all banded grouse were lost during the harvest,
October through December. The overall survival of juveniles and adults was estimated at
0.16 ﬁ'om hatch to December, which would require a mean hatch of 10.2 chicks per hen to
balance spring production with mortality. Kubisiak (1984) also showed that high harvest
rates were a potential factor depressing grouse populations at the Sandhill Wildlife Area.
He emphasized that the area studied had excellent road access, exceptionally high hunter
eﬁ‘ort and a high proportion of hunters using dogs, which are also typical characteristics
for many hunting areas in northern Michigan.

To comprehensively describe the fall and winter characteristics of a grouse
population, each of the above mentioned factors must be detailed. With the use of radio-
telemetry, there is the possibility of knowing speciﬁc causes and rates of mortality. In
addition, movements and habitat use can be measured with relative case. This study
investigates survival, habitat use and movements on two large areas of land in northern
Michigan. To observe these factors in combination with the effects of hunting pressure, 1
area has been closed to grouse and woodcock hunting for the duration of the project. I
will describe and compare two separate grouse populations, and discuss the management

implications which arise from the results of this project.

OBJECTIVES

The speciﬁc objectives of this study are to:

1) determine habitat use and movements of ruffed grouse on sites open and closed to
hunting through the fall and winter;

2) determine habitat quality and quantity on study sites;

3) determine the survival probability and the causes of ruffed grouse mortality on study
sites in fall and winter; and

4) relate measured variables to survival.

STUDY AREA

This study was conducted in the Pigeon River Country State Forest (PRCSF),
which is located in the northern part of Michigan's lower peninsula. It comprises
approximately 98,000 acres in Cheboygan, Otsego and Montmorency counties. A section
of the forest was separated into two study sites, one open to all hunting under normal
regulations and one closed to grouse and woodcock hunting for the duration of the study.
The closed site comprises approximately 91 km2 (35 miz) in Otsego county (Figure 1).
Although the rest of the PRC SF is open to hunting, we have arbitrarily drawn a boundary
around 124 km2 (48 miz) in Cheboygan county as the open study site (Figure 1).

The area is characterized by moraines, till plains, outwash plains, lake plains and
deltas, which are the results of glaciation and post glacial lakes. The dominant soil
associations in the PRCSF are Cheboygan-Blue Lake, Au gres-Rubicon Roscommon, and
Detour-Grevourt. These soils range ﬁ'om low fertility dry sands on the outwash plains to
medium-high fertility sandy [cams on till plains and moraines (Tardy 1991). The Black,
Sturgeon and Pigeon Rivers originate on the southern edge of the area and ﬂow
northward.

The average winter temperature is -7.9 C, with an average daily minimum of -10.5

C. The summer average is 17.3 C, with an average daily maximum of 25.8 C. Total

/

Cheboygan
County

 

 

 

County

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1. Location of the site open to hunting in Cheboygan County and the site
closed to hunting in Otsego County, in the northern lower peninsula of Michigan.

annual precipitation is 77.4 cm, and average seasonal snowfall is 291.1 cm (Tardy 1991).
A diversity of vegetation types are present in the PRCSF. The lowland areas are
dominated by white cedar (Thuja occidentalis), Balsam ﬁr (A bies balsamea), and alder
(Alnus spp.). Uplands consist primarily of sugar maple (Acer saccarinum), jack pine
(Pinus banksiana), quaking aspen (Populus tremuloides), bigtooth aspen (Populus

grandiadentata), white pine (Pinus strobus) and red pine (Pinus resinosa).

METHODS

Trapping and collaring

Project ﬁeld work began in May of 1993 and continued through May of 1995.
Trapping and collaring took place in the fall of 1993 and 1994. In 1993, in the site open
to hunting, trapping began on August 1 and continued through September 14 (the day
before the start of grouse hunting season). In the site closed to hunting, trapping
continued through October 2, in an attempt to collar an equal number of birds between
sites. In 1994, birds were trapped between August 7 and October 4 in the open site, and
through October 17 in the closed site. Initial trapping sites were selected based on bird
sightings by Michigan Department of Natural Resources (MDNR) personnel or project
personnel, or by their apparent quality as grouse habitat. Young aspen stands with dense
understory vegetation were assumed to support the highest density of grouse in summer
and early fall (Edminster 1947). Because of frequent grouse ﬂushes, aspen stands in
conjunction with alder and cedar swamp edges were generally selected for trapping.

Modiﬁed Cloverleaf traps (Domey and Mattison 1956) were placed in selected
forest stands with an attempt to orient the traps at right angles to suspected bird
movements. Each trap consisted of 4 trap bodies made of 2" x 4" welded wire and 2 50'
leads made of 18" tall chicken wire. The trap bodies were covered with 3' x 3' lids, and

the lids were camouﬂaged with ferns and other vegetation. The forest ﬂoor was

cleared to approximately 6" on each side of the lead to facilitate bird movements toward
the trap bodies. On the door of the trap body, chicken wire was shaped into a funnel, so
that grouse could enter the trap body but not exit.

All trapping and handling procedures were approved by the Michigan State
University Committee on Research Involving Animal Subjects (AUF # 10/93 ~400-03)
before the start of the project. Grouse were sexed according to tail and rump feather size
and coloration, as described by Domey (1966), Roussel and Ouellet (1975), and Larsen
and Taber (1980). Juveniles were identiﬁed by size during the month of August, and by
wear on the 9th and 10th primary feathers (Godin 1960) after September 1. Weight and
wing length were measured as an index of general condition. Birds weighing over 350 g
were ﬁtted with a numbered leg band, and a necklace-style radio collar which weighed
approximately 11 g. Collars were equipped with an 8-hour mortality sensor, and were
obtained from either Advanced Telemetry Systems Inc., Isanti, MN, or Lotek Engineering
Inc., Ontario, Canada (inclusion of company name does not imply endorsement).

Birds that weighed less than 350 g, or had trapping injuries around the head and
neck, were banded and released immediately. Birds that had injuries which appeared life-
threatening, or interfered with their ability to walk, were euthanized by vertebral
separation. Any animals other than ruffed grouse that were caught during trapping were
released immediately in the safest possible manner. Radio-collared birds were located and
ﬂushed the day after trapping to ensure that there were no problems related to trapping or

improper radio ﬁtting.

10

Vegetation sampling

Vegetation sampling was undertaken on 2 levels. The ﬁrst level compared the
percentages of different habitat types in the open and closed sites in the PRC SF. There
are 26 different cover types which the MDNR uses to classify state land. For sampling
purposes, speciﬁc types were combined into more general categories. In addition, aspen
was split into three age classes: 1 - 10 years, 11 - 29 years, and 30 + years. This resulted
in a total of 9 different cover types, shown in Table 1. Compartment maps and forest
inventories were used to calculate the percentage of each forest type in each site. Private
inholdings are not included in the breakdown of habitat types. They comprise
approximately 3 % of the open site, and 7 % of the closed site.

The second level of vegetation sampling quantiﬁed habitat quality for ruffed grouse
in both sites of the PRC SF, based upon the requisites described in the Habitat Model for
Ruffed Grouse in Michigan (Hammill and Moran 1986). Variables sampled correspond to
habitat requirements for fall-winter cover and winter food, which are considered limiting
factors in the northern range of ruffed grouse.

The fall-winter cover requirement in the model is related to 6 variables: stem
density of deciduous, conifer, and shrub stems; height of deciduous and shrub stems; and
low branch height of conifer stems. Between 2 and 22 stands in each habitat type in each
site were randomly chosen for sampling (Table 2). Young conifer stands were very rare in
the open site, resulting in only 2 stands being sampled. Because of standing water and
inaccessibility, lowland conifers and lowland hardwoods were sampled with only medium

intensity (4 - 8 stands). The category “Other” was not sampled, as it was assumed

11

Table 1. Michigan Department of Natural Resources cover type components used for each
sampling category in 1994 and 1995 vegetation sampling in the PRC SF.

 

 

Sampling category (Type code) Cover type
Aspen, l - 10 years (A) Aspen
Aspen, 11 - 29 years (A) Aspen
Aspen, 30 + years (A) Aspen
Upland hardwoods (B) Paper birch, (M) Northern hardwood, (0) Oak, (P)
Balsam poplar
Lowland hardwoods (E) Swamp hardwoods, (L) Lowland brush, (N) Marsh

Conifers, 1 - 30 years
Conifers, 31+ years
Lowland conifers

Jack pine
Other

(H) Hemlock, (R) Red pine, (W) White pine

(H) Hemlock, (R) Red pine, (W) White pine

(C) Cedar, (D) Treed bog, (F) Spruce - ﬁr, (Q) Mixed
swamp conifer, (S) Black spruce, (T) Tamarack, (V) Bog
or muskeg

(J) Jack pine

(G) Grass, (I) Local use, (K) Rock, (U) Upland brush,
(X) Non-stocked, (Y) Sand dunes, (Z) Water

 

12

Table 2. Number of stands sampled in each site and plot size used for each sampling
category in 1994 and 1995 vegetation sampling in the PRCSF.

 

 

 

Number of stands sampled Transect size

Sampling category Open site Closed site (m)
Aspen, l - 10 years 19 16 2x25
Aspen, 11 - 29 years 18 22 2 x 25
Aspen, 30 + years 21 18 10 x 25
Upland hardwoods 9 10 10 x 25
Lowland hardwoods 4 4 2 x 25
Conifers, l - 30 years 2 8 4 x 25
Conifers, 31+ years 10 10 10 x 25
Lowland conifers 4 8 4 x 25
Jack pine 7 10 4 x 25
Other’ - - -
Total 94 106

 

‘This category was not sampled.

13

that stem density was 0 in types such as grass, water, and rock. Because of its importance
as grouse habitat in the northern United States (Bump et al. 1947, Gullion 1970), aspen
was sampled with more intensity than other habitat types.

Within each stand, variables were quantiﬁed in 3 randomly placed belt transects of
2m, 4m or 10m x 25m in size (Table 1). Size of the plot varied depending on the density
of the habitat type, with dense types like young aspen having smaller plots than sparse
types like mature red pine.

Within each plot, stem density, average height of shrubs and trees, and average
low branch height of conifers were measured. Any woody vegetation greater than 0.9 m
tall was counted, and separated into the categories deciduous, coniferous, or shrub stems
as called for by Cade and Sousa (1985). An equivalent stem density (ESD) was calculated

for each stand using the formula:

ESD = d + 4c + 0.505

where d = deciduous stems/ha, c = coniferous stems/ha and s = shrub stems/ha (Hammill
and Moran 1986). Ten deciduous trees and 10 shrubs (by model deﬁnition) were
randomly chosen, and the height of each was measured using a meter stick or a Haga
altimeter. Ten conifer trees were also chosen and measured for lowest branch height.
These variables are considered important for fall-spring cover, inﬂuencing predation by
both mammals and avian predators.

Although not called for in the model, the woody stem category was further broken

down into percent aspen in each of three size classes: seedlings, saplings and mature

14

trees. Size class was determined using diameter at breast height (dbh): seedlings < 9 cm
dbh, saplings 9 - 18 cm dbh, mature trees > 18 cm dbh. This information was used to
ﬁrrther analyze grouse habitat, speciﬁcally in aspen stands, beyond the sc0pe of the HSI
model.

To satisfy the winter food requirement of the habitat model, a stand must contain
any canopy cover of mature aspen. If it does not, then quality decreases as the distance to
the nearest stand with an aspen canopy increases. In stands that lacked a winter food
source, the distance from the center and edge of the stand to the nearest winter food
source was measured on MDNR compartment maps.

Values for each variable in each stand were entered into the production functions
of the model (Appendix I, Figure 1- 5), and an overall habitat suitability index (HSI) value

was calculated for the stand, using the formula:

HSI = V1(d(V2)+4C(V3)+0.58(V4)) V5

d+4c+.5s

 

where V1 corresponds to ESD index value (Appendix I, Figure 1); V2 corresponds to
deciduous tree height index value (Appendix I, Figure 2); V3 corresponds to conifer
branch height index value (Appendix I, Figure 3); V4 corresponds to shrub height index
value (Appendix I, Figure 4); and V5 corresponds to distance between life requisites index

value (Appendix I, Figure 5).

15

Survival and habitat use

Radio telemetry was used to monitor the survival of grouse from late summer
through the following spring on each of the study sites. Radios were equipped with 8
hour mortality sensors, and birds were collected as soon as the mortality sensor was
known to be active. Cause of death was determined whenever possible. Predator signs
and habitat type were noted, and carcasses were collected for necropsy at the MDNR's
Rose Lake facility. Mortality causes were categorized as avian predation, mammalian
predation, hunting, stress/suffocation, illness, or unknown. Birds which survived for more
than ﬁve days after capture were used for survival analysis.

According to White and Garret (1990), transmitter failure or the inability to locate
an individual animal causes the survival time of that animal to be censored. Ifa bird was
unable to be located aﬁer an intense ground and air search, it was considered “censor ”.
If a bird died of stress or suffocation due to the radio-collar after 5 days aﬁer capture, it
also was considered censored in the survival analysis, because the natural fate of the bird
(if it had not been caught and collared) cannot be determined.

Radio locations for each bird were taken approximately every other day ﬁ'om date
of capture through the end of December. Bird locations were estimated on topographic
maps using two compass bearings from known locations, which were usually points on
county roads. The radios used in this study had a fairly strong signal up to 2 km away.
For accuracy, however, bearings were generally taken from less than 1 km from the
location of the bird. Field trials by research personnel indicated an average bearing error

of approximately 2°, which generally resulted in an error polygon of < 3 ha. The average

16

size of a forested stand in the PRCSF is 10 hectares, so the error associated with habitat
evaluation was assumed to be negligible. Universal Transverse Mercator (UTM)
coordinates were transcribed to MDNR vegetation maps and compartment number, and
the stand number and habitat type occupied by the bird were recorded. Ifa location fell
on a stand line between two habitat types, the equivalent of half of one location was
assigned to each habitat type. If a bird moved onto private land, Michigan Inventory
Resource Information System (MIRIS) maps were used to determine the habitat type
being occupied. The time of location for each bird was varied daily to provide maximum
habitat use information. Ifa bird was located 10 or more times during the fall and winter,

it was used as part of the habitat use study.

Movements

Telemetry data were also used to quantify movements of birds throughout the fall
and winter. Since birds were located for a variable number of days per week, a weekly
harmonic center (Dixon and Chapman 1980) was calculated for each bird using TELEM88
(Virginia Polytechnic Institute and State University, Dept. of Fisheries and Wildlife,
Blacksburg, VA). If a bird was located during at least 5 different weeks between time of
capture and December 31, then it was analyzed for fall movements. Weekly harmonic
centers were plotted and numbered by week as a visual representation of seasonal
movement. The straight line distance between the harmonic center of each week and the
harmonic center of the ﬁrst week located was used as a graphical analysis of movement.

These graphs, listed in Appendix II, were then used to determine whether or not

17

birds exhibited dispersal movements. Birds which never moved more than 500 m from
the ﬁrst week after capture were categorized as Type I non-dispersers (Appendix 11,
Figure l and 2). Birds which moved more than 500 m from the harmonic center of the
ﬁrst week after capture for l or 2 weeks at a time, but then returned to within 500 m,
were categorized as Type H non-dispersers (Appendix H, Figure 3 and 4). Fall/winter
home ranges were calculated for Type I and Type II non-dispersers using the convex
polygon (Mohr 1947) and harmonic mean (Dixon and Chapman 1980) methods.

Birds which moved more than 500 m from the harmonic center of the ﬁrst week
after capture and stayed 500 m or farther away for 3 or more consecutive weeks before
November 30 were considered dispersers, or Type HI (Appendix H, Figure 5 and 6).
Dispersal distance was considered to be the linear distance between the center of those
locations before dispersal to the center of those locations after dispersal. Because of the
time of year, some birds may have been caught during dispersal, causing an
underestimation of dispersal distance. Ifa bird was caught during dispersal, dispersal
distance was considered to be the linear distance between the ﬁrst location and the center
of those locations after dispersal. Some birds that were caught late in the season may have
ﬁnished dispersing before being radio-collared, which would result in some early
dispersers being miscategorized as non-dispersers. Therefore, the number of dispersers

and the distance of dispersal should be considered a conservative estimate.

18

Data analysis

Homogeneity of habitat types between sites (open and closed to hunting) was
compared using a Chi-squared test (Siegel and Castellan 1988). Using the Habitat Model
for Ruifed Grouse in Michigan (Hammill and Moran 1986), the quality of habitat was
evaluated on a stand by stand basis. Mean HSI values for each habitat type were
compared between areas using a Student’s t-test (Rosner 1990) The quality of habitat
within the sites was extrapolated by multiplying the mean HSI value for each habitat type
to the proportion of that habitat type in each site.

Grouse locations through fall and early winter were used to quantify habitat use,
which was compared to habitat availability using PREFER (Great Lakes Fishery
Laboratory, US. Fish and Wildlife Service, Ann Arbor, Michigan). A habitat use value
was calculated for each bird by multiplying the proportion of that bird’s locations in each
habitat type by the mean HSI value for that type. A mean habitat use value was calculated
for each site, and sites were compared using a Kruskal-Wallis one-way ANOVA and
Multiple Comparisons (Siegel and Castellan 1988). Fall and winter home range and
dispersal movements in 1993 and 1994 were compared between sites using an Student’s t-
test (Rosner 1990). Comparisons were also made between sex and age class, and
dispersal type.

Survival probabilities for the radioed birds on each site were calculated using the
Kaplan-Meier Product Limit estimator. This technique provides an estimate of the
survival probability at any point in the study period (Kaplan and Meier 195 8). Survival
probability curves were also generated for birds by age class and sex, and curves were

compared using the Log-rank test (Lee 1992) provided by the program SAS (SAS

19

Institute, Cary, North Carolina). Selected variables were compared to survival in both the
open and closed areas with the Cox proportional hazards model (Cox 1972), also using
the SAS system. Percent of increase or decrease in the hazard of a particular covariate
was determined by converting the risk ratio in the following manner; percent difference in

hazard = 100*(risk ratio-1).

RESULTS AND DISCUSSION

Trapping

In 1993, the most productive weeks in the open site were the 6th and 7th, which
had capture rates of 0.050 and 0.056 birds per trap night (Table 3). A total of 47 birds
(including recaptures) were caught, of which 36 individual birds were radio-collared. Two
birds were banded only (not radio-collared) due to trapping injuries, 2 were euthanized
and 2 were released without being processed. In addition, there were 5 recaptures of birds
which had already been radio-collared or banded.

In the closed area, week 8 was the most successful, With a rate of 0.066 birds per
trap night (Table 3). Thirty-three birds were caught in the closed site during 1993, of
which 28 individuals were radio-collared. Of those that were not collared, 2 were found
dead in traps, killed by unknown predators. Two birds that sustained trapping injuries
were banded only, and 1 bird was a recapture. During week 9 no birds were caught in
124 trap nights, possibly due to cold, rainy weather. In both years and both sites, daily
trapping success was generally low if rain was present before early evening. Backs et al.
(1985) found that week to week trapping success seemed to be related to prevailing
weather conditions in Indiana.

Late September to mid-October is the peak of fall dispersal for grouse in

Wisconsin (Small and Rusch 1989). In 1994, in the open site, trapping continued into

20

21

Table 3. Trapping results for 1993, PRCSF open and closed sites. Birds caught includes
recaptures and birds not collared.

 

 

 

Open site Closed site
Week Dates Trap Birds Birds per Trap Birds Birds per

nights caught night nights caught night

1 8/1 - 8/7 91 4 0.04 92 1 0.01
2 8/8 - 8/14 204 3 0.02 161 2 0.01
3 8/15 - 8/21 188 6 0.03 177 2 0.01
4 8/22 - 8/28 182 8 0.04 189 3 0.02
5 . 8/29 - 9/4 242 9 0.04 105 5 0.05
6 9/5 - 9/11 260 13 0.05 105 1 0.01
7 9/12 - 9/18 72 4 0.06 87 4 0.05
8 9/ 19 - 9/25 226 15 0.07
9 9/26 - 10/2 124 0 0.00
Total 1239 47" 0.04 1266 33" 0.03

 

'Number includes 5 recaptures, 2 birds which were handed only, 2 which were euthanized,
and 2 which were released without being processed.

I’Number includes 1 recapture, 2 birds which were banded only, and 2 which were found
dead in the trap.

22

hunting season so that we might make use of the increased movements of fall dispersal for
trapping. Trapping did not seem to cause any problems with hunters, and vandalism was
minimal. The longer trapping period and the addition of more traps resulted in more trap
nights in both sites in 1994. Weeks 5, 6 and 8 were the most productive in the open site,
and a total of 63 birds were captured during the trapping season (Table 4). Of these, 1
was found dead in the trap, l was accidentally released, 7 were recaptures and 7 were
banded only. Two of the recaptures were banded only. Researchers removed the collar of
one bird that was recaptured with injuries, but the bird was recaptured and recollared the
following week. This resulted in a total of 47 birds collared in the open area in 1994.
Weeks 4 and 11 were the most productive weeks in the closed site in 1994, with
success rates of 0.051 birds per trap night (Table 4). In week 5, there were no birds
caught in 119 trap nights, although the same week was fairly productive in the open site.
A total of 52 birds were captured, and 42 individuals were collared. One bird was found
dead in a trap, killed and eaten by a northern goshawk (A ccipiter gentilis) which was still
in the trap when researchers arrived. Four birds were banded only, and 5 were recaptures.
Success rate was higher in the open site in both years. In general, more young

aspen areas which were productive for trapping were found in the open site. The total
number of birds per trap night did not change much within sites between 1993 and 1994,
despite MDNR and hunter reports of higher population numbers in 1994. Overall success
rate in both sites was substantially lower than that reported by other researchers. Domey
and Mattison (1956) reported success rates as high as 0.768 grouse per trap night in a

small area of high grouse density. More comparable trapping efforts in Indiana

23

Table 4. Trapping results for 1994, PRCSF open and closed sites. Birds caught includes
recaptures and birds not collared.

 

 

 

Open site Closed site
Week Dates Trap Birds Birds per Trap Birds Birds per
nights caught night nights caught night

1 8/1 - 8/7 6 0 0.00 23 1 0.04
2 8/8 - 8/14 127 3 0.02 193 6 0.03
3 8/15 - 8/21 208 8 0.04 196 1 0.01
4 8/22 - 8/28 216 1 0.01 176 9 0.05
5 8/29 - 9/4 237 12 0.05 119 0 0.00
6 9/5 - 9/11 243 14 0.06 119 3 0.03
7 9/12 - 9/18 232 10 0.04 119 3 0.03
8 9/19 - 9/25 202 12 0.06 171 6 0.04
9 9/26 - 10/2 125 3 0.02 221 7 0.03
10 10/3 - 10/9 36 0 0.00 226 7 0.03
11 10/10 -10/16 0 0 176 9 0.05
12 10/17 -10/23 0 0 16 0 0.00
Total 1632 63’ 0.04 1755 52" 0.03

 

'Number includes 7 recaptures, 7 birds which were banded only, 1 which was found dead
in the trap, and 1 which was released without being processed.

l’Number includes 5 recaptures, 4 birds which were banded only, and 1 which was found
dead in the trap.

24

between 1975 - 1984 had a mean annual success rate of 0.062 grouse per trap night
(Backs et al 1985). The 2 year mean success rate in this study was 0.038 in the open site
and 0.028 in the closed site. These years, however, were assumed to be at the lowest
point of the population cycle in Michigan (John Urbain, MDNR pers. comm). Over a
10-year period, this number might increase with the inclusion of years with high
population numbers.

The age and sex ratios of radio-collared birds did not differ signiﬁcantly between
sites in either year (p > 0.10) (Table 5). In 1993 in both sites, and in 1994 in the closed,
there were slightly more adult birds radio-collared than juveniles. These results are not
consistent with prior research which showed that juveniles are more susceptible to
trapping with cloverleaf traps (DeStefano and Rusch 1986). The 2” x 4” welded wire that
was used in making trap bodies for this study may have resulted in juvenile birds escaping
early in the trapping season. It is also possible that some juveniles were mistaken for
adults, due to the difﬁculty in aging by wing molt after 1 September (Bump et al. 1947).
With the exception of the open site in 1994, the ratio of males to females approximated
1:1.

Incidental captures for both years included rabbits, raccoons, a juvenile porcupine,

an opossum and a skunk. All were released safely.

25

Table 5. Number of birds radio-collared in the PRCSF by site, year, sex and age class. No
signiﬁcant differences in age or sex distributions were detected between sites or years
(Chi-square, p > 0.10).

 

 

 

Open site Closed site
Sex Age class‘ 1993 1994 1993 1994

Female AHY 9 8 7 8
HY 10 9 7 10

UNK 0 1 0 2
Total Female 19 18 14 20
Male AHY 11 13 11 12
HY 6 11 3 4

UNK 0 l 0 3
Total Male 17 25 14 19
Unknown AHY 0 1 0 2
HY 0 3 O 1
Total birds 36 47 28 42

 

’Age class abbreviations are as follows: AHY = adult (after hatch year), HY = juvenile
(hatch year), UNK = unknown (could not be determined from feather characteristics).

26

Habitat

Before European settlement, Michigan forests were probably a mosaic of pines and
Northern hardwoods, interspersed with various sized areas of natural disturbance. These
areas, ﬁlled with thick regenerating vegetation, were probably home to a variety of wildlife
which depended on early successional species for food and cover. With the logging and
subsequent burning of much of the state (including the PRCSF) in the beginning of this
century, virtually all of the older forest types were converted to early successional species
such as aspen (Leatherbeny and Spencer 1996).

Bailey et al. (195 5) was perhaps the ﬁrst to note the similarity between the ranges
of ruffed grouse and trembling aspen in North America. Over the northern part of that
range, where snow cover persists for 4 - 6 months, aspen is an important part of the
winter diet of grouse (Bump et al. 1947). Since the 1930’s, total acreage of this
vegetation type has been on a steady decline in Michigan. This decrease has been
primarily due to the deterioration of aspen stands and natural succession of other forest
types, predominantly hardwoods (Raile and Smith 1983).

The statistical distribution of habitat types in the open site is typical of this current
trend (Table 6). In the open site, upland hardwoods predominate, comprising about 24%
of the total area. Species include maple, birch (Betula papynfera), poplar (Populus
balsamifera) and a small amount of oak (Quercus spp.). Aspen is well represented as a
species when age classes are combined (34%). However, the structural conﬁguration of
aspen changes quickly and dramatically over time, as does its quality as grouse habitat.
According to Harnmill and Visser (1984), the most important age class of aspen for

grouse through fall and winter is < 25 years, but sapling aged aspen (1 - 10 years)

Table 6. Distribution of habitat types in the Pigeon River Country State Forest open and

closed sites.

27

 

 

 

Open site’ Closed site
Habitat type Area Percent of Area Percent of

(ha) total (ha) total
Aspen 1-10 yr. 1135 9.38 642 7.62
Aspen 11-29 yr. 1862 15.38 853 10.14
Aspen 30+ yr. 1061 8.76 590 7.02
Upland hardwoods 2861 23.63 1652 19.62
Lowland hardwoods 960 7.93 375 4.46
Pine 1-30 yr. 47 .39 284 3.37
Pine 31+ yr. 1439 11.89 1764 20.96
Lowland conifers 1774 14.65 1464 17.40
Jack pine 200 1.65 470 5.59
Other 767 6.34 322 3.83
Total area 12106 8416

 

'The distribution of habitat types in the open site is signiﬁcantly different from the

distribution of habitat types in the closed site (Chi-square, 9 df, p < .001).

28

is often too dense (Gullion 1970). Eliminating the classes which are considered too old
and too young, the middle and most important age class of aspen for grouse (1 1-29 years
old) represents about 15 % of the total area of the open site. There is very little jack pine
(2%) or young pine (< 1%) in the open site, because of little or no planting of these types.

The distribution of habitat types in the closed site is signiﬁcantly different than in
the open site (p < 0.001) (Table 6). The primary difference between the sites lies in the
dominant species. Like the open site, the closed site does have a high percentage of
hardwoods (20%), but also has about an equal percentage of pine > 30 years of age
(21%). The older pine category is partially comprised of red pine or mixed pine
plantations, which have been historically planted as a timber resource in Michigan. Soil
types in the closed site may also be conducive to the succession of pine types, which are
harvested in a 60 - 70 year rotation (Joseph Jarecki, MDNR, pers. com). The
combined aspen age classes equal approximately 25% of the closed site, slightly less than
the open site. Percentages of young pine and jack pine are slightly higher in the closed
site, most likely due to planting.

The open site is fairly homogeneous with regard to the distribution of habitat types
across the landscape. The closed site, however, has a distinct shift in habitat types ﬁom
the northwest the southeast. This shift is hidden in the overall distribution (Table 6).
Birds that were trapped in the closed site tended to stay in either the northwest or the
southeast. Because of this heterogeneity, the closed site was separated on compartment
lines into 2 areas, referred to as the North area and the South area (division is shown on
maps in Appendix HI, Figure 5 - 8). The distribution of habitat types differs signiﬁcantly

between the North and South areas of the closed site (p < 0.001) (Table 7).

29

The North area is 35% pine over 30 years of age, with very few aspen dominated
stands. Much of the older pine that is present in the closed site is in the North area. The
middle age class of aspen comprises just 7% of the area, and only 11% of the area is aspen
of any age class. If clear-cutting for the regeneration of aspen stands continues at the
current rate, there will be even fewer hectares of forest in the middle age class in 10 to 20
years. Aspen 1 - 10 years of age makes up just 2% of the North area of the closed site.
Assuming the importance of the middle age class of aspen for grouse, it may be speculated
that the amount of quality habitat is on a decline.

The South area is similar to the open site in distribution of aspen types, but is still
signiﬁcantly diﬁ‘erent in overall distribution (p < 0.001). About 36% of the area is
comprised of aspen of combined age classes. Twelve percent of the South area is aspen 1
- 10 years old, indicating a high level of clearcutting in recent years. The middle age class
of aspen makes up 13% of the area, about twice the percentage of the North area. Like
the open site, only 10% of the South area consists of pine over 30 years of age. The
South area diﬂ‘ers ﬁom both the North area and the open site in its large percentage of
lowland conifers (22%).

The amount of aspen present has been and will continue to be important for ruﬁ‘ed
grouse in northern Michigan and the PRCSF. But although it is important to describe

habitat distributions, quality of each habitat type must further be evaluated.

30

Table 7. Distribution of habitat types in the North and South areas of the Pigeon River
Country State Forest closed site.

 

 

 

North area‘ South area
Habitat type Area Percent of Area Percent of

(ha) total (ha) total
Aspen 1-10 yr. 77 2.10 564 12.04
Aspen 11-29 yr. 250 6.66 606 12.92
Aspen 30+ yr. 75 2.04 515 11.00
Upland hardwoods 782 20.93 869 18.58
Lowland hardwoods 115 3.10 259 5.54
Pine 1-30 yr. 231 6.15 53 1.14
Pine 31+ yr. 1315 35.15 448 9.61
Lowland conifers 441 11.83 1022 21.84
Jack pine 321 8.55 151 3.22
Other 130 3.50 192 4.10
Total area 3737 4679

 

'The distribution of habitat types in the North area is signiﬁcantly different from the
distribution of habitat types in the South area (Chi-square, 9 (If, p < .001).

31

Habitat quality

Habitat quality within each habitat type was generally similar between the open and
closed sites (Table 8). Mean HSI values of habitat types differed signiﬁcantly between
sites in aspen 11-29 years (p < 0.10) and in aspen 30 years and older (p < 0.01). In the
open site, aspen 30 years and older had a relatively high mean HSI value of 0.495. This
value was primarily driven not by the older aspen trees, but by a young, dense understory
of conifers. Although not considered the dominant vegetation, these conifers provided the
density and height requirements of quality grouse habitat. In general, aspen types are most
commonly replaced in succession by species such as red’maple, sugar maple, white pine,
or forest openings, which are not conducive to ruffed grouse production and survival
(Hammill and Visser 1984). In this case, however, the course of succession has not
substantially decreased the value of the habitat.

Medium aged aspen had a mean HSI value of 0.524 in the open site. It was
expected that this habitat type would have a high (or the highest) value of all the types,
due to its documented value as a food and cover source for ruffed grouse (Gullion 1970).
Lowland conifers had the highest value as grouse habitat according to the model (Table
8),with a mean HSI of 0.533. The middle age-class of aspen had a slightly lower mean
HSI value, but it contributed more (0.081) to the overall HSI than the lowland conifer
type (0.078) because of its higher percentage of total area. Hardwoods had the lowest

mean value, 0.055.

32

Table 8. Overall HSI values for the PRCSF open and closed sites. Weighted HSI is the
mean HSI for the habitat type * the percent of area in that habitat type.

 

 

 

Open site Closed site
Habitat type Mean HSI Weighted Mean HSI Weighted

HSI HSI
Aspen l-10 yr. .187 .018 .165 .013
Aspen 11-29 yr. .524‘l .081 .345 .035
Aspen 30+ yr. .495b .043 .161 .011
Upland hardwoods .055 .013 .032 .006
Lowland hardwoods .123 .010 .400 .018
Pine 1-30 yr. .117 .000 .414 .014
Pine 31+ yr. .189 .022 .206 .043
Lowland conifers .533 .078 .348 .061
Jack pine .269 .004 .247 .014
Other .000 .000 .000 .000
Overall HSI .27 .21

 

'Mean HSI for aspen 11-29 years in the open site is signiﬁcantly different ﬁ'om mean HSI
for aspen 11-29 years in the closed site (Student’s t-test, p < .10).

t’Mean HSI for aspen 30 + years in the open site is signiﬁcantly different from mean HSI
for aspen 30 + years in the closed site (Student’s t-test, p < .01).

33

In the closed site, aspen 30 years and older had a lower mean HSI value than the
open site, 0.161. As in the open site, young conifers were present in this habitat type, but
the understory did not provide density and height requirements of the model (Appendix I,
Figure l and 3). Medium aged aspen in the closed site also had a lower mean HSI value
than in the open site, 0.345. Young pine had the highest HSI value of all the habitat types
in the closed site (0.414), but made up the lowest percentage of total area. The weighted
HSI value for lowland conifers had the highest numerical effect (0.061) on the overall HSI
for the closed site. This effect comes ﬁom the high percentage of total area, multiplied by
a fairly high mean HSI value of 0.348.

Overall, the open site had a weighted HSI value of 0.27, which was slightly higher
than the overall value for the closed site (0.21). Values were generally lower than
expected. Only 2 stands in each site had a perfect HSI score of 1.0. Both perfect scores in
the open site were in aspen > 30 years, owing to the large number of young conifers in
those stands. In the closed site, both perfect scores were in aspen 11-29 years, a habitat
type which was expected to be of high value for grouse. Including those which had
perfect scores, 9 stands in the open site and 8 stands in the closed had HSI scores > 0.9.

Although the HSI scale in the model is only a relative scale, it is generally assumed
that a stand with a value over 0. 5 would provide at least marginal habitat. Twenty-ﬁve
stands in the open site (26.6%) and 20 stands in the closed (18.9%) had values of 0.5 or
higher. In the open site, 88% of stands with scores over 0.5 were in aspen types. In the
closed site, only 40% of stands with scores over 0.5 were in aspen types, and all types

except hardwoods had at least one stand with a value over 0.5.

34

Weighted HSI values were also calculated for the North and South areas of the
closed site (Table 9). Although the North and South areas have different vegetation
distributions, the overall area quality is numerically similar. The South area had a slightly
higher overall HSI score than the North area. Pine over 30 years contributed 0.072 to the
overall value of 0.207 in the North area. Young and old age classes of aspen had a very
small effect on the overall value, contributing only 0.003 each. Lowland conifers added
signiﬁcantly more, at 0.041. In the South area, lowland conifers contributed 0.076 to the
overall value of 0.220. Medium aged aspen added 0.045, and young and old aspen added
more than in the North area (0.020 and 0.018).

To investigate the relationship between model variables and HSI values, Pearson
correlation coefﬁcients and related P-values were calculated. The interspersion (winter
food) variable was not used in this analysis because nearly all stands contained the
minimum requirements to receive an index value of 1.0 (Appendix I, Figure 5).
Equivalent stem density was signiﬁcantly correlated with HSI value in 7 of 8 habitat types
in the open site (p < 0.10) (Table 10). ESD was not correlated with HSI value in the
lowland hardwoods type, because all 4 stands in that habitat type had very high stern
densities. In young aspen, shrub and deciduous heights were also correlated with HSI
value. At 10 years and younger, the number of aspen stems are usually above minimum
requirements; the height of stems to provide cover is the limiting factor. Young pine was

not included in the analysis because of low sample size.

35

Table 9. Overall HSI values for the North and South areas of the Pigeon River Country
State Forest closed site. Weighted HSI is the mean HSI for the habitat type * the percent
of area in that habitat type.

 

Habitat type North area South area
Weighted HSI Weighted HSI

Aspen 1-10 yr. .003 .020
Aspen 11-29 yr. .023 .045
Aspen 30+ yr. .003 .018
Upland hardwoods .007 .006
Lowland hardwoods .012 .022
Pine 1-30 yr. .025 .005
Pine 31+ yr. .072 .020
Lowland conifers .041 .076
Jack pine .021 .008
Other .000 .000

Overall HSI .207 .220

 

36

Table 10. Pearson correlation coefﬁcients (r) and associated probabilities (P) for individual
variables versus HSI values in each habitat type in the open site of the PRCSF.

 

 

 

ESD‘ SHR. HT. DEC. HT. CON. HT.

Habitat type r P r P r P r P
Aspen 1-10 yr. .49 .033 .57 .027 .78 .0001 -.41 n.s.b
Aspen 11-29 yr. .55 .02 -.01 n.s. .09 n.s. -.01 n.s.
Aspen 30 + yr. .75 .0005 -.21 n.s. .38 n.s. -.39 n.s.
Hardwoods .74 .02 -.45 n.s. -.51 n.s. -.88 n.s.
Lowland hardwoods .07 n.s. .35 n.s. .72 n.s. - ° -
Pine 1-30 yr. - - - - - - - -
Pine 31 + yr. .91 .0002 .48 n.s. .25 n.s. -.68 .03
Lowland conifers .99 .004 .88 n.s. -.59 n.s. -.99 .001
Jack pine .93 .003 -.25 n.s. .32 n.s. -.63 n.s.

 

'Variable abbreviations are as follows: ESD = equivalent stem density, SHR. HT. = height
of deciduous shrubs, DEC. HT. = height of deciduous trees, and CON. HT. = low branch
height of conifers.

l’Not signiﬁcant (P > 0.10)
Not enough data for analysis

Table 11. Pearson correlation coefficients (r) and associated probabilities (P) for individual
variables versus HSI values in each habitat type in the closed site of the PRC SF.

 

 

 

ESD’ SHR. HT. DEC. HT. CON. HT.

Habitat type r P r P r P r P
Aspen 1-10 yr. -03 n.s. " .95 .0001 .86 .0001 -.07 n.s.
Aspen 11-29 yr. .79 .0001 .07 n.s. .33 n.s. -.19 n.s.
Aspen 30 + yr. .85 .0001 .06 n.s. -.05 n.s. -.18 n.s.
Hardwood .87 .001 -.55 n.s. -.24 n.s. -.69 n.s.
Lowland hardwoods .87 n.s. .97 n.s. -.75 n.s. -° -
Pine 1-30 yr. .87 .005 .63 n.s. .58 n.s. .07 n.s.
Pine 31 + years .97 .0001 .70 .04 .26 n.s. -.44 n.s.
Low conifer .57 n.s. .79 .04 -.37 n.s. -.92 .001
Jack pine .76 .01 -.05 n.s. .02 n.s. -.82 .004

 

'Variable abbreviations are as follows: ESD = equivalent stem density, SIR. HT. = height
of deciduous shrubs, DEC. HT. = height of deciduous trees, and CON. HT. = low branch
height of conifers.

I’Not signiﬁcant (P > 0.10)

Not enough data for analysis

37

Equivalent stem density was signiﬁcantly correlated with HSI values in 6 of 9
habitat types in the closed site (p < 0.10) (Table 11). Again, ESD was not correlated with
HSI in lowland hardwoods. ESD was also not correlated with HSI in aspen 1-10 years or
in lowland conifers. In the young aspen type, shrub height and deciduous height were
correlated with HSI values (p < 0.001).

Because of its weight in the HSI equation, it was expected that the ESD variable
would be highly correlated with ﬁnal HSI value. The Michigan model (Hammill and
Moran 1986) suggests an optimum equivalent stem density of 12085 or greater stems per
hectare (Appendix I, Figure 1). Below that density, the value of the stand decreases
linearly to 0 at about 5000 stems per hectare. Lowland conifers had the highest stem
density of all habitat types in both sites (Table 12, Table 13). In the open site, young
aspen and lowland conifers had mean equivalent stem densities greater than the optimum
of 12085 stems per hectare called for in the Michigan model.. In the closed site, young
pine, young aspen and lowland hardwoods had mean ESDs greater than the optimum. In
both sites, hardwoods have very low mean equivalent stem densities (< 5000 stems/ha).
The largest discrepancy in stem densities between the sites was in young pine stands,
which were more dense in the closed site (14288 stems/ha) than in the open site (53 75
stems/ha).

Because of the weighting process, looking at straight stem densities did not give an
accurate representation of how much each variable contributed to the ESD. The
percentage that each weighted component contributed to the ESD was calculated for each
habitat type to get a better picture of which component was driving the association. In the

young aspen type, deciduous stems (including aspen) accounted for 75.9% of the mean

38

Table 12. Mean deciduous, conifer and shrub stems/ha (SE), and equivalent stem density
for each habitat type sampled in the open site of the PRC SF.

 

 

Deciduous Conifer Shrub Equivalent
Habitat type Stems/ha Stems/ha Stems/ha stem density
Aspen 1-10 yr. 9585 (1212) 498 (171) 2102 (587) 12629 (1616)
Aspen 11-29 yr. 3981 (499) 1519 (317) 2062 (1036) 1 1086 (1614)
Aspen 30+ yr. 2570 (366) 2154 (490) 939 (367) 1 1656 (1826)
Upland hardwoods 4033 (721) 25 (21) 162 (65) 4214 (780)
Lowland hardwoods 2667 (1085) 333 (333) 19067 (10919) 13533 (4345)
Pine 1-30 yr. 1767 (1500) 900 (233) 17 (17) 5375 (2442)
Pine 31+ yr. 1 177 (295) 1404 (220) 805 (468) 7196 (73 8)
Lowland conifers 992 (147) 7242 (2548) 2350 (2162) 31133 (10628)
Jack pine 2681 (944) 1443 (313) 124 (31) 8514 (1219)

 

Table 13. Mean deciduous, conifer and shrub stems/ha (SE), and equivalent stem density
for each habitat type sampled in the closed site of the PRCSF.

 

 

Deciduous Conifer Shrub Equivalent
Habitat type Stems/ha Stems/ha Stems/ha stem density
Aspen 1-10 yr. 10146 (1216) 400 (86) 2958 (1006) 13225 (1449)
Aspen 1 1-29 yr. 4535 (446) 952 (167) 995 (206) 8841 (630)
Aspen 30+ yr. 1938 (237) 1680 (852) 774 (345) 9045 (3372)
Upland hardwoods 2579 (321) 164 (71) 200 (83) 3335 (542)
Lowland hardwoods 1617 (409) 500 (362) 20167 (12053) 13700 (5607)
Pine 1-30 yr. 1138 (554) 3233 (801) 433 (119) 14288 (3609)
Pine 31+ yr. 1222 (319) 1384 (257) 247 (84) 6882 (1158)
Lowland conifers 721 (172) 3413 (518) 4308 (1056) 16525 (2153)
Jack pine 1290 (819) 2493 (638) 207 (107) 11367 (2460)

 

39

ESD in the open site (Table 14), and 76.7% in the closed (Table 15). In aspen 11-29
years of age, conifers stems (weighted) began to contribute a signiﬁcant amount to the
mean Equivalent Stern Density. In the open site, 54.8% of the mean ESD in middle age
aspen stands were conifer stems, while deciduous stems accounted for only 35.9%.
Conifer stems made up 43.1% of the mean ESD of middle age aspen in the closed site. By
the 30+ age class, conifers provided about 74% of the ESD in both sites. Shrub stems
only contribute substantial numbers to the ESD in one habitat type, lowland hardwoods.

To assess the relationship of stem densities to HSI values even ﬁrrther, correlation
coeﬂicients and associated probabilities were calculated for separate unweighted
components of the equivalent stem density (Table 16, Table 17). Components tested were
stem densities of: aspen of all size classes and other deciduous trees combined; other
deciduous trees; aspen of all size classes; aspen < 9 cm dbh (A1); aspen 9 - 18 cm dbh
(A2); aspen > 18 cm dbh (A3); conifers of all size classes; and shrubs.

In the open site, densities of conifer stems were correlated with HSI values in all
habitat types tested except lowland hardwoods (p < 0.10)'(Table 16). The correlation in
the young aspen type between HSI and the density of large aspen stems is probably an
artifact caused by the low ﬁequency of large aspen stems. Correlations associated with
medium and total aspen stems within the hardwood habitat type are also artifacts caused
by low frequency.

In the closed site, densities of conifer stems were correlated with HSI values in all
habitat types tested except lowland hardwoods and lowland conifers (Table 17). In young

aspen, densities of deciduous stems, total aspen stems, and small and medium aspen stems

40

Table 14. Percent of mean equivalent stem density for deciduous, conifer and shrub stems
for each habitat type sampled in the open site of the PRC SF.

 

% of ESD % of ESD % of ESD

 

Habitat type Deciduous Conifer Shrub
Aspen 1-10 yr. 75.9 15.8 8.3
Aspen 11-29 yr. 35.9 54.8 9.3
Aspen 30+ yr. 22.0 73.9 4.0
Upland hardwoods 95.7 2.4 1.9
Lowland hardwoods 19.7 9.8 70.4
Pine 1-30 yr. 32.9 67.0 0.0
Pine 31+ yr. 16.4 78.0 5.6
Lowland conifers 3.2 93.0 3.8
Jack pine 31.5 67.8 0.7

 

Table 15. Percent of mean equivalent stem density for deciduous, conifer and shrub stems
for each habitat type sampled in the closed site of the PRCSF.

 

% of ESD % of ESD % of ESD

 

Habitat type Deciduous Conifer Shrub
Aspen 1-10 yr. 76.7 21.1 11.2
Aspen 11-29 yr. 51.3 43.1 5.6
Aspen 30+ yr. 21.4 74.3 4.3
Upland hardwoods 77.3 19.7 3.0
Lowland hardwoods 11.8 14.6 73.6
Pine 1-30 yr. 8.0 90.5 1.5
Pine 31+ yr. 17.8 80.4 1.8
Lowland conifers 4.4 82.6 13

Jack pine 11.3 87.7 .9

 

£3.98 new 8% .3565 «oz.

2: .o A e anemia 82..
ﬁe Eu M: A :33 n m< .ﬁe Eu M: - a :33 .1- N< doe So a v cone-a u 3. 5325* an 0.8 223505? 320 oﬂm.

 

 

 

4 .m.: mOf ©0- Vh. .md ~—. .md Mow... .md @N. .m.: FN- .m.€ m0. .md NN. QSQ u—Oﬂh
.2. 3.. 88. 8. .2. Ne. - - a... me- a... 8.- .....e 9..- ad 2.- peace geese-—
.md ON... ~O. mm.- .m-: mo. .md WG-1 .md 0?. .m.: MV. .md 3.1 .md EN. .5» +OM ”in”
- - - - - - - - - - - - - - - .- .3 a; on:
a... a _. .ne em. - - - - 8o. 8.- 8o. 8.- a... 3.- a... 3*.- meeeaea; e553
.2 en. 88. 3. - - - - 8o. em. so. a. 8. ee. 8. 8. 68263:
.md m~. NCO. Oh. .md h—f .md V—f .md O_. .m.: NO. .md ——.r .md N~.r .uhirOM Cam<
.m.: 0— . @OO. mm. .m.: MN.- .m.: 0N. .md —C. .md BO. .md no. .md V— . .5 GNI- §m<
.2 3.- 88. 2.. 88. _ .2. 3.- .2 2..- .2. 8.- .2. om. .4: en. .S 2-. :32

a a a t a t a t a t m a a t a a sea 3%:
commas
seem £28 3 ~< .2 e83 38268 9.8265

 

. mm 93 65 mo 86 some 05 E 25 ESE some 5 82.3 Em «83> Ewe-cen— 88m
Eﬂgicm 2: mo 35:03.8 36:26 88m 8838 Lou Ev 835.235 33688 ES 3 356E006 cote—oboe come—won .3 033.

42

mmmbasw 5.. 8.... .3328 .07.0

8. .o A ..V .5822. .62..

an... .5 w. A 5%.. u 2 ..Bv .5 w. - a 5%.» u N< 23.. So a v eon—8 u 3. 832—0.. 8 03 323323.. 826 35.

 

 

 

.md OO. OOO. Ow. .md me .md FO- .m.C mm. .md 0?. .m.: ONF .md ON... OSQ JON—-
.2. 8. .2. o... .2. t..- .2. 8.- .2. 8. .2. 8.- 2. me. .2. 8. 8.28 .8233
.2. $.- 88. 8. .2. 8.- .2. 8.- .2. w..- .2. S. 8. .e. 8.. 8. 2.. +8 2....
.2. N... 8... 8. - - - - .2. 8.- .2. 8.- .2. .e. .2. 8. a. 8-. 2....
.2. E. .2. 8. - - - - - - - .- .2. 8.- .2. 8.- 883...... 85.36..
8. me. 88. 8. .2. t.- - - 8. 8. 8. 8. .2. .m. 8. 2. 883.5:
.m.: NO. ~0oo. N.”- ..m: 0N. m.= ”NF ..m: ”Of ..m: ”—.I ..m: ~M... ..m: OMV RCA +OM GQQM<
.2. 2. .8... 8. .2. 8. 2. MN. .2. ...- .2. 8.- .2. 8. .2. 8. .5 8-.. :82.
.2. 8.- .8. R. - - 88. N... 8. t..- 8. 3..- So. 8. ..2. 8.- a o.-. 8%....
a . .. . .. . .. . .. . .. . .. . .. . 9.... .53....
.593.
9:8 6......8 2. ..< 5%.... 268.89 .29....8o

 

.59... o... ..o 26 38.0 o... a. cab .85.... no.8 a. mos—s, .mm 2.26., bacon 88m
82823.5 2.. ..o messages 5.2.6.. 808 8338 8.. EV 8.5536... 38.688 ...8 3 25.6568 5.3.2.8 523.. s. 63....-

43

were correlated with HSI values. In pine 30+ years, combined deciduous and aspen stems
and deciduous stems were correlated with HSI values. A correlation exists between
deciduous stems with HSI values in the hardwood type, but correlations of aspen stems
within that type are probably due to the low frequency of those stems.

Because of its historical importance as grouse habitat, aspen stands were ﬁirther
analyzed for stem densities by age. Gullion (1970) provided a range of stem densities
which had signiﬁcant use by ruffed grouse. He suggested that ruffed grouse make little
use of regenerating aspen until the sapling density has thinned to about 19700 stems per
hectare, and that heavy use continues until stem density reaches about 7400 stems per
hectare. Only a few aspen stands under 10 years of age in the open and closed sites had
the optimum value (or greater) in aspen stems alone (Figure 2, Figure 3). When other
species are added to make the equivalent stem density, substantially more stands reach the
optimum. Notice that although the density of aspen stems across ages is consistent
between the open and closed site, ESDs are higher in the open site at ages 10 - 60 years.

It has been argued whether or not conifers contribute to cover for ruffed grouse or
provide cover for predators. Young balsam ﬁr in the understory of mature aspen forests
has been shown to provide cover for drumming grouse in Wisconsin (Kubisiak et al.
1980), and conifers are considered important winter cover for ruffed grouse in New York
State (Bump et al. 1947, Edminster 1947). Yet in the original ruffed grouse model (Cade
and Sousa 1985), on which the Michigan model is based, it is suggested that the presence
of any conifers in an otherwise suitable habitat will reduce suitability of fall to spring cover

by offering concealment for predators. In that model, there is a penalty assigned to the

 

 

 

 

35000 0 o o aspen stems/ha
30000 0 equivalent stem
° '3 densit
g; Y
.3 25000 0 o
J: 20000 ....... 8 ............. 9 ................................ l 9700
3 ° 0 ° stems/ha
3‘ 15000 ° ‘ . ’ 0 12085
E.) 1 - ' '0‘ 75;:‘3'0'6.’ """" stems/ha
"’ 0000 .‘.,.;v.é... ............................... 7400
5000 o . o ‘5 . stems/ha

O

 

0 20 40 60 80 100 120 140
age (years)

Figure 2. Aspen stems per hectare and equivalent stem density in various aged
aspen stands in the open site of the PRCSF. Dashed lines indicate maximum
(19700) and minimum (7400) stems/ha called for by Gullion (1970), and optimum
(12085) stems/ha in the Michigan model (Hammill and Moran 1986).

 

35000
30000

0 aspen stems/ha

 
  
  
  

0 equivalent stem

 

 

 

a.) density

§ 25000

43': 20000 p. ................................................. 19700
K 0 stems/ha
v: 15000 0 12085
E Q o -& ----------------

3 10000 stems/ha

"’ ...... . ...................... 7400

5000 o 08, 0 ‘ stems/ha

  
  

0 20 40 60 80 100 120 140
age (years)

Figure 3. Aspen stems per hectare and equivalent stem density in various aged
aspen stands in the closed site of the PRCSF. Dashed lines indicate maximum
(19700) and minimum (7400) stems/ha called for by Gullion (1970), and optimum
(12085) stems/ha in the Michigan model (Hammill and Moran 1986).

45

HSI value which corresponds to the percentage of conifers in the stand being assessed.

The Michigan model alters this thinking dramatically, and assumes that a conifer
provides valuable cover for ruffed grouse; according to the weighting procedure, 4 times
more valuable than a deciduous tree. If conifers do indeed provide valuable cover, the
important role that they play as grouse cover in the PRC SF cannot be ignored. In most
habitat types in the study area, HSI scores are correlated with conifer density, and without
them, stem densities would not reach near optimum levels. Contrary to the customary
view of young, regenerating aspen stands, most aspen stands in the study between the ages
of 11 - 29 would be of negligible value without the contribution of conifer stems. This
may be a trend of northern Michigan forests, or unique to the circumstances at the

PRCSF.

46

Movements

In 1993, there were 15 birds in the open site and 19 birds in the closed site that
were located during at least 5 different weeks. Four birds in the open site and 5 birds in
the closed site ﬁt the into the category of Type I non-dispersers (Table 18). Type I non-
dispersers had a mean home range of 44.3 ha in the open site, and 49.7 ha in the closed
site in 1993 (home ranges included were calculated using the convex polygon method;
harmonic mean home ranges showed similar trends). Although the closed site had a
slightly higher mean, there was no signiﬁcant difference in home range sizes between sites.

Type I males in the closed site had a signiﬁcantly larger mean home range (72.7
ha) than males in the open site (42.7 ha) (p < 0.10). Closed site males also had a
signiﬁcantly larger mean home range than closed site females (34.5 ha) (p < 0.10). No
other signiﬁcant differences were detected between sites or between sex or age classes.
Because of low sample sizes, age categories were not split by sex.

Four birds in the open and 8 in the closed site were categorized as Type II non-
dispersers (Table 18). Type 11 birds had a mean home range of 148.7 ha in the open site,
and a mean home range of 163.9 ha in the closed site in 1993 (Table 18). In both sites,
Type II juvenile birds had a larger mean home ranges than Type II adults (not signiﬁcantly
different). There were no signiﬁcant differences in mean home range between sites or
between categories (p > 0.10).

In 1993, there were 7 dispersers in the open site (Table 19). Those birds had a
mean net dispersal distance of 940 m. Adults moved signiﬁcantly farther than juveniles (p

< 0.10). The 6 dispersers in the closed site in 1993 moved slightly farther (1068 m)

47

Table 18. Mean home range size (ha) for Type I and Type II non-dispersers in the PRCSF
open and closed sites, 1993.

 

 

 

Open site Closed site
)7 (SE) N 7 (SE) N
Type 1 Total 44.3 (5.0) 4 49.7 (10.0) 5
Male 42.7 (6.8)A‘I 3 72.7 (3.9)Ba 2
Female 48.9 - 1 34.5 (5.8)b 3
Sex unk. - - 0 - - 0
Adult 44.3 (5.0) 4 54.4 (11.3) 4
Juvenile - - 0 30.8 - 1
Age unk. - - O - - 0
Type II Total 148.7 (21.0) 4 163.9 (24.5) 8
Male 177.1 - 1 144.5 (17.3) 7
Female 139.2 (26.5) 3 299.6 - 1
Sex unk. - - 0 - - 0
Adult 135.2 (41.9) 2 131.3 (19.9) 5
Juvenile 162.1 (23.0) 2 218.1 (43.8) 3
Age unk. - - 0 - - 0

 

' Capital letters indicate signiﬁcant difference between sites (Student’s t-test, p < 0.10).
Small letters indicate signiﬁcant difference between sex or age categories (Student’s t-test,
p < 0.10).

Table 19. Mean dispersal distance (m) for ruffed grouse in the PRCSF open and closed
sites, 1993.

 

 

 

Open site Closed site
7((SE) N 7((SE) N
Total 939.5 (103.6) 7 1068.2 (271.2) 6
Male 1058.5 (59.0) 3 730.3 (167.8) 2
Female 850.2 (172.2) 4 1237.2 (388.1) 4
Sex unk. - - O - - 0
Adult 1064.3 (91 .3)a' 5 963.5 (360.3) 4
Juvenile 627.4 (76.5)b 2 1277.7 (508.7) 2
Age unk. - - 0 - - 0

 

'Small letters indicate signiﬁcant difference between sex or age categories (Student’s t-

test, p < 0.10).

48

than those in the open site, but the difference was not signiﬁcant (p > 0.10). Juvenile birds
in the closed site moved a mean distance of 1278 m while juveniles in the open site moved
only 627 m, but again the difference was not signiﬁcant.

In 1994, 28 birds were analyzed for movement patterns in the open site, and 26 in
the closed. There were 4 birds in the open site and 4 in the closed site which ﬁt the
description of Type I non-dispersers (Table 20). All 4 in the closed site and 3 in the open
were male birds. Similar to the 1993 data, Type I birds had a mean home range size of
45.5 ha in the open site and 39.4 ha in the closed site. Sex and age classes could not be
compared due to low sample sizes.

There were 9 Type 11 birds in the open site in 1994 with a mean home range of
106.7 ha (Table 20). Eleven Type 11 birds in the closed site had a mean home range of
105.1 ha. These values were lower than those for 1993. Closed site males had a
signiﬁcantly smaller mean home range size (63.6 ha) than closed site females (114.3 ha)
and open site males (101.5 ha) (p < 0.10). No other signiﬁcant differences were detected.

Unlike 1993 dispersers, the 15 dispersers in the open site in 1994 moved farther
(1784 m) than those in the closed site (1132 m) (Table 21). Females in the open site
moved signiﬁcantly farther than females in the closed site (p < 0.10). In the closed site,
males moved a signiﬁcantly higher net distance (1692 m) than females (65] m) (p < 0.10).
Juveniles in the open site in 1994 moved a farther mean net distance (2275 m) than any
other sex or age category in either year.

In the closed site in 1994, 4 birds exhibited dispersal movements after week 16

(N member-December). These birds were not considered dispersers. Small and Rusch

49

Table 20. Mean home range size (ha) for Type I and Type II non-dispersers in the PRCSF
open and closed sites, 1994.

 

 

 

Open site Closed site
Y (SE) N 3? (SE) N
Type I Total 45.5 (9.4) 4 39.4 (8.3) 4
Male 37.8 (7.4) 3 39.4 (8.3) 4
Female - - O - - 0
Sex unk. 68.9 - 1 - - 0
Adult 25.4 - 1 39.4 (8.3) 4
Juvenile 52.3 (9.3) 3 - - 0
Age unk. - - 0 - - 0
Type II Total 106.7 (16.7) 9 105.1 (16.2) 11
Male 101.5 (14.1)A‘I 4 63.6 (6.5) Ba 2
Female 111.6 (38.1) 4 114.3 (18.4)b 9
Sex unk. 108.4 - l - - 0
Adult 86.4 (1 1.7) 6 92.3 (20.3) 6
Juvenile 147.4 (37.5) 3 91.9 (22.2) 3
Age unk. - - 0 163.2 (51.2) 2

 

' Capital letters indicate signiﬁcant difference between sites(Student’s t-test, p < 0.10).
Small letters indicate signiﬁcant difference between sex or age categories (Student’s t-test,
p < 0.10).

Table 21. Mean dispersal distance (m) for ruﬂ‘ed grouse in the PRCSF open and closed
sites, 1994.

 

 

 

Open site Closed site
7 (SE) N 7((SE) N
Total 1784.3 (366.3) 15 1132.0 (222.1) 11

Male 1660.6 (524.4) 7 1692.4 (413.2)a 4
Female 2039.7 (601 .7)A' 7 650.6 (102.3)Bb 6
Sex unk. 861.7 - 1 1778.8 - 1
Adult 1538.3 (403.2) 9 1109.9 (347.9) 6
Juvenile 2275.4 (856.0) 5 959. 1 (288.2) 4
Age unk. 861.7 - 1 1955.8 - 1

 

'Capital letters indicate siginifcant difference between sites (Student’s t-test, p < 0.10).
Small letters indicate signiﬁcant difference between sex or age categories (Student’s t-test,
p < 0.10).

SO

(1989) also noted minor dispersal movements in December and January, and considered
them shifts in winter range. These shifts may be related to the quality of winter habitat,
density dependent factors, or some combination of the two. No birds in 1993 or in the
open site in 1994 exhibited such movements.

Birds in the closed site were split between the North and South areas in 1993 and
1994 (Table 22, Table 23). In 1993, there was only 1 bird in each dispersal type in the
North area, so no statistical comparisons could be made. However, home ranges of Type
I and 11 birds, and distance of dispersal were all larger than the means for the South area
birds. In 1994, home ranges and dispersal distances were also consistently larger, but no
statistically signiﬁcant differences were detected (p > 0.10).

Overall, there were no noticeable trends in site, sex or age classes that were
consistent between years. In a similar telemetry study, Small and Rusch (1989) measured
a mean net movement of 4.82 km for juvenile females over several years of study.
Collared juvenile males moved only half of that distance, 2.14 km. In this study in 1993,
no birds in either site moved over 4 km. In 1994, a juvenile female in the open site moved
nearly 5 km. Few birds in 1994 and none in 1993 approached these distances, and means
were considerably less. Small and Rusch (1989), however, used only birds which were
captured before September 10, to ensure that birds would be monitored for the duration of
fall dispersal. In this study, it was necessary to include birds which were caught as late as
early October.

Godfrey and Marshall (1969) also showed that juvenile female ruffed grouse move

farther than males in autumn. A compilation of research results from banding studies

51

Table 22. Mean home range size (ha) for Type I and Type 11 birds and dispersal distances
(m) for Type III birds in the North and South areas of the closed site of the PRCSF, 1993.

 

 

 

_ North area _ South area

X (SE) N X (SE) N
Type I 68.6 - 1 45.0 (11.3) 4
Type 11 205.1 - 1 158.0 (27.4) 7
Type III 1997.9 - 1 882.3 (241.8) 5

 

Table 23. Mean home range size (ha) for Type I and Type II birds and dispersal distances
(m) for Type HI birds in the North and South areas of the closed site of the PRCSF, 1994.

 

 

 

North area South area
)7 (SE) N 7 LSE) N
Type I 50.3 (11.0) 2 28.6 (8.0) 2
Type 11 121.5 (16.3) 5 91.4 (26.5) 6

TypeIII 1167.3 (334.2) 7 1070.2 (243.6) 4

 

52

provided in Bergerud and Gratson (1988) indicated that juveniles generally move farther
than adults and females moved farther than males, and that juveniles and females have
correspondingly larger home ranges. Although these trends were observed in some cases
in this study, few signiﬁcant differences were detected. Additionally, it was expected for
birds in the site open to hunting to move farther and have larger home ranges in the fall
due to increased ﬂushes by hunters. Fall movements, however, may be more related to the
quality of habitat present or density dependent factors. Low sample sizes may have
contributed to results which were not consistent between 'years. Combination of the data
across years was avoided at this point in the study, so that factors which may be

associated with yearly ﬂuctuations of the population could be assessed.

53

Habitat use

In the open site in 1993, 16 birds were located on at least 10 separate days (Table
24). There were 767 locations taken on those birds, with a mean of 48 locations per bird.
On this site, 22.2 % of all grouse locations were in aspen between 11 and 29 years of age,
and 20.2 % were in aspen between 1 and 10 years of age. In total, 56.8% of locations
were in aspen of any age class. Pine 1-29 years of age and jack pine types were used very
inﬁ'equently, 0.2% of all locations each.

In the closed site in 1993, all but 3 birds were captured and stayed within the
South area. The 3 birds in the North area had a total of 135 locations, for a mean of 45
locations per bird. Pine over 30 years of age was used heavily, with 32.6% of all locations
that type. Aspen of all age classes comprised 36.4% of the total number of locations.

Jack pine was used frequently also, with 27% of locations being in that type.

In the South area, 16 birds had a total of 741 locations, for a mean of 46 locations
per bird. More than half(56.2%) of all locations were in aspen of any age class. Pine >
30 years and hardwoods were used moderately (8-11%), while young pine and jack pine
were used rarely.

There were fewer locations per bird in 1994, due to reduced personnel and a larger
number of birds to locate. The open site had a sample size of 30 birds, with 850 daily
locations, with a mean of 28 locations per bird. As in 1993, a high percentage of locations

were in medium aged aspen (25.8%). More than half (53.8%) of locations were in aspen

types.

S4

 

 

 

 

3m 5 E m: a» E .33
2 w: 3 z E n.s. od o 2 a 3. R .25
mm 2 3 2m 3 n: o: 3m 3 2 do 2 8238.
g Q 2 3 we 8 Z 2 E. 22 3 n: 20.238.65.33
3 2m 3m .2 n: 3w 9% 3. 3. a. w: 08 @138:
no _ no 2.. 2 no 2 2 3 3 S 2 £3.25
3 n: no 2 M: 2m 3. o 0.8 m2 2 no». 38533333
2N : ad a. 3 8 3 3 4.2 m2 6: 8 $833:
9: 8 em or 0.2 m2 0.: n: 4.2 22 4.: no: .32282
M? X 9m 3m 3% 32 I: z a: em as o: £32282
as now No SN «.2 3: 0.2 n: 0.2 2: SN 2. .52-?83
.x. z .x. z .x. z .x. z .x. z .x. z 093535
58m 552 58m 552 32 :80 82 :25
82 886 82 26820

 

.39» 93 can 3 25 3:92 some 5 £83.80— omsohm Mo 8082. 98 39:52 .5 033.

55

In 1994, the North area of the closed site was trapped with more intensity than in
1993, resulting in a higher number of collared birds. The North area had 14 birds eligible
to be used in the habitat use analysis with 457 locations, for a mean of 33 locations per
bird. Again, the majority of bird locations (59.3%) were in pine 30+ years. Aspen types
were used 15.2% of the time, considerably less than in 1993. Differences may be a result
of the considerably larger sample size in the 1994 analysis.

The South had 13 birds and 346 locations, with a mean of 27 locations per bird.
Aspen 1-10 years was used more frequently in the South area in 1994 than in 1993, with
23.3% of all locations. Hardwoods were also used more in 1994, with 22.3% of all
locations in that type. Aspen of combined age classes accounted for 44.7% of all
locations.

Percent of locations in each habitat was categorized by dispersal type to explore
the possibility that movement may affect habitat selection. In the open site in both years,
the only birds that were located in jack pine were dispersers (Table 25). Dispersers also
had a lower percentage of locations in medium aged aspen in 1993 (12.3%) and in 1994
(18.2%) than did Type I and II non-dispersers.

In the North area of the closed site in 1993, only 1 bird fell into each category
(Table 26). The disperser used considerably more jack pine (58.8%) than Type I and II
non-dispersers. In 1994, all types used old pine heavily. Type 1 birds used more medium
aspen, while Type 11 birds used more jack pine.

In the South area of the closed site in 1993, Type I birds used medium aspen

heavily (36.5%) (Table 27). Type II birds used all ages of aspen with comparable

S6

 

 

 

 

24 4o... m: an «2 SN Bob
2 02 2 no a. _ 3. mm 3 : m. _ 25
4. 2 o o o o 4. 3 o o o o 2% V.3
SN 3 ”.2 ma 4.: n: 2 mm ”.2 a S : eeaoo 45.33
4.4 now 3. a 2 2 a: N4 2 4 n: 044 a» + 2 2.5
mm 2 m. 2 4. m. o o m. 2 o o i a? 8E
4.» m3 3: _4 a. _ 3 3m 2 3 E on 88322 45453
0.: m4 ”.2 a q: 2 as m4 2: on 4.2 EN 38324:
4.2 we 4.: 44 4.42 2.: <2 2m 08 3m 3 2 a» + on :83
2: 044 EN 23 3.4 mm 2. 3.4 can now 4am S .4» 3-: =82
_.: 34 a: N4 4.2 S TR 44 N: 2H 3: OR .5 o: 8&4,
.x. z ax. z .x. z .x. z .x. z .8 z 25 354:
Eoﬁoqﬂn N 09¢ _ 09¢. 4.330%5 N 09C. _ 69C.
444_ maa_

 

.30» can 25 .8836 3 8?. some 65 am 25 2332. noao E 2.288— 832» .«o .5an use c3822 .3 03am.

57

 

 

 

 

SN 42 8 04 N4 N4 .38.
4N N 3 N o o o o o o o o 850
2: EN 4.4 m.N_ Na 2 3m EN EN 2 o o 2i 6.8.
N.4 3 N m m. m. S 4.. o o _.N _ 84.228 4:233
in no: 0% on: 0.8 3 EN : NNN : 34 NN 2» + cm 2.5
3 N 2 mm m.N_ 3 o o o o N.N 2 .5 8N; 8E
3 N o o w. m. o o o o o 0 8834.2 2.4453
3 N 3 N o o o o N.m N o o 3834::
4.4 a N on S _ nN _ qNN : N.: 3 .3 + ON :25,
3 42 w.— 3. mi m N _ N.m nN MN 2: .s 3-: :83
3 n: N.m 2 S _ 2. N N2 4 4.2 3 .3 2; :83

.x. z .x. z .x. z .x. z ox. z .\. z 25 355

83.0qu N on; _ 69C. €889.25 N 33. _ 09C.

4oo_ mama

 

.3?» v5 25 5594.6 E 86 come—o 05 mo 83 5.62 05 E as “832 :68 E 25:82 836% we N583 98 $9832 6N 03am.

58

 

 

 

2: N: Nm SN EN 42 .25
5.4 m4 3 N_ 3 _ 8 t 3 n: 3 2 .25
3 4 4.4 w o o 2 ON 4.N 2 3 N Ea 6.2:
«.4 m N.N_ NN 3“ N an 5.2 M3 02 2: 2 seas 45.33
N.N_ 2: Na 2 o o N.” NN 4N. O4 _.4_ NN .5 + 8 SE
0 o o. _ o o 2 on 3 e o o .5 NN; 2.:
o o 3 3. 4.4 m 4.4 m: 3 EN 4. : 2: 883.5: 45453
N44 4 N.N_ NN 4.2 w oN. N... no 3. o c 88324:
S 3 2w : ~.NN 0: EN mm 5.2 new 3 n: .5 + 2. :83
S N 2.2 _N S _ NNN n8 «.2 8 34. S .5 NN-: :83
4.45 n: NNN 25 N.m4 3N o.4_ 2N he «No o.» m.N_ .5 2; :83

.5 z .x. z .x. z .5. z .x. z .x. z 25 35.45

383%5 N 093. N 09$. £8535 N 09¢. _ 09C.

4oo_ Nma_

 

.305 25 25 3309.26 .3 8mm come—o 2: .8 new 53m 05 E 25 «83a: :80 E 30:82 8:on mo :5an 98 53:52 SN 28¢.

59

intensity, as did dispersers. In 1994, Type I birds seemed to prefer young aspen, and
dispersers spent more time in hardwoods (44%).

The percentages mentioned above are for all birds combined, so they may be
reﬂecting the tendency of certain birds to use 1 habitat type frequently or exclusively. The
following preference analysis was based on each individual bird with habitat ranked by
usage, thus changing the deﬁnition of “use”. Availability was considered the entire open
site for open site birds, and the North or South area of the closed site for closed site birds.
In the closed site, birds which were caught in an area did not cross into the other area.

The analysis of preference is a ranking procedure, which is biased towards
infrequently used habitat types with low availability. In the open site in both years, young
pine was not used in the analysis of preference, due to its low percentage of total area. In
the South area of the closed site in both years, jack pine and young pine were not included
in the analysis. There were few locations in these habitat types in either year (Table 24).

In the open site in 1993, aspen > 30 years, “other”, young aspen, lowland
hardwoods and jack pine were used more frequently than they were available (Figure 4a).
Lowland conifers, pine > 30 years, aspen 11 - 29 years, and hardwoods were used less
than available. Trends were similar in 1994 (Figure 4b). Old aspen had the highest mean
HSI value of those habitats used frequently, but those habitats with the highest mean HSI
values of the open site (lowland conifers and aspen 11 - 29 years) were used less than
would be expected.

In the closed site in 1993, there were not enough bird locations in the North area

to carry out the analysis of preference. In the South area, the “other” category, lowland

60

353%.. Sm :o>_w 93 8mm :80 E 25 “852 :03 a8 mos—.9»
GI :82 .3832 35:35.. 5253 8— .o A a .2338.“ :SEano 29:2: 5055 8:33 “amigo no: 83 oocoaouoa 532
8:20 .voﬂ was 33 5 820m 25m E3300 52M gamma 2: E 8:03 webs.— aom cocoaomoa 53m: ~25? 28 :3 .v 833m

M721 OZHm<mEUmO

4L

mmo. mg. m2. 3%. 5.. com. oov. ooo.

 

 

 

 

mmano .52; $8-: 53? 53? 900392:
80392: 95,53 2%? 2mm? 2mm? sz 92333 HES

.83 .88 £20m .86 come—U 3

mmo. 02. gm. mmm. com. m2. 5% coo. mow.

 

 

are? ia-: amaze m2: oooBBZm .32; 93?
803% sz 2mm? 95953 MB: 92.53 2mm? «mac 2mm?

.82 .ee :25 a

61

 

825$ .v 2&3

 

 

 

 

 

 

M772 oEm<mmomo
”3. N8. 9:... me. 8m. 08. §. 08.
amazon .3 am - : .3 2 - _ .3 on A 90032: .3 cm A
95333 803923 2mm? zmam< 2.. 9,2433 2mm? mmEb
32 .83 5:8 .37. 886 G
we. 2:. an. an. 3:. 8o. mom. 32. 8c.
.3 cm A .3 a - : «Wazoo .3 S - _ m2: 803%: .3 2.. A
98392: sz 2mm? 972.53 2mm? mate V85 QZEBS 2mm?

.32 .23 :25 G

62

5.83 .3. 22mm

 

 

37:2 0233958
NS. xx. 33.. SN. 8m. 2m. 08. o8. E. 8..
39:28 .3 o... v .3sz .3 cm A .3 a - : 803%: .3 cm A .3 2 - _
803.92: 2.5/3 mza 54: 2.. zmam< «83.0 92,533 2mm? 2mm?

.32 .83 582 .23 “.820 6

63

hardwoods and pine > 30 years were used more than available (Figure 4c). Lowland
conifers and hardwoods were used less than available, and aspen types were used in
general proportion to availability. Results were similar in 1994, with the exception of
older aspen moving up in rank (Figure 4d).

Young and old aspen were used frequently in the North area of the closed site in
1994 (Figure 4e). These habitat types were rare, and used infrequently, but were not
eliminated from analysis due to the perceived value of aspen as ruﬁ‘ed grouse habitat.
Lowland hardwoods, “other”, medium aged aspen, old pine, jack pine and young pine
were used in proportion to their availability, while lowland conifers and hardwoods were
avoided. As in the open site, lowland conifers had a relatively high mean HSI value in the
closed site (0.348) and was plentiful in the South area, but was used infrequently.

It was expected that hardwood types would be avoided in both sites, due to their
low stem densities and corresponding low mean HSI values. High ranking of the “other”
category would also not be expected according to the model habitat requirements. In both
sites in both years, most bird locations which fell into the category “other” were in grass
or upland brush. Although these habitat types would have no value as fall/winter cover
according to model guidelines, openings and forest/opening edges may provide other
valuable habitat components, and were used more or in proportion to their availability.

To analyze the quality of habitat selected, habitat use values were calculated for
each bird by multiplying the percent of locations in each habitat type by the mean HSI

value for the corresponding types. Mean habitat use values were calculated for each site

64

in each year (Figure 5). Mean habitat use values were not signiﬁcantly different between
the North and South areas of the closed site in 1993 or 1994 (p > 0.10). Values were
signiﬁcantly different between the open site (0.311) and the North (0.222) and South areas
(0.232) of the closed site in 1993 (p < 0.10). Habitat use values were also different
between the open site (0.349) and the North (0.226) and South areas (0.204) of the closed
site in 1994 (p < 0.10). The values for the open site in 1993 and 1994 were slightly
higher than the overall HSI value of 0.27, but closed site values were comparable or below
overall HSIs.

A common explanation for the fall dispersal of grouse is that birds move to places
where their survival is enhanced by a more plentiful food supply or cover ﬁ'om predators.
It follows that birds would have to pass through marginal or poor quality habitat to reach
a satisfactory destination. To explore this possibility, mean habitat use values were
calculated for each dispersal type by site and year (Figure 6 - Figure 10). The North area
in 1993 was not included in this analysis, because each dispersal category consisted of only
1 bird.

In both sites in both years, habitat use values were lower for Type III birds than for
Type I birds, but no signiﬁcant differences were detected between dispersal types between
years. The highest habitat use value was for Type I birds in 1994 (Figure 7). The lowest
values occurred in the South area of the closed in 1994, because of the use of hardwoods.
Even though it was used less frequently than available, there were enough bird locations in
the hardwood type to adversely affect habitat use values.

It was assumed that in both sites, birds would use aspen types (and other habitat

types with high HSI values) in a higher frequency than their availability. Although that

65

 

0.4

N=3O

 

Habitat use value

 

 

 

Cl.No. Cl. So. Open Cl.No. Cl.So. Open
1993 1993 1993 1994 1994 1994

Site and year

Figure 5. Mean habitat use values by site and year. Error bars are one standard
error above and below the mean. Different letters indicate signiﬁcant difference
among habitat use values (Kruskal-Wallis one-way ANOVA, Multiple
Comparisons, p < 0.10).

66

0.5

 

0.4

 

 

0.3 1

0.2 -

Habitat use value

0.1 -

 

 

   

 

 

 

0 1
Type I Type II Type III

Figure 6. Mean habitat use values by dispersal type in the open site of the PRCSF
in 1993. Error bars are one standard error above and below the mean. No
signiﬁcant differences were detected (Kruskal-Wallis one-way AN OVA, p > 0.10).

0.5

 

N=4

0.4 . N=10 N=14

 

0.3 -

0.2 -

Habitat use value

0.1 -

 

 

 

 

 

 

 

Typel Typell TypeIII

Figure 7. Mean habitat use values by dispersal type in the open site of the PRCSF
in 1994. Error bars are one standard error above and below the mean. No
signiﬁcant differences were detected (Kruskal-Wallis one-way AN OVA, p > 0.10).

67

0.5

 

0.4

 

N =4 N =7 N=5

—-J':—

0.3

 

0.2 -

Habitat use value

0.1 -

 

 

 

 

 

 

 

o ..
Type I Type II Type 111

Figure 8. Mean habitat use values by dispersal type in the South area of the closed
site of the PRCSF in 1993. Error bars are one standard error above and below the
mean. No signiﬁcant differences were detected (Kruskal-Wallis one-way ANOVA,
p > 0.10).

68

 

 

 

 

 

0.5
L5 0.4
g N=2 N=5 =7
0)
g 0.3
‘5 +
;§ 0.2 - .
:l:
0.1 -
0 .

 

 

 

 

 

Type I Type II Type HI

Figure 9. Mean habitat use values by dispersal type in the North area of the closed
site of the PRCSF in 1994. Error bars are one standard error above and below the

mean. No signiﬁcant differences were detected (Kruskal-Wallis one-way ANOVA,
p > 0.10).

0.5

 

0.4

 

N=6

+ ”f4

0.3 N=

 

 

0.2

 

Habitat use value

0.1-

 

 

 

 

   

 

Type I Type II Type 111

Figure 10. Mean habitat use values by dispersal type in the South area of the closed
site of the PRCSF in 1994. Error bars are one standard error above and below the
mean. No signiﬁcant differences were detected (Kruskal-Wallis one-way AN OVA,
p > 0.10).

69

was true in some cases, grouse also frequently made use of habitats which had low values
for cover and food requirements according to the model. For example, the “high-tree”
pine forest is generally considered low security habitat, where young grouse cannot
survive. The only postulated reason for birds using such habitats is that they are unable to
compete successﬁilly for more secure habitats in the fall. Such habitats supposedly fail to
provide the quality of cover necessary to allow them to live to breeding age (Gullion
1970). The quality of old pine in the closed site was low, and birds were expected to
avoid it. Although it was not used signiﬁcantly more than available, birds in the North
area of the closed site did not seem to avoid the highly available older pine types in favor
of types with better HSI values.

One weakness in using the Michigan model is that it evaluates the forest on a stand
by stand basis, with little regard for the interspersion of habitat types across the landscape.
Many researchers have stated that optimal habitat for ruffed grouse is provided by the
interspersion of several forest age classes (Cade and Sousa 1985). This model may be
improved considerably by including weighting procedures for size of the stand and
adjoining habitat edges, an idea which has been explored by Roloﬁ‘ (1994). Although it is
beyond the scope of this study, the study sites are in the process of being digitized, at

which time landscape variables could be added to the model with relative ease.

70

Survival

Survival was analyzed ﬁ'om August 4, 1993 and 1994 through May 15 of the
following year. Past May 15, the number of censored birds increased to a point where the
accuracy of survival analysis would be compromised. In the open site in 1993, 31 birds
were used for analysis of survival. Birds which died or were censored within 5 days of
capture were not used. Excluded birds included 2 that died of stress or suffocation, 1 that
was killed by a mammalian predator, 1 that was censored, and 1 bird whose radio was
removed due to retrapping injuries. Two birds that died of stress or suffocation survived
past 5 days after capture, and were treated as censored for survival analysis. The overall
survival probability in the open site in 1993 was 0.19 (Figure 11).

In the closed site in 1993, 24 birds were used in the survival analysis. Three birds
died of stress and 1 was killed by a mammalian predator within 5 days of capture. The
overall survival probability in the closed site in 1993 was 0.63 (Figure 11). In 1993,
survival probability was signiﬁcantly higher in the closed site than the open site (Log- rank
test, p < 0.01). Because there were only 3 birds collared in the North area of the closed
site in 1993, a separate survival probability could not be calculated. The South area of the
closed site had a survival probability of 0.58 (Figure 12).

In the open site, 1 bird from 1993 survived long enough to be included in the
survival analysis for 1994. It was recaptured and recollared, and is included in the total
number of collared birds. Of the 47 birds which were collared, 4 that died of stress or
suffocation, 1 that was censored, and 1 whose radio was removed within 5 days of capture
were eliminated from survival analysis. This resulted in a total of 41 birds to be used in

the survival analysis. Overall survival was 0.29 (Figure 13), slightly higher than

71

 

Closed site (N = 31)

 

 

 

 

 

 

 

 

1- .
‘1 5 \ 5 |----Opensite(N=24)
r? 0-8 7" --‘§', ﬂ
E 0.6 J" -‘\___
8 J ‘ ---\:
o. .
u; 0.4 - ?\
Z 3 . x-
t 0.2 -- E : ‘s ...........
:1 . -
O '1 l" '1 l 1 l
I I I I f I

 

8/4/93 9/23/93 11/12/93 1/1/94 2/20/94 4/11/94 5/31/94
date
Figure 11. Survival probability for ruffed grouse radio-collared in the PRCSF

in 1993. Arrows approximate hunting season, September 15 through December
31, with a 2 week break between November 15 and November 30.

 

1 ------ ~ ---Southarea(N=21)

survival probability
P .O
1: as
‘ JI

.O
N
1
I

0 1 1 1 . 1 1 1
I l l l l l

8/4/93 9/23/93 11/12/93 1/1/94 2/20/94 4/11/94 5/31/94

 

date

Figure 12. Survival probability for ruffed grouse radio-collared in the South area of
the PRCSF closed site in 1993.

72

in 1993.

In the closed site in 1994, 52 birds were captured, of which 42 were collared.

Four birds from 1993 survived long enough to be included in the survival analysis for
1994, 2 of which were captured and recollared, and are included in the total number of 42
collared birds. One bird that died of suffocation from the radio-collar, 1 that was killed by
an avian predator, and 2 that had their radio-collars removed (all within 5 days of capture)
were not used in the survival analysis. A total of 40 birds were used in the survival
analysis. Survival probability was calculated as 0.38 (Figure 13), substantially lower than
in 1993. In 1994, there was no signiﬁcant difference in survival probabilities between the
sites (Log rank, p > 0.10). Separating the closed site, 14 birds in the North area had a
survival probability of 0.36, similar to the value of 0.38 in the South area (Figure 14).

In wildlife populations which are hunted, there is always the question as to whether
hunting is additive or compensatory to natural mortality. If hunting is additive, it is in
addition to natural deaths, and would impact overall mortality. Compensatory mortality
takes the place of natural mortality, and so would not impact overall survival across the
entire year. The signiﬁcant difference in survival in 1993 may be due to low numbers of
birds present on the sites. However, when population numbers are small, hunting may
have a higher impact on survival. To understand the impacts of hunting versus other types
of mortality, it is necessary to further examine the speciﬁc causes of mortality between
sites.

Of the birds which were used in the open site survival analysis in 1993, 9.7% were

alive on May 15 of 1994 (Table 28). The majority of the birds in the open site (32.3%)

73

 

 

  

 

 

 

 

1 ' I Closed site (N = 40)
g 0.8- ----Opensite(N=4l)
'32?
:8: 0.6 -
On
.72" 0.4 -
E 0.2 -
0 4.

 

8/4/94 9/23/94 11/12/94 1/1/95 2/20/95 4/11/95 5/31/95

date

Figure 13. Survival probability for ruffed grouse radio-collared in the PRCSF
in 1994. Arrows approximate hunting season, September 15 through December
31, with a 2 week break between November 15 and November 30.

 

North area (N = 18)
- - - - South area (N = 22)

 

   

-
~~~
~

1
I I I I I

survival probability
.9
A

p
N
1
I

 

0 1
8/4/94 9/23/9411/12/94 1/1/95 2/20/95 4/11/95 5/31/95

date

Figure 14. Survival probability for ruffed grouse radio-collared in the North and
South areas of the PRCSF closed site in 1994.

74

Table 28. Number and percent of radio-collared ruffed grouse by status in the PRC SF
open and closed sites in 1993 and 1994, as of May 15 of the following year. This table
contains only birds which were used in the analysis of survival.

 

 

 

 

1993 1994
Open site Closed site Open site Closed site
N % N % N % N %
Alive 3 9.7% 10 41.6% 6 14.6% 10 25.0%
Dead 19 61.6% 8 33.3% 19 46.3% 17 42.5%
Shot/collected 4 12.9% 0 0.0% 5 12.2% 0 0.0%
Shot/not collected 2 6.5% O 0.0% 0 0.0% 0 0.0%
Avian predator 10 32.3% 3 12.5% 9 22.0% 11 27.5%
Mammalian predator 2 6.5% 2 8.3% 2 4.9% 2 5.0%
Illness 1 3.2% 1 4.2% 0 0.0% 0 0.0%
Unknown 0 0.0% 2 8.3% 3 7.3% 4 10.0%
Censored 29.0% 6 25.0% 16 39.0% 13 32.5%

 

Table 29. Number and percent of known deaths by category in the open and closed sites
of the PRC SF in 1993 and 1994. No signiﬁcant differences were detected in non-hunting
mortalities between sites within the same year (Chi-square, p > 0.10).

 

1993 1994
Open site Closed site Open site Closed site
N % N % N % N %

Shot/collected 4 21.1% 0 0.0% 5 26.3% 0 0.0%
Shot/not collected 2 10.5% 0 0.0% 0 0.0% 0 0.0%
Avian predator 10 52.6% 3 37.5% 9 47.4% 11 64.7%
Mammalian predator 2 10.5% 2 25.0% 2 10.5% 2 11.8%
Illness l 5.3% 1 12.5% 0 0.0% 0 0.0%
Unknown 0 0.0% 2 25.0% 3 15.8% 4 23.5%

 

75

were killed by avian predators. Counting birds which had been shot, retrieved and
returned by hunters, and those which were shot by hunters and retrieved by researchers,
hunting mortalities accounted for 19.4% of birds in the open site. In the closed site,

41 .6% of birds used in the survival analysis survived to May 15, 1994 (Table 28). Avian
predators took 12.5% of birds, while mammalian predators took 8.3%.

By May 15, 1995, 14.3% of birds collared in 1994 in the open site were still alive.
Hunting mortalities accounted for 11.9% of birds collared, slightly less than the 1993
value. The rate of avian predation was slightly lower, but still accounted for the majority
of the uncensored birds (21.4%). In the closed site, 25% survived until May 15, and
27.5% were taken by avian predators.

Hunting season begins on September 15 and lasts through December 31, with a 2
week break for ﬁrearm deer season (November 15 - 30). All birds which were taken by
hunters were shot between September 15 and November 14. No collars were returned in
December. There were no known mortalities due to hunting in the closed site in 1993 or
in 1994. If there was any illegal taking of birds in the closed site, it would be expected
that the number of censored birds would be higher than in the open site. However,
percentages of censored birds were comparable between sites in 1993 and in 1994 (Table
28).

Numbers and percentages of known deaths by category, year and site were also
inspected (Table 29). Hunting accounted for 31.6% of known deaths in the open site in
1993, and 26.3% in 1994. Across both sites and years, avian predation was the leading

cause of mortality. No signiﬁcant differences were detected in non-hunting mortalities

76

Table 30. Number and percent of known deaths by category in the North and South areas
of the closed site of the PRCSF in 1994. No signiﬁcant differences were detected in non-
hunting mortalities between areas (Chi-square, p > 0.10).

 

North area South area
N % N %

Shot/collected 0 0.0% 0 0.0%
Shot/not collected 0 0.0% 0 0.0%
Avian predator 5 62.5% 6 66.6%
Mammalian predator 0 0.0% 2 22.2%
Illness 0 0.0% 0 0.0%
Unknown 3 37.5% 1 11.1%

 

77

between sites (or areas of the closed site, Table 30) within the same year. If hunting was
compensatory, there should have been a discrepancy between non-hunting mortalities
between sites. Speciﬁcally, numbers of natural mortalities should have been lower in the
site open to hunting. Although survival seems to have been lowered by such mortality in
1993, survival in 1994 does not seem to have been affected.

Small (1985) reported an average annual mortality rate of 73% for ruffed grouse
on hunted areas, which is generally consistent with the results on our open site. In a study
on hunting losses, Rusch et al. (1984) reported a mean band recovery rate of 24% ﬁom
1978 - 1981 for ruffed grouse in Wisconsin. This is slightly higher, but comparable to the
rate of return on the open site. It had long been thought that harvest rates of upwards of
50% of fall population could be tolerated by grouse populations (Bump et al. 1947).
However, loss of habitat in northern forests may make such rates dangerously high.

Keith and Rusch (1986) conclude that cyclic declines in Minnesota and Wisconsin
ruffed grouse are consistently associated with the inﬂux of raptors ﬁ'om Canada.
Additionally, Small et al. (1991) found that predation by hawks and owls was nearly
double that of hunting losses. Avian predators also accounted for the majority of
mortality in our study sites, but the association with raptor migration is not known.

Examining mortality rates by sex and age category can be of value for population
predictions. Survival was split between sex and age classes in both sites and both years
(Figure 15- 22). In the open site in 1993, the survival of juvenile birds declined to 0 at
about the end of January (Figure 15). This effect was probably due to the low sample

size, and a large number of censored birds. Survival probability for adult birds was 0.19.

78

 

1""- —AHY(N=17)J

0.8- ---HY(N=14)

  

06- ‘---~
\

0.4 -

 

 

survival probability

0.2 ‘-

"l
I
I
I
I
1

 

 

0 I I T I I 1
8/4/93 9/23/93 11/12/93 l/l/94 2/20/94 4/11/94 5/31/94

date

Figure 15. Survival probability for adult (AHY) and juvenile (HY) ruffed grouse in the
open site of the PRCSF, 1993.

 

—Male (N = 14)
- - Female (N = 17)

survival probability

 

 

 

 

8/4/93 9/23/93 11/12/93 Ill/94 2/20/94 4/11/94 5/31/94

date

Figure 16. Survival probability for male and female ruffed grouse in the open site
of the PRCSF, 1993.

79

 

 

 

 

 

 

——AHY(N=17)
---HY(N=7)

a»

$3

g t ............

g 0.4 .1

.z

E 0.2--

0 : : 1 t t +

 

8/4/93 9/23/93 11/12/93 1/1/94 2/20/94 4/11/94 5/31/94

date

Figure 17. Survival probability for adult (AHY) and juvenile (HY) ruffed grouse in the
closed site of the PRCSF, 1993.

 

—Male(N=13)
---Female(N= 11),

 

 

 

 

3:

'-§

.9

e

g 0.4 -r

.2.

g 0.2 --
0 : : : : : :
8/4/93 9/23/93 11/12/93 1/1/94 2/20/94 4/11/94 5/31/94

date

Figure 18. Survival probability for male and female ruffed grouse in the closed site
of the PRCSF, 1993.

80

 

 

 

 

 

 

1 1 —AHY (N= 19)
‘1 \ ---HY(N=18) l
31 0.8 -- 1 \
553 ' '~_
_'§ 0.6 -- ~, 1
8 1 -
Q ‘ \
a 0.4 " ..........
.2. -----
> .....
‘51 0.2 .-
0 : : : 1 + :

 

8/4/94 9/23/94 11/12/94 1/ 1/95 2/20/95 4/11/95 5/31/95
date

Figure 19. Survival probability for adult (AHY) and juvenile (HY) ruffed grouse in the
open site of the PRCSF, 1994.

 

——Male (N = 20) *

 

 

 

 

 

1 ..
g, 0.8-- - - ~Female (N: 15)
.5
,8 06*
8
o.
.2
g 0.2--
O 1 'ﬁ 1 1 1 1
8/4/94 9/23/94 11/12/94 1/1/95 2/20/95 4/11/95 5/31/95

date

Figure 20. Survival probability for male and female ruffed grouse in the open site
of the PRCSF, 1994.

survival probability

   

0.8 r

0.6 '-

0.4 4-

0.2 '-

 

0

81

 

—AHY (N = 23)
- ' 'HY (N =13)

l

1 n 1 n I
1 I I I I I

8/4/94 9/23/94 11/12/94 1/1/95 2/20/95 4/11/95 5/31/95

date

Figure 21. Survival probability for adult (AHY) and juvenile (HY) ruffed grouse in the
closed site of the PRCSF, 1994.

survival probability

   

 

 

-——Male (N = 15)
0.3. ---Female (N=21)
0.6 -- __________
0.4--
0.2 ~-
0 1 1 ‘ ‘ ' 9

 

 

8/4/94 9/23/94 11/12/94 1/1/95 2/20/95 4/11/95 5/31/95

date

Figure 22. Survival probability for male and female ruffed grouse in the closed site
of the PRCSF, 1994.

82

Female grouse (Figure 16) had a slightly higher survival probability (0.28) than males
(0.10), but the curves were not signiﬁcantly different (Log-rank, p > 0.10).

Closed site adults and juveniles had similar survival curves through most of the
1993 season (Figure 17), and ended with probabilities of 0.69 and 0.51. Male and female
birds also had similar survival curves, with females at 0.73 and males at 0.56 (Figure 18).
No signiﬁcant differences were detected.

In 1994, open site adults had a survival probability of 0.5 1, much higher than the
rate for juveniles, 0.24 (Figure 19). Survival probability was 0.41 for females (Figure 20),
and 0.22 for males. Again, no signiﬁcant differences were detected (Log-rank, p > 0.10).

In 1994 in the closed site, survival curves were fairly equal for adults and juveniles
(Figure 21). However, survival probability was signiﬁcantly higher for females (0.56) than
for males (0.19) (Log-rank, p < 0.10) (Figure 22).

With the exception of the closed site juveniles in 1994, the trends of females and
adults having higher survival rates were generally consistent and pronounced between sites
within years. The lack of signiﬁcant differences or the trends themselves may be due to
small sample sizes. Some prior research has suggested that higher mortality rates of
juveniles compared to adults are common in ruffed grouse (Gullion and Marshal 1968,
Small et al. 1991), a difference which has generally been attributed to lack of experience
(Small et al. 1991). However, other studies have failed to ﬁnd signiﬁcant differences in
survival by sex or age class (Rusch and Keith 1971, DeStefano and Rusch 1986). The
combination of yearly data (which has been done in the above mentioned studies) may be

necessary for adequate reliability of the data.

83

Proportional hazards model

The Cox proportional hazards model (Cox 1972) was used to investigate the
relationship between individual variables and survival. Because of low sample sizes within
years, data from 1993 and 1994 were combined by site for the analysis.

Movement variables were not used in this analysis. There is an inherent bias
involved in relating survival to movement variables, because a bird must survive long
enough to detect trends in movement patterns (5 weeks in this study). Birds may have
begun dispersal, and then been killed before their movements were detected as dispersal.
Or, birds which were prone to one dispersal type may have been more vulnerable to
mortality early in the season, causing their numbers to be underestimated. In future
analyses, it may be possible to restructure movement variables to remove such biases.

Ten variables were used in the analysis: age; sex; condition index (proportion of
weight to wing length); habitat use value * 100; and % locations in young, medium and
old aspen, lowland conifers, lowland hardwoods, and old pine. The other habitat types
had very few or no bird locations, and were not used in the analysis.

In the open site, the variables sex and habitat value had signiﬁcant P values (P <
0.10) (Table 31). Risk of mortality was 40.1% higher for males than for females. For
every 1 point increase in habitat use value, the risk increases 4.8%. There were no
signiﬁcant P values in the closed site.

Because it was found to be a signiﬁcant variable in the open site, birds in each site
were then stratiﬁed by sex, and the most probable proportional hazards models were

found through stepwise regression. For entry into the model and to stay in the model, P

84

Table 31. Risk ratios and associated P values for individual variables tested for association
to survival with the Cox Proportional Hazards Model (Cox 1972).

 

 

 

Open site Closed site
Risk ratio P Risk ratio P
Age 0.664 0.188 0.927 0.831
Sex 0.599 0.087 0.813 0.584
Condition index 1.099 0.604 0.802 0.330
Habitat value 1.048 0.072 1.052 0.260
% Asp 1-10 yr. 1.006 0.667 0.970 0.209
% Asp 11-29 yr. 1.008 0.424 1.018 0.259
% Asp 30+ yr. 1.016 0.366 1.010 0.576
% Lowland hardwoods 0.986 0.575 0.920 0.141
% Lowland conifers 1.018 0.266 0.974 0.466
% Pine 30+ yr. 0.974 0.233 1.011 0.213
N = 46 N = 46

 

85

values were set at 0.20. For open site males, the variables age and % locations in aspen
11-29 years were selected by the model (Table 32). Juveniles had a risk of mortality
80.1% higher than adults, and risk increased 2.4% with each increase in percent of
locations in medium aspen. In open site females, % locations in young and old aspen and
lowland conifers were used in the model. For each increase in percent of locations in
those habitat types, risk increased 9.4%, 13% and 18.1% respectively.

Age, and % locations in young and old aspen were found to ﬁt the model for
closed site males (Table 33). Juveniles had a risk of mortality 98.6% higher than adults,
and risk decreased 14.8% for each point increase in percentage of locations in young
aspen. In old aspen, however, risk increased 4.1% for each point increase in percentage
of locations. In the model for closed site females, only one variable, % locations in pine >
30 years, was used. The risk of mortality increased 4.7% with each point increase in
percentage of locations in old pine.

Age was found to be a signiﬁcant variable in males in both sites, with juveniles
having decreased survival. Some prior research, however, has shown that juveniles are
not more vulnerable to mortality (DeStefano and Rusch 1986). In the aspen variables,
results were also not as expected. In many cases, risk increased with % locations in
various age classes of aspen. These results could be due to variability of habitat quality
within individual aspen stands, or low sample size when age and sex categories were split
within sites.

In a similar telemetry study, Vispo et al. (1995) measured habitat variables at ﬂush
sites of radiomarked grouse. Researchers used the Cox model to relate those variables to

survival, and found presence of aspen and conifers to be associated with reduced

86

Table 32. Risk ratios and corresponding P values for explanatory variables found in a
stepwise regression through the Cox proportional hazards model (Cox 1972), and overall
model P values for male and female ruffed grouse in the open site of the PRCSF, 1993 and
1994.

 

 

N Risk P
ratio
Male
Age 0.199 0.026
% locs. in aspen 11-29 yr. 1.024 0.062
Overall model 23 0.053
Female
% locs. in aspen 1-10 yr. 1.094 0.121
% locs. in aspen 30+ yr. 1.130 0.086
% locs. in lowland conifers 1.181 0.003

Overall model 17

< 0.001

 

87

Table 33. Risk ratios and corresponding P values for explanatory variables found in a
stepwise regression through the Cox proportional hazards model (Cox 1972), and overall
model P values for male and female ruffed grouse in the closed site of the PRC SF, 1993
and 1994.

 

N Risk ratio P

 

Male
Age 0.014 0.007
% locs. in aspen 1-10 yr. 0.852 0.036
% locs. in aspen 30+ yr. 1.041 0.106
Overall model 21 0.002
Female
% locs. pine 30+ yr. 1.047 0.081

Overall model 18 0.036

 

88

mortality. The presence of birch in the overstory was found to be associated with
increased mortality. Such ﬁnely measured habitat variables may be the key to
understanding why some broadly generalized variables have an unexpected effect.
Sample size was of major concern in model building. As the number of birds
decreased with the splitting of categories, the number of events (deaths) decreased to a
level where the data were not very reliable (e. g. the removal of a bird would dramatically
change model output). This analysis should be considered exploratory, and further
research is warranted. A similar analysis with more collared birds, and perhaps more

detailed habitat use data would be extremely valuable.

89

Conclusions

Trapping success was generally low in both sites of the PRCSF, with the open site
having slightly higher success in both years. More adults than juveniles were caught in
1993, but age and sex ratios of trapped birds did not differ between sites. Incidences of
injuries or deaths related to trapping were minimal.

The open and closed sites were different in distribution of habitat types. The open
site had higher percentages of aspen and hardwood types, while the closed was dominated
by older pine, hardwoods and lowland conifers. The closed site was divided into the
North and South areas for analysis, due to differences in habitat types across the
landscape. Stands in the North area of the closed site were predominantly the older pine
types, while the South area had large percentages of aspen, hardwoods and lowland
conifers.

The quality of habitat in the open site was slightly higher than in the closed site in
medium and older aged aspen stands, and in overall HSI value. Although it would be
advantageous to the study for the sites to be of comparable quality, this imbalance is
preferable to the alternative of the open site being of lower quality than the closed. In this
way, if ruffed grouse survival is lower in the site open to hunting, it cannot be attributed to
poorer habitat quality.

In the open site, medium aged aspen, lowland conifers and older aspen had the
highest mean HSI values. In the closed site, young pine, lowland hardwoods and lowland
conifers all had higher mean HSI values than medium aged aspen, which was expected to

contribute the most to the overall value of the site. In both sites, conifer stems contributed

90

signiﬁcant amounts to the stem densities of most habitat types. Aspen stands were found
to be generally lacking in stem densities of aspen, but the addition of conifer stems into the
ESD (as called for in the model) brought many stands to the desired density.

Ruffed grouse in the Pigeon River showed 3 distinct movement patterns: Type I
and Type H non-dispersers, and dispersers. Non-dispersers were separated into 2 types
based on amount of movement during the fall. Type II non-dispersers showed increased
movements during the traditional dispersal period, but had no distinct dispersal from one
area to another. No consistent trends were found in the movements of ruffed grouse by
site, sex or age class. Birds in this study dispersed considerably shorter distances and had
smaller home ranges than found in other studies.

In the open site, birds seemed to show a preference for older aspen types. In the
closed site, birds used openings, lowland hardwoods, older pine and aspen types more
frequently than expected from their availability. Hardwood types were used very
infrequently in both sites in both years. Habitat use values were signiﬁcantly higher in the
open site than in either area of the closed site in both years. Habitat use values for
dispersers were consistently lower than those for non-dispersers in both sites in both years.

In 1993, there was a signiﬁcant difference in survival between sites open and
closed to hunting. No difference was detected in 1994. Most mortalities were due to
avian predation, but hunting did play a signiﬁcant role in survival in the open site in 1993.
The rate of predation by mammals was low. Adult ruffed grouse had higher survival rates
than juveniles, and females had higher survival rates than males.

The proportional hazards model showed different variables affecting survival

between sexes and sites. Age was a factor in survival for males in both areas, with adults

91

having a higher chance of survival than juveniles. Percent locations in aspen types were
negatively related to survival in many cases, results which cannot be explained at this time.
Old pine and lowland conifers were also negatively related to survival. The results of the
proportional hazards model should be taken as preliminary and exploratory, and will not

be taken into consideration for management implications.

92

Management Implications

In 1980, Hammill and Visser expressed concern that the aspen resource in
Michigan has been on a steady decline since about 1935. A low demand for aspen
products and natural succession to other species have been the main causes for the decline.
They recommended that harvest rates increase, and private landowners be educated as to
the value of aspen as a wildlife resource. The open site of the PRC SF has moderate
acreages of aspen, but the closed site (especially the North area) is lacking stands of this
type. Additionally, although older aspen continues to be of value for grouse after its
rotational age in the open site, older aspen in the closed site seems to lack adequate stem
densities. Unless cutting is increased, acreages of medium aged aspen will continue to
decline, and succeed to other types of lesser quality.

Managing aspen for ruffed grouse may be in conﬂict with some of the speciﬁc
management goals for the Pigeon River Forest. These goals include: 1) providing
favorable habitat for elk; 2) furnishing food, cover and seclusion for wildlife; and 3)
providing recreational opportunities for people in keeping with the quiet, peaceful and
wild character of the area. Management options must be weighed, considering the
beneﬁts versus the consequences to other wildlife and land uses. If it is within the
interests of the forest to sustain or increase ruffed grouse habitat, it is recommended that
cutting be increased in the North area of the closed site, and continue at the current rate in
other areas of the forest.

Ungulate browsing has been shown to have a thinning effect on regenerating aspen

stands in the PRCSF (Campa et al. 1993), and the quality of aspen for nrffed grouse may

93

be a reﬂection of this thinning. Therefore, when aspen stands are cut with the intention of
improving ruffed grouse habitat, attention should be paid to adjacent habitats, ungulate
densities in the area, site quality, and size of the cut. Although traditional ruffed grouse
management practices have stressed small cuts (< 10 acres) (Gullion 1984), the presence
of heavy ungulate browsing in the Pigeon may render small cuts inadequate. Larger cuts
(> 100 acres) may be able to sustain heavy browsing pressure on the edges, while
providing adequate ruffed grouse cover in the center. It may not be possible to provide
adequate aspen habitat for ruffed grouse throughout the Pigeon River area, but grouse
management areas away from high ungulate concentrations may be an alternative.

The importance of medium aged aspen stems for ruffed grouse in the northern
United States cannot be argued. However, many ruffed grouse in the Pigeon River
Country State Forest make use of other habitat types which are not usually considered
productive grouse habitat. Lowland hardwoods and pine types (with developed
understories) provide adequate habitat for grouse in the closed site. When considering
management on a landscape level, it would be beneﬁcial to plan such areas to be in
proximity to aspen types. A plan which takes into account the interspersion of forest
types may decrease the probability that dispersing birds must move through poor habitat,
thereby possibly increasing their chances for survival.

Historically, control of grouse populations has been by extreme changes in the
hunting regulations. Throughout the northern United States populations have continued
to cycle. The difference in survival between sites in 1993, however, may indicate that

hunting has an adverse affect when populations are at their lowest points. Since the trends

94

of the cycle are known, it may beneﬁt populations to moderately limit hunting for a year
prior to and during the low point of the cycle. Decreasing bag limits alone will not
substantially change the number of birds harvested, but shortening the season, decreasing
bag limits and educating hunters should.

In Michigan, it is widely believed that hunting in December is responsible for a
large percentage of total birds taken. In this study, no collars were returned by hunters in
December in either year. It is not recommended that a shortening of the season take place
in December, but rather in the season before November 15. If this takes place in
conjunction with a coordinated effort to encourage hunters to limit their time aﬁeld, the
low points of the cycle may not be as extreme. The possibility of a quicker recovery of the

population should encourage compliance and cooperation.

APPENDICES

95

Appendix 1. Production functions from the Habitat Model for Ruffed Grouse
in Michigan (Hammill and Moran 1986).

 

 

1 .
0.75"

0

:3

E 0.5

K

0

'6 il-

.5 0.25

0

 

0 4000 8000 12000 16000 20000
Equivalent stem density (stems/ha)

Appendix I, Figure 1. Index value for equivalent stem density in the Michigan
model. .

 

 

1 1
0.75 "
8
'3 0.5
>
K
e i
..S 0.25
0 1 ‘r 1 1 —1 1L

 

0 1 2 3 4 5 6 7 8 9
Deciduous stem height (m)

Appendix I, Figure 2. Index value for average height of deciduous trees in
the Michigan model.

96

Appendix I. (cont’d)

Index value

 

 

A
1

A
T

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Low conifer branch height (111)

Appendix I, Figure 3. Index value for height above ground of lower coniferous
branches in the Michigan model.

 

 

 

1
0.75 ‘-
D
:3
'9‘
5 0.5 w
1:
E:
0.25 ..
0 1 : 1 4. r 1 :
0 1 2 3 4 5 6 7

Shrub height (m)

Appendix I, Figure 4. Index value for average height of deciduous shrubs in
the Michigan model.

97

Appendix I. (cont’d)
l

0.75 ‘-

0.5

Index value

0.25 ..

 

 

 

g.

0 50 100 150 200 250 300 350
Distance to food source (m)

Appendix I, Figure 5. Index value for distance between life requisites in the
Michigan model.

98

Appendix 11. Examples of ruffed grouse dispersal patterns in Pigeon River Country State
Forest, 1993.

 

1N

1km

16

 

14

 

 

 

.-
J-

1km

Appendix II, Figure 1. Weekly harmonic centers for a single bird plotted spatially
and numbered by week. This bird was categorized as a Type I non-disperser.

 

 

 

2000 :
e 1500 :
A - I
32% g

0

agrooo - g
83 :
9 500 - g

M

 

00
0
ﬂ
Q
~
~
~
N
u
b.)
~
A
H
(I!
ﬂ
0‘
H
\I
_
00
ﬂ
\0
N
0

Week

Appendix 11, Figure 2. Distance moved by a single bird from the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a Type I non-disperser (same individual as in Appendix II,
Figure 1). Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at
week 16 indicates November 30, after which movements are not considered fall dispersal.

99

Appendix II. (Cont'd)

 

TN

1km

 

1.;11 15

11ng 1720
18 19

10

 

 

 

1km

Appendix II, Figure 3. Weekly harmonic centers for a single bird plotted spatially
and numbered by week. This bird was categorized as a Type II non-disperser.

 

2000

1500 '-

Dlstance from
week 1 (m)
I

§

 

 

 

8 9 10 11 12 13 14 15 16 17 18 19 20
Week

Appendix II, Figure 4. Distance moved by a single bird from the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a Type II non-disperser (same individual as in Appendix 11,
Figure 3). Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at
week 16 indicates November 30, after which movements are not considered fall dispersal.

100

Appendix II. (Cont'd)

 

T
IN 9 10
11

1km

 

20 12

'1819

15

 

 

143

 

.3
db

1km

Appendix II, Figure 5. Weekly harmonic centers for a single bird plotted spatially
and numbered by week. This bird was categorized as a disperser.

 

  

 

Drstance from
week 1 (m)
1

 

 

 

 

9 10 11 12 13 14 15
Week

Appendix II, Figure 6. Distance moved by a single bird from the harmonic center of its
locations during the ﬁrst week located to the harmonic center of each following week.
This bird was categorized as a disperser (same individual as in Appendix II,Figure 5).
Numbers indicate weeks, with week 1 = August 1 - August 7. Dashed line at week 16
indicates November 30, after which movements are not considered fall dispersal.

—.
ON

17 18 19 20

101

3836.800 AEDV
088.52 0.80555. 13.3303 .23 028m m_ 034 .30— ..8080009 - 508203 .mmUNE 05 «0 37. 0000 05 5
9000002.“ 0:08 3 commences 808m vote 258 302305 ~00 88:00 0:88.35 >333 A aha—mi .E 500009.‘
0082. 8000s 803K 088» 882. 008$ 080%
0828

 

 

ooomsm

 

 

 

_|Il.e oooEom

 

 

885m

 

 

*1:-
l-

4.
'I
d-

4.

coowuom

 

 

I-

 

oooogm

 

 

 

888m

2 L
ooowmom

.32 00m 82 5 .680K 88m 90:00
Sim coomE 05 .«0 3.7. 030—0 05 :000 05 E 800..» note 302305 00 £850 2008.8: >33? .5 5000034

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

102

0030:8000 AEDV
0030002 0803080. 38030: .23 00300 8 002 .39 00080000 - 00080000m .mmumm 05 .00 000 0000 05 8

300000000 0%.: 3 00—13035 0000.8 vote 0380.« 302208 00.0 80.000 0808.3: >200? .N 000mm,...— .5 800003..
08th 80000 0830 832. 0880 8800 8800

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 888m
— z
u .. w 82 an
e. we
a
“a. 8028
Few . 802%
0 ”I
28 e ..W e o e ,
035 6 80:8
Ea x
38 u i 802%
000 e
88 .
ES 6 8200
X
as Z
8603 808%

 

 

 

80:80 .8 0088....

103

.naaqmzooo
32.—Ad 0200002 0000300.? 300030: .23 00—000 “0 9:2 .33 00900009 - 000w:< .mmUMm 05 .«0 00?. 0000 05 0_
30000020068 .3 02.2.0005 0.30% 00.0000 038 3023—05 08 82000 £008.02 >300? .m 0.5me .5 5.000%?

Some 8895 0802. 882. 8080 093% 88%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

#4:! 888m
7v
. 888m
008 o
3.8 x . .r
83 a . u 082%
| T — ' I ~ ‘ .
on; 0 x1; 2 w...
080 0 AW.» +. . .
gﬁw I . . Ill Age—om
58 u 4
O
X 4
800 80008
28 .
3.2 a
8mm . 888m
8: a
a: .
$8 . 888m
88 .
000w0q z e
ooovmom

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6.680 .00 50503

104

 

 

 

 

00000000000 32.—H: 00.00002

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0803 8080 080% 88%
8880

«a 883m

002 + ..I.

88 u

was . m 803%

28 x

85 . o o 0 888m

:5 a so

33 0.

:3 o . 888m

00% .

8mm - 888m

:8 0

00: a . II

82 . . 888m

3: . 0H

000w0q _

 

 

 

 

 

 

 

 

 

 

 

 

coogom

02030000. 0300303 505 00—000 mm 002 .0000 000000000 - 00.92 .mmUME 05 00 000 0000 05 00 30000000000000
.3 00.000003 800% 00000.. 000. 0305—00 000 030000 03030000 00 000.000 200052 02003 .0 0.5mm"— .E ﬁ00000<

'1an

Z

00.0.80 .00 00503

105

.000 000000
0.0 00 00000 £00m 000 .0002 0.0 000300 0000000 05 00 000000— 0000080000 0.0 00000000 000 0000000 00000000000

A2030 000000002 00030000. 00000303 003 000000 00 002 .33 000000020 .. 0000000000 .00000 20 00 000 000.000 0.0
00 300000000 0000.. 00 02000000 00000» 000000 0000. 000000000 000 0.000000 00000000 0300? .m 00030 .000 0000000<

000000 000000 000000 000000 000N00 000000 000000
000N000V

 

 

“BIZ

 

_ _
__. _. as...

4
_x....__

x

 

 

0000000

 

 

 

 

 

0000000V

 

_
0..
a

 

 

000000m

 

 

 

 

 

000003
2 0
000000m

8.080 .00 005090

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

106

.000 00020

000 .00 00000 00000 000 .0002 000 0003000 020000 05 .00 000002 000000000000 05 000000000 000 000000— 0000000000
$2.03 0000002 0000000000 00000303 0005 00—000 00 002 .0000 000000000 - 0000000000 .mmumm 0.0 00 000 00020 000
00 3000000000000 00 00000000 000000 00.000 20000.0 000000000 00.0 000000 20000000 03003 .0 00000"— .5 000000000

00020 000000 000000 000000 000000 000000 000000

 

...F

f .0».

0000000

 

J0...

0000000

 

 

K

0000000

 

I
0

1.-
I}? X

 

X

 

 

 

 

 

L .I - — II

 

0000000

 

0000000

 

 

0000000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Z

 

 

0000000

 

8.080 .00 008090

.000 00020 0.0 .00 00000 .0000 000 .0002 000 0003000 020000 0.0 00 000002 000000000000 05 000000000 000 000005
0000000000 A2000 00000002 000050000 00000303 0005 02000 00 002 .0000 .0008000Q - 00000.0. 00000 0000 000 00020
05 00 90000000 2000 00 00000000 000000 00000 2000 000000000 000 0000000 20000000 000003 .0 000000 .0: 50000010

000000 000000 000000 000000 000000 000000 000000

0000000

 

 

000

 

‘4-

 

 

 

9&1

 

 

 

 

000...
0000 o
000..
000 4
:00 0
0000 o
0000 4
000 I
02 0
2000 o

107

 

 

0000

 

+3“,

 

 

0000000

 

 

 

 

 

0000000

0000000

 

 

 

00000»—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000

 

‘1an

7;

60:80 .00 00500.0

.000 00020 0.0 .00 00000 .0000 000 00002 000 0003000 020000 000 00 000002 000000000000 00002000 000 000000— 0000000000

A2030 00000002 000200000. 00000303 0005 02000 00 002 .0000 000000009 - 00000< £00.00 20 00 000 00020 0.0 00
90000000 2000 00 00000000 000000 00000 000 0300200 000 200000 000 0000000 20000000 000003 .0 0000mm .0: 00000000~

000000 000000

000000

000000

000000

000000

000000

0000000

 

+

 

+

.0

0000000

 

 

 

 

 

108

 

 

 

1P

 

0000000

 

 

TI

 

 

‘LE

 

 

 

 

.0000000

 

0000000

 

 

0000000

 

 

 

Us

 

 

 

0000000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000000

 

"BIZ

7;

6.080 .00 008000.

LITERATURE CITED

Backs, S. E., P. D. Major, C. H. Eisfelder, and J. S. Thompson. 1985. Techniques for
trapping and restocking ruﬂ‘ed grouse in Indiana. Preliminary Draft.

Bailey, W. J ., W. M. Sharp, R. B. Hazel, and G. Davis. 1955. Food habit trends of ruffed
grouse in the Centre County “Barrens.” Pennsylvania State Univ., Agric. Exp.
Sta. Bull. 604. 18pp.

Barber, H. L., R. Chambers, R. Kirkpatrick, J. Kubisiak, F. A. Servello, S. K. Stafford, D.
F. Stauffer, and F. R. Thompson HI. 1989. Cover. p 294 - 320 i_n S. Atwater and
J. Schnell, eds. Ruffed Grouse. Stackpole Books, Harrisburg.

Bergerud, A. T., and M. W. Gratson. 1988. Adaptive strategies and population ecology
of northern grouse. Vol. II. University of Minnesota Press, Minneapolis. 809pp.

Bump, G., R. Darrow, F.C. Edminster, and W.F. Crissey. .1947. The ruffed grouse: life
history, propagation, management. New York State Conservation Dept, Holling
Press, Inc., Buﬂ‘alo. 915pp.

Cade, B. S., and P. J. Sousa. 1985. Habitat Suitability Index Models: Ruﬂ‘ed Grouse.
US. Department of the Interior, Fish and Wildlife Service, Biological Report 82
(10.86). Washington DC. 31pp.

Campa, H., 111., J. B. Hauﬂer, and S. R. Winterstein. 1993. Eﬂ‘ects of white-tailed deer
and elk browsing on regenerating aspen: A ten year evaluation. Pages 304-311 in
ID. Thompson, senior editor. Proc. of the International Union of Game
Biologists XXI Congress. Forests and wildlife....Towards the 21st century.

Chambers, R. E., and W. M. Sharp. 1980. Movement and dispersal within a population
of ruffed grouse. J. Wildl. Manage. 22(3):231-239.

Cox, D. R. 1972. Regression models and life tables (with discussion). J. R. Stat. Soc. B
34: 187-220.

DeStefano, S., and D. H. Rusch. 1982. Some historical aspects of ruﬁ‘ed grouse harvests

and hunting regulations in Wisconsin. Trans. Wis. Acad. Sci., Arts, and Letters
70:27-35.

109

110

, and . 1986. Harvest rates of ruffed grouse in northeastern Wisconsin. J.
Wildl. Manage. 50:361-367.

Dixon, K., and J. Chapman. 1980. Harmonic mean measures of animal activity areas.
Ecology 61(5): 1040-1044.

Dorney, R. S. 1966. A new method for sexing ruﬁ‘ed grouse in late summer. J. Wildl.
Manage. 30(3):623-625.

, and H. M. Mattison. 1956. Trapping techniques for ruffed grouse. J. Wildl.
Manage. 20(1):47-50.

Edminster, F. C. 1947. The ruﬂ‘ed grouse: its life story, ecology, and management. New
Macmillan Press, New York. 385pp.

Godfrey, G. A., and W. H. Marshall. 1969. Brood break-up and dispersal of ruffed
grouse. J. Wildl. Manage. 33:609-20.

Godin, A. J. 1960. A compilation of diagnostic characteristics used in aging and sexing
game birds and mammals. page 156 i_n S.C. Schemnitz (ed.) Wildlife
management techniques manual, 4th ed. The Wildlife Society, Washington, DC.

Gullion, G. W. 1970. Factors inﬂuencing ruffed grouse populations. Trans. N. Am.
Wild]. Nat. Resour. Conf. 35: 93- 105.

Gullion, G. W. 1984. Managing northern forests for wildlife. The Ruffed Grouse
Society, Coraopolis, Pa. 72 pp.

, and R. W. Marshal. 1968. Survival of ruﬁ‘ed grouse in a boreal forest. Living
bird 7:117-167.

Hammill, J., and R. Moran. 1986. A habitat model for ruffed grouse in Michigan. Pages
15-18 i_n_ J. Vernor, ed. Wildlife 2000. The University of Wisconsin Press,
Madison, Wisconsin.

, and L. Visser. 1984. Status of aspen in northern Michigan as ruffed grouse
habitat. Pages 123-136 i_n W. L. Robinson (ed.) Ruffed grouse management: state
of the art in the early 1980's. Northcentral Section, The Wild]. Soc., Bathesda,
MD.

Johnsgard, P. A. 1973. Grouse and Quails of North America. Lincoln: University of
Nebraska Press, Lincoln. 553pp.

Kaplan E. L., and P. Meier. 1958. Nonparametric estimation from incomplete
observations. J. Am. Stat. Assoc. 53:457-481.

111

Keith, L. B., and D. H. Rusch. 1986. Predation’s role in the cyclic ﬂuctuations of
ruffed grouse. 19th International Ornithological Congress. Ottawa, Ontario.
V. 1 2699-732.

Kubisiak, J. F. 1984. The impact of hunting on ruffed grouse in the Sandhill Wildlife
Area. Pages 151-168 i_n W. L. Robinson (ed) Ruﬁ‘ed grouse management: state of
the art in the early 1980's. Northcentral Section, The Wildl. Soc., Bathesda.

, J. C. Moulton, and K. R. McCaffery. 1980. Ruffed grouse density and habitat
relationships in Wisconsin. Wise. Dep. Nat. Resour., Madison, Tech. Bull. 118.

16pp.

Larsen, J. S., and R. D. Taber. 1980. Criteria of sex and age. Pages 143-202 in S. D.
Schemnitz (ed.) Wildlife management techniques manual, 4th ed. The Wildlife
Society, Washington, D. C.

Leatherberry, E. C., and J. S. Spencer. 1996. Michigan forest statistics, 1993. U. S.
Dep. Agric., For. Ser. Resour. Bull. NC-67. 144pp.

Lee, E. T. 1992. Statistical methods for survival data analysis. John Wiley and Sons,
Inc., New York. 482 pp.

Mohr, C. O. 1947. Table of equivalent populations of North American small mammals.
Am. Midl. Nat. 37: 223-249

Raile, G. K., and W. B. Smith. 1983. Michigan forest statistics, 1980. U. S. Dep. Agric,
For. Ser. Resour. Bull. NC-67. 101pp.

Roloﬂ‘, G. J. 1994. Using an ecological classiﬁcation system and wildlife habitat models
in forest planning. Ph.D. Dissertation, Michigan State Univ., E. Lansing. 203pp.

Rosner, B. 1990. Fundamentals of biostatistics. Wadsworth, Inc., Boston. 655pp.

Roussel, Y., and R. Ouellet. 1975. A new criterion for sexing Quebec ruffed grouse. J.
Wildl. Manage. 39:443-445.

Rusch, D., and L. B. Keith. 1971. Seasonal and annual trends in numbers of Alberta
ruffed grouse. J. Wildl. Manage. 35(4):803-822.

, S. Destefano, and R. J. Small. 1984. Seasonal harvest and mortality of ruﬂ‘ed
grouse in Wisconsin. Pages 137-150 in W. L. Robinson (ed.) Ruffed grouse
management: state of the art in the early 1980's. Northcentral Section, The Wildl.
Soc., Bathesda.

112

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

Small, R. J. 1985. Mortality and dispersal of ruffed grouse in central Wisconsin.
Master’s Thesis, University of Wisconsin, Madison.

, J. C. Holzwart, and D. H. Rusch. 1991. Predation and hunting mortality of
ruffed grouse in central Wisconsin. J. Wildl. Manage. 55(3):512-520.

, and D. H. Rusch. 1989. The natal dispersal of ruffed grouse. The Auk
106:72-79.

Tardy, S. W. 1991. Soil Survey of Cheboygan County Michigan. US Department of
Agriculture. E. Lansing. 256pp.

Vispo, C. R., J. R. Cary, D. J. Lauten, and C. Balzer. 1995. Assessment of habitat
quality in a habitat mosaic: an example with ruffed grouse. Pages 470-474 i_n J.
Bissonette and P. Krausman (eds). Integrating People and wildlife for a
sustainable ﬁiture: Proceedings of the International Wildlife Management
Congress. The Wildlife Society, Bethesda, MD. 697pp.

White, G. C ., and R. A. Garrott. 1990. Analysis of Wildlife Radio-Tracking Data.
Academic Press, San Diego. 383pp.

IIIIIIIIII2IIIIIISIIIIIIIIIIISILIIIIIIIw