cum:

lhh

5,,m'vwinh‘ F} J

L. \

 

K ~«1VI‘

a.

 

.. nun—- v-

 

%‘
._.-

q.

1
”a
4'"!

w‘fi“:’l7"'" A
.4"
'i' ‘0
14

. :3. , .‘1
' I l. ‘ ‘

3 $31.}:
{10'1'329-

'3' 1
~31
$

93’s??? 1
g

x C
+- . ! 1 .
51 ,3 .7 52.3. 3m" ‘
h-V’fi'! ' N555: “:2 w

 

THES'S

f‘

 

[— _
"1007 LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

BROWN-HEADED COWBIRD RESPONSE TO KIRTLAND’S
WARBLER HABITAT MANAGEMENT AND COWBIRD TRAPPING
IN MICHIGAN

presented by
LAUREN A. BAILEY

has been accepted towards fulfillment
of the requirements for the

MASTER OF degree in FISHERIES AND WILDLIFE

 

 

 

SCIENCE
' Major rofessor’s Signature
513/07

Date

MSU is an affirmative-action, equal-Opportunity employer

—.—-o-o-o-u----l-n-I- -O-d-a-A

 

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

 

DATE DUE DATE DUE DATE DUE

NM? 330053 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DaleDue_indd-p.1

BROWN-HEADED COWBIRD RESPONSE TO KIRTLAND’S WARBLER HABITAT
MANAGEIVIENT AND COWBIRD TRAPPING IN MICHIGAN

By

Lauren A. Bailey

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

2007

ABSTRACT

BROWN-HEADED COWBIRD RESPONSE TO KIRTLAND’S WARBLER HABITAT
MANAGEMENT AND COWBIRD TRAPPING IN MICHIGAN

By
Lauren A. Bailey
Brown-headed cowbirds (Molothrus ater) have been of concern to wildlife
managers due to the often negative effects of their parasitic breeding practices on avian
communities. In Michigan, the cowbird host species most impacted is the endangered
Kirtland’s warbler (Dendroica kirtlandiz), whose breeding grounds comprise middle-aged
(6-20 years) jack pine forests. The Kirtland’s warbler recovery plan includes extensive
jack pine restoration and intensive cowbird trapping. The overall goals of this project
were: 1) to analyze the effects of jack pine stand sizes and ages on cowbird relative
abundance and density, 2) to determine correlations between cowbird trap captures and
surrounding land cover; and 3) to develop a habitat suitability index model for the brown-
headed cowbird. Avian point counts and vegetation surveys were performed to collect
data on cowbirds in Kirtland’s warbler habitat. Cowbird abundance was greater in large
stands (> 81 ha), had no association to stand age, and was positively correlated with host
abundance and species richness. To answer question 2, Michigan land cover data were
analyzed with cowbird trap data in ArcView GIS. Cowbird captures were positively
correlated with agricultural and urban areas, and with upland deciduous and mixed
forests, and were negatively correlated with upland and lowland coniferous forest.
Finally, I constructed a habitat suitability index (HSI) model for the brown-headed
cowbird using the following suitability index variables: distance to feeding areas,

distance to edge, percent ground cover and percent mid-story cover.

ACKNOWLEDGEMENTS

The decision to move across country to “Middle America”, enter into a new field
of study, and start graduate school were not easy ones. The last four years have been met
with some interesting learning experiences and life lessons, which I am grateful to have
been challenged by and ultimately to have muddled through.

I would like to thank my advisor, Dr. Kelly Millenbah, for accepting me into the
program and helping formulate the project idea; and my research committee, Dr. Henry
Campa, III and Dr. Catherine Lindell, for helping to guide this project into completion. I
would also like to thank Dr. Dana Infante for her invaluable help during the revisions
process.

For financial assistance for this project, I would like to thank Michigan State
University for providing me with the MSU/CANR Academic Achievement Award. I
would not have been able to complete this project without that financial support. I would
also like to thank the International Order for the Rocky Mountain Goats for awarding me
the Bill Burtness Field Research award. That contribution was invaluable in providing
funds to purchase field equipment and attend professional meetings. Finally, I would like
to thank the committee for the NASA-MSU Professional Enhancement award for
providing funding for me to attend the International Association for Landscape Ecology
Symposium in March 2005. That conference was the springboard to my professional
development and contributed greatly to shaping my future research interests.

Several agency employees were integral to the formation of this project and I
would not have had as strong a product or as interesting a graduate experience without

their help and expertise. Many thanks to Christopher Mensing, Mike DeCapita, and Phil

iii

Huber for sitting down with me in the early stages to give me the insider’s view of
cowbirds and Kirtland’s warblers. I would also like to acknowledge Rex Ennis and Phil
Huber for including me in the Kirtland’s warbler recovery team meetings and annual
censuses. I always enjoyed attending both and I thank you for giving me such unique
opportunities and perspective.

Working on a limited budget made finding field assistance tricky, but I was
lucky to know good people willing to work for free. Thanks to Sarah Panken and Adria
Van Loan for going out in the early days of season one; you both helped calm my
anxieties! Also thanks to Desiree Oliver and Kelly Thomas for their help in collecting
census data. Acknowledgements go to Mike Thalhauser for his hard work and wicked
humor. Many thanks to my favorite field assistants — Adam Mittman, Shabbi and Kika -—
for your loyalty during those long hours in the field.

I could not have done this project without the emotional and professional support
of various graduate students, faculty and staff in the department. Thanks to Sarah
Panken, Anita Morzillo and Nicole Lamp; you guys are my rocks. Many thanks to
Alexandra Felix and Daniel Linden for their help with ArcView GIS and for keeping me
in good humor. Thanks go out to Michelle Prehoda, Krishnan and Heather Sudharshan,
George and Emily Berghom, Tracy Grazia, Sarah Mayhew, and Mara Bryan for being the
wonderful friends that you are. To the “Statistics crew”—-—Veronica Lopez, Sherry
Martin, Michelle Prehoda and Tammy Giroux—thanks for your patience and help getting
me over that hurdle! I have created so many memories with the ladies of the Millenbah
lab, past and present. Thanks for the encouraging words and support. To the hard-

working ladies of administration, Jill Cruth and Mary Witchell, thanks so much for your

iv

time and effort in getting paperwork processed and problems rectified. To the custodial
crew—Mike, Lucy, and Tom—thanks for cheering me up with chocolate and
conversation during my late-night work sessions! I would also like to thank Dr. Brian
Maurer for taking time to answer my statistical questions. Thanks also to Dr. Jack Liu,
for being a supportive and encouraging force for me.

Finally, I must thank my family—Mum, Walt, Shannon, Jake, Tyuana and
Alex—for being there (by phone or in person) when I needed you. Mum, you have
sacrificed so much for me and I want you to know that I acknowledge and appreciate it.
The utmost thanks to Adam Mittman, Shabbi and Kika, my “grownup” family, for the
last 8 years together. You saw the field seasons, late nights, the tears and everything else
associated with this degree. Adam, my life with you and the girls means everything to me

and for that I dedicate this thesis to you.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES ............................................................................................................ x
CHAPTER ONE
HISTORICAL AND ECOLOGICAL REVIEW ON THE BROWN-HEADED
COWBIRD ........................................................................................................................... 1
1.1. Introduction ...................................................................................................... 2
1.2. Background ...................................................................................................... 4
1.3. Range ............................................................................................................... 4
1.4. Taxonomy ........................................................................................................ 5
1.5. Morphology and Plumage ................................................................................ 5
1.6. Diet ................................................................................................................... 6
1.7. Breeding ........................................................................................................... 6
1.8. Ecology .......................................................................................................... 10
1.9. Contemporary Issues ...................................................................................... 12
1.10.Research Needs .............................................................................................. 14
1.11.0bjectives ...................................................................................................... 16
1.12.Literature Cited .............................................................................................. 18
CHAPTER TWO
BROWN-BEADED COWBIRD RESPONSE TO VARYING PATCH SIZE AND AGE
IN KIRTLAND’S WARBLER HABITAT ....................................................................... 24
2.1 Introduction ..................................................................................................... 25
2.2 Study Area ...................................................................................................... 28
2.3 Methods ........................................................................................................... 30
2.3.1. Site selection ........................................................................................ 30
2.3.2. Avian surveys ....................................................................................... 32
2.3.3. Vegetation sampling ............................................................................ 33
2.3.4. Statistical analyses ............................................................................... 33
2.4. Results ............................................................................................................ 35
2.5. Discussion ...................................................................................................... 68
2.6. Literature Cited .............................................................................................. 75
CHAPTER THREE
THE EFFECTS OF PLACEMENT AND LAND COVER TYPE ASSOCIATIONS ON
BROWN-HEADED COWBIRD TRAP CAPTU RES ...................................................... 79
3.1. Introduction .................................................................................................... 80
3.2. Study Area ..................................................................................................... 82
3.3. Methods .......................................................................................................... 83
3.3.1. Data collection .................................................................................... 83
3.3.2. Statistical analyses .............................................................................. 89
3.4. Results ............................................................................................................ 91
3.5. Discussion ...................................................................................................... 96

vi

3.6. Literature Cited .............................................................................................. 99
CHAPTER FOUR
A HABITAT SUITABILITY INDEX MODEL FOR THE BROWN-HEADED
COWBIRD ....................................................................................................................... 102
4.1. Habitat Use Information .............................................................................. 103
4.1.1. General ............................................................................................... 103
4.1.2. Foraging ............................................................................................. 104
4.1.3. Breeding ............................................................................................. 105
4.1.4. Roosting ............................................................................................. 108
4.2. Habitat Suitability Index Model ................................................................... 108
4.2.1. Model development ........................................................................... 108
4.2.2. Model applicability ............................................................................ 109
4.2.3. Model variables .................................................................................. 109
4.2.4. Model description .............................................................................. 114
4.2.5. HSI determination .............................................................................. 117
4.3. Literature Cited ............................................................................................ 120
SUMMARY ........................................................................................ 125
5.1. Introduction .................................................................................................. 126
5.2. Discussion .................................................................................................... 126
5.3. Management Implications ............................................................................ 128
5.4. Limitations ................................................................................................... 129
5.5. Future Research ........................................................................................... 131

vii

LIST OF TABLES

Table 2.1. Mean (standard error) avian assemblages in jack pine size and age categories
in Montmorency and Presque Isle counties, Michigan, summer 2004 ..................... 36

Table 2.2. Mean (standard error) vegetation structural attributes of young, middle and
old jack pine stands sampled in Montmorency and Presque Isle Counties, MI, in
summer 2004. ........................................................................................................... 46

Table 2.3. Mean (standard error) vegetation structural attributes of small and large jack
pine stands sampled in Montmorency and Presque Isle Counties, MI, in summer
2004. ......................................................................................................................... 47

Table 2.4. Mean (standard error) vegetation structural attributes of jack pine stand
classes sampled in Montmorency and Presque Isle Counties, MI, in summer 2004.
(Small = < 81 ha, large = > 81 ha; young = 0 — 5 years, middle = 6 — 20 years, old =
> 20 years) ................................................................................................................ 48

Table 2.5. Correlations between stand vegetation characteristics and cowbird relative
abundance and density. ............................................................................................. 49

Table 2.6. Mean (standard error) values for avian assemblages in jack pine stand classes
in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small = < 81
ha, large = > 81 ha; young = 0 - 5 years, middle = 6 — 20 years, old = > 20 years).

.................................................................................................................................. 54

Table 2.7. Correlations between cowbird relative abundance and avian assemblages in
jack pine size and age classes in Montmorency and Presque Isle counties,
Michigan, summer 2004 (Small = > 81 ha, large = < 81 ha; young = O — 5 years,
middle = 6 - 20 years, old = > 20 years; Spearman rank correlation, a = 0.05). ..... 64

Table 2.8. Correlations between cowbird density and avian assemblages in jack pine size
and age classes in Montmorency and Presque Isle counties, Michigan, summer
2004 (Small = > 81 ha, large = < 81 ha; young = 0 - 5 years, middle = 6 — 20 years,
old = > 20 years; Spearman rank correlation, a = 0.05). .......................................... 65

Table 2.9. Correlations between cowbird relative abundance and avian assemblages in
jack pine stand classes in Montmorency and Presque Isle counties, Michigan,
summer 2004 (Small = > 81 ha, large = < 81 ha; young = O — 5 years, middle = 6 —
20 years, old = > 20 years; Spearman rank correlation, a = 0.05). ........................... 66

Table 2.10. Correlations between cowbird density and avian assemblages in jack pine
stand classes in Montmorency and Presque Isle counties, Michigan, summer 2004
(Small = > 81 ha, large = < 81 ha; young = 0 — 5 years, middle = 6 — 20 years, old =
> 20 years; Spearman rank correlation, a = 0.05). ................................................... 67

viii

Table 2.11. Occurrence of avian species in young (0 - 5 years), middle (6 - 20 years) and
old (> 20 years) —aged jack pine stands in Montmorency and Presque Isle counties,
Michigan, summer 2004. .......................................................................................... 69

Table 2.12. Occurrence of avian species in small (> 80 ha) and large (< 80 ha) jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. ......... 70

Table 2.13. Occurrence of avian species in jack pine stands in Montmorency and Presque
Isle counties, Michigan, during May-June (Period 1), June-July (Period 2), and July
(Period 3), 2004. ....................................................................................................... 71

Table 3.1. Reclassification for 1978, 1993 and 2001 Michigan land cover by land cover
type. Numbers in parenthesis indicate the land cover type classification code from
the year of land cover metadata. ............................................................................... 86

Table 3.2. Annual numbers of brown-headed cowbird traps used and corresponding
capture totals for 1977-79, 1992-94 and 2000-02 in northern Lower Michigan. ..... 90

Table 3.3. Mean (SE) area (ha) of land cover types within a 4 km radius of brown-
headed cowbird traps during 1977-79, 1992-94, and 2000-02 in northern Lower
Michigan. .................................................................................................................. 92

Table 3.4 . Correlations between brown-headed cowbird trap captures from 1977-79,

1992-94, and 2000-02 and associated land cover types using three years of
Michigan land cover. (Spearman rank correlation, a = 0.05) .................................. 95

ix

LIST OF FIGURES

Fig. 2.1. Clear Lake State Wildlife Area in Presque Isle and Montmorency Counties,
Michigan. .................................................................................................................. 29

Fig. 2.2. Cowbird relative abundance amongst three age classes of jack pine stands in
Montmorency and Presque Isle counties, Michigan, summer 2004. (Young = 0 - 5
years; middle = 6 - 20 years; old = > 20 years; Kruskal-Wallis nonparametric
ANOVA, a = 0.05). Same letters indicate no significant differences. Point
indicates mean, box denotes one standard error of the mean, and whiskers indicate
two standard errors of the mean. ............................................................................... 37

Fig. 2.3. Cowbird relative abundance in two size classes of jack pine stands in
Montmorency and Presque Isle counties, Michigan, summer 2004. (small = < 81 ha;
large = > 81 ha; Mann-Whitney—U, a = 0.05). Different letters indicate significant
differences. Dot indicates mean, box denotes one standard error of the mean, and
whiskers indicate two standard errors of the mean. .................................................. 38

Fig. 2.4. Figure 2.4. Cowbird relative abundance amongst six size-age classes of jack
pine stands in Montmorency and Presque Isle counties, Michigan, summer 2004.
(Small = < 81 ha; large = > 81 ha; young = 0 - 5 years; middle = 6 - 20 years; old =
> 20 years; Kruskal-Wallis nonparametric ANOVA, a = 0.05). Same letters
indicate no significant differences. Point indicates mean, box denotes one standard
error of the mean, and whiskers indicate two standard errors of the mean. ............. 39

Fig. 2.5. Cowbird relative abundance in jack pine stands for three sampling periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (1 = May-June,
2 = June — July, 3 = July). Kruskal-Wallis nonparametric analysis of variance, (1 =
0.05. Same letters indicate no significant differences. Dot indicates mean, box
denotes one standard error of the mean, and whiskers indicate two standard errors
of the mean ................................................................................................................ 40

Fig. 2.6. Cowbird density (# individuals/ha sampled) in three age classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Young
= 0 -5 years; middle = 6 - 20 years; old = > 20 years). Kruskal-Wallis
nonparametric analysis of variance, (1 = 0.05. Same letters indicate no significant
differences. Dot indicates mean, box denotes one standard error of the mean and
whiskers indicate two standard errors of the mean. .................................................. 41

Fig. 2.7. Cowbird density (# individuals/ha sampled) in two size classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small
= < 81 ha; large = > 81 ha). Mann-Whitney-U, a = 0.05. Same letters indicate no
significant differences. Dot indicates mean, box denotes one standard error of the
mean, and whiskers indicate two standard errors of the mean. ................................ 42

Fig. 2.8. Cowbird density (#Iha sampled) amongst six size-age classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small
= < 81 ha; large = > 81 ha; young = 0 - 5 years; middle = 6 - 20 years; old = > 20
years; Kruskal-Wallis nonparametric ANOVA, a = 0.05). Same letters indicate no
significant differences. Point indicates mean, box denotes one standard error of the
mean, and whiskers indicate two standard errors of the mean. ................................ 44

Fig. 2.9. Cowbird density (# individuals/ha sampled) in jack pine stands over three
sampling periods in Montmorency and Presque Isle counties, Michigan, summer
2004. (1 = May-June, 2 = June — July, 3 = July). Kruskal-Wallis nonparametric
analysis of variance, (1 = 0.05. Same letters indicate no significant differences. Dot
indicates mean, box denotes one standard error of the mean, and whiskers indicate
two standard errors of the mean ................................................................................ 45

Fig. 2.10. Host species abundance (# of individuals) in jack pine stands in three age
categories in Montmorency and Presque Isle counties, Michigan, summer 2004.
(Young = 0 -5 years, middle = 6 — 20 years, old = > 20 years). One-way analysis of
variance, (1 = 0.05). Same letters indicate no differences. Dot indicates mean and
whiskers denote 95% confidence interval from the mean. ....................................... 50

Fig. 2.11. Host species abundance (# of individuals) in jack pine stands in two size
categories in Montmorency and Presque Isle counties, Michigan, summer 2004.
(Small = < 81 ha; large = > 81 ha). One-way analysis of variance, (1 = 0.05.
Different letters indicate significant difference. Dot indicates mean and whiskers
denote 95% confidence intervals from the mean. ..................................................... 51

Fig. 2.12. Host species abundance (# of individuals) in jack pine stands over three time
periods in Montmorency. and Presque Isle counties., Michigan, summer 2004. (1 =
May-June, 2 = June — July, 3 = July). One-way analysis of variance, (1 = 0.05.
Same letters indicate no significant differences. Dot indicates mean and whiskers
denote 95% confidence interval from the mean. ...................................................... 53

Fig. 2.13. Host species density (# of individuals/ha sampled) in jack pine stands in three
age categories in Montmorency Co. and Presque Isle Co., MI, summer 2004.
(Young = 0-6 years; middle = 6-20 years; old = > 20 years). One-way analysis of
variance, Tukey’s multiple comparison test, a = 0.05. Different letters denote
significant differences; shared letters signify no significant differences. Dot
indicates mean and whiskers denote 95% confidence interval from the mean. ....... 55

Fig. 2.14. Host species density (# of individuals/ha sampled) in jack pine stands in two
size categories in Montmorency and Presque Isle counties., Michigan, summer
2004. (Small = < 81 ha, large = > 81 ha). One—way analysis of variance, (1 = 0.05.
Same letters indicate no significant difference. Dot indicates mean and whiskers
denote 95% confidence interval from the mean. ...................................................... 56

xi

Fig. 2.15. Host species density (# individuals/ha sampled) in jack pine stands during
three sampling periods in Montmorency Co. and Presque Isle Co., MI, summer
2004. (1 = May-June, 2 = June — July, 3 = July). One-way analysis of variance,
Tukey’s multiple comparison test, a = 0.05. Different letters indicate significant
differences; shared letters denote no difference. Dot indicates mean and whiskers
denote 95% confidence interval from the mean. ...................................................... 57

Fig. 2.16. Avian species richness in jack pine stands in three age categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. (Young = 0 - 5
years, middle = 6 — 20 years, old = > 20 years). One-way analysis of variance, (1 =
0.05. Same letters indicate no differences. Dot indicates mean and whiskers denote
95% confidence interval from the mean. .................................................................. 58

Fig. 2.17. Avian species richness in jack pine stands in two size categories in
Montmorency Co. and Presque Isle Co., MI, summer 2004. (small = < 81 ha; large
= > 81 ha). One-way analysis of variance, (1 =0.05. Different letters denote
significant difference. Dot indicates mean and whiskers denote 95% confidence
interval from the mean. ............................................................................................. 59

Fig. 2.18. Avian species richness in jack pines stands during three time periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (l = May-June,
2 = June — July, 3 = July). One-way analysis of variance, a = 0.05. Shared letters
indicate no significant differences. Dot indicates mean and whiskers denote 95%
confidence interval from the mean. .......................................................................... 60

Fig. 2.19. Avian species diversity in jack pine stands in two size categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. (small = < 81
ha; large = > 81 ha). Shannon-Weaver diversity indices, one-way analysis of
variance, (1 = 0.05. Different letters indicate significant difference. Dot indicates
mean and whiskers denote 95% confidence interval from the mean ........................ 61

Fig. 2.20. Avian species diversity in jack pine stands in three age categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. Shannon-
Weaver diversity indices, one-way analysis of variance, (1 = 0.05. (Young = 0 - 5
years, middle = 6 — 20 years, old = > 20 years). Same letters indicate no significant
differences. Dot indicates mean and whiskers denote 95% confidence interval from
the mean. ................................................................................................................... 62

Fig. 2.21. Avian species diversity in jack pine stands during three sampling periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (1 = May-June,
2 = June — July, 3 = July). Shannon-Weaver diversity indices, one-way analysis of
variance, (1 = 0.05. Same letters indicate no significant differences. Dot indicates
mean and whiskers denote 95% confidence interval from the mean. ....................... 63

xii

Fig. 3.1. Shaded area indicates the 12 counties in Michigan where the Cowbird Control
Program has been in effect since 1972. .................................................................... 84

Fig. 3.2. Mean (SE) cowbird trap captures in the Kirtland’s warbler management area
during 1977-79 (Period 1), 1992-94 (Period 2), 2000-02 (Period 3). One-way
ANOVA,Tukey-Kramer multiple comparison test, a = 0.05. *indicates a significant
difference within a period. Dot indicates mean and whiskers denote 95%
confidence intervals from the mean. ......................................................................... 93

Fig. 3.3. Mean cowbird captures per trap in the Kirtland’s warbler management area
during the years 1977-79 (Period 1), 1992-94 (Period 2) and 2000-02 (Period 3).
One-way ANOVA, Tukey-Kramer multiple comparison test, a = 0.05. Different
letters indicate significant differences amongst years. ............................................. 94

Fig. 4.1. Suitability index of distance to agriculture (m; variable 1) for brown-headed
cowbirds in forest systems in northern Michigan. .................................................. 112

Fig. 4.2. Suitability index of distance to edge (m; variable 2) for brown-headed cowbirds
in forest systems in northern Michigan ................................................................... 113

Fig. 4.3. Suitability index for percent (%) ground cover (variable 3) for brown-headed
cowbirds in forest systems in northern Michigan. .................................................. 115

Fig. 4.4. Suitability index model for percent (%) mid-story cover (variable 4) for brown-
headed cowbirds in forest systems in northern Michigan ....................................... 116

Fig. 4.5. Relationships between habitat variables, life requisites and cover types for the
application of the H81 model for brown-headed cowbird in northern Michigan. ..118

xiii

CHAPTER ONE

HISTORICAL AND ECOLOGICAL REVIEW ON THE BROWN-HEADED

COWBIRD (M OLOTHR US ATER)

1.1. Introduction

Since the turn of the century, there has been increased concern amongst avian
ecologists and natural resource managers regarding the effect the brown-headed cowbird
(Molothrus ater) has had, and continues to have, on songbird species throughout North
America (Mayfield 1961, Ambuel and Temple 1983, Brittingham and Temple 1983,
Rothstein 1995, Burhans et a1. 2000. The brown-headed cowbird has affected upwards
of 220 native passerine species, and is likely North America’s most well known
generalist brood parasite (Rothstein 1990, Lowther 1993, Ortega 1998). Historically
native to the Great Plains, the brown-headed cowbird began to expand its range in the
late-l9th century as forested landscapes were cleared for development and agriculture and
with the introduction of livestock (Ehrlich et a1. 1988, Ortega 1998). In the last century,
the brown-headed cowbird’s range has expanded to the entire continental United States
(Lowther 1993) and the species has become the most ubiquitous bird in North America
(Rothstein 1995). However, recent surveys suggest that brown-headed cowbird
populations are declining nationwide, except in the Great Plains region (Peterjohn et a1.
2000, Rothstein and Robinson 2000, Wiedenfeld 2000). Reasons for the downward trend
at the nationwide level are unclear.

Much of the current research on brown-headed cowbirds has focused on
endangered avian species whose populations have been subject to brown-headed cowbird
parasitism. While there have been conflicting perspectives about the degree to which
parasitism can impact such species (Mayfield 1961, Rye] 1981, Probst 1986, Rothstein
and Cook 2000), most studies agree that brown-headed cowbirds have the potential to

drive rare species to extinction, especially when parasitism is combined with a loss of

natural habitat (DeCapita 2000, Griffith and Griffith 2000, Hayden et al. 2000, Rothstein
and Cook 2000, Whitfield 2000).

As a result, several federal brown-headed cowbird control programs have been
instituted across the United States since 1972 to recover or maintain endangered avian
species (DeCapita 2000). While brown-headed cowbird trapping has reduced cowbird
population numbers and parasitism rates, few programs have resulted in increased host
populations (Rothstein and Cook 2000). C6té and Sutherland (1997) had similar findings
in their study on predator removal programs in avian communities. In addition, resilience
amongst brown-headed cowbirds is prevalent, as captures in one year do not appear to
affect the next year’s population (DeCapita 2000, Griffith and Griffith 2000, Hayden et
al. 2000).

For these reasons, continued research into brown-headed cowbird behavior and
ecology is essential for creating more effective and practical management plans. A better
understanding of the species might also contribute to a more balanced perspective on it
by scientists and the public, who have often labeled the bird with such biased and
erroneous terms as “social outcast” (Pearson et al. 1942: 354), “feathered wretch”
(Pearson and Burroughs 1923: 246) and “arch-villains” (Holmes 1993: 1514). The
brown-headed cowbird has achieved a poor reputation due to its impact on native avian
species, but also due to a lack of effective communication about its history, ecology and
status. An impartial approach combined with concrete scientific data will offer the best

possibility for resolving the ecological issues surrounding the species.

1.2. Background
Most commonly referred to as the brown-headed cowbird, the species has also
been known by several regional names, such as the “brown-headed oriole,” “cow

,9 66

bunting, sheep blackbird” and “cuckold,” among others (Ortega 1998). Originally
known as the “buffalo bird” because of its historical association with the bison (Bison
bison) that roamed the Great Plains, the brown-headed cowbird followed these herds to
forage for insects flushed during movement (Lowther 1993, Goguen and Matthews
1999). As Europeans settled the United States in the 19th century, prairies and forests
were converted to agricultural and pasture lands, which increased brown-headed cowbird
habitat and led to an expansion in its range (Ortega 1998). When domestic cows were
introduced to the altered landscape, the brown-headed cowbird formed a relationship with
them similar to the one it had previously shared with bison (Goguen and Matthews 1999).
Once rare to areas beyond the Great Plains, the brown-headed cowbird thrived under the
management practices—including widespread fire suppression and intensive logging

harvests—that characterized late-19th century American forestry, and expanded its range

throughout the continental United States (Ortega 1998).

1.3. Range

Native to the Great Plains region of North America, the brown-headed cowbird
has increased its range to cover most of the continent south of the Arctic; however, it is
rarely found outside of North America (Lowther 1993, McKay 1994, Stokes and Stokes
1996). In the United States, the brown-headed cowbird is found year round along the

Pacific coastline, throughout the southwest, the Gulf Coast and mid-Atlantic regions and

along the entire eastern seaboard, with the exception of Florida (Hoffman and
Woolfenden 1986). The brown-headed cowbird’s breeding range spans across southern
Canada, and southward through the entire continental United States, excluding the
Florida panhandle (Hoffman and Woolfenden 1986, Walkinshaw 1991). The brown-
headed cowbird winters in Baja California, southern Mexico and the Florida panhandle
(Hoffman and Woolfenden 1986). In Michigan, the brown-headed cowbird’s breeding
range has been confirmed in all but a few counties; these areas are considered probable

breeding sites (Walkinshaw 1991).

1.4. Taxonomy

The genus Molothrus includes five of the six cowbird species, including the
brown-headed cowbird, the shiny cowbird (M. bonariensis), the screaming cowbird (M.
rufoaxillaris), the bay-winged cowbird (M. badius), and the bronzed cowbird (M. aenus).
Only the giant cowbird (Scaphidura oryzivora) lies outside the genus (Ortega 1998).
There are three recognized subspecies of M. ater, including M. a. ater, which breeds in
the eastern half of the United States, from Wisconsin south to Texas, M. a. artemisiae,
which breeds from Washington east to Minnesota, and south to New Mexico and
Arizona, and M. a. obscurus, which breeds along the Pacific coast southeast to Texas and

Louisiana (Rothstein 1978, Ortega 1998).

1.5. Morphology and Plumage
The brown-headed cowbird is a relatively small blackbird, averaging

approximately 18 cm in length (Petersen 1980, Stokes and Stokes 1996). The species

displays moderate sexual dimorphism—the male is shiny, iridescent black, with a
recognizable brown head and dark-grey sparrow-like bill; in contrast, the female is grey,
lacking the brown head, and has a short, fmch-like bill (Petersen 1980). Adult females
are also approximately 10% smaller than their male counterparts (Lowther 1993).
Juvenile brown-headed cowbirds are pale grey with light-colored breast streaks; juvenile
males are commonly seen with irregular color patterns during late summer molting

(Petersen 1980).

1.6. Diet

Brown-headed cowbirds are known to feed on seeds and insects, including
grasshoppers, caterpillars and beetles (Lowther 1993, Ortega 1998). Female brown-
headed cowbirds will also intake protein by eating host eggs removed from parasitized
nests (Jackson and Roby 1992). Ortega (1998) reported that females also consume large

amounts of mollusk shells.

1.7. Breeding

The genus name Molothrus is Latin for “intruder,” an accurate description of the
breeding behavior of many cowbird species (Ortega 1998). The brown-headed cowbird
is a generalist obligate brood parasite—it is not a host specialist, but rather parasitizes a
variety of species, and parasitism is necessary because the brown-headed cowbird cannot
build its own nest (Ortega 1998). Brown-headed cowbirds are amongst the 1% of avian
species that take part in brood parasitism, including honeyguides (Indicatoridae),

cuckoos (Cuculidae), two genera of finches (Vidua and Anomalospiza, Ploceidae), one

duck (Heteronetta atricapilla, Anatidae) and four additional cowbird (Molothrus) species
(Payne 1977). Payne has suggested that brood parasitism rrright have succeeded from
“nest takeovers,” where a parental species inhabits an abandoned nest or takes over an
active nest, and subsequently cares for its offspring there. Friedmann (1929) constructed
a phylogeny that situated the bay-winged cowbird, a non-parasitic cowbird that often uses
other species’ nests, as the ancestral species. From there, the screaming cowbird, the
only brood specialist (meaning that it parasitizes a particular host species) evolves,
followed by the brood generalists (parasitizes several host species). Gill’s (1995)
research supports this model, and shows the brown-headed cowbird at the most recent
end of the phylogeny. Gill’s model, based on DNA sequencing studies, indicates that
that there is a positive relationship between how recently the particular cowbird species
has evolved and the number of host species it parasitizes. Accordingly, the brown-
headed cowbird, being the most recently evolved species in the Molothrus genus, also
parasitizes the greatest number of hosts.

In field studies, brown-headed cowbird mating behavior varies by geographic
region, and may depend on the availability of sexual partners (Lowther 1993). For
example, monogamy occurs most frequently when sex ratios are high, whereas
promiscuity between the sexes is more often exhibited when sex ratios near 1:1 (Teather
and Robertson 1986). Similarly, any pair bonds that are formed have variable lengths,
lasting from a few days to several years (Lowther 1993).

When a female cowbird is ready to lay her eggs, she will observe potential host
nests for days before she actually deposits her eggs (Norman and Robertson 1975). In

some cases, however, she may try to “flush out” hosts by flying around and vocalizing in

the vegetation (Norman and Robertson 1975). Once the female cowbird locates a host
nest, she often removes and eats one or more of the eggs in the nest—which, on some
occasions, are cowbird eggs laid earlier by other females—prior to laying her own (Harm
1941). If, however, < 2 eggs are present in the nest, she will not remove any, as the host
bird might potentially abandon an eggless nest (Harm 1941, Ehrlich et al. 1988, Wood
and Bollinger 1997).

A female cowbird can lay eggs almost daily during the breeding season,
averaging 40 eggs per year, due to her long reproductive cycle and short intervals
between clutches (Scott and Ankney 1983). While a female usually will lay only one egg
per host nest, in instances where nests are scarce she will lay more (Harm 1941, M.
DeCapita, US. Fish and Wildlife Service (USFWS), pers. commun.). The cowbird egg is
grayish-white to white, speckled with brown and/or gray spots and has a moderately
glossy surface (Lowther 1993). Walkinshaw (1991) noted that cowbirds seem to prefer
laying their eggs in nests with similar-looking eggs.

Some species are able to defend cowbird parasitism. For example, the American
robin (Turdus migratorius) and gray catbird (Dumetella carolinensis) may be able to
recognize the invasive egg and remove it; the yellow warbler (Dendroica petechia) builds
a new nest on top of the old one; or some birds desert the nest altogether (Graham 1988,
Lowther 1993, Burhans et al. 2000). Aggression toward cowbirds and nest defense
behavior by host species has also been observed (Robertson and Norman 1976, Gill et. al
1997). However, approximately 65% of cowbird targets do not identify the cowbird egg,
and consequently become hosts by incubating it along with their own (Ehrlich et al.

1988). The degree to which a host is fit for parasitism contributes considerably to the

success of the cowbird egg proceeding to adulthood; different species can be acceptors,
partial acceptors, or partial rejecters (M. DeCapita, pers.commun., Rothstein 1975).
While some hosts can adequately raise a cowbird chick into adulthood, many other
species are not suitable foster parents (Ehrlich et al. 1988). Research has suggested that
female cowbirds can recognize suitable host species from unsuitable ones, and will
sometimes purposely avoid the latter (Burhans et al. 2000, Woolfenden et al. 2003).

The impact on the host species is most substantial after the cowbird chick hatches.
Cowbird eggs hatch at least one day prior to those of the host parent (Gill 1995).
Because of this timing advantage, the cowbird chick typically is larger than its nest
mates, thereby dominating the attention of the host parent and receiving more food
(Woodward 1983). This leads to a cowbird chick that is notably larger and healthier than
its nest mates. Some evidence has shown that young cowbirds will remove host eggs or
young from the nest (Gill 1995, Dearbom 1996), but Ehrlich et al. (1988) and Ortega
(1998) maintain that such behavior is not typical. However, Kilner’s (2003) research
suggests that cowbird chicks only tend to outcompete host nestlings of much smaller
birds; nestlings of larger species such as the northern cardinal (Cardinalis cardinalis)
have a greater likelihood of faring well alongside the cowbird chick. While the
parasitization process is a regular occurrence, only about 3% of all cowbird eggs actually
result in a healthy adult bird (Scott and Ankney 1980). However, due to the high number
of eggs laid each year by individual females, this 3% is a remarkably large number,

capable of affecting many host species.

1.8. Ecology

Various studies suggest that cowbird presence and abundance is associated with
several variables, including proximity to agricultural areas, proximity to edge (distance
from stand interior to stand edge), and host abundance. Most studies cite distance to
agriculture to be the most limiting factor (Stribley and Haufler 1999, Tewksbury et al.
1999, Young and Hutto 1999). However, Donovan et al. (2000) suggested that cowbirds
experience trade-offs as the increasing availability of one variable resulted in the
decreasing availability of another. For example, increased access to agricultural areas
resulted in more feeding opportunities for cowbirds, but this also meant they had less
access to host nests, which were more often found in forested areas (Strausberger and
Ashley 1997, Robinson et al. 2000). In this case, access to a host nest was the limiting
factor, as cowbird reproductive success was dependent on it. Donovan et al. (2000) cited
a second situation that can be problematic for the species. Abundant edge contributed to
suitable habitat for cowbirds as they commuted between feeding and breeding areas, but
increasing edge was also harmful to cowbird offspring as the presence of nest predators
became more likely (Andren and Angelstam 1988).

Both spatial and temporal scales are central to evaluating cowbird breeding and
feeding habitats (Rothstein et a1. 1984, Thompson 1994). Individuals often lay eggs in
forested systems in early morning, and feed in the afternoon in grassland and agricultural
areas. Depending on the topography of a given region, the cowbird will be considerably
more likely to breed in areas that are within 4 km of agriculture (Stribley and Haufler
1999, Tewksbury et al. 1999), although Young and Hutto (1999) found cowbirds in

breeding habitat up to 10 km away from agricultural areas. In addition, Curson et al.

10

(2000) noted a commute of >11 km between breeding and feeding sites, and >18 km
between roosting and breeding sites. Goguen and Mathews (1999) concluded that the
presence of feeding sites—in their study, pasture lands—was related to increased cowbird
density. Tewksbury et al. (1999) had similar findings that cowbird potential is limited by
distance between breeding and feeding sites.

The brown-headed cowbird is considered an edge species, meaning that it favors
the area bordering two different vegetation types. Because of this, distance from
breeding areas to edge probably is an important influence on cowbird reproductive
success. Brittingham and Temple (1983) noted that brood parasitism is positively
associated with the amount of available edge, and forest openings as small as 1-2 ha can
initiate a cowbird invasion. Robinson et al. (1995) researched the effects of landscape
fragmentation on parasitism, and found parasitism increased with forest fragmentation on
nine different landscapes. Similar findings resulted from Donovan et al.’s (1997) study,
where cowbird abundance was highest in edges in highly fragmented landscapes.
Farmer’s (1999) study suggested that cowbirds prefer edge habitat and are more abundant
there. Stribley and Haufler (1999) found that edge was an important determinant of
cowbird density in a forest stand, and areas >300 m from an edge were less likely to have
a cowbird presence.

Finally, the density of potential cowbird hosts is an important indicator of the
presence and abundance of cowbird populations. Young and Hutto’s (1999) study
indicated that host density was an important signifier of cowbird presence. Tewksbury et
al. (1999) found similar results in their landscape-scale study. Findings by Donovan et

al. (1997) showed that cowbirds were positively associated with host abundance in

11

interior vegetation. However, Purcell and Vemer (1999) concluded that host species
richness (as opposed to density or abundance) was a better predictor of increased cowbird
presence in a given area.

It should be noted that vegetation type and regional variation can play a role in
which ecological variables contribute to cowbird presence and abundance. While neither
study location nor vegetation type seem to influence the importance that proximity to
agriculture has on cowbird presence (Goguen and Mathews 1999, Hejl and Young 1999,
Stribley and Haufler 1999, Tewksbury et. al. 1999, Young and Hutto 1999), this is not
necessarily the case with variables such as edge proximity and host density. Their role in
predicting cowbird presence might be dependent on what area of the country and in what
vegetation type the study takes place (Purcell and Vemer 1999, T ewksbury et al. 1999).
Furthermore, in the case of Stribley and Haufler (1999), proximity to edge was only

significant when sites were located proximal to agriculture.

1.9. Contemporary Issues

One of the species most severely affected by the cowbird’s range expansion is the
Kirtland’s warbler (Dendroica kirtlandii; Probst 1986, Huber et al. 2001). Found only in
the northern part of Michigan’s Lower Peninsula during the breeding season, the
Kirtland’s warbler has been on the federal Endangered Species list since 1973 due to its
persistently unstable population numbers
(https:l/ecos.fws.gov/species_profile/SpeciesProfile?spcode=B03I). The Kirtland’s
warbler population reached an all-time low of 167 singing males in 1987, renewing

concerns that the species was headed to extinction (Deloria 2000, Huber 2001). Due to

12

the cowbird’s ubiquity and its threat to the Kirtland’s warbler, the USFWS launched a
Michigan cowbird control program in 1972; decoyed traps designed to capture cowbirds
were placed in Kirtland’s warbler breeding areas (Deloria 2000, Huber et al. 2001). As
of 2006, the USFWS has captured and removed 137,477 cowbirds from Kirtland’s
warbler habitat (C. Mensing, USFW S, pers. commun.). Agency experts theorize that the
need for these traps will continue into the foreseeable future (Huber et al. 2001).

In conjunction with the threat of cowbirds, the warblers also have been faced
historically with the loss of breeding habitat, comprising mid-aged (6-20 years old) jack
pines (Pinus banksiana; Deloria 2000). To re-establish a sustainable Kirtland’s warbler
population, the US. Forest Service (USFS) began working in 1964 in Michigan’s
northeastern Lower Peninsula to restore jack pine forests (Huber et al. 2001). The
primary management techniques employed include whole-tree harvesting and replanting
jack pine stands on a 50-year rotation (P. Huber, USFS, pers commun., Deloria 2000).
Because jack pine is a fire-dependent species, management techniques are used to mimic
the effects of naturally occurring wildfire, which is typically suppressed in areas with
human populations (Huber et al. 2001). Management for Kirtland’s warbler habitat has
traditionally occurred on federal and state lands (Huber et al. 2001), although recently it
has also occurred in private lands.

Overall, Kirtland’s warbler populations have steadily risen through the years. The
official count in the 2006 census recorded 1,478 singing males, up from 1,415 in 2005
(http://www.michigan.gov/dnr/0,1607,7-153-10371_10402-148280--,00.html). Cowbird
control and habitat management have played key roles in this increase (DeCapita 2000).

Conversely, brown-headed cowbird populations in Michigan have decreased in recent

13

years, for reasons not yet completely understood (Sauer et al. 2003, M. DeCapita, pers
commun.). While trapping has likely been an important factor in cowbird population
declines, the resilience of this species is such that other factors, such as changes to the
landscape, may also be at work. However, experts from all three agencies (Michigan
Department of Natural Resources (MDNR), USFWS, USFS) agree that because of the
Kirtland’s warbler’s delicate population status, continued management for their breeding
habitat and brown-headed cowbird control will be necessary for an undetermined time

(Huber et al. 2001).

1.10. Research Needs

While many studies have examined the impact of cowbird parasitism on passerine
species and the factors that lend to cowbird survival and reproductive success, research
into the effects of forest management on cowbird populations is lacking. Currently, the
USFS and MDNR are managing large tracts of jack pine forest to provide ample habitat
for the Kirtland’s warbler and to meet commercial timber needs (Huber et al. 2001).
While there is a great deal of information about the impacts of habitat, recreational and
commercial management on the Kirtland’s warbler, there is a need for more knowledge
about how jack pine management affects cowbird populations.

In the Kirtland’s warbler management area, jack pine clearcuts range in size from
<1 ha to >400 ha. While the most suitable Kirtland’s warbler habitat comprises jack pine
stands of at least 400 ha (Huber et al. 2001), much smaller stands are often maintained for
the species (Rex Ennis, USFS, pers commun.). The issue of patch size becomes

important to cowbirds as more small patches are introduced across the landscape in

14

addition to, or in favor of large patches. Because an accumulation of small patches can
create more fragmentation and edge than a large patch of equivalent area, higher
abundances of cowbirds might be observed (Ambuel and Temple 1983, Gustafson and
Crow 1994, Robinson et al. 1995, Sisk and Haddad 2002). A landscape-level analysis of
the response of brown-headed cowbirds to the total available edge in the study location
would be an informative approach; however, I believe that an analysis of cowbird
response to suitable and unsuitable patch sizes in Kirtland’s warbler habitat was an
important study question. Therefore, a patch-level analysis will be used to determine
cowbird response to individual patch sizes, as was done by Saab (1999), Horn (2000),
and Johnson and lg] (2001), amongst others. Based on the cowbird ecology literature, I
would expect that cowbirds would be more abundant across a landscape that is
fragmented by numerous small patches than in a landscape with a few large patches.
However, in a given patch, cowbird abundance will likely be greater in large patches due
simply to increased area.

Numerous studies have demonstrated that cowbirds are strongly associated with
short-grass and agricultural systems (Ortega 1998) and at times, clearcut areas (Gustafson
and Crow 1994). However, they are also dependent on early-successional and forested
habitats for breeding activity (Rothstein et al. 1984, Thompson 1994). Because all of
these systems are represented over the life span of a jack pine stand managed for
Kirtland’s warblers, it is essential to determine the effects that stand age have on the
cowbird population. Knowing whether the cowbird is most highly associated with young

(clearcut), middle-aged (early successional) or old (mature forest) stands can help

15

managers to assess the current clearcut rotation cycle, as well as to consider other avian
species that depend on the different habitats.

It will also be useful to determine how location of the USFWS’s cowbird traps is
related to the capture numbers. The USFWS collects information armually about trap
locations, the number of cowbirds caught at each trap, and the total number of cowbirds
caught each year (M. DeCapita, pers commun.). However, no analysis has been done to
determine whether there is a relationship between the land cover types that surround the
trap locations and the number of cowbirds caught at each trap. The results from this type
of analysis would be helpful to the USFWS on optimal trap placement locations to
maximize the catch rate.

Finally, ecologists often use Habitat Suitability Index (HSI) models to quantify
the optimal habitat conditions for a particular species to meet its life history requirements.
Currently, no such model exists for the brown-headed cowbird, despite several studies
that have suggested factors that promote cowbird success. Such a model is necessary for
managers when attempting to assess cowbird control techniques, and when managing

areas for habitat or commercial harvesting.

1.11. Objectives
The purpose of this project is to:
1. Determine the effects that size and age characteristics of jack pine stands
managed for Kirtland’s warblers have on the relative abundance and density

of brown-headed cowbirds,

16

2. Evaluate the influence of land cover types surrounding USFWS cowbird traps
on the numbers of cowbirds captured, and

3. Build a Habitat Suitability Index (HSI) model for the brown-headed cowbird.

17

l. 12. Literature Cited

Ambuel, B., and S. A. Temple. 1983. Area-dependent changes in the bird communities
and vegetation of southern Wisconsin forests. Ecology 64 (5): 1057-1068.

Andren, H., and P. Angelstam. 1988. Elevated predation rates as an edge effect in
habitat islands: experimental evidence. Ecology 69 (2): 544-547.

Brittingham, M. C., and S. A. Temple. 1983. Have brown-headed cowbirds caused
forest songbirds to decline? Bioscience 33: 31-35.

Burhans, D. E., F. R. Thompson, 111, and J. Faaborg. 2000. Costs of parasitism incurred
by two songbird species and their quality as cowbird hosts. The Condor 102:
364-373.

Cote, I. M., and W. J. Sutherland. 1997. The effectiveness of removing predators to
protect bird populations. Conservation Biology 11(2): 395-405.

Curson, D. R., C. B. Goguen, and N. E. Mathews. 2000. Long-distance commuting by
brown-headed cowbirds in New Mexico. The Auk 117(3): 795-799

Dearbom, D. C. 1996. Video documentation of a brown-headed cowbird nestling
ejecting an indigo bunting nestling from the nest. The Condor 98: 645-649.

DeCapita, M. E. 2000. Brown-headed cowbird control on Kirtland’s warbler nesting
areas in Michigan, 1972-1995. Pages 333-341 in J. N. M. Smith, T. L. Cook, S. I.
Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management of
cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Deloria, C. M. 2000. Kirtland’s warbler diet and its relationship to warbler age, sex, and
jack pine stand characteristics. M.S. thesis, Michigan State University, East
Lansing, Michigan, USA.

Donovan, T. M., P. W. Jones, E. M. Armand, and F. R. Thompson, III. 1997. Variation
in local-scale edge effects: mechanisms and landscape context. Ecology 78(7):
2064-2075.

Donovan, T. M., F. R. Thompson, 111, and J. R. Faaborg. 2000. Cowbird distribution at
different scales of fragmentation: trade-offs between breeding and feeding
opportunities. Pages 255-264 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K.
Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

18

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The birder’s handbook: A field
guide to the natural history of North American birds. Simon and Schuster, Inc.,
New York, New York, USA.

Farmer, C. 1999. The density and distribution of brown-headed cowbirds: the Central
coastal California enigma. Studies in Avian Biology. 18: 62-67.

Friedmann, H. 1929. The cowbirds, a study in the biology of social parasitism. C. C.
Thomas, Springfield, Illinois, USA.

Gill, F. B. 1995. Ornithology. W. H. Freeman and Company. New York, New York,
USA

Gill, S. A., D. L. Neudorf, and S. G. Sealy. 1997. Host responses to cowbirds near the
nest: cues for recognition. Animal Behaviour 53: 1287-1293.

Goguen, C. B., and N. E. Mathews. 1999. Review of the causes and implications of the
association between cowbirds and livestock. Studies in Avian Biology. 18: 10-
17.

Graham, D. S. 1988. Responses of five host species to cowbird parasitism. The Condor
90: 588-591.

Griffith, J. T., and J. C. Griffith. 2000. Cowbird control and the endangered least Bell’s
vireo: a management success story. Pages 342-356 in J. N. M. Smith, T. L.
Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Gustafson, E. J ., and T. R. Crow. 1994. Modeling the effects of forest harvesting on
landscape structure and the spatial distribution of cowbird brood parasitism.
Landscape Ecology 9(4): 237-248.

Harm, H. W. 1941. The cowbird at the nest. Wilson Bulletin 53(4): 209-221.

Hayden, T. J., D. J. Tazik, R. H. Melton, and J. D. Cornelius. 2000. Cowbird control
program at Ft. Hood, Texas: lessons for mitigation of cowbird parasitism on a
landscape scale. Pages 357-370 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S.
K. Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

Hejl, S. J. and J. S. Young. 1999. Brown-headed cowbirds in ponderosa pine/Douglas-

fir dominated landscapes in the northern Rocky Mountains. Studies in Avian
Biology 18: 73-79.

19

Hoffman, W., and G. E. Woolfenden. 1986. A fledgling brown-headed cowbird
specimen from Pinellas County. Florida Field Naturalist. 14:18-20.

Holmes, B. 1993. An avian arch-villain gets off easy. Science 262: 1514-1515.

https://ecos.fws.gov/species_profile/SpeciesProfile?spcode=B03I. Retrieved September
21, 2006.

http://wwwmichigan.gov/dnr/O,1607,7-153-10371_10402-148280--,00.html. Retrieved
September 21 , 2006.

Huber, P. W., J. A. Weinrich, and E. S. Carlson. 2001. Strategy for Kirtland’s warbler :
habitat management. Publication of MDNR, USFS, and USFWS. Mio,
Michigan, USA.

Jackson, N. H., and D. D. Roby. 1992. Fecundity and egg-laying patterns of captive
yearling brown-headed cowbirds. Condor 94: 585-589.

Kilner, R. M. 2003. How selfish is a cowbird nestling? Animal Behaviour. 65: 1-8.

Lowther, P. E. 1993. Brown-headed cowbird (Molothrus ater). Pages 1-24 in A. Poole
and F. Gill, editors. The birds of North America, No. 47. The Academy of
Natural Sciences, Philadelphia, Pennsylvania, USA, The American
Omithologist’s Union, Washington, DC, USA

Mayfield, H. 1961. Cowbird parasitism and the population of the Kirtland’s warbler.
Evolution. 15: 174-179

McKay, C. R. 1994. Brown-headed cowbird in Strathclyde: new to Britain and Ireland.
British Birds 87: 284-288.

Norman, R. F., and R. J. Robertson. 1975. Nest searching behaviour in the brown-
headed cowbird. Auk 92:610-611.

Ortega, C. P. 1998. Cowbirds and other brood parasites. The University of Arizona
Press, Tucson, Arizona, USA.

Payne, R. B. 1977. The ecology of brood parasitism in birds. Annual Review of
Ecological Systems 8: 1-28.

Pearson, T. G., and J. Burroughs. 1923. Birds of America. The University Society, New
York, New York, USA.

Pearson, T. G., C. S. Brimley, and H. H. Brimley. 1942. Birds of North Carolina.
Bynum Printing Company, Raleigh, North Carolina, USA.

20

Peterjohn, B. G., J. R. Sauer, and S. Schwarz. 2000. Temporal and geographic patterns
in population trends of brown-headed cowbirds. Pages 21-34 in J. N. M. Smith,
T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Petersen, R. T. 1980. Petersen field guides: eastern birds. Houghton Mifflin Company,
Boston, Massachusetts, USA.

Probst, J. R. 1986. A review of factors limiting the Kirtland’s warbler on its breeding
grounds. American Midland Naturalist 116 (1): 87-100.

Purcell, K. L., and J. Vemer. 1999. Abundance and rates of brood parasitism by brown-
headed cowbirds over an elevational gradient in the southern Sierra Nevada.
Studies in Avian Biology. 18: 97-103.

Robertson, R. J., and R. F. Norman. 1976. Behavioral defenses to brood parasitism by
potential hosts of the brown-headed cowbird. The Condor 78: 166-173.

Robinson, S. K., F. R. Thompson 111, T. M. Donovan, D. R. Whitehead, and J. Faaborg.
1995. Regional forest fragmentation and the nesting success of migratory birds.
Science 267: 1987-1990.

Robinson, S. K., J. P. Hoover, and J. R. Herkert. 2000. Cowbird parasitism in a
fragmented landscape: effects of tract size, habitat and abundance of cowbirds and
their hosts. Pages 280-297 in J. N. M. Smith, T. L. Cook, 8. I. Rothstein, S. K.
Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

Rothstein, S. I. 1975. Evolutionary rates and host defenses against avian brood
parasitism. The American Naturalist 109(966): 161-176.

Rothstein, S. I. 1978. Geographic variation in the nestling coloration of parasitic
cowbirds. Auk 95: 152-160.

Rothstein, S. I. 1990. A model system for coevolution: avian brood parasitism.
Ecological Systems. 21: 481-508.

Rothstein, S. I. 1995. The brown-headed cowbird’s invasion of the far West: history,
causes and consequences experienced by host species. Studies in Avian Biology
15: 301-3 15.

Rothstein, S. I., and T. L. Cook. 2000. Introduction. Pages 323-332 in J. N. M. Smith,
T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

21

 

Rothstein, S. I., and S. K. Robinson. 2000. Introduction. Pages 13-20 in J. N. M. Smith,
T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Rothstein, S. 1., J. Vemer, and E. Stevens. 1984. Radio-tracking confirms a unique
diurnal pattern of spatial occurrence in the parasitic brown-headed cowbird.
Ecology 65(1): 77-88.

Ryel, L. A. 1981. The 1981 Kirtland’s warbler census. Michigan Department of Natural
Resources. Wildlife Division Report No. 2890, Lansing, Michigan, USA.

Sauer, J. R., J. E. Hines, and J. Fallon. 2003. The North American Breeding Bird Survey,
Results and Analysis 1966 - 2002. Version 2003.1, USGS Patuxent Wildlife
Research Center, Laurel, Maryland, USA.

Scott, D. M., and C. D. Ankney. 1980. Fecundity of the brown-headed cowbird in
Southern Ontario. Auk 97: 677-683.

Scott, D. M., and C. D. Ankney. 1983. The laying cycle of brown-headed cowbirds:
passerine chickens? The Auk 97: 583-592.

Sisk, T. D., and N. M. Haddad. 2002. Incorporating the effects of habitat edges into
landscape models: effective area models for cross-boundary management. Pages
208-240 in J. Liu and W. W. Taylor, editors. Integrating landscape ecology into
natural resource management. Cambridge University Press, Cambridge, United
Kingdom.

Stokes, D., and L. Stokes. 1996. Stokes field guide to birds, eastern region. Little,
Brown and Company, New York, New York, USA.

Strausberger, B. M., and M. V. Ashley. 1997. Community-wide patterns of parasitism of
a host generalist brood-parasitic cowbird. Oecologia 112: 254-262.

Stribley, J. M., and J. B. Haufler. 1999. Landscape effects on cowbird occurrences in
Michigan: Implications to research needs in forests of the inland west. Studies in
Avian Biology. 18: 68-72.

Teather, KL, and R. J. Robertson. 1986. Pair bonds and factors influencing the
diversity of mating systems in brown-headed cowbirds. Condor 88:63-69

Tewksbury, J. J., T. E. Martin, S. J. Hejl, T. S. Redman, and F. J. Wheeler. 1999.

Cowbirds in a western valley: effects of landscape structure, vegetation, and host
density. Studies in Avian Biology. 18: 23-33.

22

Thompson, III, F. R. 1994. Temporal and spatial patterns of breeding brown-headed
cowbirds in the Midwestern United States. The Auk 111(4): 979-990.

Walkinshaw, L. H. 1991. Brown-headed cowbird. Pages 508-509 in Brewer, R., G.
McPeek, and R. J. Adams, Jr., eds. The atlas of breeding birds of Michigan.
Michigan State University Press, East Lansing, Michigan, USA.

Whitfield, M. J. 2000. Results of a brown-headed cowbird control program for the
southwestern willow flycatcher. Pages 371-378 in J. N. M. Smith, T. L. Cook, S.
I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management
of cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Wiedenfeld, D. A. 2000. Cowbird population changes and their relationship to changes
in some host speces. Pages 35-46 in J. N. M. Smith, T. L. Cook, 8. I. Rothstein,
S. K. Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds
and their hosts. University of Texas Press, Austin, Texas, USA.

Wood, D. R. and E. K. Bollinger. 1997. Egg removal by brown-headed cowbirds: a field
test of the host incubation efficiency hypothesis. The Condor 99: 851-857.

Woodward, P. W. 1983. Behavioral ecology of fledgling brown-headed cowbirds and
their hosts. The Condor 85: 151-163.

Woolfenden, B. E., H. L. Gibbs, S. G. Sealy, and D. G. McMaster. 2003. Host use and
fecundity of individual female brown-headed cowbirds. Animal Behaviour.
66:95-106.

Young, J. S., and R. L. Hutto. 1999. Habitat and landscape factors affecting cowbird
distribution in the northern Rockies. Studies in Avian Biology 18: 41-51.

23

CHAPTER TWO

BROWN-HEADED COWBIRD (MOLOTHR US ATER) RESPONSE TO VARYING

PATCH SIZE AND AGE IN KIRTLAND’S WARBLER HABITAT

24

 

2.1. Introduction

The Kirtland’s warbler is a federally endangered bird that breeds only in the
northeastern part of Michigan’s Lower Peninsula. Since 1957, the USFS, and then the
USFWS and MDNR in subsequent years, have worked to restore the mid-aged (6 — 20
years old) jack pine habitat the warbler needs for successful reproduction (Huber et al.
2001). While recent censuses indicate that these management efforts have benefited the
Kirtland’s warbler population (http://www.michigan.gov/dnr/0,1607,7-153-
10371_10402-148280--,00.html), little is known about the effects on the brown-headed
cowbird, whose brood parasitic behavior might have serious ecological implications for
the warbler. The aim of this study was to address whether differences in cowbird
abundance and density were related to jack pine stand age and size classes that result
from Kirtland’s warbler habitat management.

Stand or patch size has become of increasing interest to ecologists as studies have
shown it to have an effect on the abundance, density, richness and diversity of many
organisms. Phillips and Shure’s (1990) North Carolina study found that some plant
corrrrnunities respond positively to the size of the patch that they inhabit, with increases
in both net primary production and species richness in larger patches. In a study of
bighom sheep (Ovis canadensis), Singer et al. (2001) concluded that increasing patch size
was the most significant contributor to the population’s performance and persistence.
Several papers have indicated a connection between increased patch size and avian
species richness (van Dorp and Opdam 1987, Rudnicky and Hunter 1993, Saab 1999) and
diversity (McIntyre 1995), in some instances patch size being the single best predictor

(van Dorp and Opdam 1987). Ambuel and Temple (1983) found that some forest-edge

25

bird species increase in density as patch area decreases. Bender et al. (1998) concluded
that patch size has a negative, positive or no association with the density of a particular
species, depending on what habitat (edge, interior or both) the species utilizes. However,
this association with patch size might also be influenced by stand age or it may be
vegetation-dependent, as Schieck et al. (1995) found no relation between patch size and
avian species richness. Burhans and Thompson (1999) noted that the yellow-breasted
chat (Icteria virens) favored large habitat patches, perhaps due to the lower occurrences
of nest predation that occur when edge to area ratios are low.

Several studies examining the role of patch size with regard to cowbirds in
particular has yielded varying results. Saab’s (1999) work on cottonwood (Populus
deltoides) riparian landscapes showed that cowbirds favored small patches with abundant
edge. Johnson and Igl (2001) had similar findings in their grassland study in the Great
Plains. Horn (2000) also determined that cowbirds preferred smaller patches in the
Prairie Pothole region of North Dakota. Research by Robinson et a1. (2000) in mixed
hardwood forests documented decreasing parasitism rates as tract size increased. Petit
and Petit’s (2000) results were mixed, showing cowbirds preferred small deciduous forest
fragments during Acadian flycatcher (Empidonax virescens) parasitism, but there was no
patch size association with parasitism on wood tlrrushes (Hylocichla mustelina).
Conversely, Burhans and Thompson (1999) found that cowbirds were more likely to
parasitize yellow-breasted chats in large shrubland patches rather than in small. Trine
(1998) also found cowbird parasitism high in some of the largest forest tracts in southern
Illinois. Furthermore, several other papers have documented no association between

cowbird abundance and patch size (Ambuel and Temple 1983, Hoover et al. 1995, Fauth

26

2000, Davis 2004). While the majority of studies find that cowbirds favor small,
fragmented patches, studies with contradictory results show that making this
generalization can be problematic.

Several studies have also tracked changes in bird diversity and abundance in
different stages of forest regeneration, with varying results. Imbeau et al. (1999) found
no difference in species richness among age classes, but clearcuts had the highest
abundance of birds. Schieck and Nietfeld (1995) found highest bird diversity in old
stands (120+ years) of aspen mixed-wood forest, but greatest abundance in young stands
(20-30 years). Hobson and Bayne (2000) also found species richness to be greatest in old
aspen stands (80-110 years), which they attributed to increasing structural complexity of
the stands. Johnston and Odum’s (1956) study of a pine-oak-hickory forest in different
stages of secondary succession found species richness increased with increased (100+
years) forest age. Shugart and James (1973) indicated species richness increased as
ecological succession progressed into old forest.

Prior to this study, no research has been conducted to determine the effects of
different stages of regenerating jack pine forests on cowbird abundance and density. This
study might be the first to examine cowbird response to stand size and age in jack pine
forests, in particular those managed for Kirtland’s warbler habitat. Unique management
practices, including small- and large-scale clearcuts and replantings on a 50-year rotation,
have been implemented since 1967 to create and maintain habitat for this endangered
warbler species (Huber et al. 2001). The Kirtland’s Warbler Management Area (153,000
ha), located in the northern Lower Peninsula of Michigan, afforded the opportunity to

examine jack pine habitat in three stages of growth—young clearcuts (0 — 6 years old),

27

middle-aged scrub (7 - 20 years old, also Kirtland’s warbler habitat), and old forest (> 20
years old). Within this age spectrum, cowbirds were studied to determine with which age
category they were most abundant.

This study area also allowed me to study the influence of patch size on cowbird
relative abundance and density. Despite ecological evidence that suggests that Kirtland’s
warblers will only nest and successfully breed in large (> 81 ha) jack pine stands, the
USFS continues to manage considerably smaller patches for the species (Huber et. al.
2001, Rex Ennis, USFS, pers. commun.).

The objectives for this chapter were to determine if cowbird relative abundance
and density varied with respect to a) stand size and b) stand age. Studies of cowbird
association with different patch sizes have had variables results; however, I predicted that
cowbird relative abundance would be higher in large stands due to greater available area.
However, due to evidence in previous studies that suggest cowbirds often have a positive
affiliation with small patches, I predicted that cowbird density would be higher in small
patches. In addition, since cowbirds are known to incorporate several stages of forest
regeneration—new clearcuts, mid-aged shrub scrub and mature forest—into their habitat,
I predicted that cowbirds in this study would not be related to any particular stand age

category.

2.2. Study Area
Study sites were located on state lands in Montmorency and Presque Isle Counties
north of the Clear Lake State Park (Figure 2.1), which is in the northeast range of the

Kirtland’s Warbler Management Area. The study area was comprised predominantly of

28

 
  

was]. Co. ’_ V" ' '
Clear Lake State : ' i '
Wildlife Area \

    
    

' Montmorency Co.

      
  

 
 
 
 

m...
Illg.

   

Fig. 2.1. Clear Lake State Wildlife Area in Presque Isle and Montmorency
Counties, Michigan.

29

jack pine plantations, and encompassed approximately 6,500 ha. Other tree species
present included red pine (Pinus resinosa), white pine (Pinus strobes), quaking aspen
(Populus tremuloides), various oak species (Quercus spp.) and red maple (Acer rubrum).
Understory vegetation was comprised of fire cherry (Prunus pensylvanica), serviceberry
(Amelanchier spicata), and jack pine tree saplings. Ground cover consisted of sweet fern
(Comptonia peregrine), bracken fern (Pteridium aquilinum) and blueberry (Vaccinium
augustifolium).

A small population (n = 10, P. Huber, USFS) of Kirtland’s warblers occur in the
Clear Lake State Park, and extensive habitat is managed for the species by the MDNR.
Public entry is prohibited in the breeding habitat, so access permits were obtained to enter
these areas for field research. The Cowbird Control program was not in operation in the
vicinity of the study site, with the exception of one trap in Montmorency County that was

not active in 2004.

2.3. Methods

2.3.]. Site selection. I made the assumption that recent clearcut stands (0 — 5
years old) are structurally similar to agricultural and grassland areas; mid-aged (6 — 20
years old) jack pine stands are comparable to dense scrub areas; and old (> 20 years old)
jack pine are likened to mature coniferous forests. Much of the literature on cowbirds has
shown that they are positively associated with agriculture and grasslands (Stribley and
Haufler 1999, Vander Haegen and Walker 1999, Young and Hutto 1999). Cowbirds
often will avoid tracts with dense canopy cover and high stem densities, characteristic of

older forests (Young and Hutto 1999). However, Hahn and Hatfield (1995) documented

3O

cowbird parasitism in interior mature forest rather than in nearby old-fields. Similarly,
Robinson et al. (1999) found more cowbirds in oak savannahs and forests than in
grasslands. It should be noted, however, that cowbirds are habitat generalists, and will
often use many different vegetation types to meet the habitat requirements for their
breeding and feeding behavior.

I analyzed digital spatial data in ArcView GIS (ESRI, Redlands, CA) to select
jack pine stands in two size categories: small stands, which were < 81 ha (mean = 13.8
ha) in area; and large, which were > 81 ha (mean = 122.5 ha). Spatial data also were
utilized to select age classes. Stand ages were designated as young (< 6 years), middle (6
- 20 years), and old (> 20 years). Jack pine stands in this region are routinely clearcut on
50-year rotations, although some stands in this study were older (Huber et al. 2001).
Young stands comprised clearcut areas with minimal to no jack pine growth, and
moderate to extensive fern (Comptonia perigrina and Pteridium aquilinum) and
blueberry (Vaccinium spp.) groundcover. Middle-aged stands were suitable Kirtland’s
warbler breeding habitat, and consisted of dense, shrubby jack pine 2 - 3 m tall with
blueberry understory. Old stands included mature jack pines (> 3 m), which were less
densely situated and typically had closed canopies. Both size and age designations were
made based on Kirtland's warbler habitat preferences (Huber et al. 2001). Twenty-five
stands were selected representing all size-age combinations (small-young [n = 2]; small-
rrriddle [n = 5]; small- old [n = 5]; large-young [n = 3]; large-middle [n = 5]; large-old [n
= 5]). Using the maps generated from the GIS, I ground-truthed the sites to confirm stand

size, age and vegetation type.

31

2.3.2. Avian surveys. Avian censusing took place 11 May - 17 July 2004.
Sampling periods were categorized as Period 1 (May 11 — June 15), Period 2 (June 17 —
July 1) and Period 3 (July 2 - July 17). All sites were surveyed twice; approximately half
of the sites were visited a third time. Surveys were done only on days with no
precipitation and with winds less than 16 km/hr. Sampling lasted from sunrise until 1100
hr, ending earlier on days when precipitation began or birds were unusually quiet. Six-
minute point counts, broken into intervals of 0 - 3, 3 - 5, and 5 - 6 minutes were
conducted at each point, with points systematically spaced 300 m apart. Points were
located at least 150 m from any stand edge to minimize the effects of road noise and to
reduce the instance of counting birds in bordering vegetation. The number of points
ranged from as few as one in small stands to as many as 25 in large stands. A global
positioning system (GPS) was used to navigate to all points. All birds seen or heard
within a 50 m radius of the point were recorded, except for birds flying overhead that did
not land in the stand. Data recorded at each point included species name, distance to
observer, sex (when possible), and detection type (seen or heard). Cowbird data were
summarized as relative abundance (# individuals observed) and cowbird density (#
individuals observed/ha sampled). Data for all species were quantified as avian species
richness, avian host abundance (# host species observed) and avian host density (# host
species observed/ha sampled). Host species were determined based on the cowbird host
species list compiled by Lowther (2007). For cowbird density and host species density,
the hectares sampled included only the land encapsulated in the 50 m radii around each

survey point.

32

2.3.3. Vegetation sampling. I randomly sampled five 5 x 20 m plots in each stand
to record vegetation attributes, including tree stem density (#lha), sapling density (#Iha),
shrub density (#lha), snag density (#lha), ground cover (%), mid-story cover (%), and
canopy cover (%). Woody vegetation was classified as a tree or shrub according to
species. Trees > 10 cm dbh were classified as tree stems, and trees < 10 cm dbh were
considered saplings (Young and Hutto 1999). Ground, mid-story and canopy cover were
measured using the line-intercept method (Canfield 1941). Ground cover included all
vegetation and debris from the ground to 0.5 m; mid-story cover consisted of vegetation
from 0.5 - 10 m; and canopy cover encompassed canopy vegetation over 10 m (Vickery

1981).

2.3. 4. Statistical analyses. The cowbird survey data were analyzed to determine
differences in cowbird relative abundance and density between stand sizes, stand ages
and sampling periods. Relative abundance was determined by the total number of
cowbirds detected within a stand; density referred to the total number of cowbirds
detected divided by the total area (ha) sampled. The total area sampled included the area
within the 50-m radii surrounding the sampling points in a stand. Relative abundance
and density data with regard to stand size were analyzed with a Mann-Whitney-U test (a
= 0.05). Cowbird abundance and density data with respect to stand age and sampling
period were tested with a Kruskal-Wallis nonparametric analysis of variance (ANOVA; a
= 0.05). A Dunn multiple comparison test for unequal sample sizes (a = 0.05) was used
to identify differences within categories. A Kruskal-Wallis nonparametric analysis of

variance ((1 = 0.05) was used to analyze cowbird relative abundance and density with

33

 

respect to the stand size-age combinations. A Dunn multiple comparison test (a = 0.05)
was used to detect differences between these stand combinations.

A Mann-Whitney-U test (a. = 0.05) was used to determine differences in
vegetation data between stand size classes; a Kruskal-Wallis test (a = 0.05) was used to
ascertain differences amongst age categories, and amongst stand size-age combinations.
A Dunn multiple comparison test (a = 0.05) was selected to determine where differences
lay. Vegetation variables that were tested included ground cover (GC), mid-story cover
(MS), canopy cover (CC), tree stem density (TSD), sapling density (SD), shrub density
(SI-ID) and snag density (SND). These variables were also tested at the stand-level using
a Spearman rank order correlation (a = 0.05) to determine if relationships existed with
cowbird relative abundance or density.

The abundances and densities of potential host species were calculated for each
stand. Host species abundance, host species density, and species richness were analyzed
for differences amongst stand ages, sizes, size-age combinations and sampling periods
using a one-way ANOVA (a = 0.05). Differences in stand ages, stand size-age
combinations and sampling periods were confirmed using a Tukey’s test for unequal
sample sizes (a = 0.05).

To ascertain the relative avian species diversity in the sampled areas, I calculated
a Sharmon-Weaver diversity index (Sharmon and Weaver 1949) for each stand.
Differences in host densities between size and amongst age classes, size-age
combinations and sampling periods were tested with a one-way ANOVA (a = 0.05). A
Tukey multiple comparison test (a = 0.05) for unequal sample sizes was used to

determine where differences in age categories were. Differences in diversity indices

34

between size and amongst age classes, size-age combinations and sampling periods were
tested using a one-way ANOVA (a = 0.05). Finally, a Spearman rank order correlation
(a = 0.05) was used to determine associations between cowbird relative abundance and

density and avian host abundances, densities and diversity indices ((1 = 0.05).

2.4. Results

Means, standard errors and ranges of avian characteristics were computed (Table
2.1). Cowbird relative abundance was greatest (n = 9) in large and old stands, although
all stand types had instances with no cowbirds present. Large stands had an average of
1.8 cowbirds per stand compared to small stands, which averaged 0.7 individuals per
stand. Mean cowbird relative abundance increased, although not statistically, ranging
from 0.6 individuals in young stands to 1.9 cowbirds in old stands. Host species
abundance ranged from as few as two individuals to 153 individuals in large and middle-
aged stands. Mean host abundance in large stands was more than 4 times that of small
stands; among the three stand ages, host abundance was less varied. Species richness was
relatively consistent throughout all stand types, ranging from 2 to 16 species present
amongst all stands.

No differences (p = 0.19, Figure 2.2) were detected between relative cowbird
abundance and stand age. However, relative abundance was higher (p = 0.03, Figure 2.3)
in large stands. Large old stands had higher relative cowbird abundances (p = 0.04,
Figure 2.4) than any other size-age stand class. Cowbird relative abundance was not

significantly different (p = 0.12, Figure 2.5) among the sampling periods. Cowbird

35

Table 2.1. Mean (standard error) avian assemblages in jack pine size and age classes in
Montmorency and Presque Isle counties, Michigan, summer 2004.

 

 

Small Large Young Middle Old
n=12 n=13 n=5 n=10 n=10
0.7 (0.3) 1.8 (0.4) 0.6 (0.3) 1.0 (0.3) 1.9 (0.5)
Cowbird
abundance (#) (0, 5)* (0, 9) (0, 3) (0, 5) (0, 9)
0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0)
Cowbird density
(#lha sampled) (0, 0.6) (0, 0.1) (0, 0.1) (0.0, 0.6) (0, 0.1)

9.2 (0.9) 39.6 (4.3) 22.0 (4.9) 31.2 (6.7) 23.8 (3.1)
Host abundance
(#) (2, 21) (6, 153) (6, 59) (2, 153) (4, 53)

0.6 (0.1) 0.5 (0.0) 0.6 (0.1) 0.6 (0.1) 0.4 (0.0)
Host density
(#lha sampled) (0.1, 1.3) (0.1, 1.3) (0.1, 1.3) (0.1, 1.1) (0.2, 0.7)

5.6 (0.4) 10.9 (0.5) 7.5 (0.8) 8.5 (0.9) 9.3 (0.7)
Species richness
(#) (2, 11) (5, 16) (3, 13) (2, 14) (4, 16)

 

*Numbers in parentheses denote minimum and maximum numbers of individuals
observed in stands.

36

3.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A

3.0 -

2.5 -
8
5 2.0 r ._
E A U
'8
D ——
g 1.5 -
E A
E __
.D
g 1.0 - l]
U —

D n = 10
0.5 . ~
n = 10
0.0 > —
n:
-0.5 ‘ ‘ -
Young Middle Old
Stand Age

Fig. 2.2. Cowbird relative abundance amongst three age classes of jack pine stands in
Montmorency and Presque Isle counties, Michigan, summer 2004. (Young = 0 - 5 years;
middle = 6 - 20 years; old = > 20 years; Kruskal-Wallis nonparametric ANOVA, a =
0.05). Same letters indicate no significant differences. Point indicates mean, box denotes
one standard error of the mean, and whiskers indicate two standard errors of the mean.

37

3.0

 

 

 

2.0 .

1.5 .

 

 

 

 

 

1.0 > __

 

0.5 r

Cowbird Relative Abundance

 

 

 

 

0.0 >

 

 

-l.0

 

Small Large

Stand Size

Fig. 2.3. Cowbird relative abundance in two size classes of jack pine stands in
Montmorency and Presque Isle counties, Michigan, summer 2004. (small = < 81 ha; large
= > 81 ha; Mann-Whitney-U, a = 0.05). Different letters indicate significant differences.
Dot indicates mean, box denotes one standard error of the mean, and whiskers indicate
two standard errors of the mean.

38

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B
4 r
2
<1 A
D
a —— A A
E —_ __
Ԥ __
5 1 . A D n = 5
D C]
I Cl
C]
o _ __ n = 5
n = 5 n = 5 n = 3
n = 2
-1 - - L - - -
Small Small Small Large Large Large
Young Middle Old Young Middle Old
Stand Class

Figure 2.4. Cowbird relative abundance amongst six size-age classes of jack pine stands
in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small = < 81 ha;
large = > 81 ha; young = 0 - 5 years; middle = 6 - 20 years; old = > 20 years; Kruskal-
Wallis nonparametric ANOVA, a = 0.05). Same letters indicate no significant
differences. Point indicates mean, box denotes one standard error of the mean, and
whiskers indicate two standard errors of the mean.

39

3.0

 

 

2.5 r A 1

 

2.0 - A

 

 

 

 

 

1.5 .

 

 

 

 

 

 

 

 

0.5 >

Cowbird relative abundance
S
a

 

 

 

 

 

0.0 - n=25 l

-0.5 - I

 

 

 

 

-1.o . - - - -

 

Sampling Period

Fig. 2.5. Cowbird relative abundance in jack pine stands for three sampling periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (1 = May-June, 2 =
June — July, 3 = July). Kruskal-Wallis nonparametric analysis of variance, (1 = 0.05.
Same letters indicate no significant differences. Dot indicates mean, box denotes one
standard error of the mean, and whiskers indicate two standard errors of the mean.

40

0.12 . . f

 

 

0.10 -

0.08 .

 

 

0.06 .

0.04 >

 

 

 

 

 

 

 

Cowbird density (# individuals/ha sampled)

 

 

 

 

A
0.02 . —|——
_l__

 

0.00 >

n=5 ——

n=10

 

 

-0.(Y2

 

Young Middle Old

Stand Age

Fig. 2.6. Cowbird density (# individuals/ha sampled) in three age classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Young = 0 -
5 years; middle = 6 - 20 years; old = > 20 years). Kruskal-Wallis nonparametric analysis
of variance, (1 = 0.05. Same letters indicate no significant differences. Dot indicates
mean, box denotes one standard error of the mean and whiskers indicate two standard
errors of the mean.

41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.10 . A
0.08 -
§ 0.06-
”5.
E
N
'5; a
g; 0.04 -
g A
Q I
E
"g C]
5 0.02-
n = 13
0.00 ~
11 = 12
-0.02 ‘ 4
Small Large

Stand Size

Fig. 2.7. Cowbird density (# individuals/ha sampled) in two size classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small = <
81 ha; large = > 81 ha). Mann-Whitney-U, a = 0.05. Same letters indicate no significant
differences. Dot indicates mean, box denotes one standard error of the mean, and
whiskers indicate two standard errors of the mean.

42

density was not different amongst the stand ages (p = 0.33, Figure 2.6) or between stand
sizes (p = 0.21, Figure 2.7). No differences in cowbird density were found between the
six size-age stand classes (p = 0.27, Figure 2.8). There were no significant differences in
cowbird density amongst the sampling periods (p = 0.13, Figure 2.9).

Vegetation structural variables varied across the three age categories (Table 2.2),
with the exception of sapling density and shrub density (p = 0.17, p = 0.64). Ground
cover (p = 0.03), mid-story cover (p < 0.01) and canopy cover (p < 0.01) were higher in
older stands. Tree stem density (p < 0.01) was highest in middle-aged and old stands.
Snag density was higher in older stands (p < 0.01). Differences in vegetation variables
between stand sizes (Table 2.3) included ground cover (p = 0.03), tree stem density (p =
0.01), and sapling density (p < 0.01), which were all lower in large stands. When
examining the six jack pine stand classes (Table 2.4), ground cover (p = 0.03) and
rrridstory cover (p = 0.03) were greater in small old and large old stands than in large
small stands. Canopy cover (p < 0.00) was highest in small old and large old stands;
however, small young stands were comparable to other stands with little or no canopy
cover, yet showed no differences from small old and large old stands. This is likely due
to the process of ranking done in a Kruskal-Wallis test, which might not have detected
this difference. Tree stem density (p < 0.00) was higher in small middle and large middle
stands than in small young or large young stands. Neither sapling density (p = 0.08) nor
shrub density (p = 0.81) were different amongst the stand classes. Snag density (p <
0.00) was higher in small old and large old stands than in large young or large middle
stands. No correlations (p > 0.05) existed between any of the vegetation attributes in

stands and cowbird relative abundance or density (Table 2.5).

43

0.20

0.18 .

0.16 .

0.14

0.10

0.04

Cowbird Density (#lha sampled)

0.02

0.00 .

-0.02 >

-0.04

Figure 2.8. Cowbird density (#lha sampled) amongst six size-age classes of jack pine
stands in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small = <
81 ha; large = > 81 ha; young = 0 - 5 years; middle = 6 - 20 years; old = > 20 years;
Kruskal-Wallis nonparametric ANOVA, a = 0.05). Same letters indicate no significant
differences. Point indicates mean, box denotes one standard error of the mean, and

 

0.12 .

0.08 .

0.06 -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A
U
. A
A A |
A A _|_ D
I D] 1 D l
U U l n = 5
_L "I— _L_
n : 5 1'1 = 3 II = 5
n=2 I n=5
Small Small Small Large Large Large
Young Middle Old Young Middle Old
Stand Class

whiskers indicate two standard errors of the mean.

44

 

 

0.12 . ,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A

0.10 -
3g 0.08 -
E
N
E
-8 0.06 b
25' D
.2
1*.
g» 0.04 »
g A A
.0 ——
'U
5 0.02 - D

__I__ D
0.00 - n=25
n=25
n = 10
-0.02
l 2 3
Sampling period

Fig. 2.9. Cowbird density (# individuals/ha sampled) in jack pine stands over three
sampling periods in Montmorency and Presque Isle counties, Michigan, summer 2004.
(1 = May-June, 2 = June — July, 3 = July). Kruskal-Wallis nonparametric analysis of
variance, a = 0.05. Same letters indicate no significant differences. Dot indicates mean,
box denotes one standard error of the mean, and whiskers indicate two standard errors of
the mean.

45

Table 2.2. Mean (standard error) vegetation structural attributes of young, middle and

old jack pine stands sampled in Montmorency and Presque Isle Counties, MI, in summer
2004.

 

Young (0 — 5 yrs) Middle (6 — 20 yrs) Old (> 20 yrs)

Vegetation variable 11 = 5 n = 10 n = 10

 

GroundCover(%) 78.2 (4.1)A 84.6 (2.1)AB 91.0 (1.8)B

Mid-story Cover (%) 15.8 (4.3)A 38.8 (4.4) B 57.6 (3.2)C

Canopy Cover (%) 0.0 (0.0) A 0.0 (0.0) A 36.5 (4.4) B

Tree Stern Density A B B
(#lha) 0.2 (0.2) 18.6 (3.1) 13.7 (1.4)

Sapling Density A A A
(#lha) 28.1 (7.3) 26.5 (3.4) 19.2 (4.1)

Shrub Density (#lha) 1.2 (0.5) A 1.8 (0.6) A 1.1 (0.4) A

Snag Density (#lha) 0.0 (0.0) A 0.0 (0.0) A 0.9 (0.2) B

 

*Within a row, vegetation attributes with the same letter are not significantly different
(Kruskal-Wallis nonparametric analysis of variance, (1 = 0.05).

46

Table 2.3. Mean (standard error) vegetation structural attributes of small and large jack
pine stands sampled in Montmorency and Presque Isle Counties, MI, in summer 2004.

 

Vegetation Variable

Small (< 81 ha)

Large (> 81 ha)

 

n = 12 n = 13
Ground Cover (%)* 88.0 (1.9) 83.6 (2.4)
Mid-story Cover (%) 44.3 (4.6) 38.3 (4.6)
Canopy Cover (%) 14.5 (5.0) 14.0 (4.1)
Tree Stem Density (#lha)* 13.5 (2.9) 12.0 (1.8)
Sapling Density (#/ha)* 30.7 (4.2) 17.9 (2.7)
Shrub Density (#lha) 1.4 (0.4) 1.4 (0.5)
Snag Density (#lha) 0.4 (0.1) 0.4 (0.1)

 

*Asterisks denote significant differences between stand size (Mann-Whitney-U, a = 0.05)

47

Amod u d .88 8.8.5980 032:8 SEQ .858?
Mo mam—8:8 05088890: mfiwgaexmavc 58386 bugomawa Ho: 08 Amy—0:2 083 05 8:5 883th :oufiowo.» .38 a 55:5...

 

.88 8.8 <88 88 <88 8.8 88.8 8.8 $28 3 $28 3 888883 8%
<88 3 <88 3 <88 8.8 <8.8 8.8 <88 2 <28 2 8888888 858

<8.8 8.2 <8.8 mom <88 8.: <8.8 3N <88 m: <88: 8.? 88888508588

. . . . . . . . . . . . 85$
a 8.8 Em <88 8.8 <88 8.8 a 8.8 8.8. <88 8.8 $8.8 8.8 88 880 825

<88 0% $8.8 we. m8.8 8.8 <88 38 $8.8 8.2 $8.8 88m $388882:

<8.8 8.8 $8.8 new 88.8 2: <8.8 8.8 $8.8 S8 $8.8 8.88 $881,885.80

 

mu: mu: mua mu: mu: Nu:

bomb no :2 Some
20 owned 0622 ems w55> owed..— EO =~Em 062—2 :35 950% :35 u .8 >

 

Amado» om A M So .889» on I c n 2288 8on m I o u was?» ”a: 5 A u own: an S v n =«Emv .38 8883 E .34 805550
03 0:305 was 5:895:32 5 @2958 momma—o 6538 05m xoam mo 82358 3.5683 553%? Coho 33533 cue—2 .v.~ 033.

48

Table 2.5. Correlations between stand vegetation characteristics and cowbird relative
abundance and density.

 

 

Cowbird Relative Abundance Cowbird density
Ground Cover (%) -0.05 -0.08
3173;18th Cover 0.14 0.14
Canopy Cover (%) 0.26 0.21
gfjsist;e$/ha) 0.16 0.16
183323.12? (#lha) -0.25 -0.20
82:11.. W) 0.19 0.20
:22.-.

 

*Denotes significant correlations (Spearman rank correlation 01 = 0.05).

49

45 . . fi

 

: >

35, A

30- «

25>

 

20 ’ “—

II
C

 

15-

Host abundance (# of individuals)

10’

 

 

 

 

Young Middle Old

Stand Age

Figure 2.10. Host species abundance (# of individuals) in jack pine stands in three age
categories in Montmorency and Presque Isle counties, Michigan, summer 2004. (Young
= 0 -5 years, middle = 6 — 20 years, old = > 20 years). One-way analysis of variance, (1 =
0.05). Same letters indicate no differences. Dot indicates mean and whiskers denote
95% confidence interval from the mean.

50

55

 

50- B
45-
40- .

35-

 

25-
20- A

15-

Host abundance (# of individuals)

10» . 4

 

 

 

 

Small Large

Stand Size

Fig. 2.11. Host species abundance (# of individuals) in jack pine stands in two size
categories in Montmorency and Presque Isle counties, Michigan, summer 2004. (Small =
< 81 ha; large = > 81 ha). One-way analysis of variance, 0. = 0.05. Different letters
indicate significant difference. Dot indicates mean and whiskers denote 95% confidence
intervals from the mean.

51

Host species abundance was not different amongst stand ages (p = 0.46; Figure
2.10), but was considerably higher in large stands (p < 0.01; Figure 2.11). Host species
abundance did not differ by sampling period (p = 0.18; Figure 2.12). Large middle
stands had higher host species abundances (p < 0.00, Table 2.6) than small young, small
middle and small old stands. Host species density was higher in middle aged stands than
in old stands (p = 0.02; Figure 2.13), but did not differ between stand sizes (p = 0.83;
Figure 2.14). Host density (p = 0.04, Table 2.6) was slightly higher in large middle
stands than in small old and large old stands. Host density increased from Period 1 to
Period 2 (p < 0.05; Figure 2.15). Species richness was not different amongst stand ages
(p = 0.41; Figure 2.16) but was higher in large stands (p < 0.01; Figure 2.17). Species
richness (p < 0.00, Table 2.6) was highest in large middle stands. Species richness was
not different amongst sampling periods (p = 0.14; Figure 2.18). Shannon-Weaver
diversity indices were higher in large stands (p < 0.00; Figure 2.19), but had no
significance in relation to stand age (p = 0.61; Figure 2.20). Diversity was not different
amongst the six stand classes (p = 0.23, Table 2.6) or amongst sampling periods (p =
0.39; Figure 2.21).

Cowbird relative abundance was positively correlated with host abundance in old
stands (r = 0.52, a = 0.05; Table 2.7), and across all stands (r = 0.32). Relative
abundance was not correlated to host density in any stand size or age class, but was
significantly correlated to species richness across all stands (r = 0.26). Correlations
between relative abundance and species diversity were not significant. Cowbird density
was not correlated with host abundance, host density, species richness or species diversity

in any stand size or age class (Table 2.8). All six stand classes were tested for

52

 

55

50' _

45-

35 » <5

30 ' q- +

25-

 

Host abundance (# of individuals)

20 > o

 

15> - . 1

 

 

 

 

Sampling Period

Fig. 2.12. Host species abundance (# of individuals) in jack pine stands over three time
periods in Montmorency. and Presque Isle counties., Michigan, summer 2004. (1 = May-
June, 2 = June — July, 3 = July). One-way analysis of variance, (1 = 0.05. Same letters
indicate no significant differences. Dot indicates mean and whiskers denote 95%
confidence interval from the mean.

53

.Amod n a .58 5889.80 23:58

ESQ 60:38., no £822“ 03089898: mam 3.135an 388.28 DES—paws 8: 8a @832 2:3 2: 8:3 829 .38 a 55:5...

 

 

<88 mo <88 .3 <88 .3 <28 .3 <28 3 <28 .3 gagameoam
.88 a: 08.8 mi .288 c... 88.8 B <68 3. .22.: o... Atvmmofiocmosam
<88 to .28 3 ..<8.8 no <28 to 82.8 8o 88.8 .3 803:8.3888038:
38 8 3m .88 3% 822.8 v.8 88.: M2: .26.: 3 0.28.8 no. 888585528:
mu:
WHG C—UUJZ MHG WHG WHH— NHC
so .83 owg was» .83 20 8.5 .822 :25 38> 8.5

 

$39» cm A u 20 .23» om I o n 0825 $.36» m - o u wane» ”an E A u emu: a: E V n =~Emv doom SEES .qmwfiog
.8958 03 0:365 98 5:895:02 E 8320 USN; 05m #02 5 89.3833 936 H8 mos—S, Coho 6.88533 :82 dd 2an

54

0.3 . .

 

 

 

 

 

 

 

 

A
AB
_F—

0.7 -
G
O
E.
g 0.6 ,. .....
d

B

i
g __ __
3'2
.2 0.5 - , , - n =10
'U
.E
8
>5
.3 _L x
g 0.4 -
3; n = 5
:8

0.3 - n = 10

0.2 ‘ A .

Young Middle Old
Stand Age

Fig. 2.13. Host species density (# of individuals/ha sampled) in jack pine stands in three
age categories in Montmorency Co. and Presque Isle Co., MI, summer 2004. (Young = 0-
6 years; middle = 6-20 years; old = > 20 years). One-way analysis of variance, Tukey’s
multiple comparison test, a = 0.05. Different letters denote significant differences;
shared letters signify no significant differences. Dot indicates mean and whiskers denote
95% confidence interval from the mean.

55

0.70

 

 

 

 

 

 

A
0.65 - A
8
g 0.60 -
I3
8
'3
:3
.5 0.55
:2 .
.8
43' 0.50
8
I:
0.45 -
n = 12 n = 13
0.40
Small Large

Stand Size

Fig. 2.14. Host species density (# of individuals/ha sampled) in jack pine stands in two
size categories in Montmorency and Presque Isle counties., Michigan, summer 2004.
(Small = < 81 ha, large = > 81 ha). One-way analysis of variance, (1 = 0.05. Same letters
indicate no significant difference. Dot indicates mean and whiskers denote 95%
confidence interval from the mean.

56

0.9

 

 

 

 

 

 

 

AB
0.8 .
B

3 0.7 _ “F"
g 0.6 » A
.2. __
E
8 05 _L
.8 25 _._
W n:
g # I] = 10
g 0.4

0'3 ' n=25

0.2 4 ‘

l 2 3
Sampling period

Fig. 2.15. Host species density (# individuals/ha sampled) in jack pine stands during
three sampling periods in Montmorency Co. and Presque Isle Co., MI, summer 2004.
(l = May-June, 2 = June — July, 3 = July). One-way analysis of variance, Tukey’s
multiple comparison test, a = 0.05. Different letters indicate significant differences;
shared letters denote no difference. Dot indicates mean and whiskers denote 95%
confidence interval from the mean.

57

 

 

 

 

 

 

 

12 r
A
11 >
1_
A
A
10 - 2 , --
.T.
J.

9 - 2
8
5 2
.g 8 ’
.3 _-
r?! 9 n = 10

7 _L .

n = 10

6 .

5 n = 5

4

Young Middle Old
Stand Age

Fig. 2.16. Avian species richness in jack pine stands in three age categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. (Young = 0 - 5 years,
middle = 6 — 20 years, old = > 20 years). One-way analysis of variance, (1 = 0.05. Same
letters indicate no differences. Dot indicates mean and whiskers denote 95% confidence
interval from the mean.

58

13 . -

 

12>

11- ' '

 

10 - ——

Specia richness
co

 

 

 

 

Small Large

Stand size

Fig. 2.17. Avian species richness in jack pine stands in two size categories in
Montmorency Co. and Presque Isle Co., MI, summer 2004. (small = < 81 ha; large = > 81
ha). One-way analysis of variance, (1 =0.05. Different letters denote significant
difference. Dot indicates mean and whiskers denote 95% confidence interval from the
mean.

59

l4

 

12’

ll- A

10+

Species richnas
>

 

 

 

 

 

 

Sampling Period

Fig. 2.18. Avian species richness in jack pines stands during three time periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (1 = May-June, 2 =
June — July, 3 = July). One-way analysis of variance, 0 = 0.05. Shared letters indicate no
significant differences. Dot indicates mean and whiskers denote 95% confidence interval
from the mean.

60

 

1.00

 

 

 

 

 

 

 

0.95 . B
0.90 . 0
0.85 >
g n = 13
.E
s? 0.80 »
8
.2
o A
0.75 -
0.70 >
O
0.65 ~
_L_
n = 12
0.60 -
Small Large

Stand Size

Fig. 2.19. Avian species diversity in jack pine stands in two size categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. (small = < 81 ha;
large = > 81 ha). Shannon-Weaver diversity indices, one-way analysis of variance, (1 =
0.05. Different letters indicate significant difference. Dot indicates mean and whiskers
denote 95 % confidence interval from the mean.

61

0.95

 

 

 

 

 

 

 

A
0.90 » -r
A
A —_
0.85 .
0
g 0.80 » <>
.5
.8
«S 2r __
'5 0.75 . -
n = 10
070 "=10
0.65 — n = 5
0.60
Young Middle Old
Stand Age

Fig. 2.20. Avian species diversity in jack pine stands in three age categories in
Montmorency and Presque Isle counties, Michigan, summer 2004. Shannon-Weaver
diversity indices, one-way analysis of variance, (1 = 0.05. (Young = 0 - 5 years, middle =
6 - 20 years, old = > 20 years). Same letters indicate no significant differences. Dot
indicates mean and whiskers denote 95% confidence interval from the mean.

62

1.00

 

 

 

 

 

 

 

A
0.95 -
A
0.90 » __
g 0.85 » A .
.s
.8 "' <>
'5 0.80 >
0 __
n = 10
0.70 -
n = 25
0.65 .
1 2 3
Sampling period

Fig. 2.21. Avian species diversity in jack pine stands during three sampling periods in
Montmorency and Presque Isle counties, Michigan, summer 2004. (l = May-June, 2 =
June — July, 3 = July). Shannon-Weaver diversity indices, one-way analysis of variance,
(1 = 0.05. Same letters indicate no significant differences. Dot indicates mean and
whiskers denote 95% confidence interval from the mean.

63

 

80:22.80 88$ch 08:20 *

 

 

2 .o ado 2.8 2 .o :8 wad 8.8% 8.8%

. . . . - . . A8 8236:

.8 o 8.. o 2 o S o 8 o E o 8.8%

. - . - . . . - . - 8388 .58

2 0 mm o 3 o 2 o 2 o 2 o 8.58 86:

00:6 G: a

80o .de Ed 80 mod mod 8 8.8.?
an: en: en: mu: 2n. an:
.85. =< 20 28:2 88> 683 88m

 

Amod u a down—8.80 088 §§8m $80» om A n
20 .880.» cm I o .1. 0:88 .850» m I o u was?» a: fix v u 033 .0: 5 A n =«Emv voom H0883 .Swwanz £03880 0E 0:305 28
8:085:82 5 8830 0mm 98 05m 0:8 M08. 5 8838080 588 0:0 005888. 038—0.. @5938 8053 28320.80 gum 030B

64

08022.80 0:002:20 08:00 *

 

6S

 

2.0 2.0 8.0- 2.0 2.0- ~00 006% 800%
. . . . . . 00 32:00

t o 00 o 3 o 3 o 8 0 8 0 802m

0 0200 0

2 .o- 000. 8.0 8.0 8.0- 2 .o- snag “mum
. . . . . . 00 8:00:30

om 0 mm o S o z o S o S o 08:

WNHC O~NG Ofiflc Wflfl Mfiufl Nfiflfl

€50 =< 20 2022 was» 003 =25

 

Amod u 0 20022.80 2:0: 5200mm ”200% on
A u 08 .200» om I 0 u 200:: .200» m I o u mg?» ”0: E v n 032 .0: 5 A n =2:@ voom 88:80 £02222 £00580 22 0030.5
0:0 8:295:02 E 000020 0w0 0:0 020 0:8 202. S. 0033:8000 :2>0 0:0 38:00 020300 :002500 2802280 .w.N 030B

20022.80 0:002:20 08:00 ._.

 

 

0.0 to- do- No 3 0.0 022% 028%

0.0 no :0- 0.0 :0- 0.0 at 32:0: 302m
.00- 0.0 0.0 do- No. no- 62:83 «£0 0200 mo:

«.0 no 3 no 0.0. 0.0 00 8.00030 08:
mn: nu: mu: mu: mu: NH:

20 00:3 2022 00.3 was» 00.3 20 :05 202.2 :25 28> :25

 

.806 u 0 2002280 2:0: 520025 ”200» on
A n 08 .200» ON I 0 H 0.000: .200» m I o .I. 0.58» ”0: S v u 022 .0: S A n =02mv voom 002200 202222 £00880 02 0:32am
0:0 »0:20:0:02 E 00220 0:02 0:8 200» E 0023:2020 :2>0 0:0 00:00::00 0>022 0.00300 :002000 2002280 dd 030,—.

66

80022.80 0:002:20 08:00 *

 

 

0.0- no- N o- 0.0 _ o 0.0 00030 8020
:0- no S- 0.0 :0- 0.0 00 32:00 8020
.00- to 0.0 0.0- S- no- 82080... :50 0080 .8:
0.0 0.0 :o no 0.0. 0.0 00 8505.0 .8:
mun mu: mu: mu: mu: Nu:
20 00:3 2022 000: 28> 003 20 :05 2022 :35 28> :85

 

.306 u 0 8002280 292 52.00%

”200» cm A n 08 .200» 8 I 0 n 2000: .200» m I o n w::0» ”0: 3 v u 022 .0: S A n 0029 voom 82:80 £02022 000980
22 0:328 0:0 »0:20:::02 E 00220 0:02 0:8 200» 2 00838020 0000 0:0 26:00 020300 :003000 0802280 .ofim 030,—.

67

correlations between cowbird relative abundance (Table 2.9) and density (Table 2.10)
and avian assemblage data. Only host density was correlated with cowbird relative
abundance (r = -0.5) and cowbird density (r = -0.6), and only in large old stands.
Individual species data were analyzed for occurrence information, and presented for stand
age (Table 2.11), stand size (Table 2.12) and sampling period (Table 2.13). In particular,
Kirtland’s warblers, whose habitat is characterized by large, middle-aged jack pine

stands, occurred in only these types of stands in this study.

2.5. Discussion

Consistent with my predictions, cowbird abundance and density did not differ
with respect to stand age. This was expected, since cowbirds are a generalist species,
using many different vegetation and stand types to comprise their habitat. However, host
species density was lower in old stands than in mid-aged stands. This was consistent
with findings by Johnston and Odum (1956), who showed that avian species density
decreased from pine shrub areas to pine forests. They concluded that this might have
been due to a lack of suitable understory for birds in pine forests, because avian species
density increased again once a hardwood understory developed.

There were no differences among stand ages for overall species richness or
diversity. These results are interesting, because previous studies have shown increased
species diversity (Schieck and Nietfeld 1995) and species richness (Johnston and Odum
1956, Shugart and James 1973, Hobson and Bayne 2000) as stand age increased.
However, in these studies old stands ranged from 80 — 150+ years, and represented

climax communities. While this study categorized jack pine > 20 years old as old stands,

68

..t_. —;mq

Table 2.11. Occurrence of avian species in young (0 - 5 years), middle (6 - 20 years) and
old (> 20 years) -aged jack pine stands in Montmorency and Presque Isle counties,

 

 

Michigan, summer 2004.

Common Name Scientific Name Young Middle Old
American crow Corvus brachyrhynchos X X X
American goldfinch“ Carduelis tristis X X X
American redstart* Setophaga ruticilla X
American robin* Turdus migratorius X
Black-and-white warbler* Mniotilta varia X
Black-capped chickadee* Poecile atricapilla X X X
Black-throated green warbler* Dendroica virens X

Blue jay Cyanocitta cristata X X X
Brewer's blackbird* Ephagus carolinus X

Brown thrasher* Toxostoma rufum X X X
Cedar waxwing* Bombycilla cedrorum X X X
Chestnut-sided warbler* Dendroica pensylvanica X

Chipping sparrow* Spizella passerina X X X
Clay-colored sparrow* Spizella pallida X X X
Common grackle* Quiscalus quiscula X
Common raven Corvus corax X X
Common yellowthroat* Geothlypis trichas X
Dark-eyed junco"I Junco hyemalis X X X
Eastern bluebird* Sialia sialis X X

Eastern phoebe* Sayornz's phoebe X
Eastern towhee* Pipilo erythrophthalmus X X

Eastern woodpewee* Contopus virens X X
Field sparrow* Spizella pusilla X X X
Grasshopper sparrow* Ammodramus savannarum X X

Gray catbird* Dumetella carolinensis X

Hermit thrush“ Catharus guttatus X X X
Indigo bunting* Passerina cyanea X X

Kirtland's warbler* Dendroica kirtlandii X

Lincoln's sparrow* Melospiza lincolnii X X

Nashville warbler* Vermivora ruficapilla X X X
Northern flicker Colaptes auratus X X X
Nuthatch spp Sitta spp. X X
Olive-sided flycatcher Contopus cooperi X
Ovenbird* Seiurus aurocapillus X
Red-eyed vireo“ Vireo olivaceus X X
Rose-breasted grosbeak* Pheucticus ludovicianus X X X
Ruffed grouse Bonasa umbellus X
Savannah sparrow* Passerculus sandwichensis X

Song sparrow* Melospiza melodia X X X
Tree swallow Tachycineta bicolor X X

Upland sandpiper Bartramia longicauda X X
Vesper sparrow* Pooecetes gramineus X X X
Vireo spp.* Vireo spp. X
White-throated sparrow* Onotrichia albicollis X

Wood thrush“ Hylocichla mustelina X
Yellow-bellied sapsucker Sphyrapicus vaius X
Yellow-rumped warbler* Dendroica coronata X X X

 

* Asterisk denotes potential host species

69

Table 2.12. Occurrence of avian species in small (> 80 ha) and large (< 80 ha) jack pine

stands in Montmorency and Presque Isle counties, Michigan, summer 2004.

 

 

Common Name Scientific Name Small Large
American crow Corvus brachyrhynchos X X
American goldfinch* Carduelis tristis X X
American redstart“ Setophaga ruticilla X

American robin* Turdus migratorius X
Black-and-white warbler* Mniotilta varia X
Black-capped chickadee* Poecile atricapilla X X
Black-throated green warbler* Dendroica virens X

Blue jay C yanocitta cristata X X
Brewer's blackbird* Ephagus carolinus X
Brown thrasher“ T oxostoma rufum X X
Cedar waxwing* Bombycilla cedrorum X X
Chestnut-sided warbler* Dendroica pensylvanica X

Chipping sparrow* Spizella passerina X X
Clay-colored sparrow* Spizella pallida X X
Common grackle* Quiscalus quiscula X

Common raven Corvus corax X
Common yellowthroat* Geothlypis trichas X

Dark-eyed junco* Junco hyemalis X X
Eastern bluebird* Sialia sialis X
Eastern phoebe* Sayorm's phoebe X
Eastern towhee* Pipilo erythrophthalmus X X
Eastern woodpewee* Contopus virens X X
Field sparrow* Spizella pusilla X X
Grasshopper sparrow* Ammodramus savannarum X
Gray catbird* Dumetella carolinensis X
Hermit thrush* Catharus guttatus X X
Indigo bunting* Passerina cyanea X X
Kirtland's warbler* Dendroica kirtlandii X
Lincoln's sparrow* Melospiza lincolnii X
Nashville warbler* Vennivora ruficapilla X X
Northern flicker Colaptes auratus X X
Nuthatch spp Sitta spp. X X
Olive-sided flycatcher Contopus cooperi X
Ovenbird* Seiurus aurocapillus X X
Red-eyed vireo* Vireo olivaceus X X
Rose-breasted grosbeak* Pheucu'cus ludovicianus X X
Ruffed grouse Bonasa umbellus X

Savannah sparrow* Passerculus sandwichensis X X
Song sparrow* Melospiza melodia X X
Tree swallow Tachycineta bicolor X X
Upland sandpiper Bartramia longicauda X
Vesper sparrow* Pooecetes gramineus X X
Vireo spp.* Vireo spp. X X
White-throated sparrow* Onotrichia albicollis X

Wood thrush* Hylocichla mustelina X
Yellow-bellied sapsucker Sphyrapicus vaius X
Yellow-rumped warbler* Dendroica coronata X X

 

*Asterisk denotes potential host species.

70

Table 2.13. Occurrence of avian species in jack pine stands in Montmorency and Presque
Isle counties, Michigan, during May-June (Period 1), J une-J uly (Period 2), and July

 

 

(Period 3), 2004.

Common Name Scientific Name Period 1 Period 2 Period 3
American crow Corvus brachyrhynchos X X X
American goldflnch* Carduelis tristis X X X
American redstart* Setophaga ruticilla X

American robin* Turdus migratorius X
Black-and-white warbler* Mniotilta varia X

Black-capped chickadee* Poecile atricapilla X X X
Black-throated green warbler* Dendroica virens X

Blue jay Cyanocitta cristata X X X
Brewer's blackbird* Ephagus carolinus X X X
Brown thrasher“ Toxostoma rufum X X X
Cedar waxwing* Bombycilla cedrorum X X X
Chestnut-sided warbler* Dendroica pensylvanica X

Chipping sparrow* Spizella passerina X X X
Clay-colored sparrow* Spizella pallida X X X
Common grackle* Quiscalus quiscula X

Common raven Corvus corax X X
Common yellowthroat* Geothlypis trichas X

Dark-eyed junco* Junco hyemalis X X X
Eastern bluebird* Sialia sialis X X X
Eastern phoebe* Sayomis phoebe X X
Eastern towhee* Pipilo erythrophthalmus X X

Eastern woodpewee* Contopus virens X X X
Field sparrow* Spizella pusilla X X X
Grasshopper sparrow* Ammodramus savannarum X X
Gray catbird“ Dumetella carolinensis X

Hermit thrush* Catharus guttatus X X X
Indigo bunting* Passerina cyanea X X X
Kirtland's warbler* Dendroica kirtlandii X X X
Lincoln's sparrow* Melospiza lincolnii X

Nashville warbler* Vermivora ruficapilla X X X
Northern flicker Colaptes auratus X X X
Nuthatch spp Sitta spp. X X X
Olive-sided flycatcher Contopus cooperi X

Ovenbird* Seiurus aurocapillus X X X
Red-eyed vireo* Vireo olivaceus X X X
Rose-breasted grosbeak* Pheucticus ludovicianus X X X
Ruffed grouse Bonasa umbellus X

Savannah sparrow* Passerculus sandwichensis X X

Song sparrow* Melospiza melodia X X X
Tree swallow Tachycineta bicolor X X
Upland sandpiper Bartramia longicauda X X X
Vesper sparrow* Pooecetes gramineus X X X
Vireo spp.* Vireo spp. X X
White-throated sparrow* Onotrichia albicollis X

Wood thrush“ Hylocichla mustelina X

Yellow-bellied sapsucker Sphyrapicus vaius X
Yellow-rumped warbler* Dendroica coronata X X X

 

*Asterisk denotes potential host species.

71

this probably is not comparable to the old stands in the previously mentioned studies, for
a few reasons. First of all, jack pine is an early-successional species whose aging, under
natural conditions, is limited by wildfire and the appearance of late-successional forest
communities. Second, in this study area, jack pine is harvested on a 50-year rotation to
mimic the effects of natural wildfire (Huber et al. 2001). While there were some forest
parcels in old stands that were older, most stands in the ‘old stand’ category were < 50
years old. Therefore, the species richness and diversity of old stands in this study would
be very different from that of old stands in the previously mentioned studies.

One surprising result was the positive relationship between cowbird abundance
and large stands, given the strong evidence for cowbirds associating with small patches
(Saab 1999, Horn 2000, Johnson and Igl 2001). However, studies by Burhans and
Thompson (1999) and T rine (1998) support these findings, as they also documented more
cowbirds in large stands. Two possible explanations for this include a) a large stand
simply has more area (and more edge) than a small stand to provide habitat for more
cowbirds, or b) a landscape-level approach is necessary to determine how an increase in
edge from a patchwork of small stands compared with a patchwork of large stands affects
cowbird abundance. While cowbirds are considered an edge species (Ortega 1998),
favoring small, fragmented patches over large, contiguous ones, it might be erroneous to
categorize them as such. Their historical habitat comprised large, uninterrupted
landscapes of grassland areas; cowbird preference for small, fragmented patches appears
to be a response to their relatively recent acquaintance with new vegetation types, such as
forests and urban areas. The factors that influence cowbird habitat selection are likely

more complicated than simply patch size, and are probably based on a combination of

72

preferences for region, vegetation type, and host abundance. In this study, host
abundance and species richness (van Dorp and Opdam 1987, Rudnicky and Hunter 1993,
Saab 1999) were considerably higher in large stands . This increased access to nests for
parasitism might account for cowbirds favoring large stands.

The inconsistencies between cowbird abundance and avian assemblage
correlations were difficult to explain. Cowbird abundance was positively correlated with
host abundance and species richness, which is logical since more species equates to more
opportunities for a greater number of cowbirds to successfully breed. However, cowbird
abundance had no relationship to host density or species diversity. This is puzzling
because intuitively a greater diversity of avian species could signify a greater opportunity
for cowbird parasitism; however, if greater diversity means increasing numbers of non-
cowbird hosts, then there would not necessarily be an association between them. The
lack of relationship between cowbird abundance and host density did not support the
findings of Tewksbury et al. (1999), whose study showed host density to be a strong
predictor of cowbird abundance.

One interesting observation was that host density increased in successive months,
while neither cowbird abundance nor cowbird density changed. This seeming
inconsistency between cowbirds and their avian hosts might be explained by the fact that
cowbirds are less vocal in the later parts of the summer, while other avian species are as
conspicuous or moreso than earlier in the summer. Ortega (1998) suggested that the
typical breeding season for cowbirds is early May through June. Unlike other birds,
cowbirds do not have a nest to defend or clutch with which to communicate, and so they

might be less conspicuous once their breeding season has ended.

73

From a management perspective, the increase in host density in mid-aged stands
was surprising but very interesting. This age category is the most intensively managed
by state and federal agencies, because it is the optimal age range for Kirtland’s warbler
habitat. While there has been some concern amongst scientists and managers in recent
years about the effects of “single species management” in Kirtland’s warbler habitat, the
findings from this study indicate that cowbird host abundance, species richness and
species diversity are greater in large stands, and host density is high in mid-aged stands.
These results might help alleviate the worry that habitat management for Kirtland’s
warblers are not benefiting, or worse yet are discouraging the presence of other avian
species.

Because I did not perform nest sampling in this study, it was impossible to know
with any certainty whether cowbirds were parasitizing the avian host community and
which host species were affected. Since research has shown that many species can be
parasitized differently depending on vegetation composition and structure, host density
and availability, host species abundance and regional effects, I was unable to conduct any
analyses between cowbirds and individual host species to determine whether correlations
existed (Barber and Martin 1997, Burhans 1997, Spautz 1999, Carello and Snyder 2000,
Koford et al. 2000). To better determine whether cowbirds are avoiding host-dense sites
in this study area, or vice versa, a much more intensive field study including nest

searching for parasitism rates would need to be performed.

74

2.6. Literature Cited

Ambuel, B., and S. A. Temple. 1983. Area-dependent changes in the bird communities
and vegetation of southern Wisconsin forests. Ecology 64(5): 1057-1068.

Barber, D. R., and T. E. Martin. 1997. Influence of alternate host densities on brown-
headed cowbird parasitism rates in black-capped vireos. Condor 99: 595-604.

Bender, D. J ., T. A. Contreras, and L. Fahrig. 1998. Habitat loss and population decline:
a meta-analysis of the patch size effect. Ecology 79 (2): 517-535.

Burhans, D. E. 1997. Habitat and microhabitat features associated with cowbird
parasitism in two forest edge cowbird hosts. Condor 99:866-872.

Burhans, D. E., and F. R. Thompson. 1999. Habitat patch size and nesting success of
yellow-breasted chats. Wilson Bulletin 111 (2): 210-215

Canfield, R. H. 1941. Application of the line-intercept method in sampling range
vegetation. Journal of Forestry 39: 388-394.

Carello, C. A., and G. K. Snyder. 2000. The effects of host numbers on cowbird
parasitism of red-winged blackbirds. in J. N. M. Smith, T. L. Cook, S. I.
Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management of
cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Davis, S. K. 2004. Area-sensitivity in grassland passerines: effects of patch size, patch
shape, and vegetation structure on bird abundance and occurrence in southern
Saskatchewan. Auk 121 (4): 1130-1145.

Fauth, P. T. 2000. Reproductive success of wood thrushes in forest fragments in
northern Indiana. Auk 117 (1): 194-204

Hahn, D. C., and J. S. Hatfield. 1995. Parasitism at the landscape scale: cowbirds prefer
forests. Conservation Biology 9 (6): 1415-1424.

Hejl, S. J ., and J. S. Young. 1999. Brown-headed cowbirds in ponderosa pine/Douglas
fir—dominated landscapes in the northern Rocky Mountains. Studies in Avian
Biology 18: 73-79.

Hobson, K. A., and E. Bayne. 2000. The effects of stand age on avian communities in
aspen-dominated forests of central Saskatchewan, Canada. Forest Ecology and

Management 136: 121-134.

Hoover, J. P., M. C. Brittingham, and L. J. Goodrich. 1995. Effects of forest patch size
on nesting success of wood thrushes. Auk 112 (1): 146-155.

75

Horn, D. J. 2000. The influence of habitat features on grassland birds nesting in the
Prairie Pothole Region of North Dakota. Ph.D. Dissertation, Iowa State
University, Ames, IA.

http://www.michigan.gov/dnr/O,1607,7-153-10371_10402-148280--,00.html, retrieved
November 10, 2006.

Huber, P. W., J. A. Weinrich, and E. S. Carlson. 2001. Strategy for Kirtland’s warbler :
habitat management. Publication of MDNR, USPS, and USFWS. Mio, Michigan,
USA

Imbeau L., J. P. L. Savard, and R. Gagnon. 1999. Comparing bird assemblages in
successional black spruce stands originating from fire and logging. Canadian
Journal of Zoology 77 (12): 1850-1860.

Johnson, D. H., and L. D. Igl. 2001. Area requirements of grassland birds: a regional ’
perspective. Auk 118 (1): 24-34.

Johnston, D. W., and E. P. Odum. 1956. Breeding bird populations in relation to plant
succession on the piedmont of Georgia. Ecology 37 (1): 50-62.

Koford, R. R., B. S. Bowen, J. T. Lokemoen, and A. D. Kruse. 2000. Cowbird
parasitism in grassland and cropland in the northern Great Plains. Pages 229-235
in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy,
editors. Ecology and management of cowbirds and their hosts. University of
Texas Press, Austin, Texas, USA.

Lowther, P. E. 2007. Lists of victims and hosts of the parasitic cowbirds (Molothrus).
Field Museum, Chicago, Illinois, USA.
(httpzllfml.fieldmuseum.org/aa/Files/lowther/CBList.pdf), retrieved April 17,
2007.

McIntyre, N. E. 1995. Effects of forest patch size on avian diversity. Landscape Ecology
10(2): 85-99.

Ortega, C. P. 1998. Cowbirds and other brood parasites. The University of Arizona
Press, Tucson, Arizona, USA.

Petit, L. J ., and D. R. Petit. 2000. Brown-headed cowbird parasitism of migratory birds:
effects of forest area and surrounding landscape. Pages 265-270 in J. N. M.
Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors.
Ecology and management of cowbirds and their hosts. University of Texas Press,
Austin, Texas, USA.

Phillips, D. L., and D. J. Shure. 1990. Patch-size effects on early succession in southern
Appalachian forests. Ecology 71 (1): 204-212.

76

Robinson, S. K., J. D. Brawn, S. F. Morse, and J. R. Herkert. 1999. Use of different
habitats by breeding brown-headed cowbirds in fragmented midwestem
landscapes. Studies in Avian Biology 18: 52-61.

Robinson, S. K., J. P. Hoover, and J. R. Herkert. 2000. Cowbird parasitism in a
fragmented landscape: effects of tract size, habitat and abundance of cowbirds and
their hosts. Pages 280-297 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K.
Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

Rudnicky, T. C., and M. L. Hunter, Jr. 1993. Reversing the fragmentation perspective:
effects of clearcut size on bird species richness in Maine. Ecological Applications
3 (2): 357-366.

Saab, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian
forests: a hierarchical analysis. Ecological Applications 9 (1): 135-151.

Schieck, J ., and M. Nietfield. 1995. Bird species richness and abundance in relation to
stand age and structure in aspen mixedwood forests in Alberta. Pages 115-157 in
J .B. Stelfox, editor. Relationships between stand age, stand structure, and
biodiversity in aspen mixedwood forests in Alberta. Jointly published by Alberta
Environmental Centre, Vegreville, Alberta, Canada, and Canadian Forest Service,
Edmonton, Alberta, Canada.

Schieck, J ., K. Lertzman, B. Nyberg, and R. Page. 1995. Effects of patch size on birds in
old-growth montane forests. Conservation Biology 9 (5): 1072-1084.

Shannon, C. E., and W. Weaver. 1949. The mathematical theory of communication.
University of Illinois Press, Urban, Illinois, USA.

Shugart, H. H., and D. James. 1973. Ecological succession of breeding bird populations
in northwestern Arkansas. Auk 90:62-77.

Singer, F. J ., L. C. Zeigenfuss, and L. Spicer. 2001. Role of patch size, disease and
movement in rapid extinction of bighom sheep. Conservation Biology 15 (5):
1347-1354.

Spautz, H. 1999. Common yellowthroat brood parasitism and nest success vary with
host density and site characteristics. Studies in Avian Biology 18: 218-228.

Stribley, J. M., and J. B. Haufler. 1999. Landscape effects on cowbird occurrences in

Michigan: implications to research needs in forests of the inland west. Studies in
Avian Biology 18: 68-72.

77

Tewksbury, J. J ., T. E. Martin, S. J. Hejl, T. S. Redman, and E. J. Wheeler. 1999.
Cowbirds in a western valley: effects of landscape structure, vegetation, and host
density. Studies in Avian Biology 18: 23-33.

Trine, C. L. 1998. Wood thrush population sinks and implications for the scale of
regional conservation strategies. Conservation Biology 12 (3): 576-585.

Vander Haegen, W. M., and B. Walker. 1999. Parasitism by brown-headed cowbirds in
the shrubsteppe of Eastern Washington. Studies in Avian Biology 18: 34-40.

van Dorp, D., and P. F. M. Opdam. 1987. Effects of patch size, isolation and regional
abundance on forest bird communities. Landscape Ecology 1 (1): 59-73.

Vickery, W. L. 1981. Habitat use by northeastern forest rodents. American Midland
Naturalist 106 (1)2111-118.

Young, J. S., and R. L. Hutto. 1999. Habitat and landscape factors affecting cowbird
distribution in the Northern Rockies. Studies in Avian Biology 18: 41-51.

78

CHAPTER THREE

THE EFFECTS OF PLACEMENT AND LAND COVER TYPE ASSOCIATIONS ON

BROWN-HEADED COWBIRD (M OLOTHR US ATER) TRAP CAPTURES

79

3.1. Introduction

Federal brown-headed cowbird removal programs have been instituted across the
United States since 1972 (Rothstein and Cook 2000). One of the most notable programs
was implemented in Fort Hood, Texas, in 1987 for the recovery of the endangered black-
capped vireo (Vireo atricapillus; Hayden et al. 2000; Colonel Randall J. Butler,
Department of the Army, USFWS Memo, July 15, 2004) and those begun in California in
1983 to protect the endangered least Bell’s vireo (Vireo bellii pusillus; Griffith and
Griffith 2000) and southwestern willow flycatcher (Empidonax traillii extimus; Whitfield
2000). The success of these trapping programs has been tempered by factors such as
decreasing habitat availability and high predation levels (Whitfield et al. 1999, Winter
and McKelvey 1999, Griffith and Griffith 2000). However, most researchers agree that
the continued use of cowbird trapping is necessary to maintain or increase populations of
endangered bird species parasitized by cowbirds.

The oldest and perhaps most well-known cowbird trapping program was
implemented in northern Michigan in 1972 by the USFW S for recovery of the Kirtland’s
warbler, which received endangered status in 1973 under the Endangered Species Act
(httpszllecos.fws.gov/species_profile/SpeciesProfile?spcode=B03I). The program,
combined with jack pine habitat restoration, has assisted in the recovery of the Kirtland’s
warbler population, which has increased from 167 singing males in 1987 (DeCapita 2000,
Huber et al. 2001) to 1,478 in the 2006 census (httpz/lwww.michigan.gov/dnr/0,1607,7-
153-10371_10402-148280--,00.html).

Several studies of removal programs have concentrated on the impacts of trapping

on parasitism rates and host species recovery (DeCapita 2000, Griffith and Griffith 2000,

80

Hayden et al. 2000, Whitfiield 2000); to my knowledge, none have looked at how the
locations of traps affect the number of cowbirds captured.

Based on previous research (Saab 1999, Stribley and Haufler 1999, Tewksbury et
al. 1999, Young and Hutto 1999, Curson et al. 2000, Coker and Capen 2000), it has been
determined that cowbirds often are positively associated with proximity to agricultural,
suburban and urban areas, which they use to access food resources. Presumably,
agricultural areas (particularly those with grazing livestock) mimic the grassland
ecosystems which both bison and the cowbirds that associated with them, historically
inhabited (Friedmann 1929). Beutler’s (2000) study in northeastern Washington supports
this hypothesis, as cowbirds strongly favored riparian meadows over clearcuts,
regenerating forests and mature forests. However, in Illinois, Robinson et al. (1999)
found that cowbirds were most abundant in savanna ecosystems, followed by shrublands
and forests, but were least abundant in grasslands. Furthermore, Hahn and Hatfield
(1995) reported higher parasitism levels in interior forested systems than in adjacent old
fields. However, both Rothstein et al.’s (1984) California study and Thompson’s (1994)
midwestem survey examined the temporal variation in cowbird use of different
vegetation types. Both studies concluded that cowbirds used host-rich forested habitats
for morning breeding behavior, and commuted to grasslands, agricultural lands and
human habitations for afternoon feeding. These studies, conducted in different regions of
the United States, yet having come to similar conclusions, suggest that cowbirds likely
incorporate multiple vegetation types into their habitat to fulfill the requirements of their

reproductive and feeding behaviors.

81

This chapter will examine the brown-headed cowbird removal program in
northern Lower Michigan during three periods: 1977-79, 1992-94, and 2000-02. The
goals of this study are to determine if land cover type is correlated with brown-headed
cowbird trap captures; and, using these results, to make suggestions to managers for
future trap locations to increase captures of cowbirds. Based on previous studies, I
expect that cowbird captures will be associated with land cover types that provide them
with essential habitat components. In particular, I predict that cowbird trap captures will
be highly associated with proximal agricultural and urban areas, which provide the
feeding component of cowbird habitat (Stribley and Haufler 1999, Tewksbury et al. 1999,
Young and Hutto 1999), and deciduous forest, where females perform their egg-laying

behavior (Hahn and Hatfield).

3.2. Study Area

Cowbird traps were located on federal and state lands in 12 counties throughout
the northeastern region of Michigan’s Lower Peninsula (Figure 3.1). The landscape is
predominantly coniferous forest, with deciduous forest interspersed throughout.
Agricultural land and open fields are uncommon and relatively scattered throughout the
study region. Major cities located in or near the study area include Grayling (pop. 1943),
Gaylord (pop. 3730), Oscoda (pop. 992) and Clare (pop. 3233; year 2005 census,
http://www.census. gov). Camp Grayling, the nation’s largest US. Army National Guard
base, is located within the city limits of Grayling, and maintains several brown-headed
cowbird traps for protection of a small Kirtland’s warbler population that resides there

(https://www.mi.ngb.army.mil/grayling/default.asp, Huber et al. 2001).

82

 

Traps were situated in 6 - 20 year old jack pine stands, the sole breeding grounds
for Kirtland’s warblers. Stands were chosen based on predictions of warbler habitat
preferences and confirmed Kirtland’s warbler presence (Huber et al. 2001). The USFS
has actively managed jack pine stands for Kirtland’s warblers in this region since 1964,
primarily with prescribed burning, clearcutting, seeding and replanting (Huber et al.
2001). However, prescribed burning all but ceased in 1980 after the Mack Lake Fire,
where a controlled burn escaped and destroyed thousands of acres of private and federal
land (Mann 1980). Since that incident, clearcutting and replanting have been the
dominant management techniques employed (Huber et al. 2001). Kirtland’s warbler
breeding habitat is closed to the public every May-September (Huber et al. 2001), and

access to cowbird traps is highly restricted.

3.3. Methods

3. 3. I . Data collection. Land cover data existed for the northern lower peninsula
of Michigan for the years 1978, 1993 and 2001. The 1978 data were derived from air
photo interpretation using the Michigan Resource Information System (MIRIS) land
cover classification system (M. Donovan, Michigan DNR, pers commun.). The 1993
land cover data were developed from a computer classification of Landsat satellite
imagery, based on human interpretation of a selected sample of polygons. The
classification scheme was developed specifically for the Michigan Gap Analysis Program
(MiGAP; M. Donovan, pers commun.). The 2001 land cover data were derived from
Landsat satellite imagery, and classified according to the Integrated Forest Monitoring

Assessment and Prescription (IFMAP; MDNR, Forest, Mineral and Fire Management

83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{
i:
/ l . ‘

Fig. 3.1. Shaded area indicates the 12 counties in Michigan where the Cowbird
Control Program has been in effect since 1972.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Division 2003). Due to the variation in the methods used to create and classify these
three years of land cover, comparisons could not be made to determine any differences
amongst the three time periods; analyses were limited to correlations between land cover
and cowbird trap captures within periods. Because the land cover had been classified
differently for each of these years, I created a standardized classification scheme, to
which all three years of land cover data adhered, and that resulted in 12 land cover
categories (Table 3.1).

Historic and recent cowbird trap locations were obtained from the USFW S.
Because the older data (19708 and 1990s) were not always marked with specific .,
geographic coordinates, I used agency maps, field notes and data sheets to assist in
finding trap locations. These points were then associated with X and Y coordinates in
ArcView GIS (ESRI, Redlands, CA). Some of the 19903 and all of the 20003 trap points
had been previously marked with X and Y coordinates in the field using a Global
Positioning System (GPS). All trap points were uploaded into a GIS. Trap data were
used from three-year periods corresponding to each of the three land cover years. Three
years of trap data were selected to lessen the effects of any single-year stochastic event.
Therefore, the 1978 land cover corresponded with 1977-79 (Period 1) trap data, 1993
land cover with 1992-94 (Period 2) trap data, and 2001 land cover with 2000-02 (Period
3) trap data. I made the initial assumption that land cover did not change within the
three-year period. All traps from each of the nine years were used for analysis. Within a
period, some traps were used for trapping in one, two or all three years, based on the
population of Kirtland’s warblers and the cowbird parasitism rate in a particular year. I

treated a trap that occurred in more than one year as a unique trap for each year of

85

 

 

 

 

E05005 w:_:000 x00 :0 0:5 AmmV

 

 

 

600::
- 050336 80 0500.0: 256 as
56:00 30:00:06 0:0—:3 AN: 0:0—:000 000000803 va 0:0—0mg: 6000008005 29 0:0—:30 Amv
0:0—:000 000000003 A03 0:0—0m00M A C
0:060: ::0:0:::0m Avmv
b00520 :0 05200 00:0 Amy 6:000:30 m:_000h_ 000:0:00 Ammv
00::“_H:HMW0%MM0MMM::M: A V 0:0.A0:_>\0:0:0:O A8 0:0—000:0: _0::0::0::0 0:: 02% A v
0 =.: 6 30:0 300.5 Abv 000:0 .05::06m< Avv 0:0 00:03:; 0:35-600 £00080 ANNV : . < N
m 0 8:58 05 05.020 08
0:0—8:5 00:0:0w0>-:0z Amv _0:::_:0::w< Amv
%
:050 0:0 :000 A0:
0>_:00::xm ASV
0m: 0::
606500 “Ema—Sm Am: 0:0—603w :00:D A03 6x :0 : _0_:_::m:00:“ mew
:0 06: :00000M A3 wEEE . 3:00:05
_ .0. 00.: A v : :0: I 02 um .x A v 60055 0 0 0
:: .< m rm 0 .: 0 m m .m:0_:00_::E::0U 0000000600; Av: \: 0: 0 m
:00:D 36:85 :wE ANV :00:D 56:85 33 AS . 050005 A m: E0055 A:
50:3 E6505 33 A: :00:D ~A6:85 :33 A3 0:058:65 .600_>:0m ._0_0:§00 ANS
_0_::0060m ASV
:0 as: 0.0 :35 A:
E Sou 30.9952 mag £52 mum: ~A0:6 6:;

 

0:000:08 :0>00 0:0— .:0 :00» 0:0 805 0000 800056003 093 :0>00 0:0—
05 00005 6060550: 5 6:09:02 .093 :0>00 0:0— 3 :0>00 0:0— :0w5052 Sow. 0:0 .33 £03 :0: 00000560305 .56 030.5.

 

 

 

 

 

 

 

:0000 0:23 0:05:02 A200 :0:::00 0:0—305 Ammvv 60:0&

60:0: 30:00:00 0:0—305 Ammv 008% 0005 A00 30:05:00 Ava 30:00:00
2:: 08.: 05.33 060 05. 02o: 00 05:33 00

6000300: 0:0—30.1300 60:0 2

6:0: :00

:85: $8280 2233 0.0: 08322 0232 03:2 :5 00053 :20: “28“....on A: :0 82.280
2.2 020: 0.: 20:33 E

60:02

2:: v0.00000 :8
60:0 05:: :0 0 . . 0x:
0 0 . 0 _ D ANS 00000000300: E00602 Am: HERA“ A8
50080—0
0008:0080 as
00 6086 0w0:0

8:68 05:0: 02:2 :00 8:: :8: 22:: $0: 0:02:80“:st 00$: MW 008:

6:03:00 0:03: :050 A08 05: 00% A08 . 0:5 2va 30:00:00
80:: a: 8:: 223 as :85: 0:05.080 80 Baa: 5

60:0,: 0:800:00 :050 ANS 000— 60: 0 m A 00
32.28:. 22% 03:2 6: :00 a: 800830 025-802 6:0
60000000 0:03: :050 A50 0000:0003: A03 000300: 00:00 ANGV 60:0 m
0000:0060 :006< A0: 003 :0 E0 0 A 000300: E00602 A: Sb 0:0: :00

:0000060 0.00 An: 0 0 a :0: Z .0: 60000500 0. Q

. . 60:0: 60000600 . 0:0—:3 AE
:0006060 0000—0005 050 Am: 3:00:03 60:00 0030—0005 CE
000300: E0560 2 A30 0:0_ 60:02 A00
A052": 58 m<0\mgz mom: man—2 mg: .006 6:2.

 

32580 :0 030:.

87

 

 

 

 

 

 

 

00:0:0m0> 306005200 :050 Ammv 0.00: 0060000 A20
6:00 0:2 89 0::0 000m Amt 0:0_
:60: 0060000 Ammv 002 0003910 0:000:00: .0000m Amt 00::0m AN:
006008 A30 00::0m AC
600.05 00:0 A08
6090602 Ammv
. m0: 0
.5003 AMNV .0803 A03 wens ANWV 20.0”“ C C
903:0:03 0:0 08025 A20
60:000 :80? AD
0:06 :00:m:0>0
003020—000: 000.305 A03
000003 00608: :050 Ann
000003 60:00-00: 0002 A00 0:06 50:30:30
000003 :00w:0::m A08 00>00_-000:0 000—33 A00 000003 000003-002 ANS 000003 A0:
0:06 0:0—30x— Awmv 0:06 0:0:2000 000003 000003 A08
000:3 0:08.: :5 832-085 0533 60 20:03 60
0:06 0:330. :050 ANS
300000
03000003 003000 A 0 O
60:02
60:0: 00:00 000305 A08 0003000 00:02
. :0 :000 :0 30 0x: .
0. 0 _ _ 0 .2 Amvv 0:0—30..— A00
A2": SON 00.9952 mag 92 mg: .A0:6 602.

 

08508 0.0 0300

 

88

analysis. Land cover associated with the trap was created for each year the trap was used.
A total of 474 traps were analyzed (Table 3.2).

Spatial data layers were manipulated in ArcView GIS and projected in Michigan
GeoRef (http://www.michigandnr.com/spatialdatalibrary/Map_Projections.htm). Buffers
(radius = 4 km) were constructed around each trap, as previous studies have suggested
that cowbirds will commute that average distance to access agricultural lands (Young and
Hutto 1999). The buffer was used to clip the land cover layer, resulting in a clipped
coverage that included all of the associated land cover within 4 km radii of the traps for
the given year. The total area Ora) of each land cover type present was tabulated for each

clip and associated with the corresponding capture totals for each cowbird trap.

3. 3. 2. Statistical analysis. The cowbird catch total for each trap within a period
was treated as a unique trap. A one-way ANOVA (a = 0.05) was used to detect
differences in captures amongst the three time periods (Periods l - 3). A Tukey-Kramer
post hoc test (a = 0.05) determined which years were different with respect to trap
captures.

Land cover data were not normally distributed, therefore a Spearman rank test (a
= 0.05) was used to determine correlations between cowbird trap captures and the total
area of each land cover type within each 4 km radiated clip. Correlation tests were run
separately for each period because of the differences in data collection methods between

the three years of respective land cover.

89

Table 3.2. Annual numbers of brown-headed cowbird traps used and corresponding
capture totals for 1977—79, 1992-94 and 2000-02 in northern Lower Michigan.

 

 

 

Year # of active traps Total # of cowbirds captured
1977 39 3284
1978 40 341 1
1979 37 3691
1992 49 3737
1993 51 4614
1994 56 3109
2000 70 4345
2001 68 3906
2002 65 3722
Total 474 33819

 

90

In many instances, the 4 km buffers constructed around traps overlapped
substantially, potentially introducing questions about the independence of the traps and
the land cover that surrounded them. However, traps were not randomly positioned, but
rather strategically placed based on the age of the jack pine stand, the parasitism rates in
the stand, and the number of Kirtland’s warblers present. Since traps were not randomly
located, inherent biases existed in their capture rates. However, based on my assumption
that cowbirds were associating with traps based on the surrounding land cover, I
determined that proximity of traps to each other did not confound the statistical

associations between trap captures and land cover.

3.4. Results

Comparisons of the land cover across the three time periods could not be made
due to the variability in methods by which the years of land cover were created.
However, it remained important to demonstrate, albeit not statistically, how the
associated land cover changed through the periods as the number and locations of traps
also changed (Table 3.3).

There were generally no differences in cowbird captures within a period, except
for the years 1993-94, which exhibited a substantial decrease (p < 0.01; Figure 3.2).
Analyses showed that cowbird captures decreased across the three time periods (Figure
3.3).

Cowbird trap captures were positively correlated with urban areas in 2001 (Table

3.4). A positive correlation existed between trap captures and agriculture areas in 1993.

91

Table 3.3. Mean (SE) area (ha) of land cover types within a 4 km radius of brown-
headed cowbird traps during 1977-79, 1992-94, and 2000-02 in northern Lower

 

 

Michigan.
Land cover type 1977-79 1992-94 2000-02
n=116 n=156 n=202
Urban 55.6 (5.1) 190.8 (31.3) 91.5 (7.6)
Agriculture 8.0 (2.9) 55.3 (9.6) 16.9 (2.7)
Openland 590.5 (27.7) 904.6 (106.4) 1 146.5 (30.1)
Upland deciduous forest 1466.5 (53.0) 797.3 (43.4) 670.4 (27.5)
Upland coniferous forest 2594.1 (69.0) 2840.8 (48.9) 2001.7 (40.6)
Upland mixed forest 0.0 (0.0) 18.7 (2.3) 580.9 (13.9)
Lowland deciduous forest 134.9 (16.1) 18.4 (3.1) 60.9 (5.4)
Lowland coniferous forest 114.7 (13.9) 98.6 (7.7) 2559.1 (1363.5)
Lowland mixed forest 0.0 (0.0) 198.4 (18.9) 3.3 (0.5)
Wetland 91.3 (8.6) 66.3 (6.9) 188.6 (11.1)
Water bodies 26.4 (2.6) 40.2 (5.4) 44.7 (6.1)
Barren land 0.0 (0.0) 0.0 (0.0) 11.1 (1.4)

 

92

.0008 00.. 0000 003000 00000030000 003 000000 00006003 000
0008 000000000 009 00600 0 5003 000000.60 00000.00w6 0 000000000... .36 H 0 .600 000009000

002:8 00000000000403 003-000 .6. 8000 8-800 .00 00000 3.002 .0 000000 00
-030 m06:0 0000 000E0w00000 00—0003 0.0000000 05 5 6003000 000 000300 Ammv 0002 .N.m .wE

a
NOON room ooom 0am? mam? «map one 059. RE.

J . . a . . q . i on

 

 

 

 

o
co
SGHIEIMOC) # NVEIW

 

 

l

 

oo_.

. 0:.

 

. omw

 

. on?

93

.0000» 030000 0000000000 00000.00me 00000000 000000 00000005 .36 u 0 .600 000609000

0:00:00 000009-303. .<>OZ< 033-000 .3 000.00% Noéoom 000 AN 030$ 3&02 .3 00605 on
-030 0000» 00 m0t:0 0000 000E0w00000 00—0003 0.00000M 0.0 :0 000 000 0003000 000300 0002 .m.m .wE

moan

 

 

ID
[x
SGHIHMOC) # NV EIW

 

. 00—.

 

. mo_.

94

Table 3.4. Correlations between brown-headed cowbird trap captures from 1977-79,
1992-94, and 2000-02 and associated land cover types using three years of Michigan land
cover. (Spearman rank correlation, a = 0.05)

 

 

Land cover type 1978 1993 2001
Urban 0.03 -O.13 017*
Agriculture 0.05 0.29* 0.13
Openland -0. 16 0.07 -0.04
Upland Deciduous Forest 0.16 024* 0.09
Upland Coniferous Forest 000 -0.35* -0. 17*
Upland Mixed Forest ------- ** -O.20* 0.32*
Lowland Deciduous Forest -0. 14 0.18* -0.03
Lowland Coniferous Forest 029* -0. 19* -0.04
Lowland Mixed Forest ------- 0.24* -0.10
Wetland -0.18* 0.10 -0.14*
Water bodies 0.09 -0.03 0.21*
Barren land ............. 0.190

 

* Denote significant correlations between captures and cover type. ** Dotted lines

indicate land cover classes that were not classified for a particular year, or did not occur

in study site in that year.

95

There were no significant relationships between cowbirds and openlands for 1978, 1993
or 2001. Upland deciduous forest was positively associated with trap captures in all
years, but was only significant in 1993. Upland coniferous forest, which included jack
pine forest, was negatively correlated with cowbirds in all years, with significance in
1993 and 2001. Cowbirds were negatively related to upland mixed forest in 1993, but
positively correlated in 2001. Lowland deciduous forest was positively associated with
trap captures in 1993. Lowland coniferous forest, which included jack pine forest, was
negatively correlated with cowbirds in all three years, with significance in 1978 and
1993. Lowland mixed forest was not classified in 1978, so no analyses were performed
on it for that year.

However, trap captures were positively correlated with lowland mixed forest in
1993. Wetlands were negatively associated with trap captures in 1978 and 2001, with a
non-significant positive association in 1993. Water bodies were positively correlated
with cowbirds in 1978 and 2001, but only significant in the latter year. Barren land was
not represented in 1978 or 1993, but a positive correlation with cowbirds did occur in

2001.

3.5. Discussion

The decline in cowbird captures over the three time intervals is concurrent with
the decline in cowbird populations nationwide, and in Michigan specifically (Sauer et al.
2003, M. DeCapita, pers commun.). While there is little knowledge as to why cowbird

populations have decreased, Michigan’s intensive Cowbird Control program might play a

96

role; so too, could the increase in forest cover and decrease in agriculture that has typified
the northern Midwest region in recent decades (Parody et al. 2001).

Cowbirds were negatively correlated with upland and lowland coniferous forests
but positively associated with upland and lowland deciduous forest. This is consistent
with findings by Tewksbury et al. (1999) and Beutler (2000), who showed cowbirds
avoided upland and riparian coniferous forests in favor of riparian deciduous forests.
Tewksbury et al. (1999) suggested that while host density in riparian coniferous forests
tends to be high, cowbirds are less able to locate and parasitize nests due to the dense
vegetation. This might also account for the consistent negative correlations between trap
captures and wetland areas, which are often characterized by densely distributed trees,
reeds and grasses. Young and Hutto (1999) also found cowbirds to be abundant in
deciduous riparian areas; however, in their study, cowbirds also inhabited upland
coniferous forests, a finding which was not supported by this study.

While no direct correlation analyses were conducted specifically for jack pine
cover and cowbird captures, it is worth noting the negative correlation between cowbird
trap captures and upland coniferous forest because jack pine falls into this cover class.
This might explain the lack of differences between stand sizes and among stand ages in
my field study (Chapter 2). These findings also have implications for the Cowbird
Control Program. Placing traps in jack pine stands might not be the best strategy, since
cowbird captures are negatively associated with coniferous forest cover, both in this
study and in previous studies (Tewksbery 1999, Beutler 2000).

Cowbirds were positively correlated with urban and agriculture, despite the fact

that both land cover types comprised a relatively small proportion of the study sites.

97

These results were consistent with several previous studies of cowbird associations with
agriculture (Goguen and Matthews 1999, Saab 1999, Stribley and Haufler 1999,
Tewksbury et al. 1999, Young and Hutto 1999, Curson et al. 2000) and urban areas
(Coker and Capen 2000). Perhaps most interesting is that the amount of available
agriculture within trap proximity was greatest in 1993; this was also the only year in
which a significant positive correlation existed between cowbirds and agriculture. This
result supports the general hypothesis that cowbirds respond positively to increasing
agriculture due to increased access to food sources, such as grain seeds and insects.

Based on results from this study, traps would be most advantageously located in
agricultural areas or upland deciduous forests, as both of these cover types were
consistently positively correlated with trap captures. This study showed that upland
coniferous forest is not an advantageous location to place traps, as cowbird captures were
consistently negatively associated with this cover type. However, if traps must be
located in jack pine stands, a modified trap placement scheme should be considered.
Managers should consider placing cowbird traps in jack pine stands that are in proximity
to deciduous forest or agricultural areas to attract more cowbirds.

There were limitations on the extent to which comparisons could be made
between the three years of land cover, due to the variability in methods by which they
were created. Therefore, associations could only be made between cowbird captures and
individual years of land cover, but no analysis to detect trends across the three years
could be done. This is unfortunate, because examining how the changing locations of
traps have affected cowbird captures could be helpful in making management decisions

for future trap placement.

98

3.6. Literature Cited

Beutler, D. K. 2000. Distribution of the brown-headed cowbird in northeastern
Washington: local and landscape factors. Doctoral dissertation, Washington
State University, Pullman, Washington, USA.

Coker, D. R., and D. E. Capen. 2000. Distribution and habitat associations of brown-
headed cowbirds in the Green Mountains of Vermont. Pages 236-243 in J. N. M.
Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors.
Ecology and management of cowbirds and their hosts. University of Texas Press,
Austin, Texas, USA.

Curson, D. R., C. B. Goguen, and N. E. Mathews. 2000. Long-distance commuting by
brown-headed cowbirds in New Mexico. Auk 117(3): 795-799.

DeCapita, M. 2000. Brown-headed cowbird control on Kirtland’s warbler nesting areas
in Michigan, 1972-1995. Pages 333-341 in J. N. M. Smith, T. L. Cook, S. I.
Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management of
cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Friedmann, H. 1929. The cowbirds: a study in the biology of social parasitism. Charles
C. Thomas, Springfield, Illinois, USA.

Goguen, C. B., and N. E. Mathews. 1999. Review of the causes and implications of the
association between cowbirds and livestock. Studies in Avian Biology 18: 10-17

Griffith, J. T., and J. C. Griffith. 2000. Cowbird control and the endangered Least Bell’s
vireo: a management success story. Pages 342-356 in J. N. M. Smith, T. L.
Cook, S. l. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Hahn, D. C., and J. S. Hatfield. 1995. Parasitism at the landscape scale: cowbirds prefer
forests. Conservation Biology 9(6): 1415-1424.

Hayden, T. J ., D. J. Tazik, R. H. Melton, and J. D. Cornelius. 2000. Cowbirds control
program at Fort Hood, Texas: Lessons for mitigation of cowbird parasitism on a
landscape scale. Pages 357-370 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S.
K. Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

https://ecos.fws.gov/species_profile/SpeciesProfile?spcode=BO3I, retrieved November 5,
2006.

https://www.mi.ngb.army.mil/grayling/default.asp, retrieved November 5, 2006.

99

http://www.michigan.gov/dnr/O,1607,7-153-10371_10402-148280--,00.html, retrieved
April 17, 2007

http://www.michigandnr.com/spatialdatalibrary/Map_Projections.htm, retrieved April 17,
2007

http://www.en.wikipedia.org, retrieved November 5, 2006.

Huber, P. W., J. A. Weinrich, and E. S. Carlson. 2001. Strategy for Kirtland’s warbler
habitat management. Publication of MDNR, USFS, and USFWS. Mio,
Michigan, USA.

IFMAP; MDNR, Forest, Mineral and Fire Management Division 2003

Mann, W. K. 1980. Mack Lake fire briefing. U. S. Forest Service, Huron-Manistee
National Forest, Mio, Michigan, USA.

Parody, J. M., F. J. Cuthbert, and E. H. Decker. 2001. The effect of 50 years of
landscape change on species richness and community composition. Global
Ecology and Biogeology 10: 305-313.

Robinson, S. K., J. D. Brawn, S. F. Morse, and J. R. Herkert. 1999. Use of different
habitats by breeding brown-headed cowbirds in fragmented midwestem
landscapes. Studies in Avian Biology 18: 52—61.

Rothstein, S. 1., J. Vemer, and E. Stevens. 1984. Radio-tracking confirms a unique
diurnal pattern of spatial occurrence in the parasitic brown-headed cowbird.
Ecology 65(1): 77-88.

Rothstein, S. I., and T. L. Cook. 2000. Introduction. Pages 323-332 in J. N. M. Smith,
T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and

management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Saab, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian
forests: a hierarchical analysis. Ecological Applications 9(1): 135-151.

Sauer, J. R., J. E. Hines, and J. Fallon. 2003. The North American breeding bird survey,
results and analysis 1966 - 2002. Version 2003.1, USGS Patuxent Wildlife
Research Center, Laurel, MD.

Stribley, J. M., and J. B. Haufler. 1999. Landscape effects on cowbird occurrences in
Michigan: Implications to research needs in forests of the inland west. Studies in
Avian Biology 18: 68-72.

100

Terpening, K. K. 1999. Golden-checked warbler fatality in a cowbird trap. Studies in
Avian Biology 18: 290-291.

Tewksbury, J. J ., T. E. Martin, S. J. Heijl, T. S. Redman, and F. J. Wheeler. 1999.
Cowbirds in a western valley: effects of landscape structure, vegetation, and host
density. Studies in Avian Biology 18: 23-33

Thompson, F. R. 1994. Temporal and spatial patterns of breeding brown-headed
cowbirds in the Midwestern United States. Auk 111(4): 979-990.

Whitfield, M. J. 2000. Results of a brown-headed cowbird control program for the
Southwestern willow flycatcher. Pages 371-380 in J. N. M. Smith, T. L. Cook, S.
I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management
of cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Whitfield, M. J ., K. M. Enos, and S. P. Rowe. 1999. Is brown-headed cowbird trapping
effective for managing populations of the endangered southwestern willow
flycatcher? Studies in Avian Biology 18: 260-266.

Winter, K. J. and S. D. McKelvey. 1999. Cowbird trapping in remote areas: alternative
control measures may be more effective. Studies in Avian Biology 18: 282-289.

www.census. gov, retrieved April 18, 2007.

Young, J. S., and R. L. Hutto. 1999. Habitat and landscape factors affecting cowbird
distribution in the Northern Rockies. Studies in Avian Biology 18: 41-51.

101

CHAPTER FOUR

A HABITAT SUITAB ILITY INDEX MODEL FOR THE BROWN—HEADED

COWBIRD (MOLOTHR US ATER)

102

4.1 HABITAT USE INFORMATION

4.1.1. General

Brown-headed cowbirds are distributed throughout all 48 contiguous United
States, with breeding grounds as far north as southern Canada, and wintering grounds
south to southern Mexico (Lowther 1993, Stokes and Stokes 1996). However, cowbirds
rarely occur outside of North America (McKay 1994). Although they are considered a
short grass edge species (Ortega 1998), cowbirds are commonly dispersed throughout a
large range of habitat types, including, but not limited to, deciduous forests, coniferous
forests, shrub steppes and riparian areas (Robinson et al. 1999, Tewksbury et al. 1999,
Vander Haegen and Walker 1999). They are closely associated with the presence of
grazing livestock, human developments, and highly fragmented tracts of land with
abundant edge (Donovan et al. 2000, Goguen and Mathews 1999, Goguen and Mathews
2000). Cowbirds are Opportunistic and are able to adapt to many different types of
ecosystems, often in spite of variable vegetation composition and structure.

Because cowbirds are temporally and spatially variable in their habitat utilization
(breeding in forests in the morning, foraging in agricultural areas in the afternoon and
roosting in openlands in the evening), their home ranges are often very large and can
differ between regions. For example, Thompson and Dijak (2000) found a breeding
range of 43 - 253 ha, a non-breeding range of 89 - 242 ha, and a total home range of 261 -
845 ha in a study of cowbirds in Illinois and Missouri. A study done by Rothstein et al.
(1984) in the Sierra Nevada forest concluded the mean breeding range was 68 ha, and the
total home range averaged 442 ha. Gates and Evans’ (1998) study in Maryland reported

a mean home range of 1,592 ha, although only 3.5 % of it was used for daily activities.

103

However, Teather and Robertson (1985) reported home ranges of only 10 ha for female
cowbirds in Ontario. Differences in breeding ranges might be attributed to the host
density in that area, as female cowbirds could be forced to travel further to find suitable
hosts (Ortega 1998). Total home range could vary considerably depending on the
distance between breeding and foraging sites.

Brown-headed cowbirds can alter the nesting success rates for many of the
species in the avian communities to which they belong. Cowbirds are generalist brood
parasites, and are known to parasitize over 220 avian species (Lowther 1993, Ortega
1998). While many bird species have developed defenses against cowbird parasitism,
several species will raise cowbird chicks to the detriment of their own. This pattern has
resulted in cowbird populations maintaining resiliency, while many other avian

populations have steadily declined.

4.1.2. Foraging

Cowbirds are ground foragers, and are strongly associated with grazing ungulate
species (Goguen and Mathews 1999, Goguen and Mathews 2000, Kostecke et al. 2003).
Historically, cowbirds followed the bison herds of the Great Plains to feed on insects that
were disturbed by their movement (Lowther 1993). Presently, cowbirds continue this
behavioral pattern with grazing livestock that occur in agricultural areas.

Cowbirds are dependent upon agricultural areas and human developments to
fulfill their foraging needs (Coker and Capen 2000). Young and Hutto (1999) concluded
that the strongest predictor of increased cowbird presence in the northern Rockies of

Montana was decreased distance to agriculture. Similar conclusions were made by

104

Stribley and Haufler (1999) for cowbirds in northern Michigan and by Tewksbury et al.
(1999) in their study in Montana. Stribley and Haufler (1999) reported cowbirds were
three to three and a half times more likely to occur when sites were within 3 km of
agriculture. Tewksbury et al. (1999) never encountered cowbirds beyond 4 km from
agricultural areas in their study in Montana. However, in their New Mexico studies,
Goguen and Mathews (1999) and Curson et al. (2000) found that cowbirds will commute
well beyond 10 km to access agricultural sites. Cowbirds are also attracted to human
developments, such as horse corrals, pack stations and birdfeeders (Rothstein et al. 1980,
Wright 1999, Coker and Capen 2000). Therefore, access to feeding sites appears to be
the limiting factor for cowbird occurrence.

Cowbirds are known to feed on weed seeds and insects, including grasshoppers
and beetles (Lowther 1993). Female cowbirds will also intake protein by eating host
eggs removed from parasitized nests (Jackson and Roby 1992). Ortega (1998) reported

that females also consume large amounts of mollusk shells.

4.1.3. Breeding

Brown-headed cowbirds do not build their own nests, and therefore are dependent
upon suitable hosts to rear their chicks. Cowbirds sometimes are able to differentiate
rejecter species (bird species that can recognize and discard the foreign egg) from
acceptor species (bird species that do not have defenses against cowbird parasitism and
act as hosts), and will breed in areas where accepters are more abundant (Strausberger
and Ashley 1997, Woolfenden et al. 2003), with some exceptions (Hahn and Hatfield

1995). Suitable host density, therefore, appears to be an important variable for predicting

105

whether and to what extent cowbirds will occur (Vemer and Ritter 1983, Strausberger
and Ashley 1997, Hejl and Young 1999).

Host species density can vary considerably between vegetation types (T ewksbury
et al. 1999), which can have an effect on whether cowbirds are associated with these
areas. While cowbirds will use many different forest types, they are more highly
associated with deciduous forests than with coniferous forests (Hejl and Young 1999,
Tewksbury 1999, Beutler 2000). This is perhaps due to decreased host densities in
coniferous forests, as well as denser vegetation, which can reduce nest visibility for
cowbirds. Results I obtained from an analysis of cowbird traps in Michigan showed that
cowbird captures are positively associated with upland and lowland deciduous forest, and
negatively correlated with upland and lowland conifers (Chapter 3). Regional differences
in climate, topography and vegetation also appear to play a role in which vegetation types
cowbirds select more often for breeding (Hahn and Hatfield 1995).

While the cowbird’s historical range was limited to the grasslands of the Great
Plains, the species now appears to favor host-rich forests and riparian areas as breeding
habitat (Strausberger and Ashley 1997, Hejl and Young 1999, Robinson et al. 1999,
Tewksbury et a1. 1999, Young and Hutto 1999). Cowbirds still occur in grassland areas
and parasitize host species there at high rates, especially as these areas have become more
fragmented in North America (Davis and Sealy 2000). However, when several
vegetation types are available, cowbirds often favor forested and riparian areas over
grasslands.

Female cowbirds typically parasitize ground and mid-story nesting birds (Briskie

et al. 1990, Hahn and Hatfield 2000, Howe and Knopf 2000). To identify and access

106

these nests, there must be limited obstructions, such as large trees and shrubs (Staab and
Morrison 1999). However, dead trees and snags can be convenient perching spots for
female cowbirds to locate and observe potential host nests. Snags were significant at the
a = 0.10 level for cowbird abundance in my field study (Chapter 2). In the field, I
observed females and many singing males (presumably soliciting or communicating with
a mate) perched in tall snags. Snags are also potential sites for host nests, such as those
of the black-capped Chickadee (Poecile atricapilla) or Eastern bluebird (Sialia sialis), for
cowbirds to parasitize.

Cowbirds are positively associated with ground and mid—story cover, presumably
because ground-searching cowbirds are more likely to find nests in such structurally
complex areas (Stribley and Haufler 1999, Young and Hutto 1999, Uyehara and
Whitfield 2000). However, while positively associated with deciduous forests, cowbirds
tend to avoid mature, old-growth stands with high canopy cover (Vemer and Ritter 1983,
Young and Hutto 1999).

Many studies have documented heightened cowbird parasitism with increasing
landscape fragmentation and proximity to stand edge (Brittingham and Temple 1983,
Robinson et al. 1995, Thompson et al. 2000). Stribley and Haufler (1999) documented
that census points > 300 m from an edge had a lower occurrence of cowbirds. Davis and
Sealy (2000) noted the highest rates of parasitism occurred on the narrowest site in their
study, indicating that the decreased distance between interior and edge allowed the
cowbirds to be more successful in penetrating the interior. Donovan et al. (2000) also
reported increased cowbird abundance and parasitism levels in small, fragmented tracts.

However, sensitivity to edge might be a regional phenomenon, because Davis (2004)

107

 

reported no response to edge by cowbirds, and Burhans and Thompson (1999)
documented a positive association between cowbird parasitism and increasing patch size

(increased distance from interior to edge).

4.1.4. Roasting

While many studies address the requirements for cowbird breeding and foraging
habitat, there has been relatively little documentation for roosting habitat. Curson et al.
(2000) documented individual females roosting in a variety of habitats, including a
riparian area and a forested valley, which were in or near breeding grounds. They also
reported that large congregations of cowbirds roosted with other blackbird species in a
cattail marsh in open prairie. Thompson (1994) found large communal roosts of
cowbirds in a stand of flooded silver maples, while Rothstein et al. (1984) documented
cowbirds roosting on their breeding grounds. Vemer and Ritter (1983) reported large
flocks of cowbirds roosting in willow (Salix spp.) thickets. These various findings

suggest that roosting habitat might not be as specific as breeding and foraging habitat.

4.2 HABITAT SUITABILITY INDEX (HSI) MODEL
4.2.]. Model development

Information used to develop this model has been garnered from various research
studies throughout the midwest and from field surveys performed in the northern Lower
Peninsula of Michigan. While regional differences sometimes arise in terms of habitat
usage and parasitism rates, most studies throughout the country have had similar results

regarding habitat requirements for the cowbird. Due to a scarcity of publications on

108

cowbirds in Michigan, this model will use information from studies conducted
throughout the Midwest, in conjunction with the surveys I conducted in my field study
(Chapter 2). The studies used were all conducted within the last 25 years, which is
important because cowbird range and distribution throughout the United States has

changed substantially in the last 50 - 100 years.

4.2.2. Model applicability

This model is applicable to Michigan and the Upper Great Lakes region, which
have large agricultural and forested areas. While the majority of the publications used to
build this model are from the Midwest region, caution should be applied in using this
model outside of the specified area. This is particularly the case for determining habitat
suitability for the Great Plains grassland and prairie region, where cowbirds are native
and increasingly abundant; however, studies of these systems were not a strong
consideration for constructing this model. Within the Michigan and Upper Great Lakes
region, this model can be applied to forested systems, riparian areas and agricultural and

human-developed areas with a forest component.

4.2.3. Model variables

The most significant variable in determining the presence and abundance of
brown-headed cowbirds is proximity to foraging areas (agricultural lands and human
developments). Distance to foraging sites has been cited as the most important predictor
of cowbird occurrence in several studies (Hejl and Young 1999, Stribley and Haufler

1999, Tewksbury et al. 1999, Young and Hutto 1999). Cowbird captures in my trap

109

analysis were also positively associated to surrounding agricultural cover (Chapter 3).
However, this variable only accounts for meeting the food needs of the cowbird; it does
not account for the species’ reproductive requirements. Distance between forest interior
and edge is important in cowbird breeding habitat since the cowbird is an edge species
(Brittingham and Temple 1983, Ortega 1998), preferring proximal access to forest edges
for nest observation and parasitic behavior. This is being considered the primary
breeding variable because it allows cowbirds to access both edge and interior host nests.
Finally, habitat conditions necessary for breeding include understory (ground and mid-
story) cover (Brittingham and Temple 1983, Stribley 1993, Stribley and Haufler 1999)
that will support nests but will not shield cowbirds from finding them.

Variable 1 (Distance to foraging sites)

Brown-headed cowbirds are thought to be most highly associated with areas that
are in close proximity to foraging sites such as agriculture and human developments.
Analyses of data from my study in jack pine forests of northern Michigan (Chapter 2)
were limited because no survey points were located beyond a distance of 4 km from
agricultural or urban areas. However, in my analysis of cowbird traps in northern
Michigan (Chapter 3), despite comparatively small amounts of agricultural (55 ha) and
urban (190 ha) land cover surrounding the traps, cowbirds were positively associated
with both. Studies in the Midwest have shown cowbirds will commute a maximum of 7
km between breeding and foraging sites, but average a distance of 1.2 - 4.4 km
(Thompson 1994, Stribley and Haufler 1999, Thompson and Dijak 2000). Therefore, the
optimal distance between breeding and foraging sites is S 4 km; sites that are within this

range receive a suitability index (SI) value of 1.0. Suitability of habitat decreases as

110

 

distances to foraging areas increase to 7 km, after which habitat is considered unsuitable
and is assigned an SI value of 0 (Figure 4.1).

Variable 2 (Distance to edge)

The second variable for this model is distance between the stand interior and
stand edge. Stribley and Haufler (1999) determined that cowbird abundance was highest
< 300 m from the stand edge, and Howell et al. (2007) found 90% of cowbird
occurrences within 150-350 m of forest edge. Brittingham and Temple’s (1983)
Wisconsin study concluded that parasitism by cowbirds decreased as distance from edge
increased from O to > 300 m. Thompson et a1. (2000) found that cowbirds can be found
more than 400 m from edge, but theorized edge is only a factor when other constraints,
such as host density, are present. Gustafson et al.’s (2002) Midwest study found an
increase in cowbird parasitism as distance to edge increased, until 500 m, when
indications of parasitism ceased, except for one case at 800 m. My surveys yielded no
associations between cowbirds and distance to stand edge (Chapter 2). For the purpose
of this model, distance to edge should be no more than 300 m for a suitability index of
1.0. Distances more than 300 m between interior and edge will decrease in suitability,
and distances > 600 m will be considered unsuitable, with a score of 0 (Figure 4.2).

Variable 3 (Vertical cover: ground cover)

Brown-headed cowbirds need ground cover (0—1 m tall) in great enough
percentages to support host nests, but not so dense so as to prevent nest detection. My
surveys did not detect a correlation between cowbirds and ground cover (Chapter 2).
However, Stribley (1993) indicated cowbird abundance was greater in stands with

increased percent vertical cover (x = 0.29) in the 0 - 1 m height strata. Brittingham and

111

 

 

1.2

.o
oo

Suitability
O
O

0.4

0.2

 

010002000300040005000600070008000
Distance to agriculture (m)

 

Fig. 4.1. Suitability index of distance to agriculture (m; variable 1) for brown-
headed cowbirds in forest systems in northern Michigan.

112

 

 

1.2

Suitabllity
.o .o .o
-b O) on

9
N

 

0 100 200 300 400 500 600 700 800
Dlmnoo to edge (in)

 

 

 

Fig. 4.2. Suitability index of distance to edge (in; variable 2) for brown-headed cowbirds
in forest systems in northern Michigan.

113

Temple’s (1996) Wisconsin study found that cowbird parasitism was higher when
percent vertical cover (0 - 0.5 m) was greater (x = 0.53). For cowbirds in northern
Michigan, ground cover must occur in the range of 30 - 55% to be considered optimal
and to receive a score of 1 (Figure 4.3). Ground cover decreasing from 30% or
increasing from 55% declines in suitability; a suitability index of 0 is assigned to 0% and
100% ground cover.

Variable 4 (Vertical cover: midstory cover)

Midstory cover (1 - 7 m) is necessary for providing observation points for
locating nests in the ground cover below (Stribley 1993). My vegetation sampling found
no relationship between midstory cover and cowbird abundance (Chapter 2). However,
Stribley and Haufler (1999) found that cowbird abundance was greater in stands with
substantial midstory cover (x = 0.55; 1 - 7 m). However, Brittingham and Temple (1996)
reported cowbird parasitism in stands with decreased percent midstory cover (3 - 10 m; x
= 0.46), compared with non-parasitism in stands with increased percent midstory cover (x
= 0.58). Midstory cover will be considered optimal in the range of 45 - 55%, and receive
a score of 1 (Figure 4.4). Midstory cover decreasing from 45% and increasing from 55%

declines.

4.2.4. Model description
This model attempts to portray the food and reproductive needs of the brown-
headed cowbird in forest systems. While the cowbird will inhabit many different

vegetation types, including grassland and riparian areas, forested systems (or riparian,

114

 

1.2

.0
on

Suitability
O
'0

0.4

0.2

 

0 20 40 60 80 100 120
% Ground Cover

 

 

 

Fig. 4.3. Suitability index for percent (%) ground cover (variable 3) for brown-headed
cowbirds in forest systems in northern Michigan.

115

 

1.2

.0
co

Suitability
p
O,

 

0.4
0.2
0
O 20 40 60 80 100 120
% Mldaory Cover

 

 

 

Fig. 4.4. Suitability index model for percent (%) mid-story cover (variable 4) for brown-
headed cowbirds in forest systems in northern Michigan.

116

agricultural and urban systems with a forest component) are the ecosystems around which
this model is being based. Habitat variables have been categorized by how they
contribute to a cowbird’s life requisite, or requirement for survival (Figure 4.5). Distance
to agriculture is defined as a food requirement, while distance to edge, percent ground
cover, and percent mid-story cover are all reproductive requirements. The combination
of these four variables comprises the habitat suitability for cowbirds in upland forest
systems.

4. 2. 5. HS! determination

This model was constructed based on HSI models developed previously for other
songbird species (Schroeder and Sousa 1982, Sousa 1983, Raymer 2000). The four
variables outlined in this model will contribute to the overall habitat suitability for the
brown-headed cowbird in forests in the Michigan and Great Lakes region.

In utilizing the formula for the H81, it should be noted that only distance to
agriculture (S11) is noncompensatory; that is, if a SI of 0 is achieved for this variable, the
values of other variables cannot compensate for this, and therefore the overall HSI will
equal 0. However, distance to edge (SIZ), % ground cover (813) and % mid-story cover
(S 14) are compensatory variables; that is, if any one of them receives a SI value of 0, the
H81 will not necessarily equal 0, as the values of the other variables can make up for the
0 rating. In this model, of the three compensatory variables, distance to edge is
considered more important than either of the % cover measures, or than their summed
measure. Therefore, distance to edge is weighted twice as important as these cover
measures. The % cover measures were considered of equal importance, and therefore,

were weighted equally.

117

6000.0

“000000000

00008 0000 0003 00>0U

 

0080002 0000000 8 000300 000000
03000 000 0008 0mm 05 00 000000000 05 000 0003 00>00 000 00068000 000 0000060.» 000000 000300 6006000000 0.0 .mE

000 0006-008 00

 

0000000000

080 ll

 

0068000 005

00.300 00:00 00
A80 0w00 00 000065

A80 08000000 00 000065

I038? 000%:

00508 000000080000
00008 0000008005
m5 0 8 080002

.0100 0 8 080002

000008 0: 80m

118

Food and reproduction were the only life requisites addressed in this model;
therefore, the HS] for the brown-headed cowbird is equal to the life requisite values for

food and reproduction. The formula is:

HSI = 311 x (2 (312)"2 + (S13 + 314)”)
4

119

4.3. Literature Cited

Beutler, D. K. 2000. Distribution of the brown-headed cowbird in northeastern
Washington: local and landscape factors. Doctoral dissertation, Washington
State Uinversity, Pullman, Washington, USA.

Briskie, J. V., S. G. Sealy, and K. A. Hobson. 1990. Differential parasitism of least
flycatchers and yellow warblers by the brown-headed cowbird. Behavioral
Ecology and Sociobiology 27: 403-140.

Brittingham, M. C., and S. A. Temple. 1983. Have cowbirds caused forest songbirds to
decline? Bioscience 33(1): 31-35.

Brittingham, M. C., and S. A. Temple. 1996. Vegetation around parasitized and non-
parasitized nests within deciduous forest. Journal of Field Ornithology 67(3):
406-413.

Burhans, D. E., and F. R. Thompson, HI. 1999. Habitat patch size and nesting success of
yellow-breasted chats. Wilson Bulletin 111(2): 210-215.

Coker, D. R., and D. E. Capen. 2000. Distribution and habitat associations of brown-
headed cowbirds in the Green Mountains of Vermont. Pages 236-243 in J. N. M.
Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors.
Ecology and management of cowbirds and their hosts. University of Texas Press,
Austin, Texas, USA.

Curson, D. R., C. B. Goguen, and N. E. Mathews. 2000. Long-distance commuting by
brown-headed cowbirds in New Mexico. Auk 117(3): 795-799.

Davis, S. K. 2004. Area sensitivity in grassland passerines: effects of patch size, patch
shape and vegetation structure on bird abundance and occurrence in southern
Saskatchewan. Auk 121(4): 1130-1145.

Davis, S. K., and S. G. Sealy. 2000. Cowbird parasitism and nest predation in
fragmented grasslands of southwestern Manitoba. Pages 220-228 in J. N. M.
Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors.
Ecology and management of cowbirds and their hosts. University of Texas Press,
Austin, Texas, USA.

Donovan, T. M., F. R. Thompson III, and J. R. Faaborg. 2000. Cowbird distribution at
different scales of fragmentation: trade-offs between breeding and feeding
opportunities. Pages 255-264 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K.
Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

120

Gates, J. E., and D. R. Evans. 1998. Cowbirds breeding in the central Appalachians:

spatial and temporal patterns and habitat selection. Ecological Applications 8(1):
27-40.

Goguen, C. B. and N. E. Mathews. 1999. Review of the causes and implications of the
association between cowbirds and livestock. Studies in Avian Biology 18: 10-17.

Goguen, C. B., and N. E. Mathews. 2000. Local gradients of cowbird abundance and
parasitism relative to livestock grazing in a western landscape. Conservation
biology 14(6): 1862-1869.

Gustafson, E. J., M. G. Knutson, G. J. Niemi, and M. Friberg. 2002. Evaluation of
spatial models to predict vulnerability of forest birds to brood parasitism by
cowbirds. Ecological Applications 12(2): 412-426.

Hahn, D. C., and J. S. Hatfield. 1995. Parasitism at the landscape scale: cowbirds prefer
forests. Conservation Biology 9(6): 1415-1424.

Hahn, D. C., and J. S. Hatfield. 2000. Host selection in the forest interior: cowbirds
target ground-nesting species. Pages 120-127 in J. N. M. Smith, T. L. Cook, S. I.
Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and management of
cowbirds and their hosts. University of Texas Press, Austin, Texas, USA.

Hejl, S. J ., and J. S. Young. 1999. Brown-headed cowbirds in ponderosa pine/Douglas
fir-dominated landscapes in the northern Rocky Mountains. Studies in Avian
Biology 18: 73-79.

Howe, W. H., and F. L. Knopf. 2000. The role of vegetation in cowbird parasitism of
yellow warblers. Pages 200-203 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S.
K. Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

Howell, C. A., W. D. Dijak, and F. R. Thompson, 111. 2007. Landscape context and
selection for forest edge by breeding brown-headed cowbirds. Landscape
Ecology 22: 273-284.

Jackson, N. H., and D. D. Roby. 1992. Fecundity and egg-laying patterns of captive
yearling brown-headed cowbirds. Condor 94: 585-589.

Kostecke, R. M., J. A. Kolszar, and D. C. Dearbom. 2003. Effect of a reduction in cattle

stocking rate on brown-headed cowbird activity. Wildlife Society Bulletin 31(4):
1083-1091.

121

Lowther, P. E. 1993. Brown-headed cowbird (Molothrus ater). Pages 1-24 in A. Poole
and F. Gill, editors. The birds of North America, No. 47. The Academy of
Natural Sciences, Philadelphia, Pennsylvania, USA; The American
Omithologist’s Union, Washington, DC, USA.

McKay, C. R. 1994. Brown-headed cowbird in Strathclyde: New to Britain and Ireland.
British Birds 87: 284-288.

Ortega, C. P. 1998. Cowbirds and other brood parasites. University of Arizona Press,
Tucson, Arizona, USA.

Raymer, D. F. N. 2000. Effects of elk and white-tailed deer browsing on aspen
communities and wildlife habitat quality in northern Lower Michigan: an l8-year
evaluation. Dissertation, Michigan State University, East Lansing, Michigan,
USA.

Robinson, S. K., F. R. Thompson 111, T. M. Donovan, D. R. Whitehead, and J. Faaborg.
1995. Regional forest fragmentation and the nesting success of migratory birds.
Science 267: 1987-1990.

Robinson, S. K., J. D. Brawn, S. F. Morse, and J. R. Herkert. 1999. Use of different
habitats by breeding brown-headed cowbirds in fragmented midwestem
landscapes. Studies in Avian Biology 18: 52-61.

Rothstein, S. 1., J. Vemer, and E. Stevens. 1980. Range expansion and diurnal changes
in dispersion of the brown-headed cowbird in the Sierra Nevada. Auk 97: 253-
267.

Rothstein, S. I., J. Vemer, and E. Stevens. 1984. Radio-tracking confirms a unique
diurnal pattern of spatial occurrence in the parasitic brown-headed cowbird.
Ecology 65(1): 77-88.

Schroeder, R. L., and P. J. Sousa. 1982. Habitat suitability index models: eastern
meadowlark. U. S. Fish and Wildlife Service publication, Fort Collins, Colorado,
USA.

Sousa, P. J. 1983. Habitat suitability index models: field sparrow. U. S. Fish and
Wildlife Service publication, Fort Collins, Colorado, USA.

Staab, C. A., and M. L. Morrison. 1999. Managing riparian vegetation to control
cowbirds. Studies in Avian Biology 18: 18-22.

Stokes, D., and L. Stokes. 1996. Stokes field guide to birds, eastern region. Little,
Brown and Company, New York, New York, USA.

122

Strausberger, B. M., and M. V. Ashley. 1997. Community-wide patterns of parasitism of
a host “generalist” brood-parasitic cowbird. Oecologia 112: 254-262.

Stribley, J. M. 1993. Factors influencing cowbird distributions in forested landscapes of
northern Michigan. Masters thesis, Michigan State University, East Lansing,
Michigan, USA.

Stribley, J. M., and J. B. Haufler. 1999. Landscape effects on cowbird occurrences in
Michigan: implications to research needs in forests of the inland west. Studies in
Avian Biology 18:68-72.

Teather, K. L., and R. J. Robertson. 1985. Female spacing patterns in brown-headed
cowbirds. Canadian Journal of Zoology 63(2): 218-222.

Tewksbury, J. J ., T. E. Martin, S. J. Hejl, T. S. Redman, and F. J. Wheeler. 1999.
Cowbirds in a western valley: effects of landscape structure, vegetation, and host
density. Studies in Avian Biology 18: 23-33.

Thompson III, F. R. 1994. Temporal and spatial patterns of breeding brown-headed
cowbirds in the Midwestern United States. Auk 111(4): 979-990.

Thompson III, F. R., and W. D. Dijak. 2000. Differences in movements, home range
and habitat preferences of female brown-headed cowbirds in three Midwestern
landscapes. Pages 100-109 in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K.
Robinson, and S. G. Sealy, editors. Ecology and management of cowbirds and
their hosts. University of Texas Press, Austin, Texas, USA.

Thompson 111, F. R., S. K. Robinson, T. M. Donovan, J. R. Faaborg, D. R. Whitehead,
and D. R. Larsen. 2000. Biogeographic, landscape and local factors affecting
cowbird abundance and host parasitism levels. Pages 271-279 in J. N. M. Smith,
T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy, editors. Ecology and
management of cowbirds and their hosts. University of Texas Press, Austin,
Texas, USA.

Uyehara, J. C. and M. J. Whitfield. 2000. Association of cowbird parasitism and
vegetative cover in territories of southwestern willow flycatchers. Pages 204-209
in J. N. M. Smith, T. L. Cook, S. I. Rothstein, S. K. Robinson, and S. G. Sealy,
editors. Ecology and management of cowbirds and their hosts. University of
Texas Press, Austin, Texas, USA.

Vander Haegen, W. M., and B. Walker. 1999. Parasitism by brown-headed cowbirds in
the shrubsteppe of eastern Washington. Studies in Avian Biology 18: 34-40.

Vemer, J ., and L. V. Ritter. 1983. Current status of the brown-headed cowbird in the
Sierra National Forest. Auk 100(2): 355-368.

123

 

Woolfenden, B. E., H. L. Gibbs, S. G. Sealy, and D. G. McMaster. 2003. Host use and
fecundity of individual female brown-headed cowbirds. Animal Behaviour
66:95-106.

Wright, A. L. 1999. Distribution and abundance of brown-headed cowbirds in the
wilderness of central Idaho. Studies in Avian Biology 18: 94-96.

Young, J. S., and R. L. Hutto. 1999. Habitat and landscape factors affecting cowbird
distribution in the northern Rockies. Studies in Avian Biology 18: 41-51.

124

SUMMARY

125

5.1. Introduction

The issues surrounding the impact of brown-headed cowbird parasitism on avian
communities have been of interest to researchers and managers for several decades.
Specific to Michigan, cowbird parasitism on the Kirtland’s warbler population has been
of concern, due to the latter’s endangered species status. A Kirtland’s warbler recovery
plan was enacted in 1967, which has resulted in jack pine management as well as an
intensive cowbird trapping program throughout northern Lower Michigan. This project
is possibly the first to have examined the effects of Kirtland’s warbler habitat
management on the brown-headed cowbird, as well as the first to have investigated which
landscape factors might be contributing to the capture success of cowbird traps. In
addition, I constructed a habitat suitability index (HSI) model for the cowbird, which to

my knowledge, had not previously been done.

5.2. Discussion

Brown-headed cowbirds did not respond to the different age classes of managed
jack pine stands. This is not unusual, since cowbirds are known to use different
vegetation types and structures for their habitat requirements. Each jack pine age
category represented a stage of succession that cowbirds typically respond to—open
fields (young), shrub scrub (mid-aged) and forest (old). Open fields might serve as part of

their reproductive habitat, but would primarily provide food resources in the form of

126

 

grass and weed seeds and insects. Shrub scrub and forest would provide access to host
nests for cowbirds to reproduce.

Cowbirds were more abundant in large stands than in small. This was
unexpected, since as an edge species, cowbirds tend to prefer smaller, fragmented
patches. However, host species abundance was also higher in large patches; therefore,
the potential for more nests to parasitize could have drawn cowbirds into these areas.
This study only surveyed stand sizes; there was no landscape-level analysis of cowbird
response to fragmentation levels or edge amounts. A large-scale analysis investigating
these correlations might provide a better account of cowbird relationships to Kirtland’s
warbler habitat management. It should also be noted, however, that cowbirds were not
overly abundant in any given stand size-age combination. Previous studies have
demonstrated that cowbirds often avoid coniferous forests in favor of deciduous stands.
This might have been at issue in this study, which took place entirely in regenerating jack
pine.

My analysis of cowbird traps used by the USFWS showed that cowbird captures
were positively associated with surrounding urban, agriculture and deciduous forest.
These findings are consistent with previous studies, which have concluded that cowbirds
access urban areas for food resources such as bird feeders, agricultural lands for their
supplies of seeds and insects; and deciduous forests for host nests to meet their
reproductive requirements. Cowbird captures were negatively correlated with coniferous
forests and wetlands; other researchers have suggested that vegetation structure in

coniferous areas might impede high host densities and the nest-searching abilities of

127

 

cowbirds. I would suggest that the dense vegetation characteristics of many wetland

areas might also explain why cowbird captures were negatively correlated with them.

5.3. Management Implications

Results from the field component of this study suggest that management of
Kirtland’s warbler habitat is introducing cowbirds into large stands, but that the mid-aged
jack pine stands that Kirtland’s warblers depend on is not attracting cowbirds. Since
Kirtland’s warblers are known to nest only in large patches of jack pine forest, it is my
belief that managers will continue to manage these sized stands, regardless of the
increased number of cowbirds. This strategy is likely advantageous, since despite
increased cowbird abundance, there is also evidence of higher species richness, species
diversity and host abundance in these stands.

This study has, in part, addressed the uncertainty over how the avian community
responds to Kirtland’s warbler habitat management. Managers from the Kirtland’s
Warbler Recovery Team have been concerned by the “single species management”
practices surrounding the Kirtland’s warbler recovery plan, and have questioned how
other bird species are affected. This study has found that the avian community, in
general, has responded positively to the size and age categories maintained for Kirtland’s
warbler habitat. Species richness, species diversity and host abundance were higher in
large stands, while host density was high in mid-aged stands. Large, mid-aged stands are
suitable Kirtland’s warbler habitat, and appear to be beneficial for the avian community

. in general. However, a more in-depth study should be done to examine how the jack pine

128

ecosystem—addressing soil quality, tree growth and wildlife response— is responding to
a management strategy utilizing clearcuts in place of natural wildfires.

Findings from the trap and land cover analysis have implications for the trapping
program, since the coniferous (jack pine) forests where traps are placed are negatively
correlated with trap captures. I recommend that traps be placed in urban areas,
agricultural fields or in deciduous forests that are adjacent to Kirtland’s warbler habitat to
improve capture efficiency. However, if traps must be placed in jack pine stands because
of the requirements of the Kirtland’s warbler recovery plan, I suggest that they be in close

proximity to any of the above land cover types.

5.4. Limitations

There were limitations to the field-based portion of this project. Jack pine is an
early successional species whose growth is naturally limited by wildfires or by
replacement by late-successional tree species like oak and hickory. Therefore, comparing
regenerating jack pine to the age classes of other forest types (late-successional species)
documented in the literature was somewhat limited. Other studies have demonstrated
increased avian richness in old, mature forests; however, these tend to be upland
hardwood forests that can exceed 150 years in age. Old jack pine stands do not grow
much older than 50-75 years, due to wildfire, succession to other tree species, or forest
management. In this study, old jack pine was not as old as the old forests discussed in the
literature. This might account for why host density decreased in old jack pine and why

no other characteristic of the avian community responded to stand age.

129

 

Another limitation of this project was the avian survey method used to detect bird
species in different stand types. I used a standard point-count census method, observing
in early - late mornings. However, brown-headed cowbirds are known to use different
vegetation types to meet their habitat requirements at different times of the day. The
results of this study might have been different had avian surveys been conducted into the
afternoon and evening hours. I might have detected increased use of young stands in the
afternoon for feeding and either young or mid-aged stands in the evening for roosting. I
would recommend that such a study be done to better determine how cowbirds are
responding to the stand types throughout the course of their day.

There were two issues related to the cowbird trap analysis. The first was that the
4 km buffers that surrounded the traps overlapped considerably amongst them. This
raised the question of whether traps were independent, since traps in proximity to each
other could essentially “take” cowbirds from each other, which could confound—by
decreasing the influence of—associations with land cover. I chose to assume that traps
were independent, based on the non-random, strategic method in which traps were
placed, and on the assumption that cowbirds were attracted to traps due to their location
within certain land cover types; however, there is no assurance that this assumption is
true. I would suggest a spatial analysis that determines trap independence, and that is
able to then contend with any non-independence issues. The second constraint of this
study was the inability to compare capture rate associations with land cover over time.
This was due to the fact that the 1978, 1993 and 2001 land cover for Michigan were
created using different methods under different criteria, making them impossible to

compare. Therefore, a separate cowbird capture-land cover correlation analysis had to be

130

done for each period, but no comparisons could be made amongst the three periods. In
addition, traps in different periods were situated in different locations. This might have
had an effect on the composition of surrounding land cover (which is difficult to
determine since the land cover classes amongst the three periods are not comparable),
which in turn, might have influenced the number of trap captures between the three
periods. Since trap captures declined over the three periods, an ideal study might be able
to show an association between these trap declines and changes in land cover
composition surrounding the traps. However, since the 1978, 1993 and 2001 Michigan
land cover data are the only ones available for these time periods, I do not see an
alternative to this situation.

While the habitat suitability index model I constructed for the brown-headed
cowbird is legitimate, it has not yet been validated. This is the next step in determining
whether this model is accurate in determining suitable habitat for the brown-headed
cowbird. Limitations of the model included the fact that few of the results gained from
my field study were helpful in constructing the model, presumably because a) my
fieldwork was conducted in coniferous forest, which has been determined to deter
cowbirds, and b) my study sites were all within 4 km of agricultural areas. Therefore, I
would suggest that this model be revised using field data from deciduous forests, and that
consideration be given to incorporating a ratio of percent deciduous forest to percent

coniferous forest as a suitability index.

5.5. Future Research

131

 

This project has also raised questions that might be worthy of being explored in
future studies. USFWS managers have documented the escape of the live cowbird
decoys that they use to attract individuals to traps. In recent years, managers have
noticed that increasingly more decoys have learned how to escape from the traps. This
suggests a need to build improved trap structures to adequately contain both decoys and
enticed cowbirds. In addition, this escape phenomenon could be studied in more detail,
to determine if cowbirds a) learn to escape from traps via trial-and-error, b) learn how to
escape from other individuals, or c) whether individuals who escape from traps are
genetically different from those that do not, and pass on this “intelligence” phenotype to
their offspring. This latter scenario would suggest that successive generations of
cowbirds might become “more intelligent” and avoid being captured. An intensive mark-
recapture, monitoring, and genetics study would need to be conducted to accurately test
these questions.

Another issue for future inquiry is to determine to what degree different
subspecies of cowbirds are hybridizing, especially with populations from some
subspecies decreasing due to control programs. Molothrus ater obscuras, found mainly
in the southwest US, has been penetrating the rest of the country and interbreeding with
Molothrus ater ater and Molothrus ater artemis, which are found in the eastern half of
the US. and the Pacific northwest, respectively. With M. a. ater being subjected to mass
removal by the control program in Michigan, determining whether M. a. obscura is
occurring in higher proportions in the eastern US. would be important from a genetic

standpoint. This could also raise a further question of whether M. a. obscure, which is

132

subjected to the trapping programs at Ft. Hood, Texas, and throughout California, is
moving eastward as a response.

Finally, several studies have indicated that climate change could cause the jack
pine ecosystem in northern Michigan to recede to more northern latitudes, since the
species requires cooler climates with low humidity. If this were to happen, Kirtland’s
warblers might not be able to access these areas for breeding, which could create a
decline in their population. This scenario puts into question the costly, intensive
management efforts to sustain the Kirtland’s warbler population. A predictive model
would be the most accurate assessment of the future of jack pine ecosystems, and
therefore, Kirtland’s warbler populations. However, despite evidence that might suggest
future habitat loss for Kirtland’s warblers, realistically, the USFWS should and will

continue to put forth the best effort to maintain the population of this rare species.

133

   

illlilllllllllllll