Accepted by the Graduate Faculty, Indiana University, in partial

fulfillment of the requirements for the degree of Doctor of Philosophy.

Doctoral Committee

_____________________________________

Donald Whitehead, Ph.D.

_____________________________________
Ellen Ketterson, Ph.D.

_____________________________________
J.C. Randolph, Ph.D.

_____________________________________

David Parkhurst, Ph.D.

21 January 2003

Donald Edward Winslow

For Asheesh and Diane and the little babies in 94IN4OVEN2.

ACKNOWLEDGMENTS

I would like to thank the present and former members of my Graduate Committee; Don Whitehead, Ellen Ketterson, J.C. Randolph, Les Real, and Dave Parkhurst; for their guidance and patience. Professor Val Nolan, although not on my committee, helped shape my thinking while I wrote the manuscript that became Chapter Two.

This research was funded by the National Fish and Wildlife Foundation, the U.S. Fish and Wildlife Service, the U.S. Forest Service, and many Audubon chapters and other organizations. The personnel at Yellowwood State Forest and Hoosier National Forest were helpful in coordinating site preparation, logging operations, post-harvest timber stand improvement, and openings maintenance with our research.

I would like to thank my colleagues in Don Whitehead's lab; especially Tom Ford, Peter Hurley, Pat Doran, Rebecca Flaten, Matt Koukol, Steve Koch, Bernadette Slusher, Beth Geils, Grant Greenberg, and Carolyn Frazer Whyte.

Great appreciation is due to the many field assistants who worked early and long hours to find and monitor nests and the birds who endured our intrusions into their lives.

I also thank my family and friends for their support and patience. In particular, my mother, Miriam Vessels, and my father, Charles Winslow, gave me a computer which greatly facilitated the writing of this dissertation. My dear friend Feather-Byahe Maka Washte Speaking Truth Sebree woke up with me at insane hours, fixed meals, gave me a home, and suffered my idiosyncratic compulsions while I was completing this dissertation.

Donald Winslow

THE EFFECTS OF FOREST MANAGEMENT PRACTICES ON

REPRODUCTIVE SUCCESS OF NEOTROPICAL MIGRANT BIRDS

IN SOUTH-CENTRAL INDIANA

ABSTRACT

Broad-scale fragmentation of Midwestern forests has increased avian nest predation and parasitism of nests by Brown-headed Cowbirds (Molothrus ater). Local disturbances may also attract cowbirds and elevate parasitism. Chapter One presents cowbird abundance and brood parasitism data from various forest contexts in south-central Indiana. Cowbird abundances were similar at forest edges and interior forest. Parasitism levels were high near external and internal edges. Host species varied in parasitism frequency and in patterns among contexts. Increasing forest core habitat and eliminating interior edges should benefit forest-breeding birds.

Ornithologists have studied edge effects in many systems but have devoted little attention to underlying mechanisms. It is thus unclear why edge effects occur in some but not all situations. Edge-elevated nest densities may attract predators and cowbirds, but density-dependent responses may not always increase proportions of nests depredated or parasitized. Chapter Two models interrelationships among edge proximity, cowbird density, host density, and parasitism level. These models suggest elevated nest densities are unlikely to increase parasitism unless cowbirds respond both functionally and numerically. I describe an empirical approach to study relationships among these variables.

Logging generates edges and may reduce nest success, but few studies have measured success before timber extraction. Chapter Three describes an experiment evaluating the effects of logging on survival and parasitism of Acadian Flycatcher (Empidonax virescens) nests in Yellowwood State Forest, Indiana. Four tracts were selectively logged between two monitored breeding seasons. Four tracts were controls. Nest success at control sites increased from 1995 to 1996, but was similar at treatment sites before and after logging. Cowbird parasitism at control sites decreased from 1995 to 1996, but was similar at treatment sites before and after logging. If factors responsible for between-year variation operate similarly at all sites, the interactions between year and condition may indicate logging decreases success. Forest managers should limit logging where productivity of late-successional birds is a priority until we better understand factors influencing variation in success. Longer-term studies before and after logging are necessary to understand how logging affects success. It is especially important to investigate how management practices affect species restricted to regenerating forest.

TABLE OF CONTENTS

INTRODUCTION 1

CHAPTER ONE: Within-landscape Variation in Patterns of Cowbird Parasitism in the Forests of South-central Indiana 4

ABSTRACT 4

INTRODUCTION 6

STUDY SITES AND METHODS 13

A. Census Methods 13

B. Landscape Patterns of Parasitism 15

Interior Forest 16

Exterior Edge 17

Forests near Clearcuts 17

Forests near Forest Openings 18

Timbercuts in State Forests 19

Old Field 20

C. Methods for Determining Nesting Success 20

RESULTS 23

A. Spatial Distribution of Cowbirds 23

B. Landscape Patterns of Brood Parasitism 25

C. Parasitism in Relation to Edge Proximity 31

DISCUSSION 31

CHAPTER TWO: Mechanistic Approaches to the Investigation of Edge Effects on Avian Productivity 40

ABSTRACT 40

POPULATION DECLINES IN NEOTROPICAL MIGRANT BIRDS 41

DIRECT EFFECTS OF HABITAT LOSS 42

EDGE EFFECTS 43

MECHANISMS OF EDGE EFFECTS 47

A. Null Models Based on Nest Density 48

B. Functional Responses of Encounter Rate 52

C. Avian Community Composition 60

D. Functional Responses of Spatial Activity Patterns 69

E. Structure-directed Responses 80

F. Use of Adjacent Habitat by Predators and Cowbirds 85

G. Numerical Responses 87

H. Alterations in Vegetative Structure near Edges 89

I. Edge Effects on Nest Predation 91

AN APPROACH FOR TESTING THE MODELS 94

A. Interrelations among the Models and Hypothesis-testing 94

Predictions of the Models 95

Estimating Reproductive Parameters from Nest Data 99

Estimating Parameters using Regression 101

B. Testing Alternative Mechanisms of Edge Effects 104

Functional Response to Nest Density 105

Variation in Avian Community Composition 109

Pseudo-numerical Response 111

Structure-directed Response 114

Response to Adjacent Habitat 116

Numerical Response 117

Alterations in Vegetative Structure 117

C. Path Analysis 119

CONCLUSIONS: IMPLICATIONS FOR RESEARCH AND CONSERVATION 122

CHAPTER THREE: An Experimental Investigation of the Effects of Logging on Success of Acadian Flycatcher (Empidonax virescens) in Yellowwood State Forest, Indiana 124

ABSTRACT 124

INTRODUCTION 126

A. Variation in Nesting Success Over the Course of a Season 127

B. History of Nest Survival Analysis 128

C. Objectives 131

METHODS 133

A. Field Methods 133

B. Analysis of Variation in Predation Within a Season 135

C. Analysis of Effects of Logging on Nest Survival 136

D. Analysis of Effects of Logging on Brood Parasitism 137

RESULTS 137

A. Variation in Predation over the Course of a Season 137

B. Effects of Timber Extraction on Nest Survival 142

C. Effects of Timber Extraction on Cowbird Parasitism 142

DISCUSSION 145

CONCLUSIONS AND RECOMMENDATIONS FOR MANAGEMENT 148

APPENDIX: Glossary of Terms 150

LITERATURE CITED 159

TABLES

Table 1.1 Percent brood parasitism for all species for each year summed over all landscape contexts. 21

Table 1.2 Levels of parasitism for selected species, summarized by landscape context. 26

Table 1.3 Loglinear modeling of proportion of parasitized nests with respect to landscape context and host species. 28

Table 2.1 Statements and predictions for models of parasitism limitation. 96

Table 2.2 Predictions of alternative mechanisms for edge effects of parasitism. 106

Table 3.1 Eight tracts in Yellowwood State Forest that were monitored in 1995 and 1996. Timber was extracted from Treatment sites between the two breeding seasons; Control sites remained unlogged. 133

Table 3.2 Reproductive statistics for nests of 19 species monitored in Yellowwood State Forest (all sites combined) during 1995. 138

Table 3.3 Daily predation rate of Acadian Flycatcher nests in 1995 in eight tracts in Yellowwood State Forest in the egg and nestling stages. 141

Table 3.4 Survival rate of Acadian Flycatcher nests in Yellowwood State Forest. 143

Table 3.5 Overall success rate of Acadian Flycatcher nests in eight tracts in Yellowwood State Forest. 144

Table 3.6 Resampling analysis of overall success rate of Acadian Flycatcher nests in Yellowwood State Forest. 145

Table 3.7 Cowbird parasitism data for Acadian Flycatcher nests in eight tracts in Yellowwood. 146

FIGURES

Figure 1.1 Forest cover of Indiana, adapted from Landsat imagery taken during the period 1973-1981. 10

Figure 1.2 The Pleasant Run Unit of the Hoosier National Forest, showing the location of internal disturbances. 11

Figure 1.3 Map of study sites in and around the Pleasant Run Unit of the Hoosier National Forest. 12

Figure 1.4 Numbers of cowbirds detected during morning censuses in 1990, 1991, and 1992, in interior forest, 150 m from edge, and 50 m from edge. 24

Figure 1.5 Proportion of nests parasitized in 1992 and 1993 among nests of four species of forest-breeding Neotropical migrants in interior forest (IN), exterior edge (EX), and forests near clearcuts (CC; includes 6 nests within clearcuts.) 27

Figure 1.6 Parasitism as a function of distance from clearcut edge for Acadian Flycatcher nests in forests near clearcuts in 1992,all sites pooled. 32

Figure 2.1 (a.) Null model of brood parasitism (P = aC). (b.) Level of parasitism as a function of nest availability under the null model. (c.) The encounter-rate limitation model of brood parasitism (P = gNC). (d.) Level of parasitism as a function of nest availability under the encounter-rate limitation model. 50

Figure 2.2 (a.) Finding-rate limitation model of brood parasitism [P = bC + g(N - b/r)C]. (b.) Level of parasitism as a function of nest availability under the finding-rate limitation model. 56

Figure 2.3 (a.) Ideal free distribution model of pseudo-numerical or numerical response of cowbird density (C) to nest availability (N), incorporating egg limitation (P = aC) and encounter rate limitation (P = gNC) parasitism functions. (b.) Level of parasitism as a function of nest availability when cowbirds follow an ideal free distribution with regard to nest availability. 76

Figure 2.4 Path diagram showing relationships among variables that may affect level of parasitism. 119

Figure 3.1 Map of nest monitoring plots in Yellowwood State Forest. 134

Figure 3.2 Number of nests known to be active and monitorable each day in 1995 in eight tracts in Yellowwood State Forest, pooling all species. Also shown are the percentage of nests taken by predators on each day. 139

Figure 3.3 Number of nests known to be active and monitorable each day in 1995 in eight tracts in Yellowwood, for all species pooled (squares) and for Acadian Flycatchers (diamonds). Also shown are the proportion of Acadian Flycatcher nests depredated each day. 140

Figure 3.4 (a.) Daily survival rate of Acadian Flycatcher nests during the egg stage at logged (treatment) and uncut sites in Yellowwood State Forest. (b.) Daily survival rate of Acadian Flycatcher nests during the nestling stage at logged (treatment) and uncut sites in Yellowwood State Forest. 141

Figure 3.5 Overall success rate of Acadian Flycatcher nests in eight tracts in Yellowwood State Forest. 144

Figure 3.6 Resampling analysis of overall success rate of Acadian Flycatcher nests in Yellowwood State Forest. 147