Tree Species Selection and Use by Foraging

Insectivorous Passerines in a Forest Landscape by

Peter S. McKinley

M.S.E.S., Indiana University, 1991

B.A. Colby College, 1987

A Thesis Submitted in Partial Fulfilment of

the Requirements for the Degree of

Doctor of Philosophy in the graduate academic unit of Biology

Supervisor: Antony W. Diamond, Ph.D., Department of Biology and

Faculty of Forestry and Environmental Management,

University of New Brunswick.

Examining Board: John Spray, PhD, Department of Geology

Paul Arp, Faculty of Forestry and Environmental

Management

Steven Heard, Department of Biology

External Examiner: Richard Holmes, Ph.D., Department of Biology, Dartmouth

College, New Hampshire, USA.

This thesis is accepted.

______________________

Dean of Graduate Studies

THE UNIVERSITY OF NEW BRUNSWICK

July, 2003

Peter S. McKinley,

ii

Dedication

Dedicated to my parents, Robert S. and Carol E. McKinley, an ever present source of

passion, curiosity, encouragement, and love, and in loving memory of my brother,

Andrew S. McKinley, Andy, 1966-1996. Your kindness, courage, and love are a bright

light and constant source of motivation, strength, and will. You are remembered and

loved always.

iii

Abstract

Tree species were selectively used or avoided by foraging songbirds in a northern

temperate mixed-wood forest. American beech (Fagus grandifolia) was selected out of

proportion to its abundance by the black-throated green warbler (Dendroica virens)

during several years of the four year study. Black-throated blue warblers (Dendroica

caerulescens), American redstart (Setophaga ruticilla), and least flycatcher (Empidonax

minimus) also showed selective foraging use of the American beech. Yellow birch,

(Betula alleghaniensis) was usually avoided by these same birds. Selective foraging

occurred in sugar maple in a stand with a history of intensive forest management that had

resulted in the virtual absence of American beech.

Rates of foraging differed significantly among tree species in some years for the

black-throated blue and black-throated green warblers, but the results were less consistent

than the selective use results. The black-throated green warbler showed significantly

higher rates of foraging in the American beech versus yellow birch in two years. The

black-throated green warbler also showed a higher ratio of attacks per flight in the

American beech. This measure is interpreted as the number of attacks per foraging patch.

I defined the patch as the length of branch searched by a bird hopping in between short

flights within the same tree.

I collected data on the birds’ stomach contents to confirm that the avian diet was

represented in the arthropod samples that I collected from the various tree species. The

arthropod hypothesis predicts that differences in arthropod density or diversity among

tree species drives selective foraging behaviour. Arthropod density and diversity did not

iv

differ among tree species in a pattern that would explain the selective foraging behaviour

of the focal bird species. The tree species that tended to have higher abundances of

arthropods (yellow birch) was usually avoided by the foraging birds. The tree

morphology hypothesis focuses upon an optimal relationship between a tree species’

structure and a bird’s foraging behaviour. This process is probably responsible for the

differential patterns of use among tree species in this study. This research offers the first

evidence for the sole function of one of the explanatory hypotheses in a deciduous forest

system.

Based on the role played by American beech as a foraging substrate, its

availability was regressed on avian presence/absence data. American beech basal area

predicted landscape patterns of distribution of one focal species, black-throated green

warbler, and two non-focal species, northern parula and magnolia warbler.

Finally, patterns of foraging behaviour were compared with landscape patterns of

productivity for the black-throated green warbler. American beech basal area was a

significant predictor of avian productivity in a companion study. This suggests that

patterns of foraging behaviour by individual birds can be “scaled up” to predict spatial

patterns in distribution of productivity across the landscape.

Whereas previous work has established the importance of tree species diversity to

avian foraging ecology, this research documents the importance of morphological

diversity within deciduous tree species. Previous ambiguity about causal processes

responsible for selective foraging patterns in tree species has been eliminated. This work

also demonstrates the dynamic nature of avian foraging ecology across a region that

v

includes the same bird and tree species. Caution is warranted in extrapolation of results,

even within similar forest systems.

vi

Acknowledgements

First and foremost, I wish to express deep gratitude to my advisor, Tony Diamond. His

patience, humanity, and science are exceptional and have helped to sustain me on this

long trip. I also wish to thank my committee members, Graham Forbes, Marc-Andre

Villard, and Gerry Parker for their insight and guidance throughout this project. Thanks

also to the University of New Brunswick Biology Department and especially Tim

Dilworth, Tillmann Benfey, Dion Durnford, and Linda Allen for their support and

assistance. I also wish to thank my field assistants, Daniel Mazerolle, Trista Michaud,

Jaime Glenn, and Karen Gosse for their hard work and companionship through many

hours in the field and field camp. Thanks also to Rhonda Hudson and again to Jaime

Glenn for their help on this project in association with their honours theses. I would also

like to express my appreciation to Fraser Papers Inc. and specifically to Gilles Couturier

and Steve Young. Thanks also to John Gunn, Jeff Bowman, and Mark Edwards for their

many contributions including friendship and moral support. This study was funded by

grants to A.W. Diamond from the Sustainable Forest Management Network of Centres of

Excellence and Natural Sciences and Engineering Research Council of Canada

(NESERC), and the Atlantic Cooperative Wildlife Ecology Research Network

(ACWERN)

1

Table of Contents

Abstract…………………………………………………………………………………..iii

List of Tables………………………………………………………………………….….3

List of Figures………………………………………………………………………….…6

List of Plates…………………………………………………………………………. .…9

List of

Abbreviations………………………………………………………………….………....10

Chapter 1: Introduction………………...……………………………………………….. 11

Research Questions…………………………………………………………………..11

Macro-scale (Habitat) Vegetation Structure…………………………………………13

Floristic Diversity and Micro-Scale Vegetation Structure…………………………..14

Landscape Pattern and Process………………………………………………………21

Chapter 2: Methods…………………………………….………………………………...27

Study Area and General Sampling Design…………………………………………..27

Grid Vegetation Sampling…………………………………………………………...30

Foraging Behaviour………………………………………………………………….31

Diet Composition………………………………………………………………….…36

Arthropod and Tree Species Associations…………………………………………..38

Landscape Distribution and Productivity…………………………………………...40

Chapter 3: Foraging Behaviour Results………………………………………………...47

Foraging Behaviour Overview……………………………………………………..47

Foraging Site Selection ……………………………………………………………48

Forage Rate………………………………………………………………………..51

Forage Manoeuvre ……………………………………………………………….53

2

Diet Composition…………………………………………………………………54

Chapter 4: Arthropod Prey Density and Diversity Results…………………………...73

Chapter 5: Landscape Distribution and Productivity Results…………………………91

Chapter 6: Discussion, Consclusions, and Management Implications……………….97

Foraging Site Selection and Success………………………………………………97

Evaluation of the Arthropod Hypothesis…………………………………………102

Evaluation of the Morphology Hypothesis………………………………………104

Landscape Pattern and Process…………………………………………………..111

Regional Comparison……………………………………………………………112

Forest Management……………………………………………………………..116

Future Research…………………………………………………………………118

Appendix I………………………………………………………………………….125

References Cited……………………………………………………………………128

Vita…………………………………………………………………………………141

3

List of Tables

Table 2.1. Relative Proportions of the four focal tree species on each grid………….…43

Table 3.1. Number of observation sequences by grid, bird species, and year

with mean duration of sequence and standard error (SE)…………………………..56

Table 3.2. Number of attacks each season by bird and tree species, and mean

number of attacks per observation sequence with standard error SE by grid

and year……………………………………………………………………………..57

Table 3.3. Goodness of fit analysis for foraging use versus availability of

the four focal tree species combined……………………………………………….58

Table 3.4. Multiple comparison of forage attacks per minute by btgw

(Nemenyi test p<0.05) for differences among tree species, 1996

reference grid……………………………………………………………………..59

Table 3.5. Multiple comparison of forage attacks per minute by btgw

(Mann-Whitney p < 0.05) for differences among tree species,

1999 reference grid……………………………………………………….………59

Table 3.6. Multiple comparison of forage attacks per minute by btbw

(Mann-Whitney test p<0.05) for differences among tree species,

1996 reference grid………………………………………………………………60

4

Table 3.7. Multiple comparison of forage attacks per flight by btgw

(Mann-Whitney U-test p<0.05) for differences among

tree species, 1999 reference grid…………………………………………….………60

Table 3.8. Multiple comparison of forage attacks per flight by btgw

(Mann-Whitney U-test p<0.05) for differences among

tree species, 1996 reference grid…………………………………………………..60

Table 3.9 Attack manoeuvres by tree species and bird species,

all years combined, reference grid only……………………………………………61

Table 3.10. Goodness of fit analysis for forage manoeuvres among tree

species………………………………………………………………………………61

5

Table 4.1. Number, mean, and standard error of occurrence (SE) of

arthropod orders for reference grid 1998…………………………. ………………. 75

Table 4.2. Number, mean, and standard error of occurrence (SE) of

arthropod orders for reference grid 1999………………………………. …..……...76

Table 4.3. Number, mean, and standard error of occurrence (SE) of

arthropod orders for managed grid 1999………………………….. .………………77

Table 4.4. Significant (p<0.05) Bonferonni post-hoc pair-wise comparisons of

arthropod variables, data based on number per unit leaf area………………………78

Table 4.5 Significant (p<0.05) Bonferonni post-hoc pair-wise comparisons of

arthropod variables, data based on clip sample……………………………………..79

Table 4.6 Number and area of leaves per clip sample………………………………….80

Table 5.1. Logistic regression of bird presence on tree basal area…………………... 96

6

List of Figures

Figure 2.1. New Brunswick study site location………………………………………...44

Figure 2.2. Schematic diagram of research grid design…………………………………45

Figure 2.3. Schematic diagram of vegetation sampling plots………………………….. 46

Figure 3.1. Least flycatcher and black-throated blue warbler expected and observed

foraging use, im-grid……………………………………………………………...…62

Figure 3.2. Black-throated green warbler expected and observed foraging use,

im-grid……………………………………………………………………………….63

Figure 3.3. American redstart expected and observed foraging use, mm-grid………….64

Figure 3.4. Least flycatcher expected and observed foraging use, mm-grid……………65

Figure 3.5. Black-throated blue warbler expected and observed foraging use,

mm-grid……………………………………………………………………………66

Figure 3.6. Black-throated green warbler expected and observed foraging use,

mm-grid…………………………………………………………………………...67

Figure 3.7. Black-throated green warbler expected and observed use of hover and glean

forage manoeuvre…………………………………………………………………68

Figure 3.8. Black-throated blue warbler expected and observed use of hover forage

manoeuvre…………………………………………………………………………69

Figure 3.9. American redstart expected and observed use of hover and glean forage

manoeuvres…………………………………………………………………….….70

Figure 3.10. Least flycatcher expected and observed use of hover and hawk forage

manoeuvre…………………………………………………………………………71

Figure 3.11. Proportional occurrence of arthropod orders in stomach contents……...72

7

Figure 4.1. Araneida individuals per clip sample by tree species

on the reference grid in 1999………………………………………………………..81

Figure 4.2. Coleoptera individuals per clip sample

by tree species on the reference grid in 1999………………………………...…81

Figure 4.3. Total arthropod density per clip sample

by tree species on the reference grid in 1999…………………………………..82

Figure 4.4. Araneida density per square centimetre of

leaf area by tree species on the reference grid in 1999…………………………82

Figure 4.5. Coleoptera density per square centimetre of

leaf area by tree species on the reference grid in 1999…………………………83.

Figure 4.6. Total arthropod density per square centimetre of

leaf area by tree species on the reference grid in 1999…………………………83

Figure 4.7. Coleoptera density per square clip sample

by tree species on the managed grid in 1999…………………………………..84

Figure 4.8. Total arthropod density per clip sample

by tree species on the managed grid in 1999……………………………….…84

Figure 4.9. Coleoptera density per square centimetre

of leaf area by tree species on the managed grid in 1999……………………...85

Figure 4.10. Hymenoptera density per square centimetre

of leaf area by tree species on the managed grid in 1999…………………..…85

Figure 4.11. Total arthropod density orders per square

centimetre of leaf area by tree species on the managed grid in 1999…………86

Figure 4.12. Coleoptera density per clip sample

by tree species on the reference grid in 1998…………………………………86

8

Figure 4.13. Order richness per clip sample by tree species on

the reference grid in 1998……………………………………………………….87

Figure 4.14. Arthropod density per clip sample by tree species

on the reference grid in 1998…………………………………………………….87

Figure 4.15. Araneida density per square centimetre of leaf area

by tree species on the reference grid in 1998…………………………………....88

Figure 4.16. Coleoptera density per square centimetre of leaf area

by tree species on the reference grid in 1998………………………………..…..88

Figure 4.17. Hymenoptera density per square centimetre of leaf area

by tree species on the reference grid in 1998……………………………….…...89

Figure 4.18. Diptera density per square centimetre of leaf area

by tree species on the reference grid in 1998……………………………………89

Figure 4.19. Arthropod richness per square centimetre of leaf area by

tree species on the reference grid in 1998………………………………………90

Figure 4.20. Arthropod density per square centimetre of leaf area by

tree species on the reference grid in 1998………………………………………90

9

List of Plates

Plate 6.1. American beech branch and twig morphology within approximately 1 m3

foraging patch, total leaf area: 3224 cm2..………………………………………….119

Plate 6.2. Yellow birch branch and twig morphology within approximately 1 m3 foraging

patch, total leaf area: 1550 cm2……………………………………………………..120

Plate 6.3. Sugar maple branch and twig morphology within approximately 1 cubic meter

foraging patch, total leaf area: 2860 cm2…………………………………………..121

Plate 6.4 Yellow birch (left) and American beech branch, twig, and leaf morphology

removed from tree……………………………………………………………….…122

Plate 6.5 Sugar maple (left) and American beech branch, twig, and leaf morphology

removed from tree………………………………………………………………….122

Plate 6.6 American beech branch morphology on tree………………………..……….123

Plate 6.7 Yellow birch branch morphology on tree………………………………..….123

Plate 6.8 Sugar maple branch morphology on tree………………………..…………..124

10

List of Abbreviations

Please refer to the following guide for interpretation of abbreviations used in the tables

and graphics:

Tree Species: ambe – American beech (Fagus grandifolia)

yebi – yellow birch (Betula alleghaniensis)

suma – sugar maple (Acer saccharum)

bafi – balsam fir (Abies balsamifera)

Bird Species: amre – American redstart (Setophaga ruticilla)

lefl – least flycatcher (Empidonax virens)

btbw – black-throated blue warbler (Dendroica caerulescens)

btgw – black-throated green warbler (Dendroica virens)

mnwa – magnolia warbler (Dendroica magnolia)

scta – scarlet tanager (Piranga olivacea)

yrwa – yellow rumped warbler (Dendroica coronata)

revi – red-eyed vireo (Vireo olivaceus)

bhvi – blue-headed vireo (Vireo solitarius)

cswa – chestnut-sided warbler (Dendroica pensylvanica)

nopa – northern parula (Parula americana)

phvi – Philadelphia warbler (Vireo philadelphicus)

Arthropod Orders:

aran – Araneida

cole –Coleoptera

hyme – Hymenoptera

dipt – Diptera

lepi – Lepidoptera

tric – Trichoptera

homo – Homoptera

hemi – Hemiptera

tric – Trichoptera

Research Grids:

mm-grid – moderately managed grid

im-grid – intensively managed grid

11

Chapter 1: Introduction

Research Questions

Historic concentration on the importance of habitat or vegetation macro-structure

has left the realm of avian foraging behaviour in relation to floristic diversity still

relatively poorly understood, especially with respect to causal processes (Whelan 2001

and see reviews below). This dissertation examines how tree species affect arboreal

insectivore foraging behaviour. Previous work (Holmes and Robinson 1981, Robinson

and Holmes 1984) has not been able to distinguish between the two primary causal

processes proposed to explain selective foraging in deciduous tree species.

The ‘arthropod’ hypothesis predicts that abundance or diversity of arthropod prey

on a tree species determines avian selection of that tree species. The ‘morphology’

hypothesis predicts that the foraging strategy of a bird species is adapted to the fine-scale

differences in structure of particular tree species (Morse 1976, Greenberg and Gradwohl

1980, Holmes and Robinson 1981, Morrison et al. 1985). One of my goals is to better

understand the relative contributions of these two processes to tree species-specific

selective foraging patterns among birds. No clear distinction between these two

hypotheses has been established for selective avian foraging patterns in deciduous tree

species.

This dissertation also explores for the first time how a birds’ use of a tree

species as a foraging patch may be scaled up to explain landscape patterns of

avian productivity and distribution. This type of work is an answer to calls made

for a clearer understanding of the relationship between landscape pattern and

Comment [TD1]: Does ‘landscape

distribution’ mean distribution of landscape

elements (e.g. habitat patches)? OR

distribution of birds about the landscape? If

the latter, I suggest deleting the word

‘landscape here; then its clear your

comparing bird productivity & distribution

at the landscape scale.

12

process (Wiens 1995, Lima and Zollner 1996, South et al. 2001, Morales and

Ellner 2002, Wu and Hobbs 2002). Recent work has examined behavioural

processes in the context of landscape questions (Lambert and Hannon 2000,

Mortberg 2001, Belisle and Desrochers 2002, Gobeil and Villard 2002, Belisle

and Desrochers 2002, Robichaud et al. 2002, and see reviews below), but no other

work attempts to scale foraging behaviour up to landscape pattern and process.

My work allows an examination of avian foraging ecology at an even

larger spatial scale when I compare and contrast it to the work of Holmes and

Robinson (1981) in New Hampshire, several hundred kilometres away from my

New Brunswick study sites. Parrish (1995b) offered a similar comparison in his

study of coastal Maine and inland New Hampshire birds. Such comparisons

highlight behavioural differences possibly due to genetic variation within the

region-wide meta-population (as with Parrish 1995b) or due to regional

differences in the arthropod prey base.

Other components of my work examine the impact of stand diversity

simplification due to commercial forest management on avian foraging behaviour.

Previous work in avian foraging ecology (Holmes and Robinson 1981, Whelan

2001) has suggested, but not tested the potential impacts of reducing tree species

diversity on avian foraging behaviour.

13

Macro-scale (Habitat) Vegetation Structure

Wiens (1989) argued that most work on avian habitat selection and use

emphasizes the importance of macro-scale vegetation structure rather than species

composition. Neglect of vegetation floristic composition is evident in influential work

spanning four decades (MacArthur and MacArthur 1961, Karr and Roth 1971, Anderson

and Shugart 1974, Willson 1974, Cody 1981, James and Wamer 1982, Wahlter 2002).

This emphasis can be traced to the early work of MacArthur and MacArthur (1961), who

found that plant species diversity and foliage height diversity were good predictors of

bird species diversity. Plant species diversity and foliage height diversity were correlated

and they saw no independent role for plant species diversity apart from its contribution to

foliage height diversity, though they could plausibly have argued the opposite. Morse

(1985) concluded that though foliage height distributions are often variables in

multivariate models, they contribute little to explaining the variance in distributions of

birds. Wiens (1989) suggested that the relative ease of measurement of structural

variables over floristic variables may have further contributed to the dominance of this

trend in past research.

14

Floristic Diversity and Micro-Scale Vegetation Structure

Other work has nevertheless demonstrated that plant species composition is

important to generalized selection and use of habitat by birds (Morse 1977, 1978,

Franzreb 1978, Holmes et al. 1979, Beedy 1981, James and Wamer 1982, Anderson et al.

1983, Morrison and Meslow 1983, Rice et al. 1983, Abbott and Van Heurck 1985,

Virkkala 1988). More specifically, a few studies have demonstrated selective use of tree

species by foraging birds (Morse 1971, 1976, Holmes and Robinson 1981, Robinson and

Holmes 1984, Airola and Barrett 1985, Cumming 1994, Robichaud and Villard 1999,

Greenberg et al. 2001, Whelan 2001).

Morse (1968, 1971, 1976) observed warblers foraging in white spruce (Picea

glauca) and red spruce (Picea rubra). His data set included arthropod abundance in the

tree species, avian diet composition, avian foraging behaviour, and examination of fine-

scale foliage structure. He found that black-throated green warblers (Dendroica virens)

selectively foraged in red spruce in spite of lower insect biomass in that tree species.

Morse (1976) proposed that the longer red spruce branches would provide a greater

surface of substrate to search in between inter-branch flights. Red spruce branches might

also be easier to traverse while foraging due to acute angle needle orientation as opposed

to right angle orientation of white spruce (Morse 1976).

Holmes and Robinson (1981) found that seven bird species including the black-

throated green warbler foraged selectively in yellow birch (Betula allegheniensis) on

their New Hampshire study site. They also demonstrated selective avoidance of sugar

maple (Acer saccharum), and American beech (Fagus grandifolia) by most of the same

bird species. They found higher arthropod abundance in the yellow birch than in

15

American beech and sugar maple. They also they also argued that yellow birch had a leaf

and branch morphology most conducive to efficient avian foraging. There are no cases

where selective foraging occurred in a tree species with sub-optimal (according to the

investigators’ definitions) leaf and branch arrangement coupled with optimal arthropod

resources. Similarly there were no cases where selective foraging occurred in a tree

species with optimal (according to the investigators’ definitions) leaf and branch

arrangement coupled with sub optimal arthropod resources (Holmes and Robinson 1981,

Robinson and Holmes 1984, Holmes and Schultz 1988).

Previous research on foraging manoeuvres has addressed the relationship between

forage manoeuvre and time of season (Miles 1990, Lovette and Holmes 1995), forage

manoeuvre differences among congeners (James 1976, Holmes and Robinson 1984),

forage manoeuvre in breeding vs. wintering grounds (Lovette and Holmes 1995), and the

relationship between forage tactics and vegetation community structure (Maurer and

Whitmore 1981). Differential use of forage manoeuvres by insectivorous passerines in

particular tree species has been documented in a northern hardwood forest similar to my

own study sites and for some of the same bird species (Robinson and Holmes 1982,

Robinson and Holmes 1984, Holmes and Schultz 1988).

I selected black-throated blue warbler (Dendroica caerulescens), black-throated

green warbler, American redstart (Setophaga ruticilla), and least flycatcher (Empidonax

minimus) for my research. The four neotropical migrant species are arboreal insectivores

that demonstrate varying degrees of reliance upon the various foraging attack strategies

(Morse 1993, Briskie 1994, Holmes 1994, Sherry and Holmes 1997). Such varied

responses to micro-scale vegetation structure are likely to provide a good model system

16

for elucidation of selective foraging causal processes in relation to micro-scale vegetation

differences.

Morse (1993) found that black-throated green warblers glean over 80% of the

time, and use the hover manoeuvre only 10% of the time in predominantly softwood

forests. Robinson and Holmes (1982) found black-throated green warblers using the

glean and hover manoeuvres 54% and 35% of the time respectively in their

predominantly hardwood study site and recorded the hawk manoeuvre only 2% of the

time. Black-throated blue warblers used the glean and hover manoeuvres 33% and 61%

of the time respectively, and used the hawk manoeuvre 4% of the time (Robinson and

Holmes 1982). These congeners offer a potential contrast in their use of foliage micro-

structure due to their foraging strategy differences. The third warbler, American redstart,

used the glean manoeuvre 23 % of the time and used the hover manoeuvre 53 % of the

time (Robinson and Holmes 1982). Use of the glean by American redstarts was less

frequent than either Dendroica species and its use of the hover was intermediate.

American redstarts used the hawk manoeuvre 6% of the time, slightly more often than the

two Dendroica species (Robinson and Holmes 1982). Least flycatchers used the glean

manoeuvre the least of all, 3% of all forage manoeuvres, and used the hover and hawk

manoeuvres more than the other species, 81% and 10% of the time respectively

(Robinson and Holmes, 1982).

The patterns above represent responses pooled over all tree species, implying an

overall optimal strategy for each bird species. A finer scale of analysis reveals that

foraging strategies can be affected by tree species. Holmes and Schultz (1988) reported

that black-throated green warblers used the glean manoeuvre most frequently on

17

American beech and least frequently on sugar maple, while it used the hover manoeuvre

most frequently on sugar maple and approximately equally on American beech and

yellow birch. Black-throated blue warblers used the glean most frequently on yellow

birch and approximately equally on sugar maple and American beech, while it used the

hover manoeuvre least frequently on yellow birch and approximately equally on sugar

maple and American beech. American redstarts used the hover and glean manoeuvres

approximately equally in American beech, sugar maple and yellow birch. Robinson and

Holmes (1984) showed similar results in their analysis that differentiated the vegetation

into canopy, subcanopy, and understory layers.

More recent field-based work has continued to confirm patterns of selective

foraging in tree species but has not addressed causal processes. Robichaud and Villard

(1999) demonstrated selective foraging and singing by black-throated green warblers in

white spruce (Picea glauca) and selective foraging in trembling aspen (Populus

tremuloides) and balsam poplar (Populus balsamifera). Greenberg et al. (2001) observed

that black-throated green warblers did not select particular tree species for their foraging

activity on their wintering sites in Mexico. This species did, however, show the greatest

degree of specialization in its foraging strategies. The other two species in his study,

Townsend’s warbler (Dendroica townsendi) and hermit warbler (Dendroica occidentalis)

demonstrated strong selective use of pine and weak selective use of oak (Greenberg et al.

2001). Their study did not measure the prey available, so determination of the process

driving the foraging site selection was not possible. The authors speculated that bird and

tree morphology probably played strong roles in the evolution of this genus.

18

Parrish (1995a) offered experimental tests of some of Morse’s (1976) predictions

concerning white and red spruce needle and branch morphology effects on bird foraging

behaviour. He found that manipulating white spruce needle angles to resemble red

spruce angles resulted in higher proportions of prey taken from this substrate versus

controls. Decreasing white spruce needle length to approximate red spruce needles

resulted in increases in prey proportions taken and time spent in this species versus

controls (Parrish 1995a).

In other experimental aviary work Parrish (1995b) took advantage of regional

differences in foliage use by black-throated green warblers. Individuals from New

Hampshire select deciduous foliage, while coastal Maine individuals select coniferous

foliage for foraging and displaying. He experimentally increased the prey for each

species on the less favourable tree type. The birds from each regional population selected

the less favourable tree type for foraging while maintaining the original tree type for

display activities. He took a variety of morphological measurements from individuals of

each population and found significant differences in maxillar width, tibiotarsi,

tarsometatarsi, humeri, and radii. He concluded that morphological variation within a

species could be responsible for intrinsic microhabitat preferences across a region

(Parrish 1995b).

Gunnarson (1996) employed experimental manipulation of spruce needle density

to demonstrate the importance of fine scale tree morphology to foraging insectivores.

One particularly interesting effect might have implications for my own work with

naturally varying leaf densities among deciduous trees. In one trial, he observed lower

19

effects of bird predation on arthropod density on tree branches with needle density

experimentally reduced.

Whelan (2001) employed experimental methods with broad-leaved trees and

insectivorous passerines to demonstrate the influence of fine-scale tree morphology to

foraging insectivores. Black-throated blue warblers, black-throated green warblers, and

American redstarts captured more prey from the upper surfaces of yellow birch (Betula

alleghaniensis). Black-throated blue warblers and American redstarts captured more

prey from lower leaf surfaces of sugar maple (Acer saccharum), while black-throated

green warblers captured more prey from upper leaf surfaces of sugar maple. Yellow

birch has leaves within the same plane as the branch, while sugar maple has leaves

elevated above the branch (Whelan 2001). Further experimentation demonstrated that

leaf dispersion is responsible for leaf surface use, and distance to prey is responsible for

whether the prey manoeuvre is aerial or non-aerial.

In related work, Whelan (1989a) demonstrated that black-throated green warblers

displayed a greater learning efficiency for locating prey on leaf undersides. Black-

throated blue warblers displayed a greater learning efficiency for locating prey on leaf

upper sides . Interestingly, the foliage structure of yellow birch and sugar maple had no

effect on the birds’ ability to locate prey.

Whelan (1989b) used an experimental design to control prey available on

simulated branches and foliage of sugar maple and yellow birch. Black-throated

blue warblers selected distal sugar maple foliage over yellow birch when prey

biomass was equal. Black-throated blue warblers selected yellow birch when the

biomass on this foliage was increased. Black-throated green warblers showed no

Comment [TD2]: Scientific name here

Comment [TD3]: A bad case of noun-

stringing! Try ‘the birds’ ability to learn the

location of prey’

20

selection between these same two foliage patterns when biomass was equivalent,

but did select yellow birch when biomass was increased. However, the increase

in biomass needed to shift the black-throated green warbler selection was greater

than the biomass needed for black-throated blue warblers.

21

Landscape Pattern and Process

I started this project at a time when landscape ecology was replete with studies of

pattern rather than causal processes (see review in Wiens 1995). Lima and Zollner

(1996) suggested that models of animal behaviour should be used to explain landscape

patterns and processes. South et al. (2001) echoed this need for more landscape research

directed at behavioural mechanisms. Morales and Ellner (2002) concluded that

movement models that incorporate behavioural heterogeneity were more powerful in

predicting dispersion than models of landscape structure. Wu and Hobbs (2002) report a

better understanding of landscape process in relation to landscape pattern among the

“Top 10 List for Landscape Ecology in the 21st Century” developed at the 16

th Annual

Symposium of the US Regional Association of International Association of Landscape

Ecology.

Fragmented landscapes have been well studied in the interest of

conservation and the opportunities they offer for experimental approaches to basic

research (Lynch and Whitcomb 1978, Villard et al. 1993, Hagan et al. 1996,

Wiens 1996, Diamond 1999, Lambert and Hannon 2000, Belisle and Desrochers

2002, Robichaud et al. 2002). Early work on landscape fragmentation often

involved applied concepts of island biogeographic theory (MacArthur and

Wilson 1967). Patterns of species diversity and abundance in fragmented

landscapes have been related to patch area (Forman et al. 1976, Galli et al. 1976,

Ambuel and Temple 1983, Freemark and Merriam 1986, Robbins et al. 1989,

Freemark and Collins 1992, Wiens 1996), and patch isolation (MacClintock et al.

1977, Lynch and Whitcomb 1978, and Urban et al. 1988, Litwin and Smith 1992,

22

Villard et al. 1992, and Gilpin 1996).

A common theme in early studies of landscape process focused on habitat edge

dynamics and avian nest predation effects upon reproductive success (Gates and Gysel

1978, Kroodsma 1982, Brittingham and Temple 1983, Wilcove 1985, Harris 1988,

Robinson et al. 1995, Lidicker and Koenig 1996, Bayne and Hobson 1997). These

studies represent an examination of causal process in a landscape pattern, but do not

delve into behavioural processes of a single species.

The study of pairing success in fragmented landscapes represents a

behavioural approach focused on the autecology of a single species. This process

is associated with isolation and reduced area of individual patches (Porneluzi et

al. 1993, Villard et al. 1993, Hagan et al. 1996, Lambert and Hannon 2000). A

variety of studies have addressed landscape processes and resource quality,

though they have not explicitly related foraging behaviour to landscape pattern or

process. Holmes et al. (1992) demonstrated that population recruitment can be

influenced by the availability of arthropod prey at Hubbard Brook, thus

establishing a process whereby arthropod availability might influence productivity

and population distribution across the landscape. Mazerolle and Hobson (2001)

were interested in the relative physiological stresses experienced by territorial

birds in fragments and contiguous forest. Their work focused on territorial

defence rather than foraging ecology. They found that larger and more

competitive males settled more densely in the contiguous forest and experienced

higher stress in defence of their territories. This is a good example of the

interplay between behaviour and landscape pattern.

23

Johnson and Sherry (2001) related arthropod prey distribution to

landscape distribution of warblers on their winter grounds in Jamaica. They did

not relate these patterns to forest fragmentation and foraging behaviour, and could

not relate the patterns to reproductive success. Zanette et al. (2002) documented

lower densities of arthropod prey available to an insectivorous passerine and

subsequent lower reproductive success in fragments vs. contiguous forest in

Australia. While this study simultaneously addressed resource quality,

fragmentation, and reproductive success, it did not include a study of foraging

behaviour. Mortberg (2001) investigated the contribution of tree species to

habitat quality and landscape pattern as perceived by nuthatches. The birds

showed a positive association with oak in landscape patches. The association was

based on census data and did not address such causal processes as foraging.

Belisle and Desrochers (2002) experimented with gap crossing behaviour

of forest birds and not only demonstrated gap avoidance, but quantified effective

gap size. Hinsley (2000) related gap crossing to body mass, flight speed and

brood size of the great tit (Parus major). Matthysen (2002) related great tit and

blue tit (Parus caeruleus) dispersal to natal territory position within a patch.

Robichaud et al. (2002) demonstrated the dynamic reversible quality of a gap

through time as a clearcut forest regenerated. Movement rates of adults increased

in adjacent buffer strips following the harvest. After only four years of

regeneration, movement rates of dispersing juveniles increased in the clearcut.

Gobeil and Villard (2002) performed experimental translocations of two

bird species to examine landscape permeability with respect to forest cover. The

24

ovenbird was selected for its association with closed canopy hardwood stands

while the white-throated sparrow (Zonotrichia albicollis) uses a variety of age

classes and stand compositions. Ovenbird movement was facilitated by the

presence of forest cover while white throated sparrow movement was optimal in

harvested and agricultural landscapes.

In spite of significant advances that establish linkages among behavioural

processes and landscape pattern, and independent work examining the role of

floristic composition in foraging ecology, the two approaches have remained

separate. Studies of landscape pattern and habitat selection by birds continue to

describe vegetation cover types in broad terms and fail to investigate causal

processes (Freedman et al. 1981, Spies et al. 1994, McGarigal and McComb

1995, Willson and Comet 1996, Farina 1997, Hagan et al. 1997, Jansson and

Angelstam 1999, Howell et al. 2000, Penhollow and Stauffer 2000, Jones and

Robertson 2001, Wolter and White 2002).

Such work documenting extremes on a continuum of predictor variables

such as hardwood vs. softwood are valuable first approximations. Personal

involvement in one of these studies (Hagan et al. 1997) leaves me with no doubt

about the need to simplify the experimental design when looking for broad

patterns. The next logical step is to incorporate more natural complexity into our

work as we attempt to understand causal landscape processes.

This dissertation relates selective foraging in tree species to songbird

landscape distribution and productivity to establish the importance of local

behavioural processes in landscape pattern and dynamics. Previous work has

Comment [TD4]: How did WTSP

respond? Having mentioned them, you

should not leave us in suspense..!

25

related foraging behaviour and habitat selection, but at limited temporal and

spatial scales within homogeneous landscape patches rather than across the

mosaic (Sherry and Holmes 1979) or across isolated islands in a coastal land-

seascape rather than a strictly terrestrial mosaic (Morse 1976). Sherry and

Holmes (1979) used ordination techniques to relate avian community structure to

intra-patch distributions of those tree species selected by foraging birds. Sherry

and Holmes (1985) describe the scale of their work as encompassing one habitat

type. This work did not address larger temporal and spatial scales associated with

seasonal landscape settlement. Orians (1991) makes distinctions among different

types of habitat selection based upon the spatial scale of an animal’s behaviour.

At the largest scale, he describes habitat selection as a function of emigration,

immigration, and migration. I use local foraging patterns to predict a landscape

settlement pattern.

I continue my investigation of the role local behavioural processes play in

landscape dynamics in conjunction with that of my colleague working

simultaneously on the same study site (Gunn 2002). I relate local foraging

behaviour and tree species composition to landscape patterns of reproductive

success documented by Gunn (2002). While this component of my work is more

suggestive than conclusive, such a scaling up of a behavioural process to

landscape pattern of productivity has not previously been demonstrated.

Whelan (2001) and Holmes and Robinson (1981) argue for the

maintenance of native tree species diversity in managed forests in light of the

unique foraging opportunities offered by different tree species. I was able to

26

compare bird foraging behaviour in moderately managed and intensively

managed landscapes to document the effects of simplifying the tree species

composition. My work also offers the first look at a region-wide comparison of

foraging behaviour differences similar to the work of Parrish (1985b) comparing

tree species selection differences between coastal and inland populations of a

warbler species. The bird and tree species that are discussed in this dissertation

included the same species studied in the most comprehensive work to date on this

subject (Holmes and Robinson 1981, Robinson and Holmes 1984, Holmes and

Schultz 1988), yet the results of their studies and mine differ in important ways.

The only other work (Block 1988) that made a regional comparison of tree

species use did not compare the two causal processes.

27

Chapter 2 Methods

Study Area and General Sampling Design

My work was conducted in the Appalachian highlands of northwestern

New Brunswick, Canada. The study site (Figure 2.1) is located west of the

Tobique River (47 N, 67 W) on freehold land managed by Fraser Papers Inc.

The field work for this dissertation was conducted in the upland stands.

We selected two 4900 ha study landscapes to represent the two extremes found in

commercially managed forest. One of these landscapes was chosen for its maximal area

of mature secondary mixed-wood forest not scheduled to be harvested for the duration of

the project, approximately 4 years. Less than 15% of this landscape was in young

plantation or recent clearcuts. This served as a reference landscape (mm-grid) for

comparison with the second 4900 ha landscape, the intensively managed area (im-grid).

Over 50% of this landscape was covered by young regenerating clearcut or softwood

plantation.

American beech (Fagus grandifolia), yellow birch, and sugar maple were

the dominant upland overstory species in the mm-grid, while yellow birch and

sugar maple were the dominant upland overstory species in the im-grid. American

beech was a minimal component of the im-grid (Table 2.1). Please see Table 2.1

for relative proportions of tree species in the two grids. Balsam fir (Abies

balsamifera) and white spruce, occurred in low densities in the mid to upper

canopy layers. The upland understory was composed of a mixture of regenerating

saplings of the hardwood overstory, and hobblebush (Viburnum alterniflora),

striped maple (Acer spicatum), beaked hazelnut (Corylus cornuta), and mountain

28

maple (Acer pennsylvanicum). My work focused upon avian use of yellow birch,

sugar maple, American beech, and balsam fir because these trees account for 92-

97% of the trees on the study grids.

This study landscape was shared by several other research projects from the

Universities of New Brunswick and Universite dé Moncton in conjunction with the

Sustainable Forest Management Network of Centres of Excellence (NCESFM) and

Natural Sciences and Engineering Research Council of Canada (NSERC). Various

projects included landscape influences on mammal population structure (Bowman 2000

Bowman et al. 2000, Bowman et al. 2001), and avian productivity (Gunn 2002, Gunn et

al. 2000, Bourque and Villard 2001).

We employed systematic sampling after the methods of Fortin et al. (1989) and

Roland and Taylor (1997). Systematic sampling provides statistical properties

comparable to random sampling (McPherson 1990, Hayak and Buzas 1997, Bart et al.

1998), but avoids a priori imposition of perceived patterns upon the landscape. Wiens

(1989) cites the work of Taylor et al. (1984), who employed both a stratified random

sampling and systematic sampling design. They found that systematic sampling is better

suited for investigation of floristic composition effects. Wiens (1989) argues that

previous studies documenting the significance of vegetation structure over floristics

might have been biased by reliance upon stratified random sampling considering the

results of Taylor et al. (1984).

We established a 64 point grid system on the two separate 4900 ha landscapes

(Figure 2.2). Points were 1 kilometre apart on these macro-grids. Within each of these

two landscapes, two smaller 64 point grid systems were established. Points were 250

29

meters apart on these meso-grids. My work was conducted on the four meso-grids within

the two macro-grids. One meso-grid in each macro-grid was composed of mature, closed

canopy, upland forest dominated by hardwood. The other meso-grid in each landscape

differed in that the stands had been selectively harvested within the past 15 years. The

habitat requirements of the focal bird species precluded use of other cover types such as

regenerating clearcuts and plantations.

30

Grid Vegetation Sampling

We sampled vegetation at each grid point using a modification of the methods

employed by Hagan et al. (1996). We established three separate belt transects that

radiated from each grid point (Figure 2.3). We used these to sample all woody and

herbaceous plants and a variety of physical parameters. The length and width of the belt

varied according to the parameters being measured. I used only the tree data in the

analyses for this work as my primary interest was in differential use of tree species as a

foraging substrate. While understory vegetation might have an influence on tree species

use through a community pathway, I designed this study to examine avian predator,

arthropod prey, and tree species substrate interactions first.

Tree transects were 10 x 20 meters. All trees greater than or equal to 8 cm

diameter at breast height (dbh) and located within the belt transect were identified to

species and their dbh recorded. The density and dbh allowed calculations of basal areas

for each species of tree at each grid point and for the entire grid. These measurements

were necessary for calculation of availability of tree species to foraging songbirds.

31

Foraging Behaviour

There has been confusion over the concepts of resource use, selection and

preference, so it is worthwhile to outline here how these terms are interpreted and

applied. I tested for selective use in this study, as have most field observational studies in

spite of their claim to have documented resource preference. Preference is better

determined in a strictly experimental setting where resources are equally available

(Litvaitis et al. 1996).

Bart et al. (1998) presented a slightly different interpretation of preference. They

claim the proportions of available resources may vary and that preference is demonstrated

if the use of the resources remains unchanged in spite of varying availability. This

definition is closer to the concept of selective use that most investigators have employed

in field-based studies. If a resource is used out of proportion to its availability to an

animal, then that use is said to be selective (Johnson 1980). If a resource is used in

proportion to its availability then the use is indiscriminate or random. The distinction

made by Bart et al. (1998) is that in order to claim a resource is preferred, use of it must

remain consistent in spite of adjustments to its availability, whereas Litvaitis maintains

that preference is demonstrated only if resources remain equally available.

My design, as with most field-based work, cannot adjust resource proportions

temporally, nor does it provide for equal proportions. I elected to study resource use in

terms of selective use, selective avoidance, or random use, and simply avoid the concept

of preference.

Robinson and Holmes (1984) measured number of attacks on prey per minute for

4 species of forest songbird, including two of the species I studied, black-throated blue

32

warbler, and American redstart. I have computed two different foraging rates, attacks

per unit time and attacks per forage flight. Attacks per unit time is the same as the rate

calculated by Robinson and Holmes (1984). Both of our rate variables are measures of

efficiency, one in terms of temporal units, the other in terms of spatial units. Attacks per

forage flight is the prey yield in a given unit of branch and or volume of foliage searched

by the bird before flying to the next region of branch, or to a completely different branch.

Robinson and Holmes (1982) report that this search area, which I will call a foliage

volume, has a radius up to 1 meter.

I conducted observations on black-throated blue warbler, black-throated green

warbler, American redstart, and least flycatcher (Empidonax minimus), found within a

100 meter radius of the meso-grid points. I used all-occurrences sampling in conjunction

with focal-animal sampling (Lehner 1996) to collect a specific set of observations on

individuals of the focal species. All-occurrences sampling is alternatively known as

event sampling (Hutt and Hutt 1970) and complete-record sampling (Slater 1978). I

selected a set of specific behaviours related to foraging and display. Under all-

occurrences sampling every occurrence of these behaviours is recorded along with a

record of elapsed time. The period of time a bird was under continuous observation in

an individual tree is defined here as an observation sequence. I located focal birds

usually through their territorial song prior to first visual contact. In this manner, there is

not likely any bias toward any particular manoeuvre being associated with a tree species.

My analyses are confined to observations collected on males only. In general, more

effort was devoted to observation of black-throated blue and black-throated green

warblers and this is reflected in the sample sizes obtained.

33

Remsen and Robinson (1990) present a five part model of foraging behaviour

that includes search movements, attack movements at prey or substrate containing prey,

foraging site, food items, and food handling. I focused efforts on the first three, finding

the latter two observations occurring too infrequently for analysis. It is well established

that though forage attacks might not necessarily be successful, they are not initiated

unless prey is detected (Holmes and Robinson 1981, Remsen and Robinson 1990,

Lovette and Holmes 1995). While foraging manoeuvres such as hops and flights do not

necessarily indicate the presence of prey, attacks do.

My use and interpretation of the various foraging variables is similar to that

employed by Holmes and Robinson (1981). Recorded behaviours included the foraging

manoeuvre (e.g. glean, hover, hawk, hover-glean), frequency of hops on a branch,

frequency of flights, and duration of the observation sequence. Location of behaviour

was also recorded, including tree species, size of tree, zone of tree, and tree condition.

The behaviours I recorded included the attack manoeuvre (e.g. glean, hover, hawk,

hover-glean), foraging manoeuvre (e.g. frequency of hops and flights), and duration of

the observation sequence. our, including tree species, size of tree, and vertical zone of

the forest.

Hops are leg-propelled movements along the tree branch or twig in search of prey

items and flights are usually short rapid flights along a branch or to a new branch within

the same tree followed by another sequence of hopping. I defined glean, hover, and

hover-glean respectively as attacks directed toward prey on a leaf from a perched

position, attacks directed toward prey on a leaf while hovering underneath, and attacks

34

directed at foliage from above while in flight. The hawk is an attack manoeuvre directed

at prey in the air column rather than on leaves.

I recorded the tree species and vertical zone of the forest for all behavioural

observations. The focal species forage primarily in two vertical zones in the forest.

Black-throated green warbler concentrates its activity in the middle to upper strata

(Morse 1993), while black-throated blue warbler concentrates in the lower to middle-

strata (Robinson and Holmes 1984). Male black-throated blue warblers foraged at a

height of 5.9 +/- 3.8 SD m (Holmes 1994). American redstarts forage from the ground to

the top of the canopy, with foraging heights for males averaging 10.9 m +/- SD (Sherry

and Holmes 1997). Sherry (1979) reported that least flycatchers foraged in the upper

canopy at heights averaging 12-15 m.

Calculation of tree species use was based on number of attacks on potential prey

items in relation to the availability of a tree species (Holmes and Robinson 1981). This

calculation is derived from the generally accepted method of calculating use of a resource

in proportion to its availability (Litvaitis et al. 1996, Bart et al. 1998). The resource here

is the leaf area of each tree species determined on a relative basis from the basal areas

(Holmes and Robinson 1981, Dodge et al. 1990) of American beech, yellow birch, sugar

maple and balsam fir. Two different measures of foraging rate were calculated. Both

variables use the number of attack manoeuvres. For one variable the number of attacks is

related to units of time, and for the other the number of attacks is related to the number of

flights. The former is reported as the attacks per minute and the latter as attacks per

flight. Attacks per flight is a spatial variable as attacks are conducted in a defined area

35

(e.g. along a branch) bounded by a flight to that section of branch and away from that

section of branch.

My use and interpretation of the various foraging variables is similar to that

employed by Holmes and Robinson (1981). The behaviours I recorded included the

attack manoeuvre (i.e. glean, hover, and hawk). I defined glean and hover respectively as

attacks directed toward prey on a leaf from a perched position and attacks directed toward

prey on a leaf while hovering underneath. The hawk is an attack manoeuvre directed at

prey in the air column rather than on leaves.

36

Diet Composition

I required avian diet data from focal bird species to confirm that arthropod

samples from the trees included the prey items consumed by the birds observed foraging.

It is conceivable that an arthropod sampling technique might not capture crucial

components of a bird’s diet. This would render conclusions about foraging site selection

based upon arthropod community data questionable. This component of my work was

designed to ensure that the arthropods sampled from the arboreal community included

those arthropods consumed by the focal bird species. Rosenberg and Cooper (1990)

comment that most studies on avian foraging ecology overlook this important logical

connection between foraging behaviour and the prey items collected from the substrate.

(For a review of collection means see Diamond and McKinley appendix 1).

Rosenberg and Cooper (1990) cite three methods to quantify diet items. (1)

Percent occurrence is a measure of the presence of a prey item without any statement

about abundance. It is useful in partially digested samples containing a mix of arthropod

body parts. (2) Frequency requires a count of individual prey items in their entirety, not

practical in a digested sample. (3) Percent volume and percent weight estimates require

identification of all body parts without assigning them to a specific number of prey items.

This method is not practical in stomach samples containing hundreds of exoskeletal parts.

I selected the first method of prey item quantification, and calculated percent occurrence

of each arthropod order across all stomach samples for each bird species separately. I

recorded the presence of an arthropod, either through the presence of an intact individual

or the presence of at least one distinctive body part indicative of a particular order. This

measure was essentially one of presence or absence of an order in the sample obtained

37

from each individual bird. The percent tally shows the proportion of individuals by bird

species that had a particular arthropod order in its stomach sample.

I captured black-throated blue warblers and black-throated green warblers on the

mm-grid and im-grid during the 1998 and 1999 breeding seasons. Territorial males were

lured into mist nets using playback of male song in conjunction with model birds placed

near the mist net. I used syrup of ipecac to induce emesis (Diamond and McKinley,

Appendix 1) in the captured birds.

38

Arthropod and Tree Species Associations

I designed this component of my study to test the arthropod hypothesis.

Ornithologists have addressed the question of arthropod availability to birds by a

variety of methods (Wolda 1990). As with sampling other taxonomic groups one may

wish to derive an estimate of relative abundance (index), or an estimate of absolute

abundance (density). Methods used to obtain an index include the use of sticky traps,

malaise traps, shake-cloth methods, sweep-net sampling, and pitfall traps. Methods used

to estimate density include collection of vegetation, stationary suction traps, portable

vacuum sampling, and counts of individuals from direct observation (Cooper and

Whitmore 1990).

I equipped a pole pruner with a cloth basket to clip and retrieve foliage samples

(Majer et al. 1990) from American beech, sugar maple, yellow birch, and balsam fir. I

mounted a bag 1 meter deep and 0.5 meters in diameter was mounted on a stiff wire

frame which was mounted on the cutting head of a set of pole pruners. I slipped the

pole pruning bag over each foliage sample slowly to minimize disturbance of arthropods

present. Foliage, branches and twigs were not distorted to fit the bag, and movement of

the bag over the end of the branch was stopped when the branch reached the bottom. I

clipped the branch and lowered the bag immediately to search foliage, branches and bag

for arthropods. I placed the arthropods in 70% solutions of ethanol, and sealed the

foliage samples in plastic bags for storage at 4C. I later searched the foliage samples for

additional arthropods while counting leaf numbers.

39

I collected foliage samples at grid points on the mm-grid in 1998 and at points on

the mm-grid and im-grid in 1999. I developed a sampling scheme to provide an estimate

of arthropod density and diversity per unit of tree foliage area from the two primary

vertical feeding zones used by the focal birds. These zones were located at 3 meters

(zone 1) and 12 meters (zone 2) above the ground. These zones are based upon the

previously described foraging strata for the focal bird species. I cut foliage samples from

zone 1 from the nearest representative of each tree species >4cm dbh that possessed

foliage within reach of the pole pruner cutting head and basket. I cut foliage samples

from zone 2 from the nearest representative of each tree species >8cm dbh that possessed

foliage within reach of the pole pruner cutting head and basket. The tree selection criteria

were further constrained by selecting trees of similar height within each zone across all

points sampled so that age of foliage was controlled. The region sampled from trees of

both zones was the middle to upper middle portion of the tree.

I calculated arthropod densities per unit area of leaf and per clip sample. I used a

polar planimeter to calculate average areas for the leaves of each tree species and counted

the number of leaves in each sample to yield a density of arthropods per unit area of leaf.

The number of arthropods per clip sample is simply the count obtained for each order for

a given clipping in a tree species in a given zone. As the densities from zones 1 and 2 did

not differ (anova p<0.05), no distinction between them was made in subsequent analyses.

40

Landscape Distribution and Productivity

If there is a relationship between the abundance of a tree species and either the

presence or productivity of a bird species, that relationship should be strongest between

black-throated green warblers and American beech. I predict a similar relationship

between black-throated blue warbler and American beech. These predictions are based

upon the results from the selective foraging component of this study that demonstrated

higher foraging success for these bird species in American beech. American redstart may

also exhibit this relationship, but it is more likely for the two Dendroica warblers because

these species foraged in American beech more often. The lack of an association between

the abundance of a tree species and either the presence of that species or the productivity

of that species does not necessarily imply that the tree species is not important. It could

mean that a minimum threshold density for a particular tree species has been reached. I

also predict that species recorded during the point counts that possess foraging strategies

similar to the Dendroica warblers will be more likely to demonstrate an association

between their presence at a point and American beech abundance.

Predictions concerning tree species distribution in relation to avian distribution

and productivity stem from my foraging ecology work discussed earlier in this thesis.

The predictor variable I selected was the basal area of American beech, yellow birch, and

sugar maple, the three dominant hardwood tree species. I used logistic regression to

examine the relationships among these variables and the presence of foliage gleaning

birds (including my focal species) in the independent avian landscape census.

41

We used point counts to conduct the independent avian census in the reference

landscape during the 1996 and 1997 field seasons (NCESFM, New Brunswick,

unpublished data). All species within 100 meters of a point were recorded within a five

minute period during two separate visits within each season (Blondel et al. 1981,

Morrison et al. 1981, Bibby et al. 1993, Hagan et al. 1996, and Hagan et al. 1997).

I used the point count data to determine the presence or absence of a bird species

at a grid point. I elected to use presence/absence data rather than an estimate of

abundance since many of these species were registered only once or twice during a single

point count. A bird recorded at least once at a point on either round or in either year of

the point count data collection counted as a registration. Multiple registrations within a

round or over both rounds were not given extra weight, a species was either present or

absent at a point for the breeding season. I determined presence and absence of the four

focal bird species, American redstart, black-throated blue warbler, black-throated green

warbler, and least flycatcher at each grid point. I also determined presence and absence

of the following species: northern parula Parula americana, red-eyed vireo Vireo

olivaceus , blue-headed vireo Vireo solitarius, Philadelphia vireo Vireo philadelphicus,

yellow-rumped warbler Dendroica coronata, magnolia warbler Dendroica magnolia,

chestnut-sided warbler Dendroica pensylvanica, and scarlet tanager Piranga olivacea. I

selected these species because they forage from tree foliage and might exhibit patterns of

presence and absence with respect to tree species similar to the focal species.

42

I used results from a colleague’s study (Gunn 2002) conducted on the same study

plots during the same field seasons to examine avian productivity across the landscape in

terms of stand composition. His analyses included three of my focal species and several

additional foliage foraging species.

43

Table 2.1. Relative proportions of the four focal tree species on each grid.

Tree Species mm-

grid

im-grid

American beech 0.41 0.02

balsam fir 0.1 0.23

sugar maple 0.37 0.40

yellow birch 0.13 0.34

44

Figure 2.1. New Brunswick study site location.

45

Figure 2.2. Schematic diagram of research grid design.

46

Figure 2.3. Schematic diagram of vegetation sampling plots.

47

Chapter 3 Foraging Results

Foraging Behaviour Results Overview

Table 2.1 presents the number of observation sequences and the mean duration in

seconds of each observation sequence by bird species, year, and research grid.

Observations in the im-grid were limited to the years reported, 1997 and 1999. There are

no data reported on the im-grid in 1999 for the American redstart as reduced collection

efforts for this species yielded only three observation sequences. The number of

observation sequences for least flycatcher and American redstart in the latter two years of

the study reflect the lower effort devoted to observation of these two species.

Conversely, there are more observation sequences for black-throated blue and black-

throated green warblers as I directed more effort toward collecting data on these species.

Following the concerns of Hejl et al.(1990) and Bell et al. (1990) I examined the

data for autocorrelation prior to applying the chi-square goodness of fit tests to the

sequential observations. I found an insignificant degree of autocorrelation at a maximum

temporal lag of two observation sequences. The concern in applying chi-square tests to

such data is that the standard errors will be underestimated. I resampled the data using

the jackknife procedure to estimate the standard errors free of any degree of

autocorrelation. The standard error (SE = 0.0114) was the same value in the original data

set and the jackknifed data set. I concluded that chi-square goodness of fit tests could be

applied to the sequential data.

Table 2.2 reports the sum of foraging attacks made by each bird species in each

tree species for the four years of this study. As mentioned previously, there were only

48

two years (1997 and 1999) of data collected on the intensively managed grids. During

years 1996 and 1997 I used 3 person field crews and 2 person field crews for the years

1998 and 1999. Consequently, there were more observation sequences (Table 2.1) and

higher attack numbers (Table 2.2) recorded overall in 1996 when the intensively

managed grid was not under study and a 3 person field crew was at work. In 1997 the

efforts of the three person field crew were divided over two grids (managed and

reference) and this is reflected in the lower numbers of observation sequences and total

attack numbers. In 1998 I used a two person field crew, but did not work on the

intensively managed grid. The low numbers of observation sequences and attack

numbers for 1999 reflect the two person field crew now covering managed and

moderately managed grids. The standard errors for these various years and bird species

remain comparable.

Foraging Site Selection

It is commonly considered that expected values in a chi-square goodness of fit test

must be at least 5.0 (Zar 1996). There were only three combinations of bird species/year/

/grid where there were expected values below 5 where there were also significant

differences in expected versus observed tree species usage. These three cases were:

American redstart/mm-grid/1997, least flycatcher/mm-grid/1997, and least

flycatcher/mm-grid/1998 (see Tables 1.1 and 2.2).

Data were available for 23 of the 24 combinations of bird species/year/grid (Table

2.2). Twenty cases out of 23 showed a significant result (p<0.05) for use of all four tree

species versus their availability (Table 2.3). The results in 14 of these cases were

49

significant at p< 0.001 (Table 2.3) . Thus it is apparent that tree species differed in the

foraging opportunities they offered to birds.

After determining that use of the tree species was not random in terms of overall

attacks versus availability of all tree species, I tested the statistical significance of the

proportional use versus availability of each pairwise comparison of tree species. I

constructed confidence intervals with the alpha level adjusted for multiple simultaneous

comparisons (Litvaitis et al. 1996). The dominant trend in the mm-grid for all four bird

species in all four years (Figures 2.1-2.6) is selective use of beech and selective

avoidance or use in proportion to availability of yellow birch, sugar maple, and balsam fir

(p<0.05).

It was selectively avoided in one case, (American redstart in 1999), and used in

proportion to its abundance in the remaining 4 cases. In only one year and for one bird,

(American redstart in 1999), was birch selected out of proportion to its abundance. In 9

of the cases, birch was selectively avoided, and in the remaining 6 cases use was in

proportion to availability. Sugar maple was selectively used in only one year by one

species of bird, the least flycatcher in 1999. In 8 cases it was avoided and in the

remaining 7 cases, use was in proportion to availability. Fir was selectively used in one

case, (black-throated green warbler in 1999), selectively avoided in 10 cases, and used in

proportion to its abundance in the remaining 4 cases.

Birch, maple and fir were selectively used in only 1 of 16 cases each while beech

was selectively used in 11 cases. Beech was selectively avoided in only one case while

birch, maple and fir were avoided in 9, 8, and 10 cases respectively. The selection of

50

beech is one striking result. Another striking pattern here is the consist avoidance of the

other three species.

Figures are presented only for those years and bird species where some significant

difference in use was found. Review of Figures 2.1- 2.6 show similar patterns of use and

availability for the four tree species for black-throated blue warbler, least flycatcher, and

American redstart for years 1996 through 1998. In year 1999 these three bird species

show a relatively lower use of the beech and either higher use of the sugar maple as with

the flycatcher and black-throated blue warbler, or higher use of the birch as with the

American redstart. The years 1996, 1997, and 1999 for black-throated green warbler

show similar patterns of selective use and avoidance as those shown by the other three

species of bird. The departure from the trend for the black-throated green warbler occurs

in 1998.

51

Forage Rate

The foraging rate data measured both in terms of time and number of flights were

not normally distributed. Consequently, I applied the nonparametric Kruskal-Wallis test

to evaluate the differences in the two foraging rate variables among the four tree species.

I used two different post-hoc tests to determine the significance of multiple pairwise

comparisons of both foraging rate variables for each tree species. I used the

nonparametric post-hoc Nemenyi test (Zar 1996) and multiple Mann-Whitney tests

(Sheskin 1997). The Nemenyi test adjusts alpha for the multiple comparisons. These

results are reported in Table 2.4. As it was my original intent to test the foraging rates

among tree species a priori, I concur with Sheskin that use of the Mann-Whitney test is

justified as long as the specific comparisons made were formulated a priori. These

results are reported in tables 2.5-2.8. In both cases, only those combinations of year and

bird that yielded significant results are reported.

The more conservative approach of the Nemenyi post-hoc test shows significant

results (p<0.05) for one bird species in one year, the black-throated green warbler in

1996, for attacks per unit time only. Three pairwise comparisons are significantly

different in this test including higher foraging rates in beech and maple and lower rates in

fir and birch. Multiple Mann-Whitney tests reveal a similar pattern, but in several more

cases as this post-hoc approach is less conservative than the Nemenyi which adjusts the

alpha level. Foraging rates did not differ among tree species on the intensively managed

grid. As beech was the tree associated with higher foraging rates on the moderately

managed grid, yet in only a limited number of cases, perhaps it is not surprising that the

52

intensively managed grid, lacking the beech component, showed no foraging rate

variation among tree species.

Foraging rates measured as attacks per unit time are different (p<0.05) in 1996

and 1999 for black-throated blue and black-throated green warblers respectively (Tables

2.4-2.5) as determined by multiple Mann-Whitney post-hoc tests. Foraging rate

measured as attacks per flight are different (p<0.05) in 1996 and 1999 for black-throated

green warbler only (Tables 2.6-2.7) as determined by multiple Mann-Whitney post-hoc

tests. The post-hoc Nemenyi test shows a difference for one bird species, black-throated

green warbler in only one year, 1996. The dominant trend revealed by either post-hoc

test is for higher foraging rates in beech over any other species of tree.

53

Forage Manoeuvre Results

Table 3.1 reports attack manoeuvres by bird and tree species, these are the

observed values used for the chi-square tests. American redstart, least flycatcher, black-

throated blue warbler and black-throated green warbler total forage attacks in the four

focal tree species combined were 77, 240, 177, and 147 respectively (Table 3.1).

Expected values were derived from the proportion of a particular forage manoeuvre

expected to occur in a tree species based upon that tree species abundance in the

landscape if forage manoeuvres were being conducted randomly with respect to tree

species abundance. All birds showed a significant difference (chi-square p < 0.05) in

their use of the hover among tree species (Table 3.2). The only species to show a

significant difference (chi-square p <0.05) in the use of the hawk was least flycatcher

(Table 3.2). Two species, black-throated green warbler and American redstart, exhibited

a significant difference (chi-square p < 0.05) in their use of the glean manoeuvre (Table

4.2).

Post-hoc multiple comparisons were made after adjustment of alpha levels so that

I might draw inferences central to the hypothesis concerning the use of the hover

manoeuvre in relation to beech and birch leaf and branch morphology. These results are

reported in Figures 3.1-3.4.

54

The black-throated blue warbler and American redstart hovered in higher

proportions in American beech (Figures 3.2 and 3.3 respectively, chi-square

p<0.05,). Hovers by black-throated green warbler in American beech were not

significantly different (Figure 3.2, p>0.05). Proportions of hovers performed by

black-throated green warbler and American redstart in yellow birch were

significantly lower (Figures 3.1 and 3.3 respectively, chi-square p<0.05) than in

other tree species. Proportion of hovers by the black-throated blue warbler in

yellow birch were not significantly different (Figure 3.2, chi-square p>0.05).

Neither black-throated blue (Table 3.2, chi-square p<0.05,) nor black-

throated green warbler (Figure 4.2, chi-square p<0.05) performed higher

proportions of gleans in the broad leaved trees; American beech, yellow birch,

and sugar maple. American redstart performed a higher proportion of gleans in

beech (Figure 3.3, chi-square p < 0.05). Black-throated green warbler did show a

higher proportion of gleans in balsam fir (Figure 3.1, chi-square p<0.05).

Diet Composition

Representatives of 9 arthropod orders were recovered from the stomach samples

of the two bird species. The nine arthropod orders occurred in statistically similar (chi-

square p>0.05) proportions in the samples obtained from both Dendroica species.

Members of the order Araneida (spiders) were present in more samples than any other

order and in similar proportions for both bird species. Exoskeletal remains of several

arthropod orders were recovered from all samples. Orders identified included araneida,

trichoptera, coleoptera, diptera, hemiptera, lepidoptera, thysanoptera, homoptera, and

hymenoptera.

55

The orders most frequently encountered in the samples from both warbler species

were Araneida, Coleoptera, Diptera, and Hymenoptera. Araneida was found in > 75% of

the samples for both warblers, and Coleoptera was in > 60 % of the samples for both

birds. Diptera was in > 20 % of black-throated green warblers and > 60 % of black-

throated blue warblers. Hymenoptera was in >50% of black-throated green warblers and

>30% of black-throated blue warblers (Figure 3.21). There were no significant

differences (chi-square p>0.05) between the Dendroica species in the proportional

occurrence of an arthropod order based upon the following two assumptions: 1) no orders

are missed due digestion of a diagnostic body part, and 2) the sample collected included

all stomach contents.

56

Table 3.1. Number of observation sequences by grid, bird species, and year with mean

duration of sequence and standard error (SE).

Bird Year Grid* N Sequences Mean Duration

secs.

SE

amre 1996 mm 136 19.8 1.2

amre 1997 mm 383 28.1 2.6

amre 1998 mm 264 11.6 1.1

amre 1999 mm 14 25.7 6.1

lefl 1996 mm 377 37.0 2.8

lefl 1997 mm 67 32.0 8.8

lefl 1998 mm 131 24.9 4.4

lefl 1999 mm 34 30.6 6.9

btbw 1996 mm 675 33.2 2.4

btbw 1997 mm 261 34.3 3.2

btbw 1998 mm 452 18.4 1.5

btbw 1999 mm 201 25.8 3.7

btgw 1996 mm 223 69.2 10.7

btgw 1997 mm 238 27.2 2.0

btgw 1998 mm 360 17.1 1.1

btgw 1999 mm 134 20.9 5.5

amre 1997 im 28 35.7 8.0

lefl 1997 im 27 28.0 6.0

lefl 1999 im 38 25.0 8.8

btbw 1997 im 118 26.0 4.4

btbw 1999 im 172 13.0 1.1

btgw 1997 im 302 17.0 1.9

btgw 1999 im 83 27.0 4.0

*mm signifies moderately managed grid, im signifies intensively managed grid

57

Table 3.2. Number of attacks each season by bird and tree species, and mean number of

attacks per observation sequence with standard error SE by grid and year.

moderately managed grid 1996

amre lefl btbw btgw

N Mean SE N Mean SE N Mean SE N Mean SE

ambe 55 0.57 0.1 161 0.3 0.05 280 0.57 0.04 113 0.62 0.07

bafi 3 0.38 0.26 26 0.46 0.09 14 0.58 0.16 19 0.38 0.1

suma 20 0.57 0.14 0 0 0 56 0.36 0.06 52 0.78 0.1

yebi 4 1 0.58 0 0 0 15 0.52 0.13 0 0 0

moderately managed grid 1997

ambe 134 0.62 0.07 22 0.46 0.12 92 0.71 0.1 89 0.79 0.11

bafi 6 0.86 0.71 0 0 0 33 0.89 0.46 19 0.7 0.21

suma 26 0.42 0.1 8 0.73 0.38 50 0.63 0.13 28 0.42 0.16

yebi 15 0.38 0.1 0 0 0 26 1.1 0.35 21 0.72 0.22

moderately managed grid 1998

ambe 31 0.25 0.04 13 0.19 0.06 65 0.28 0.05 38 0.27 0.05

bafi 0 0 0 0 0 0 15 0.31 0.1 24 0.41 0.11

suma 17 0.38 0.12 1 0.09 0.09 35 0.25 0.05 32 0.25 0.05

yebi 2 0.29 0.18 1 0.33 0.33 16 0.55 0.19 12 0.34 0.16

moderately managed grid 1999

ambe 0 0 0 3 0.12 0.14 29 0.4 0.09 27 0.49 0.14

bafi 0 0 0 0 0 0 4 0.22 0.22 3 0.18 0.01

suma 0 0 0 8 1 0.4 39 0.40 0.08 9 0.18 0.07

yebi 8 0.89 0.42 0 0 0 5 0.39 0.18 1 0.08 0.08

intensively managed grid 1997

ambe 4 0.75 0.48 0 0 0 0 0 0 2 0.33 0.21

bafi 7 0.86 0.55 0 0 0 0 0 0 18 1 0.38

suma 12 1.42 0.40 8 0.33 0.16 35 0.52 0.1 134 0.57 0.09

yebi 5 0.8 0.58 0 0 0 38 0.76 0.15 32 0.74 0.15

intensively managed grid 1999

ambe na* na na 0 0 0 0 0 0 1 0.5 0.05

bafi na na na 0 0 0 0 0 0 10 1.7 0.16

suma na na na 12 0.39 0.12 34 0.24 0.04 29 0.51 0.16

yebi na na na 0 0.6 0.25 11 0.37 0.11 12 0.67 0.24 *

not available

58

Table 3.3. Goodness of fit analysis for foraging use versus availability of the four focal

tree species combined, chi-square p<0.05*, P<0.01**, P<0.001***, ns if not significant.

a mm signifies moderately managed grid, im signifies intensively managed grid.

Bird species Year grida

P

amre 1996 mm ***

amre 1997 mm ***

amre 1998 mm **

amre 1999 mm ***

lefl 1996 mm ***

lefl 1997 mm **

lefl 1998 mm *

lefl 1999 mm ns

btbw 1996 mm ***

btbw 1997 mm ***

btbw 1998 mm ***

btbw 1999 mm *

btgw 1996 mm ***

btgw 1997 mm ***

btgw 1998 mm ***

btgw 1999 mm ***

amre 1997 im ns

lefl 1997 im **

lefl 1999 im *

btbw 1997 im ***

btbw 1999 im ***

btgw 1997 im ***

btgw 1999 im ns

59

Table 3.4. Multiple comparison of btgw forage attacks per minute (Nemenyi test p<0.05)

for differences among tree species, 1996 moderately managed grid.

ambe bafi suma yebi

ambe - - - -

bafi ns*

- - -

suma ns (suma>bafi) - -

yebi (ambe>yebi) ns (suma>yebi) - *p values from Kruskal-Wallis tests are denoted as ns if not significant.

Table 3.5. Multiple comparison of btgw forage attacks per minute (Mann-Whitney p <

0.05) for differences among tree species, 1999 moderately managed grid.

ambe bafi suma yebi

ambe - - - -

bafi (ambe>bafi) - - -

suma (ambe>suma) ns* - -

yebi (ambe>yebi) ns ns - *p values from Kruskal-Wallis tests are denoted as ns if not significant.

60

Table 3.6. Multiple comparison of btbw forage attacks per minute (Mann-Whitney test

p<0.05) for differences among tree species, 1996 moderately managed grid.

ambe bafi suma yebi

ambe - - - -

bafi ns*

- - -

suma (ambe>suma) ns - -

yebi ns ns ns - *p values from Kruskal-Wallis tests are denoted as ns if not significant.

Table 3.7. Multiple comparison of btgw forage attacks per flight (Mann-Whitney test

p<0.05) for differences among tree species, 1999 moderately managed grid.

ambe bafi suma yebi

ambe - - - -

bafi (ambe>bafi) - - -

suma (ambe>suma) ns* - -

yebi (ambe>yebi) ns ns - *p values from Kruskal-Wallis tests are denoted as ns if not significant.

Table 3.8. Multiple comparison of btgw forage attacks per flight (Mann-Whitney test

p<0.05) for differences among tree species, 1996 moderately managed grid.

ambe bafi suma yebi

ambe - - - -

bafi (ambe>bafi) - - -

suma (ambe>suma) (bafi>suma) - -

yebi ns* ns (suma>yebi) -

*p values from Kruskal-Wallis tests are denoted as ns if not significant.

61

Table 3.9. Number of attack manoeuvres by tree species and bird species, all years

combined, moderately managed grid only.

Amre Lefl Btbw Btgw

gln1 hwk2 hov3 gln hwk hov gln hwk hov gln hwk hov

ambe 12 8 31 11 66 119 34 4 67 28 8 37

bafi 2 5 12 5 18 20 17 3 14 20 2 26

suma 0 0 1 1 0 0 8 5 6 5 0 1

yebi 0 2 4 0 0 0 9 2 8 16 2 2

1glean

2hawk

3hover

Table 3.10. Goodness of fit analysis for forage manoeuvres among tree species, chi-

square p < 0.05.

Bird species Glean Hawk Hover

btbw ns ns *

btgw * ns *

amre * ns *

lefl ns * *

* ns = not significant.

62

Figure 3.1. Least Flycatcher and black-throated blue warbler observed foraging use and

expected foraging use based upon availability of each of the four focal tree species in the

intensively managed grid. Significant difference between use and availability (chi-square

p<0.05) after adjustment of alpha for multiple comparisons indicated by (*).

Least Flycatcher IM-Grid 1999

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ambe bafi suma* yebi

Tree Species

obs.

exp.

Least Flycatcher IM-Grid 1997

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ambe bafi suma* yebi

Tree Species

obs.

exp.

Black-throated Blue Warbler

IM-Grid 1997

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ambe bafi suma yebi*

Tree Species

obs.

exp.

Prop

ort

ion

s of

Ob

served

an

d E

xp

ecte

d F

ora

ge A

ttack

s

Att

ack

s

Black-throated Blue Warbler

IM-Grid 1999

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ambe bafi suma* yebi

Tree Species

obs.

exp.

63

Black-throated Green Warbler

IM-Grid 1997

00.10.20.30.40.50.60.70.80.9

1

ambe bafi suma* yebi

Tree Species

obs.

exp.

Figure 3.2. Black-throated green warbler observed foraging use and expected foraging

use based upon availability of each of the four focal tree species in the intensively

managed grid. Significant difference between use and availability (chi-square p<0.05)

after adjustment of alpha for multiple comparisons indicated by (*).

Prop

ort

ion

s of

Ob

served

an

d

Exp

ect

ed

Forage A

ttack

s

64

American Redstart MM-Grid 1996

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma* yebi*

Tree Species

obs.

exp.

American Redstart MM-Grid 1997

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma* yebi

Tree Species

obs.

exp.

American Redstart MM-Grid 1998

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ambe* bafi* suma yebi*

Tree Species

obs.

exp.

American Redstart MM-Grid 1999

00.10.20.30.40.5

0.60.70.80.9

1

ambe* bafi* suma* yebi*

Tree Species

obs.

exp.

Figure 3.3. American redstart observed foraging use and expected foraging use based upon

availability of each of the four focal tree species in the moderately managed grid. Significant

difference between use and availability (chi-square p<0.05) after adjustment of alpha for

multiple comparisons indicated by (*).

Prop

ort

ion

s of

Ob

served

an

d E

xp

ecte

d F

ora

ge A

ttack

s

65

Least Flycatcher MM-Grid 1998

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma yebi

Tree Species

obs.

exp.

Least Flycatcher MM-Grid 1999

00.10.20.30.40.50.60.70.80.9

1

ambe bafi* suma* yebi*

Tree Species

obs.

exp.

Figure 3.4. Least flycatcher observed foraging use and expected foraging use based

upon availability of each of the four focal tree species in the moderately managed grid.

Significant difference between use and availability (chi-square p<0.05) after adjustment

of alpha for multiple comparisons indicated by (*).

Pro

port

ion

s of

Ob

serv

ed a

nd

Exp

ecte

d F

ora

ge

Att

ack

s

Least Flycatcher MM-Grid 1996

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma* yebi*

Tree Species

obs.

exp.

Least Flycatcher MM-Grid 1997

00.1

0.20.3

0.40.50.6

0.70.8

0.91

ambe* bafi* suma yebi*

Tree Species

obs.

exp.

66

Black-throated Blue Warbler

MM-Grid 1996

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma* yebi*

Tree Species

obs.

exp.

Black-throated Blue Warbler

MM-Grid 1997

00.10.20.30.40.50.60.70.80.9

1

ambe bafi suma* yebi

Tree Species

obs.

exp.

Black-throated Blue Warbler

MM-Grid 1998

0

0.2

0.4

0.6

0.8

1

ambe* bafi suma* yebi*

Tree Species

obs.

exp.

Figure 3.5. Black-throated blue warbler observed foraging use and expected foraging

use based upon availability of each of the four focal tree species in the moderately

managed grid. Significant difference between use and availability (chi-square p<0.05)

after adjustment of alpha for multiple comparisons indicated by (*).

Pro

port

ion

s of

Ob

serv

ed a

nd

Exp

ecte

d F

ora

ge

Att

ack

s

67

Black-throated Green Warbler

MM-Grid 1996

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi suma yebi*

Tree Species

obs.

exp.

Black-throated Green Warbler

MM-Grid 1997

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi suma* yebi

Tree Species

obs.

exp.

Black-throated Green Warbler

MM-Grid 1998

00.10.20.30.40.50.60.70.80.9

1

ambe bafi* suma yebi

Tree Species

obs.

exp.

Black-throated Green Warbler

MM-Grid 1999

00.10.20.30.40.50.60.70.80.9

1

ambe* bafi* suma yebi*

Tree Species

obs.

exp.

Figure 3.6. Black-throated green warbler observed foraging use and expected foraging

use based upon availability of each of the four focal tree species in the moderately

managed grid. Significant difference between use and availability (chi-square p<0.05)

after adjustment of alpha for multiple comparisons indicated by (*).

Pro

port

ion

s of

Ob

serv

ed a

nd

Exp

ecte

d F

ora

ge

Att

ack

s

68

Hover Manoeuvre

0

20

40

60

80

100

120

140

ambe bafi* suma yebi*

Tree Species

Nu

mb

er

of

Ho

ve

rs

obs

exp

Glean Manoeuvre

0

5

10

15

20

25

30

ambe bafi* suma yebi

Tree Species

Nu

mb

er

of

Gle

an

s

obs

exp

Figure 3.7. Forage manoeuvres performed in each tree species by black-throated green

warbler all years combined; *signifies a significant difference between hovers and

availability of the tree species (chi-square p<0.05) after adjustment of alpha for multiple

comparisons.

69

Hover Manoeuvre

0

20

40

60

80

100

120

140

ambe* bafi suma* yebi

Tree Species

Nu

mb

er

of

Ho

vers

obs

exp

Figure 3.8. Forage manoeuvres performed in each tree species by black-throated blue

warbler all years combined; *signifies a significant difference between hovers and

availability of the tree species (chi-square p<0.05) after adjustment of alpha for multiple

comparisons.

70

Hover Manoeuvre

0

20

40

60

80

100

120

140

ambe* bafi suma yebi*

Tree Species

Nu

mb

er

of

Ho

vers

obs

exp

Glean Manoeuvre

0

2

4

6

8

10

12

14

ambe* bafi suma yebi

Tree Species

Nu

mb

er

of

Gle

an

s

obs

exp

Figure 3.9. Forage manoeuvres performed in each tree species by American redstart all

years combined; *signifies a significant difference between hovers and availability of the

tree species (chi-square p<0.05) after adjustment of alpha for multiple comparisons.

71

Hover Manoeuvre

0

20

40

60

80

100

120

140

ambe* bafi suma* yebi

Tree Species

Nu

mb

er

of

Ho

ve

rs

obs

exp

Hawk Manoeuvre

0

10

20

30

40

50

60

70

ambe* bafi suma* yebi

Tree Species

Nu

mb

er

of

Haw

ks

obs

exp

Figure 3.10. Forage manoeuvres performed in each tree species by least flycatcher all

years combined; *signifies a significant difference between hovers and availability of the

tree species (chi-square p<0.05) after adjustment of alpha for multiple comparisons.

72

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

aran

cole

hym

edipt

hem

i

hom

oth

ystri

ch lepi

Arthropod Order

Pro

po

rtio

nal

Occu

rren

ce

btbw

btgw

Figure 3.11. Proportional occurrence of arthropod orders in regurgitated stomach

samples, N=19 for black-throated blue warbler and N=25 for black-throated green

warbler.

73

Chapter 4 : Arthropod Prey Density and Diversity Results

I limited my analyses to the 4 most common arthropod orders in the stomach

samples (Figure 3.11), and lepidoptera. These four orders, araneida, coleoptera,

hymenoptera and diptera were the most common orders in the arboreal clip samples as

well. Other arthropod orders found in the stomach contents (Figure 3.11) occurred

infrequently.

Basic summary statistics from the densities calculated per unit of leaf area may be

found in Tables 3.1-3.4 for both grids over the two years that I gathered arthropod data.

As the im grid was not sampled in 1998, there are no results reported for this year. The

tables report the sum, mean, and standard error of the mean for the raw number of

arthropods from each taxonomic order collected in all clip samples from all species of

tree in each year. Table 1.1 reports the basal area of each tree species on the mm-grid

and im-grid. As this table indicates, American beech was almost absent from the im-grid.

74

I analyzed a subset of the arthropod orders collected from the trees. This subset

consisted of the five most common orders of arthropods found in the data pooled from

the stomach samples of black-throated green warbler and black-throated blue warbler.

Figures 3.1-3.20 show the results of those analyses of variance where a significant

(p<0.05) difference was detected among tree species for the arthropod parameters

calculated based upon density estimates corrected for units of leaf area and based upon

density estimates per clip sample. Based upon these analyses of variance it was far more

likely that a significant difference would be demonstrated using the arthropod densities

corrected for leaf area. Arthropod densities per clip sample were more likely to be

statistically similar.

Yellow birch usually had higher densities of all arthropod prey combined, higher

densities of many individual orders, and higher overall taxonomic richness (Tables 4.1-

4.5). Table 4.4 reports the results of post-hoc Bonferonni tests of the multiple

comparisons made based upon total unit area of leaves sampled. Table 4.5 reports the

results of post-hoc Bonferonni tests of the multiple comparisons made based upon the

clip sample. The dominant pattern observed with both methods of calculating density

was a higher density of most arthropod orders in yellow birch. This pattern prevailed

regardless of year, period within season, vertical zone within a tree, and managed versus

moderately managed grid. Overall taxonomic richness was usually higher in yellow

birch as well.

75

Table 4.1. Total number, mean (per cm2), and standard error (SE) of occurrence of

arthropod orders for moderately managed grid 1998.

American Beech

order N mean standard error

araneida 77 2 0.23

coleoptera 22 1.6 0.29

hymenoptera 21 1.3 0.15

diptera 12 1 0

lepidoptera 6 2 0.56

Balsam Fir

order N mean standard error

araneida 121 2.1 0.2

coleoptera 10 1.1 0.11

hymenoptera 49 1.7 0.22

diptera 17 1.31 0.13

lepidoptera 39 3.54 1.76

Sugar Maple

order N mean standard error

araneida 83 1.93 0.17

coleoptera 11 1 0

hymenoptera 28 1.4 0.2

diptera 24 1.2 0.09

lepidoptera 18 2 1

Yellow Birch

order N mean standard error

araneida 108 2.20 0.24

coleoptera 126 2.93 0.40

hymenoptera 37 1.76 0.30

diptera 30 1.58 0.28

lepidoptera 13 1.44 0.34

76

Table 4.2. Number, mean (per cm2), and standard error (SE) of occurrence of arthropod

orders for moderately managed grid 1999.

American Beech

order N mean standard error

araneida 12 0.38 0.09

coleoptera 9 0.28 0.11

hymenoptera 16 0.50 0.19

diptera 10 0.31 0.11

lepidoptera 5 0.16 0.09

Balsam Fir

order N mean standard error

araneida 22 0.65 0.16

coleoptera 13 0.38 0.10

hymenoptera 27 0.79 0.23

diptera 7 0.21 0.08

lepidoptera 4 0.12 0.06

Sugar Maple

order N mean standard error

araneida 20 0.63 0.15

coleoptera 7 0.22 0.09

hymenoptera 13 0.41 0.12

diptera 6 0.19 0.08

lepidoptera 6 0.19 0.08

Yellow Birch

order N mean standard error

araneida 10 0.29 0.09

coleoptera 116 3.41 0.62

hymenoptera 4 0.12 0.07

diptera 10 0.29 0.10

lepidoptera 3 0.09 0.05

77

Table 4.3. Number, mean (per cm2), and standard error (SE) of occurrence of arthropod

orders for intensively managed grid 1999.

American Beech

order N mean standard error

araneida 0 0 na

coleoptera 1 1 na

hymenoptera 0 0 na

diptera 1 1 na

lepidoptera 0 0 na

Balsam Fir

order N mean standard error

araneida 25 0.81 0.14

coleoptera 11 0.36 0.12

hymenoptera 16 0.52 0.13

diptera 5 0.16 0.07

lepidoptera 4 0.13 0.06

Sugar Maple

order N mean standard error

araneida 33 0.89 0.17

coleoptera 19 0.51 0.14

hymenoptera 11 0.30 0.10

diptera 6 0.16 0.06

lepidoptera 8 0.22 0.09

Yellow Birch

order N mean standard error

araneida 37 0.97 0.51

coleoptera 253 6.66 1.10

hymenoptera 4 0.11 0.05

diptera 8 0.21 0.07

lepidoptera 5 0.13 0.06

78

Table 4.4. Significant (p<0.05) Bonferonni post-hoc pair-wise comparisons of arthropod

variables, data based on number per unit leaf area.

arthropod variable grid year comparison

rich1

mm 1998 yebi>ambe

rich mm 1998 yebi>bafi

rich mm 1998 yebi>suma

dens2

mm 1998 yebi>ambe

dens mm 1998 yebi>bafi

dens mm 1998 yebi>suma

aran3

mm 1998 yebi>ambe

aran mm 1998 yebi>bafi

aran mm 1998 yebi>suma

hyme4

mm 1998 yebi>ambe

hyme mm 1998 yebi>suma

cole5

mm 1998 yebi>ambe

cole mm 1998 yebi>bafi

cole mm 1998 yebi>suma

dens im 1999 yebi>suma

dens im 1999 bafi>yebi

cole im 1999 yebi>suma

cole im 1999 yebi>bafi

dens mm 1999 yebi>ambe

dens mm 1999 suma>yebi

dens mm 1999 bafi>yebi

aran mm 1999 bafi>ambe

aran mm 1999 bafi>suma

aran mm 1999 bafi>yebi

cole mm 1999 yebi>bafi

1order richness per unit area,

2density of all arthropods per unit area,

3density of Araneida

per unit area, 4density Hymenoptera per unit area,

5density of coleptera per unit area.

79

Table 4.5. Significant (p<0.05) Bonferonni post-hoc pair-wise comparisons of arthropod

variables, data based on total clip sample.

arthropod variable grid year comparison

cole

mm 1998 yebi>suma

rich1

mm 1998 yebi>ambe

dens2

mm 1998 yebi>ambe

dens mm 1998 yebi>suma

cole3 im 1999 yebi>suma

cole im 1999 yebi>bafi

dens im 1999 yebi>bafi

cole mm 1999 yebi>bafi

dens

mm 1999 yebi>ambe

dens mm 1999 yebi>suma

dens mm 1999 yebi>bafi

1order richness per clip sample,

2density all arthropods combined per clip sample,

3density of Coleoptera per clip sample.

80

Table 4.6. Mean leaf number, mean leaf area, and standard errors by tree species.

Tree Species Mean leaf

number per

clip sample

SE leaf

number

Mean area

per leaf

(cm2)

SE leaf

area

Mean leaf

area per

clip sample

cm2

ambe 62 5.1 52 3.28 3224

suma 52 6.1 55 3.57 2860

yebi 50 6.8 31 1.90 1550

bafi* 7500 1039 0.27 0.02 2025

* A mean number of needles per cm of branchlet (se = 1.05) was used to calculate the

mean number of needles per clip sample.

81

Araneida Density per Clip R-Grid 1999

ambe bafi suma yebi

Tree Species

0

1

2

3

4

5A

raneid

a / C

lip

Figure 4.1. Araneida density per branch and leaf clip sample by tree species. Moderately

managed grid 1999, anova p<0.05.

Coleoptera Density per Clip R-Grid 1999

ambe bafi suma yebi

Tree Species

0

1

2

3

4

5

Co

leo

pte

ra /

Clip

Figure 4.2. Coleoptera density per branch and leaf clip sample by tree species.

Moderately managed grid 1999, anova p<0.05.

82

Arthropod Density per Clip R-Grid 1999

ambe bafi suma yebi

Tree Species

0

1

2

3

4

5A

rth

rop

od

s /

Clip

Figure 4.3. Total arthropod density per branch and leaf clip sample by tree species.

Moderately managed grid 1999, anova p<0.05.

Araneida Density per square cm R-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010

Ara

neid

a / s

quare

cm

Figure 4.4. Araneida density per square centimetre of leaf area by tree species.

Moderately managed grid 1999, anova p<0.05.

83

Coleoptera Density per square cm R-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010C

ole

op

tera

/ s

qu

are

cm

Figure 4.5. Coleoptera density per square centimetre of leaf area by tree species.

Moderately managed grid 1999, anova p<0.05.

Arthropod Density per square cm R-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010

Art

hro

pod

s /

sq

ua

re c

m

Figure 4.6. Total arthropod density per square centimetre of leaf area by tree species.

Moderately managed grid 1999, anova p<0.05.

84

Coleoptera Density per Clip M-Grid 1999

ambe bafi suma yebi

Tree Species

0

2

4

6

8

10

Co

leo

pte

ra /

Clip

Figure 4.7. Coleoptera density per branch and leaf clip sample by tree species.

Intensively managed grid 1999, anova p<0.05.

Arthropod Density per Clip M-Grid 1999

ambe bafi suma yebi

Tree Species

0

2

4

6

8

10

Art

horp

od

s /

Clip

Figure 4.8. Total arthropod density per branch and leaf clip sample by tree species.

Intensively managed grid 1999, anova p<0.05.

85

Coleoptera Density per square cm M-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010

Co

leo

pte

ra /

sq

ua

re c

m

Figure 4.9. Coleoptera density per square centimetre of leaf area by tree species.

Intensively managed grid 1999, anova p<0.05.

Hymenoptera Density per square cm M-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010

Hym

enopte

ra / s

quare

cm

Figure 4.10. Hymenoptera density per square centimetre of leaf area by tree species.

Intensively managed grid 1999, anova p<0.05.

86

Arthropod Density per square cm M-Grid 1999

ambe bafi suma yebi

Tree Species

0.0000

0.0003

0.0007

0.0010

Art

hro

pod

s /

sq

ua

re c

m

Figure 4.11. Total arthropod density per square centimetre of leaf area by tree species.

Intensively managed grid 1999, anova p<0.05.

Coleoptera Density per Clip R-Grid 1998

ambe bafi suma yebi

Tree Species

0

2

4

6

8

10

Co

leo

pte

ra /

Clip

Figure 4.12. Coleoptera density per branch and leaf clip sample by tree species.

Moderately managed grid 1998, anova p<0.05.

87

Order Richness per Clip R-Grid 1998

ambe bafi suma yebi

Tree Species

0

2

4

6

8

10O

rde

r R

ich

ne

ss /

Clip

Figure 4.13. Order richness per branch and leaf clip sample by tree species. Moderately

managed grid 1998, anova p<0.05.

Arthropod Density per Clip R-Grid 1998

ambe bafi suma yebi

Tree Species

0

2

4

6

8

10

Art

hro

pod

s /

Clip

Figure 4.14. Total arthropod density per branch and leaf clip sample by tree species.

Moderately managed grid 1998, anova p<0.05.

88

Araneida Density per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Ara

ne

ida

/ s

qu

are

cm

Figure 4.15. Araneida density per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

Coleoptera Density per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Cole

opte

ra / s

quare

cm

Figure 4.16. Coleoptera density per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

89

Hymenoptera Density per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Hym

en

op

tera

/ s

qu

are

cm

Figure 4.17. Hymenoptera density per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

Diptera Density per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Dip

tera

/ s

quare

cm

Figure 4.18. Diptera density per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

90

Order Richness per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Ord

er

Ric

hn

ess /

squ

are

cm

Figure 4.19. Order richness per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

Arthropod Density per square cm R-Grid 1998

ambe bafi suma yebi

Tree Species

0.0000

0.0017

0.0033

0.0050

Art

hro

pods / s

quare

cm

Figure 4.20. Total arthropod density per square centimetre of leaf area by tree species.

Moderately managed grid 1998, anova p<0.05.

91

Chapter 5 Landscape Distribution and Productivity Results

Presence of three focal species was significantly and positively associated with

the basal area of a deciduous tree species in a logistic regression (p<0.05) (Table 1).

Black-throated green warbler presence was associated with basal area of American beech.

Black-throated blue warbler presence was associated with basal area of yellow birch.

American redstart presence was associated with basal area of sugar maple. Least

flycatcher showed no association between presence and the basal area of a tree species.

Two non-focal species, magnolia warbler and northern parula, demonstrated a significant

positive association between their presence and basal area of American beech (p<0.05,

Table 5.1).

Gunn (2002) measured reproductive success of some of these bird species on the

same study area using an index he developed. He based this index on observations of

indicators of breeding activity. These indicators included paired adults, adults carrying

nesting material or food, and the presence of fledglings. He related reproductive success

to basal area of various tree species using Canonical Correspondence Analysis (CCA).

He found black-throated green warbler, black-throated blue warbler, and red-eyed vireo

reproductive success associated with above average basal area of American beech. The

only other species reported in Table 5.1 that showed an association between reproductive

success and any environmental variable were American redstart and northern parula.

Reproductive success for these two species was associated with above average basal area

of softwood understory.

92

Black-throated green warblers exhibited the strongest foraging associations with

American beech. This led to my prediction that if any bird species were to exhibit a

positive relationship between beech basal area and its own presence or beech basal area

and reproductive success, it would be black-throated green warbler. This was indeed the

case with this species being the only focal bird to exhibit a significant (p<0.05) result.

Following this through to the data of Gunn (2002), black-throated green warbler emerges

as one of the species showing reproductive success associated with above average beech

basal area. These lines of evidence suggest a relationship between foraging behaviour,

landscape distribution, and reproductive success.

Black-throated blue warbler also exhibited strong foraging associations with

American beech. While its presence was not associated with beech basal area,

reproductive success data (Gunn 2002) demonstrated an association with higher than

average beech basal area. A variety of explanations might apply here, but a particularly

interesting consideration is the existence of a dominance hierarchy. Morse (1989)

suggests that of northern hardwood bird species, black-throated green warbler holds the

dominant position in the hierarchy. Such a dynamic might explain the random

relationship between black-throated blue warbler presence and beech basal area in spite

of selective use of beech and higher reproductive success associated with higher basal

area of beech. Their settlement and presence on these sites might be inhibited by the

black-throated green warbler.

93

An alternative explanation might stem from the more generalized habitat

requirements of black throated blue warblers (Sherry and Holmes 1985). They have

observed that black-throated blue warblers have a less patchy distribution than black-

throated green warblers due to less specific habitat requirements. This might lead to

random settlement by black-throated blue warblers with respect to beech distribution.

Such a priori selection would not preclude a possible advantage of American beech with

respect to foraging success and reproductive success should adequate beech be available.

The results are more ambiguous for American redstart. Chapter 2 demonstrates

some selective foraging in American beech by this species depending upon the year.

This species also either avoids sugar maple as a foraging site or selection is neutral. Its

presence is associated with sugar maple basal area, and its reproductive success is

associated with basal area of softwood understory. Sherry and Holmes (1985) found that

this species, like the black-throated blue warbler, has less specific habitat requirements

and a less patchy distribution than black-throated green warblers.

It was not surprising that least flycatcher did not exhibit an association between

the abundance of any tree species and either its own presence at a point or reproductive

success. I included this species in my focal group because its primary foraging strategy

(mid-air hawking) was the least likely to exhibit a dependence on a particular species of

tree, thus providing a contrast to the foliage dependent foragers.

94

Two non-focal species, northern parula and magnolia warbler, demonstrated an

association (p<0.05) between their presence at a point and the basal area of beech. Gunn

(2002) found that northern parula reproductive success was associated with softwood

understory rather than beech basal area. While it might have selected territories based

upon beech abundance it is possible that a threshold value was reached and not limiting

to productivity. There were no other significant relationships between the presence of

any bird species and the abundance of any tree species at a point.

In addition to black-throated green and black-throated blue warbler reproductive

success being related to beech abundance, Gunn (2001) found the same to be true for red-

eyed vireo. This species did not show an association between presence at a point and

beech abundance. It did demonstrate a tendency to forage in American beech in the one

year that I recorded incidental observations for this non-focal species. Eight of ten

foraging observations made on this species were in beech. With respect to presence at a

point and productivity, this species represents a relationship that is opposite to northern

parula. Although red-eyed vireo might not select a point with higher abundance of beech,

it does show higher reproductive success in association with this species.

The lines of evidence from my work and the work of Gunn (2002) are particularly

striking for black-throated green warbler. Black-throated green warbler forage

selectively (p<0.05) in American beech and was the only focal bird species to

demonstrate a significant (p<0.05) relationship between its presence and beech

abundance. Black-throated green warbler also showed a significant association between

reproductive success and beech abundance in a Canonical Correspondence Analysis

(Gunn 2002).

95

Previous findings that black-throated green warblers are more patchily distributed

due possibly to more specific habitat requirements (Holmes and Sherry 1985) including

tree species selection (Morse 1976) are interesting in light of my own results. It would

have been predicted that the least plastic species with respect to foraging requirements

should have exhibited the stronger association with a key resource in its habitat selection,

foraging behaviour, and reproductive success. As a socially dominant warbler (Morse

1989), this species should have had the ability to exercise this site specificity over birds

such as black-throated blue warbler. As noted previously, black-throated blue warblers

selected beech for foraging, enjoyed higher reproductive success in areas with higher

abundance of beech, yet did not settle in areas with high abundance of beech, perhaps due

to displacement.

The evidence for a behavioural process influencing landscape pattern is

compelling, but does result from a-posteriori relation of temporally and spatially parallel

studies. The next step would be a study designed to test landscape patterns of settlement

and productivity in terms of selective foraging behaviour among the same individual

birds using the predictions proposed in this synthesis.

96

Table 5.1. Logistic regression of bird presence on tree basal area, significant values

(p<0.05) in bold, significant values for focal bird species in bold and italics.

Bird Species American beech sugar maple yellow birch

amre 0.61 0.01 0.82

mnwa 0.03 0.71 0.17

scta 0.60 0.60 0.60

yrwa 0.20 0.74 0.78

btbw 0.63 0.62 0.04

revi 0.69 0.74 0.57

bhvi 0.56 0.68 0.55

btgw 0.04 0.85 0.70

cswa 0.65 0.06 0.22

nopa 0.02 0.35 0.60

phvi

lefl

0.69

0.88

0.29

0.36

0.93

0.16

97

Chapter 6: Scientific Conclusions and Management Implications

Foraging Site Selection and Success

Tree species use by four species of foraging songbirds in a northern

temperate mixed-wood forest was observed over four consecutive breeding

seasons. American beech on the mm-grid was selected out of proportion to its

availability by black-throated green warbler, black-throated blue warbler,

American redstart, and least flycatcher, in 11 of the possible 16 cases involving

these 4 bird species over 4 years (Figures 3.3-3.6). Yellow birch, sugar maple,

and balsam fir, were selectively avoided by these same birds in 9, 8, and 10 cases

respectively (Figures 3.1-3.6).

Rates of foraging differed significantly among tree species in some years for the

black-throated blue and black-throated green warblers, but the results were less consistent

than the selective use results (Tables 3.4-3.8). Attacks per unit time is a straight forward

measure of rate in temporal units. Attacks per flight is a spatial measure, as these are

attacks performed within a foraging patch (a unit of branch and foliage) following the

arrival flight and prior to the departure flight. The more conservative post-hoc Nemenyi

test shows that the black-throated green warbler was the only bird species in all years to

demonstrate differential foraging success (Tables 3.4 –3.5). This forage rate was in terms

of units of time only. The less conservative multiple application of post-hoc Mann-

Whitney-U tests (Sheskin 1997) shows significant differences in foraging rate measured

in terms of time for black-throated green warbler in 1999 and in terms of flight numbers

in 1999 and 1996. The general trend for these post-hoc tests shows higher foraging rates

in beech. Black-throated blue warbler is the only other species to demonstrate any

98

difference in a foraging rate measure among tree species. In 1996, attacks per minute

were higher in beech than maple using the less conservative Mann-Whitney post-hoc test

for multiple comparisons. Holmes and Robinson (1981) found no difference in foraging

rates for black-throated blue warbler or American redstart. They did not present results

for black-throated green warbler.

Beech basal area was very low due to its complete absence at most grid

points on the im-grid and consequently is unlikely to be available for selection.

Maple seemed to replace beech as the selected tree, and birch was still either

avoided or selection was neutral (Figures 3.1-3.2) . Least flycatcher selected

maple in both years that I conducted work in the intensively managed grid, 1997

and 1999. Black-throated blue warbler selected maple in 1999 and black-throated

green selected maple in 1997. There were no significant differences among tree

species for either measure of foraging rate for any of the bird species on the im-

grid.

In spite of general patterns of selective use of beech, there was some variation

among years that would not have been captured in an analysis that pooled years. The

black-throated blue warbler, American redstart, and least flycatcher broke this pattern in

1999 on the mm grid. Black-throated blue warbler neither selected nor avoided

American beech. American redstart selected yellow birch and avoided American beech,

and least flycatcher selected sugar maple and neither selected nor avoided American

beech (Figures 3.3-3.4). Black-throated green warbler broke the general pattern of

American beech selection in 1998, neither selecting nor avoiding this species (Figure

3.6). In summary, black-throated green warbler departed from the use patterns seen in

99

the other three bird species in two different years. The Black-throated green warbler was

selecting American beech in 1999 when the other three bird species were avoiding or

randomly using this species. The black-throated green warbler was using American

beech randomly in 1998 when the other three species were selectively using this species.

The latter pattern is especially interesting in light of the arthropod data and the

following observations. Balsam fir was selectively used in only one year (1998) by one

bird species (black-throated green warbler) considering both the im and mm-grids (Figure

3.6). This species has the strongest association with coniferous trees (Morse 1993),

therefore, it is not surprising that of all focal species this would demonstrate a positive

association with a balsam fir. While selective use of balsam fir in one year of four not

strong support for this pattern, one interesting conjecture is there might be a structural

similarity shared by beech and fir or a behavioural trait on the part of the black-throated

green warbler’s foraging repertoire more suited to the morphology of these two tree

species.

Another conjecture is related to the overall lower densities of arthropods on the

mm-grid in 1999 versus 1998 (Tables 4.1-4.2). Perhaps it is only in years of low

arthropod density that there is some advantage to foraging in American beech.

Unfortunately, I do not have arthropod data from 1996 and 1997 to compare with the

typical pattern of selective foraging in American beech. My post-hoc prediction would

be that arthropod densities were atypically high (considering the four year span of my

work) in 1998 and that densities of 1999 would be comparable to 1996 and 1997.

Araneida is unique in several aspects. It is the most abundant arthropod overall in

tree foliage (Tables 4.1-4.5), it is the most abundant arthropod in stomach samples

100

(Figure 3.11), and it is a predator of many other arthropod prey. It is also the only

arthropod to vary significantly in its relative density between American beech and balsam

fir over the two years. Araneida density in terms of units of foliage area was higher in

balsam fir than in American beech, sugar maple, and yellow birch (p<0.05) in 1999

(Table 4.4). Araneida density was equal in American beech and balsam fir in 1998 and

higher (p<0.05) in yellow birch compared with sugar maple, balsam fir, and American

beech (Table 4.4).

The black-throated green warbler selected balsam fir as a foraging site in a year

(1988) when overall arthropod availability was higher across all tree species and araneida

density in balsam fir was lower relative to all other tree species. One explanation for the

avian foraging pattern might be related to easier location of arthropod prey by the

foraging bird in an environment free of a competing predator, araneida. Lepidopteran

larvae were more abundant in balsam fir than in the other tree species in 1999 on the mm-

grid (Table 4.1), though this difference was not statistically significant (p>0.05). This is

worth consideration given its know importance in the diet of insectivorous passerines

(Morse 1993, Homes 1994, Sherry and Holmes 1997).

I propose a general model with three categories of tree species utility for foraging

birds. At one extreme is American beech is most often selected for foraging by 4 species

of forest passerine. One of these passerines, black-throated green warbler, is particularly

sensitive to beech showing higher foraging success in this tree species. The next

category of foraging utility is filled by sugar maple. When beech is not available, maple

is selected out of proportion to its abundance, but foraging success does not differ at any

time in the maple versus the other tree species. It is still possible that there is an

101

advantage in maple in terms of energy expended even in the absence of higher prey

yields. The third category of utility is filled by birch and fir, both species that are rarely

selected as a foraging site even in the intensively managed grid where beech is not

readily available.

102

Evaluation of the Arthropod Hypothesis

All arthropod orders preyed upon by the focal songbirds are found in the clip

samples obtained from the trees. This allays the concern that a critical prey item might

not be sampled from one of the tree species thus leading to erroneous conclusions about

selective foraging causality. While probability of detection among stomach contents may

vary among arthropod orders, it is less likely that exoskeletal remains of one particular

order would consistently have been missed in 42 stomach samples.

These results help ensure that arboreal arthropod sampling did not exclude

a component of the warblers’ diet. The potential does exist that a critical prey

species is not represented in the arthropod community sample as both the prey

data and community data are identified to order. Wolda (1990) states that

taxonomic pooling of the arthropod prey base is justified if those groups pooled

are not likely to exhibit physical or behavioural differences from the perspective

of the foraging bird. Holmes and Robinson (1981) performed similar work on

some of the same bird and tree species in New Hampshire and found

identification of the arthropod prey base to taxonomic order adequate.

103

I found higher densities of several orders of arthropod and higher order richness

in yellow birch compared with American beech and sugar maple (Table 4.4). Araneida

densities ranged from a mean of 2.2 per cm2

in yellow birch to 2.0 per cm2 in American

beech, and 1.93 per cm2 in sugar maple. These density results are based on arthropods

per total unit area of leaf sampled in the tree. I also calculated densities based upon the

arthropods found in the clip samples, and found fewer cases of significant differences

among tree species (Table 4.5).

I am confident that I can reject the arthropod hypothesis for selective foraging

observed in American beech. Densities and diversities of arthropod prey are consistently

higher on yellow birch using either method (per unit of leaf area or foliage volume) of

calculation. In spite of this, yellow birch was usually avoided by foraging songbirds.

Lepidopteran larvae are notably absent from the stomach contents data and the

arthropod community data as well. I would be more concerned with my methods if one

of these components of my work had demonstrated a significant presence of this

important insectivorous passerine prey item (Morse 1993, Homes 1994, Sherry and

Holmes 1997). Quite to the contrary, I interpret this as a potentially significant result

with possible explanatory power in relation to regional differences (discussed below) in

foraging patterns.

104

Evaluation of the Morphology Hypothesis

This research introduces the first comprehensive study of selective foraging in

deciduous tree species by insectivorous birds that is able to isolate the two dominant

explanatory hypotheses from one another. The American beech had lower abundances of

all arthropod prey found in the focal birds’ diets, while yellow birch had higher

abundances of these prey items. Yellow birch also had the higher arthropod richness and

total arthropod abundance outside of primary prey items than American beech and sugar

maple. I reject the arthropod hypothesis in those cases where birds foraged selectively or

at higher rates of success in American beech and sugar maple, while avoiding yellow

birch.

Previous work has sought the causal process responsible for selective foraging

behaviour by birds in particular species of trees, and has found mixed results (see review

in Chapter 1). Holmes and Robinson (1981) present evidence for both the arthropod and

tree morphology hypotheses. While their evidence supports the arthropod hypothesis, the

tree morphology hypothesis is not excluded. In contrast, my own results do exclude one

of the hypotheses (the arthropod hypothesis), which leads me to invoke the tree

morphology hypothesis for explanation of the primary foraging patterns that I observed.

Rejection of this primary hypothesis has led me to consider the morphology

hypothesis alternative hypothesis involving leaf and branch morphology and foraging

behaviour since beech and birch do have contrasting morphologies (Horn 1971).

105

The morphology hypothesis is based in part upon a description of northern

temperate tree morphology given by Horn (1971). He determined that the genus Betula

(including yellow birch) differs from American beech and sugar maple. Birch has

branches arranged in multiple layers, while beech and maple branches are arranged in a

monolayer. Put in other terms, the foliage of beech and maple is condensed in a plume of

leaves in one layer of branches, and the foliage of birch is arranged in multiple layers on

branches spaced vertically along the tree trunk.

The terms monolayer and multilayer can be misleading if viewed at the wrong scale.

Horn classifies beech and maple as monolayered trees and birch as a multilayered tree.

This is correct when speaking of branches, but it is not an appropriate classification for

leaf arrangement at the within-branch scale. As Plates 6.1-6.8 indicate the

“monolayered” American beech and sugar maple are actually multilayered at the scale of

the branch while the “multilayered” yellow birch is actually mono-layered at this scale.

Holmes and Robinson (1981) further confound the situation in their paper by

incorrectly stating that the leaves, not just the branches, of beech and maple are

monolayered while the leaves of birch are multilayered. These are very important scalar

distinctions, because I contend that the within branch scale is more relevant to the

efficiency of a foraging bird.

Considering the scale of a foraging bird, a mono-layer would functionally be a

multiple layered offering of substrate, while. A mono-layered tree species has its leaves

in several planes within reach of a foraging bird advancing by hops down a branch or

twig

106

Considering the spatial scale of a bird foraging from a branch, the monolayer

actually has multiple layers of leaves due to overlapping growth of twigs and branches

through several planes. Multiple foraging patches of foliage exist in three dimensions

relatively short within-branch flights or hops away. The branches of a multilayered tree

have only a single layer of leaves available to a bird foraging along that branch (see

Plates 6.1-6.8). The multi-layered (at the whole tree scale) yellow birch possesses layers

reachable only through a relatively longer flight to another branch or foraging patch.

The data of Table 4.6 reflect this condition with much higher total surface area of

leaves available along a unit length of branch within the 1 meter search area of a foraging

insectivorous passerine (Robinson and Holmes 1982). Table 4.6 reports the mean

numbers of leaves per unit length of branch sampled with the pole pruners, the mean

surface area per leaf, and the mean leaf area per unit length of branch sampled for each

tree species. Yellow birch had the smallest leaves and the fewest per unit length of

branch sampled. These factors combined to produce a mean leaf area per unit length of

branch sampled that was approximately half of the mean leaf area per unit length of

branch sampled for American beech. American beech leaves were slightly smaller than

sugar maple leaves on average, but the greater number of them per unit length of branch

compensated for this and gave beech the highest overall leaf area per unit length of

branch sampled.

107

Horn (1971) also demonstrated that birch leaves are randomly spaced while maple

and beech leaves are evenly spaced. The vertical separation of branches in a multilayer

tree allows for random spacing of the leaves as light is able to penetrate to each layer at

an angle. Monolayered trees have higher densities of leaves along a given length of

branch than multilayered trees, therefore, leaves are spaced evenly in the dense foliage of

a monolayer to minimize overlap and shading. I propose that even spacing of the

monolayer leaves would provide more a predictable foraging environment than the

random leaf arrangement along the branch of a multilayered tree. Another benefit of the

dense cluster of leaves in a monolayer might be related to the ability of a leaf gleaning

bird to approach prey without causing them to flush prematurely. American beech

foliage might also offer a safer place to forage for the birds, though higher foraging rates

of foraging success in American beech (Tables 3.4-3.8) suggest that foraging in this

species is not a by product of another use of the tree such as a source of refuge or a place

or a place to use for display. An advantage to foraging in American beech could still be

related to avian predator avoidance if the bird is able to devote more energy and time to

foraging, thus resulting in greater foraging efficiency. If this were the case, the

morphology hypothesis would still be supported.

As previously discussed, American beech has a higher volume of foliage in

multiple layers (see Table 4.6, Plates 6.1-6.8, and Horn 1971). A bird capable of

foraging in multiple planes while hopping along a branch would be better able to exploit

the resources in American beech. While the glean manoeuvre is suited for yellow birch,

the addition of the hover manoeuvre would be needed to exploit resources within the

three dimensional foraging patches of American beech. I predicted a priori higher

108

proportions of the hover manoeuvre in American beech than in yellow birch. My focal

birds showed either a lower (p<0.05) use of the hover in yellow birch (where leaves are

located below a bird), or a greater (p<0.05) use of the hover in American beech (where a

greater proportion of leaves are located above a foraging bird rather than below it) (Table

3.10 and Figures 3.7-3.10).

The glean manoeuvre was used to a similar extent across the three broad-leaved

tree species by the two Dendroica species. Black-throated blue warblers and American

redstarts used the hover manoeuvre in greater proportions than expected by chance in

American beech. Black-throated green warblers and American redstarts used the hover

manoeuvre in proportions lower than expected by chance in birch (Table 3.10 and

Figures 3.7-3.10).

It was particularly striking that least flycatcher used a higher proportion of hovers

in American beech. This species uses the hawk manoeuvre more than any of the other

focal birds, yet it not only foraged selectively in American beech (3.4), but it also adjusts

its foraging strategy in the same fashion as do the other focal species. I view this as more

inferential evidence for the morphology hypothesis in explaining selective foraging in

American beech.

I propose that within the uni-planar foraging patch in yellow birch, it is primarily

the topside of the leaves that are available, while within multiple-planar foraging patch on

beech, both sides are available simultaneously at a relatively low cost to the foraging bird

as hovers are short within-branch manoeuvres. This could explain the significant use of

the hover and glean in beech and only the use of the glean in yellow birch . This

explanatory process requires a novel view of what constitutes a patch for these foraging

109

birds and how is this patch best exploited while a bird is searching. Dunning (1990)

urges just such consideration of patch definition in avian foraging studies.

Whelan (2001) found that black-throated green warblers, black-throated blue

warblers, and American redstarts foraged disproportionately on the upper surfaces of

yellow birch. Whelan (2001) also found the black-throated blue warbler and American

redstart foraged disproportionately on the upper surfaces of sugar maple leaves.

Although Whelan did not study American beech, sugar maple foliage structure is more

similar to American beech structure than yellow birch. Whelan’s results, therefore, lend

support to my contention that the yellow birch might be less suited to the hover

manoeuvre as this strategy is best used for leaf surfaces above a bird. His work

represents patterns of use of the tree species consistent with my own hypotheses and

interpretations of my data.

Whelan (2001) speaks in terms similar to my own when he refers to distinct tree

species offering specialized foraging microhabitats that enhance niche diversity. While

some of my hypotheses here clearly need further testing, perhaps in an experimental

design that includes aviary work, I have ample evidence that tree species morphologies

offer different foraging opportunities and are likely the cause for the patterns of selective

use and avoidance that I observed.

110

Holmes and Robinson (1981) and Holmes and Schultz (1988) did not eliminate

the tree morphology hypothesis as a potential explanation. Based upon their

interpretation of Horn’s results, they contend that the multilayered birch would actually

offer more leaves to a bird foraging from any given location on a branch. This

interpretation does not make sense in light of Horn’s description of monolayer and

multilayer forms, therefore, I tested this assertion in the course of gathering our foliage

clip samples. As Plates 6.1-6.8 and Table 4.6 suggests, American beech, not yellow birch

has an optimal morphology for leaf-surface feeding passerines. Interestingly, in this

light, Holmes and Schultz would have actually demonstrated a case for the arthropod

hypothesis alone had their interpretation of the optimal morphologies matched my own.

Based upon their interpretation of the tree morphologies Holmes and Robinson (1981)

and Holmes and Schultz (1988) concluded that for some species of bird foraging in

yellow birch, both hypotheses 1 and 2 are responsible for selective use of that tree

species. It was my goal to isolate a situation where either one or the other but not both

could be invoked as the explanation for selective use of a particular tree species

111

Landscape Pattern and Process

My work in conjunction with the companion study performed by John Gunn

(2002) offers the first evidence for a local behavioural process influencing population

dynamics in a landscape. My work also offers the first evidence for a local behavioural

process explaining landscape patterns of species distribution. American beech was

selected as a foraging site by the focal birds in my own study and the bird with the

strongest association, black-throated green warbler, showed a strong association between

reproductive success and the presence of American beech across the landscape.

American beech was also a strong predictor of the presence of black-throated green

warbler in the landscape.

It is not hard to imagine that such a basic life process as foraging would have such

widespread and profound influence across multiple spatial scales. Our evidence suggests

that behavioural choices made at the organismal level do indeed ripple through multiple

scales of space and time across the landscape. I acknowledge that the evidence is not a

formal test, but rather a suggestive link worthy of further testing.

Discussion of different types of habitat selection, and mention here that this is a

major of apriori selection of landscape based upon beech and that it was by Green which

had the strongest associations with beech.

112

Regional Comparison

My work contrasts with similar work in New Hampshire that showed higher

selective use of yellow birch and higher arthropod densities in yellow birch versus

American beech. Holmes and Schultz (1988) concluded that differential arthropod

density among tree species is responsible for the higher selective use of yellow birch in

New Hampshire. I conclude that it is the morphology of the American beech coupled

with foraging manoeuvres that drives the selective use of the beech. Perhaps there is a

regional difference in the processes responsible for foraging site selection and success.

My work contrasts with the work of Holmes and Robinson (1981) with respect to

avian foraging patterns in American beech and yellow birch in a mature secondary

mixed-wood forest. I consistently found selective use of American beech and selective

avoidance or neutral use of yellow birch over 4 field seasons for the 4 species of bird I

studied in the moderately managed grid. Holmes and Robinson (1981) found the

opposite pattern with strong selective use of yellow birch and avoidance of American

beech. My results for the sugar maple on the moderately managed grid agree with the

results of Holmes and Robinson. This tree species was avoided or selection was neutral.

113

Holmes and Robinson (1981) documented higher densities (14 per 10,000 cm2) of

arthropod prey on yellow birch versus 8 per 10,000 cm2 for sugar maple and 5 per 10,000

cm2 for American beech. They cite this as support for the arthropod hypothesis regarding

the selective foraging they observed in yellow birch. Holmes and Schultz (1988)

evaluated the abundance of a particularly important prey item, Lepidoptera larvae.

Though they documented higher overall densities of larvae on beech, they determined

that the yellow birch had higher densities (approximately 1 per 10,000 cm2) of free living

larvae, the life stage available to leaf gleaning birds. Again this is cited as evidence in

favour of the arthropod hypothesis.

114

The arthropod densities per unit of leaf area of all tree species on my study sites

were over 30 % higher than the densities observed in New Hampshire. An interesting

speculation is that the more northerly latitude of my study site might support higher

densities in a shorter period of time due to the shorter summer period. Perhaps once an

arthropod prey population has exceeded a threshold level across all tree species, the only

way for a foraging bird to increase its capture rate is to maximize foraging efficiency in

relation to tree species leaf and branch morphology. Two previously mentioned

observations within this research might lend credibility to this hypothesis. The apparent

lack of a potentially critical resource, lepidopteran larvae, suggests an environment poor

in arthropod prey. Perhaps the New Brunswick forests were below a threshold value for

this prey item alone leading to a selective mechanism for increased foraging efficiency

through an optimal fit between foraging strategy and tree morphology. A microcosm for

this model might exist within my own work with the selective use of the balsam fir over

American beech by black-throated green warblers in 1988 (Figure 3.6). A higher overall

arthropod abundance in 1988, or conversely, a switch back to an optimal efficiency in

American beech in 1999 might be a demonstration of the morphology hypothesis.

115

If New Hampshire arthropod densities were calculated using leaf volume, the

disparity between the foraging patterns in the two areas would be even more evident.

Calculation of arthropod density using leaf volume within a bird’s search area would

deflate the effective availability of arthropods in American beech relative to yellow birch.

This would strengthen the speculation that there is some regional difference between the

study sites and that the arthropod hypothesis is the best predictor of the patterns in New

Hampshire, while the morphology hypothesis is the best predictor of the patterns in my

study.

Another comparison between the two study sites is the proportion of tree species.

American beech, sugar maple, and yellow birch were the dominant tree species on my

mm-grid and in the Hubbard Brook study area. Yellow birch basal area is, however,

approximately 1/3 of the basal areas of American beech and sugar maple on the mm-grid

in New Brunswick while it occurs in approximately equal proportions as American beech

and sugar maple in New Hampshire. Yellow birch is more abundant than American

beech and sugar maple on the im-grid in New Brunswick, with American beech

practically nonexistent. The proportions found on the mm-grid might suggest that part of

the pattern of yellow birch avoidance could be due to it being below a threshold value.

This is not likely given the relative proportions of yellow birch, American beech, and

sugar maple on the im-grid and the same pattern of yellow birch avoidance by foraging

birds as seen on the mm-grids.

116

Forest Management

I concur with the forest management implications suggested by Holmes and

Robinson (1981) and Whelan (2001). They propose that forest management practices

that either cull or selectively harvest particular tree species compromise the resource base

and niche diversity available to forest passerines.

My proximate management recommendation is for the maintenance of American

beech in our industrial forests as it appears to offer unique and important foraging

resources to insectivorous passerines. In addition to its use by songbirds it is well known

that the beech offers an important mast crop to a variety of small and large mammals

(Degraaf and Yamasaki 2001). Unfortunately, American beech is intentionally removed

by forest managers throughout its range in Canada and the United States due to its low

commercial value and susceptibility to disease. This proximate recommendation is not a

novel suggestion given other work across other taxonomic groups already suggesting the

importance of this tree species.

117

In light of this, I offer a more theoretical, ultimate management recommendation

that should transcend the particular forest type I studied. I have demonstrated that it is

not merely important to maintain American beech, nor is it merely important to maintain

a diversity of tree species. By linking reproductive success to fine-scale morphological

differences I have demonstrated that it is important to maintain a diversity of tree

morphological types in the landscape. Moreover, our perception of what constitutes a

novel tree morphological type should be tuned to perhaps previously unnoticed

seemingly minor differences in micro-scale vegetation structure. I have also

demonstrated the potential folly in extrapolating results even within seemingly similar

forest types separated by only several hundred kilometres.

118

Future Research

At the first opportunity available, I would like to construct aviary experiments

with the focal birds after the methods of Whelan (1989a, 1989b, 2001) and Parrish

(1995a and 1995b). Whelan’s work especially was remarkably close to my own work

with tests of forage site selection and forage manoeuvres on sugar maple and yellow

birch by the same focal bird species. His work was tantalizing in its similarity combined

with what was to me a conspicuous absence of American beech. Sugar maple foliage is

similar to American beech, or at least more so than yellow birch, therefore, relevant

arguments already exist within his work.

An experimental design adjusting prey availability above and below physiological

thresholds would be an elegant test of my hypothesis that low prey availability will cause

a bird to switch to an optimal foraging strategy based upon the morphology hypothesis.

Prior to that I would like to confirm or refute selective foraging in American beech by

birds from the New Brunswick study sites and birds from New Hampshire and coastal

Maine. This would allow the type of regional comparisons made by Parrish (1995a and

1995b).

119

Plate 6.1. American beech branch and twig morphology within approximately 1 m3

foraging patch, total leaf area: 3224 cm2.

120

Plate 6.2. Yellow birch branch and twig morphology within approximately 1 m3 foraging

patch, total leaf area: 1550 cm2.

121

Plate 6.3. Sugar maple branch and twig morphology within approximately 1 cubic meter

foraging patch, total leaf area: 2860 cm2.

122

Plate 6.4. Yellow birch (left) and American beech branch, twig, and leaf morphology

removed from tree.

Plate 6.5. Sugar maple (left) and American beech branch, twig, and leaf morphology

removed from tree.

123

Plate 6.6. American beech branch morphology on tree.

Plate 6.7. Yellow birch branch morphology on tree.

124

Plate 6.8. Sugar maple branch morphology on tree.

125

Appendix 1

Use of Syrup of Ipecac (Ipecacuanha) as an Emetic for Obtaining

Diet Samples of Several North American Passerines *

* authors are Antony W. Diamond and Peter S. McKinley

A non-toxic and non-lethal means to acquire such a sample is important for

ethical, legal, and study design considerations. Methods previously used in other studies

have included analysis of fecal samples (Davies 1977), flushing the digestive tract with a

saline solution (Moody 1970), forced regurgitation with water (Rosenberg and Cooper

1990), and forced regurgitation with an emetic, most notably, tartar emetic (antimony

potassium tartrate) (Tomback 1975, Zach and Falls 1976, Gavett and Wakeley 1986,

Poulin et al 1993).

Most methods are either inefficient with respect to both time and quality of

sample, or are toxic and potentially lethal to the bird. For example, Tomback (1975)

observed a response period between 10 and 25 minutes depending upon the dosage level

of tartar emetic. Zach and Falls (1976) captured 41 Ovenbirds and administered tartar

emetic to 20. They were all maintained in captivity, and at the end of one week, 50% of

the treated birds had died while only 14.3% of the untreated birds had died. While Poulin

et al (1993) reported the seemingly low mortality of 2.0% with the use of tartar emetic,

this amounted to 70 birds in their study.

Syrup of Ipecac (Ipecacuanha) is an infrequently used, but promising alternative

to tartar emetic. Ipecac is a natural extract from the roots of the rubiaceous plants

Cephaelis ipecacuanha or C. acuminata (Martindale 1977). It is used as an emetic on

children as young as six months to induce emesis after consumption of a poisonous

126

substance. Ipecac induces emesis by two means. It irritates the stomach lining and

directly stimulates chemoreceptors that induce the emesis response as well. Ipecac does

not produce systemic toxicity (Canadian Pharmaceutical Association, 1997).

Diamond and Fayad compared the mortality and effectiveness of tartar emetic

and ipecac on landbirds in Kenya (unpublished manuscript). They found ipecac to be as

effective as tartar emetic, while only the birds treated with ipecac showed no mortality.

Radke and Frydenhall (1974) used syrup of ipecac on the granivorous House Sparrow,

Passer domesticus, with no success. A concentration of 50% produced no emesis, while

higher concentrations resulted in death due to suffocation from the high viscosity of the

syrup.

We used a dose to body weight ratio of 0.1 cc per gram of body mass based on the

work of Diamond and Fayad (unpublished manuscript). Syrup of ipecac was delivered

with a syringe attached to 2mm physiological tubing. The tubing was inserted into the

stomach through the oesophagus. The bird was placed in an opaque plastic container

with a lid immediately following this treatment.

In the summer of 1998 the emetic was administered to 2 American Redstarts, 5

Black-throated Blue Warblers, and 7 Black-throated Green Warblers. One of the Black-

throated Green Warblers was recaptured approximately two weeks following its first

capture. The bird was treated with the emetic again so that for 1998 there were 15 total

treatments. With the exception of one female Black-throated Blue Warbler, all birds

captured in 1998 were male. Stomach contents were recovered 100% of the time.

During the 1999 field season the emetic was administered to 19 Black-throated

Blue Warblers and 11 Black-throated Green Warblers. Two of the Black-throated Blue

127

Warblers were recaptured approximately two weeks after their first capture. These birds

were treated with the emetic again so that for 1999 there were 32 total treatments. All

birds captured in 1999 were male. Stomach contents were recovered 100% of the time.

Treatments for both years totaled 47.

Stomach contents were regurgitated within 30 seconds. Identifiable prey remains

in the stomach samples included arthropod wings, legs, and other parts of the

exoskeleton. In most cases identification to order was possible. Exoskeletal remains of

several arthropod orders were recovered from all samples. Orders identified include

Coleoptera, Diptera, Hemiptera, Lepidoptera, Thysanoptera, and Hymenoptera.

The birds required virtually no recovery time. All handled individuals flew from

the hand and resumed normal activities immediately, including foraging and singing.

The recaptured birds showed similar quick recovery and behavior.

128

References Cited

Abbott, I., and P. Van Heurck. 1985. Tree species preferences of foraging birds in Jarreh

Forest in Western Australia. Australian Wildlife Research 12: 461-466.

Airola, D.A., and M.H. Barrett. 1985. Foraging and habitat relationships of insect-

gleaning birds in a Sierra Nevada mixed-conifer forest. Condor 87: 205-216.

Ambuel, B., and S.A. Temple. 1983. Area-dependent changes in the bird communities

and vegetation of southern Wisconsin forests. Ecology 64: 1057-1058.

Anderson, B.W., M.D. Ohmart, and J. Rice. 1983. Avian and vegetation community

structure and their seasonal relationships in the lower Colorado River valley. Condor

85: 392-405.

Anderson, S.H., and H.H. Shugart Jr. 1974. Habitat selection of breeding birds in an east

Tennessee deciduous forest. Ecology 55: 828-837.

Bart, J., M.A. Fligner, and W. I. Notz. 1998. Sampling and statistical methods for

behavioural ecologists. Cambridge University Press, New York, NY., USA.

Bayne, E. M. and K.A. Hobson. 1997. Comparing the effects of landscape fragmentation

by forestry and agriculture on predation of artificial nests. Conservation Biology

11:1418-1429.

Beedy, E.C. 1981. Bird communities and forest structure in the Sierra Nevada of

California. Condor 83: 97-105.

Belisle, M., and A. Desrochers. 2002. Gap-crossing decisions by forest birds: an

empirical basis or parameterizing spatially-explicit, individual-based models.

Landscape Ecology. 17: 219-231.

Bell, G.W., S.J. Hejl, and J. Verner. 1990. Proportional use of substrates by foraging

birds: model considerations on first sightings and subsequent observations. Pages

161-165 in M.L. Morrison, C.J. Ralph, J. Verner, and J.M. Jehl Jr., editors. Avian

Foraging: Theory, Methodology and Applications. Studies in Avian Biology No. 13.

The Cooper Ornithological Society. Los Angeles, CA., USA.

129

Bibby, C.J., N.D. Burgess, and D.A. Hill. 1993. Bird Census Techniques. Academic

Press. New York, N.Y., USA.

Block, W.M. 1988. Biogeographic variation in foraging ecologies of breeding and

nonbreeding birds in oak woodlands. Pages 264-269 in M.L. Morrison, C.J. Ralph, J.

Verner, and J.M. Jehl Jr., editors. Avian Foraging: Theory, Methodology and

Applications. Studies in Avian Biology No. 13. The Cooper Ornithological Society.

Los Angeles, CA., USA.

Blondel, J., C. Ferry, and B. Frochot. 1981. Point counts with unlimited distance. Pages

414-420 in C.J. Ralph and J.M. Scott, editors. Estimating Numbers of Terrestrial

Birds. Studies in Avian Biology no. 6. The Cooper Ornithological Society. Los

Angeles, CA., USA.

Bourque, J., and M.-A. Villard. 2001. Effects of selection cutting and landscape-level

harvesting on the reproductive success of two Neotropical migrant bird species.

Conservation Biology 15: 184-195.

Bowman, J. 2000. The spatial structure of small-mammal populations in a managed

forest. Ph.D. thesis. University of New Brunswick, Fredericton, NB., Canada.

Bowman, J., G. Forbes, and T. Dilworth. 2000. Landscape context and small-mammal

abundance in a managed forest. Forest Ecology and Management 140: 253-259.

Bowman, J. and G. Forbes. and T. Dilworth. 2001. The spatial component of variation in

small-mammal abundance measured at three scales. Canadian Journal of Zoology 79:

137-144.

Briskie, J.V. 1994. Least flycatcher (Empidonax minimus). In The birds of North

America, No. 99 (A. Poole and F. Gill, eds.). The Academy of Natural Sciences,

Philadelphia, PA, and the American Ornithologists’ Union, Washington, D.C., USA.

Brittingham, M.C., and S.A. Temple. 1983. Have cowbirds caused forest songbirds to

decline? Bioscience 33: 31-35.

Cody, IM.L. 1981. Habitat selection in birds: the role of vegetation structure,

competitors, and productivity. Bioscience 31: 107-111.

Cooper, M.J., and M.C. Whitmore. 1990. Arthropod sampling methods in ornithology.

Pages 29-37 in M.L. Morrison, C.J. Ralph, J. Verner, and J.M. Jehl Jr., editors. Avian

foraging: theory, methodology, and applications. Studies in avian biology no. 13.

Cooper Ornithological Society, Los Angeles, CA., USA.

130

Cumming, E.E. 1994. Songbird habitat use in Saskatchewan old-growth boreal mixed

wood forest. M.Sc. thesis, U. of Saskatchewan, Saskatoon, Sask., Canada.

DeGraaf, R.M., and M. Yamasaki. 2001. New England Wildlife: habitat, natural history,

and distribution. University Press of New England. Hanover, NH., USA.

Diamond, A.W. 1999. Concluding remarks: content versus context in forest bird research.

Pages 139-143 in A.W. Diamond and D.N. Nettleship, editors. Biology and

Conservation of Forest Birds, Society of Canadian Ornithologists Special Publication

no. 1. Fredericton, NB., Canada.

Dodge, K.M., M.C. Whitmore, and E.J. Harner. 1990. Analyzing foraging use versus

availability using regression techniques. Pages 319-324 in M.L. Morrison, C.J. Ralph,

J. Verner, and J.M. Jehl Jr., editors. Avian Foraging: Theory, Methodology and

Applications. Studies in Avian Biology No. 13. The Cooper Ornithological Society.

Los Angeles, CA., USA.

Dunning, J.B. 1990. Meeting assumptions of foraging models: an example using tests of

avian patch choice. Pages 462-470 in Morrison, M.L., C.J. Ralph, J. Verner, and

J.MM. Jehl jr., editors. Avian Foraging: Theory, Methodology and Applications.

Studies in Avian Biology No. 13. The Cooper Ornithological Society. , Los Angeles,

CA., USA.

Faaborg, J., M. Brittingham, T. Donovan, and J. Blake. 1993. Habitat fragmentation in

the temperate zone: a perspective for managers. Pages 331-338 in D.W. Finch and

P.W. Stangel, editors. Habitat fragmentation in the temperate zone: a perspective for

managers. U.S. Department of Agriculture Forest Service Rocky Mountain Forest

Range Experimental Station, General Technical Report Rm- 229. Fort Collins CO.,

USA.

Farina, A. 1997. Landscape structure and breeding bird distribution in a sub-

Mediterranean agro-ecosystem. Landscape Ecology. 12: 365-378.

131

Forman, R.T.T., A.E. Galli, and C.F. Leck. 1976. Forest size and avian density in New

Jersey woodlots with some land use implications. Oecologia 26: 1-8.

Fortin, M.-J., P. Drapeau, and P. Legendre. 1989. Spatial autocorrelation and sampling

design in plant ecology. Vegetatio 83: 209-222

Franzreb, K.E. 1978. Tree species used by birds in logged and unlogged mixed-

coniferous forests. Willson Bulletin 90: 221-238.

Freedman, B., C. Beauchamp, I.A. McLaren, and S.I. Tingley. 1981. Forestry

management practices and birds in a hardwood forest in Nova Scotia. Canadian Field

Naturalist 95: 307-311.

Freemark, K. and B. Collins. 1992. Landscape ecology of birds breeding in temperate

forest fragments. Pages 443-454 in J.M. Hagan III and D.W. Johnston, editors.

Ecology and conservation of Neotropical migrant landbirds. Smithsonian Institution

Press, Washington, D.C., USA.

Freemark, K.E. and H.G. Merriam. 1986. Importance of area and habitat heterogeneity to

bird assemblages in temperate forest fragments. Biological Conservation 36: 115-141.

Galli, A.E., C.F. Leck, and R.T.T. Forman. 1976. Avian distribution patterns in forest

islands of different sizes in New Jersey. Auk 93: 356-364.

Gates, J.E., and L.W. Gysel. 1978. Avian nest dispersion and fledging success in field-

forest ecotones. Ecology 59: 871-883.

Gilpin, IM. 1996. Metapopulations and wildlife conservation: Approaches to modelling

spatial structure. Pages 11-27 in D.M. McCullough, editor. Metapopulations and

wildlife conservation. Island Press, Washington, D.C.,USA.

Gobeil, J-F., and M.-A. Villard. 2002. Permeability of three boreal forest landscape types

to bird movements as determined from experimental translocations. Oikos. 98: 447-

458.

Greenberg, R., C.E. Gonzales, P. Bichier, and R. Reitsma. 2001. Nonbreeding habitat

selection and foraging behavior of the black-throated green warbler complex in

southeastern Mexico. The Condor. 103: 31-37.

132

Greenberg, M., and J. Gradwohl. 1980. Leaf surface specializations of birds and

arthropods in a Panamanian forest. Oecologia 46: 115-124.

Gunn, J.A., 2002. Landscape and Local Effects of Industrial Forestry on the

Reproductive Activity of Forest Songbirds in Northwestern New Brunswick, Canada.

Ph.D. thesis, University of New Brunswick, Fredericton, NB., Canada.

Gunn, J.A., A. Desrochers, M.-A. Villard, J. Bourque, and J. Ibarsabal. 2000. Playbacks

of mobbing calls of black-capped chickadees as a method to estimate reproductive

activity of forest birds. Journal of Field Ornithology 71(3): 473-483.

Hagan, J.M. III, P.S. McKinley, A.L. Meehan, and S.L. Grove. 1997. Diversity and

abundance of landbirds in a north-eastern industrial forest landscape. Journal of

Wildlife Management 61: 718-735.

Hagan, J.M., W. M. Vander Haegen, and P. S. McKinley. 1996. The early development

of fragmentation effects on birds. Conservation Biology 10: 188-202.

Hanski, I. 1994. Patch-occupancy dynamics in fragmented landscapes. Tree 9: 131-195.

Harris, L.D. 1988. Edge effects and conservation of biotic diversity. Conservation

Biology 2: 330-332.

Hayak, L.-E. C., and M.A. Buzas. 1997. Surveying Natural Populations. Columbia

University Press, New York, NY., USA.

Hejl, S.J., J. Verner, and G.W. Bell. 1990. Sequential versus initial observations in

studies of avian foraging. Pages 166-173 in M.L. Morrison, C.J. Ralph, J. Verner, and

J.M. Jehl Jr., editors. Avian Foraging: Theory, Methodology and Applications.

Studies in Avian Biology No. 13. The Cooper Ornithological Society. Los Angeles,

CA., USA.

Hinsley, S. A. 2000. The costs of multiple patch use by birds. Landscape Ecology. 15:

765-775.

Holmes, R.T. 1994. Black-throated blue warbler (Dendroica caerulescens). In The

birds of North America, No. 87 (A. Poole and F. Gill, eds.). The Academy of Natural

Sciences, Philadelphia, PA, and the American Ornithologists’ Union, Washington,

D.C., USA.

Holmes, R.T., R.E. Bonney Jr., and S.W. Pacala. 1979. Guild structure of the Hubbard

Brook bird community: a multivariate approach. Ecology. 60: 512-520.

133

Holmes, R.T., and S. K. Robinson. 1981. Tree species preferences of foraging

insectivorous birds in a northern hardwoods forest. Oecologia 48: 31-35.

Holmes, R.T. and J.C. Schultz. 1988. Food availability for forest birds: effects of prey

distribution and abundance on bird foraging. Canadian Journal of Zoology 6: 720-

728.

Holmes, R.T., T.W. Sherry, P.P. Marra, and K.E. Petit. 1992. Multiple brooding and productivity of a neotropical migrant, the Black-throated Blue Warbler Dendroica

caerulescens, in an unfragmented temperate forest. Auk 109: 321-333.

Horn, H.S., 1971. The adaptive geometry of trees. Monograph in Population Biology.

Princeton University Press, Princeton, NJ., USA.

Howell, C.A., S.C. Latta, T. M. Donovan, P.A. Porneluzi, G.R. Parks, and John Faaborg.

2000. Landscape effects mediate breeding bird abundance in midwestern forests.

Landscape Ecology. 15: 547-562.

Hutt, S.J., and C. Hutt. 1970. Direct observation and measurement of behaviour. Charles

C. Thomas, Springfield IL.

Hutto, MM.L. 1990. Measuring the availability of food resources. Pages 20-28 in M.L.

Morrison, C.J. Ralph, J. Verner, and J.MM. Jehl Jr., editors. Avian Foraging: Theory,

Methodology and Applications. Studies in Avian Biology No. 13. The Cooper

Ornithological Society. Los Angeles, CA., USA.

James, F.C., and N.O. Wamer. 1982. Relationships between temperate forest bird

communities and vegetation structure. Ecology 63: 159-171.

Jansson, G., and P. Angelstam. 1999. Threshold levels of habitat for the presence of the Long-tailed tit (Aegithalos caudatus) in a boreal landscape. Landscape Ecology. 14:

283-290.

Jones, J., and R.J. Robertson, 2001. Territory and nest-site selection of cerulean warblers

in eastern Ontario. The Auk. 118: 727-735.

Lambert, D. A., and S.J. Hannon. 2000. Short-term effects of timber harvest on

abundance, territory characteristics, and pairing success of ovenbirds in riparian

buffer strips. The Auk. 117: 687-698.

134

Johnson, D. H. 1980. The comparison of usage and availability measurements for

evaluating resource preference. Ecology 61: 65-71.

Johnson, IM.D., and T.W. Sherry. 2001. Effects of food availability on the distribution of

migratory warblers among habitats in Jamaica. Journal of Animal Ecology 70: 546-

560.

Karr, J.MM., and MM.MM. Roth. 1971. Vegetation structure and avian diversity in

several New World areas. American Naturalist 105: 423-435.

Kroodsma, MM.L. 1982. Edge effect on breeding forest birds along a power-line

corridor. Journal of Applied Ecology 19: 361-370

Lehner, P. N. 1996. Handbook of ethological methods. Cambridge University Press,

Cambridge, U.K.

Lidicker, W.S. Jr. and W. D. Koenig. 1996. Responses of terrestrial vertebrates to habitat

edges and corridors. Pages 85-109 in D.MM. McCullough, editor. Metapopulations

and wildlife conservation. Island Press, Washington, D.C., USA.

Lima, S.L., and P.A. Zollner. 1996. Towards a behavioural ecology of ecological

landscapes. Tree 11: 131-135.

Litvaitis, J.A, K. Titus, and E.IM. Anderson. 1996. Measuring vertebrate use of terrestrial

habitats and foods. Pages 254-274 in T.A. Bookout, editor. Research and

management techniques for wildlife and habitats. The Wildlife Society, Bethesda,

MD., USA.

Litwin, T.S., and C.MM. Smith. 1992. Factors influencing the decline of Neotropical

migrants in a north eastern forest fragment: Isolation, fragmentation, or mosaic

effects? Pages 483-496 in J.M. Hagan III and D.W. Johnston, editors. Ecology and

conservation of Neotropical migrant landbirds. Smithsonian Institution Press,

Washington, D.C., USA.

Lovette, I.J., and MM.T. Holmes. 1995. Foraging behaviour of American redstarts in

breeding and wintering habitats: implications for food availability. Condor 97: 782-

791.

Lynch, J.F., and MM.F. Whitcomb. 1978. Effects of insularization of the eastern

deciduous forest on avifaunal diversity and turnover. Pages 461-489 in A.

Marmelstein, editor. Classification, inventory, and analysis of fish and wildlife habitat.

Publication OBS-78716. U.S. Fish and Wildlife Service, Washington, D.C., USA.

MacArthur, MM. H., and John W. MacArthur. 1961. On bird species diversity. Ecology

42: 594-598.

135

MacClintock, L., MM.F. Whitcomb, and B.L. Whitcomb. 1977. Island biogeography and

habitat islands of eastern forest. II. Evidence for the value of corridors and

minimization of isolation in the preservation of biotic diversity. American Birds 31:

6-12.

Majer, J.D., H.F. Recher, W.S. Perriman, and N. Achuthan. 1990. Spatial variation of

invertebrate abundance within the canopies of two Australian eucalypt forests. Pages

65-72 in M.L. Morrison, C.J. Ralph, J. Verner, and J.M. Jehl Jr., editors. Avian

foraging: theory, methodology, and applications. Studies in avian biology no. 13.

Cooper Ornithological Society, Los Angeles, CA., USA.

Matthysen, E. 2002. Boundary effects on dispersal between habitat patches by forest birds (Parus major, P. caeruleus). Landscape Ecology. 17: 509-515.

Maurer, B.A., and MM.C. Whitmore. 1981. Foraging of five bird species in two forests

with different vegetation structure. Wilson Bulletin 93: 478-490.

Mazzerolle, D.F., and K.A. Hobson. 2001. Physiological ramifications of habitat

selection in territorial male ovenbirds: consequences of landscape fragmentation.

Oecologia 130: 356-363.

McGarigal, K., and W.C. McComb. 1995. Relationships between landscape structure and

breeding birds in the Oregon coast range. Ecological Monographs. 65: 235-260.

McPherson, G. 1990. Statistics in scientific investigation. Springer-Verlag, New York,

NY., USA.

McIntyre, N.E., and J.A. Wiens. 1999. Interactions between landscape structure and

animal behavior: the roles of heterogeneously distributed resources and food

deprivation on movement patterns. Landscape Ecology. 14: 437-447.

Miles, D. 1990. The importance and consequences of temporal variation in avian

foraging behaviour. Pages 210-217 in IM.L. Morrison, C.J. Ralph, J. Verner, and

J.MM. Jehl Jr., editors. Avian foraging: theory, methodology, and applications.

Studies in avian biology no. 13. Cooper Ornithological Society, Los Angeles, CA.,

USA.

Morales, J.M., and S.P. Ellner. 2002. Scaling up animal movements in heterogeneous

landscapes: the importance of behavior. Ecology. 83: 2240-2247.

136

Morrison, M.L., M.W. Mannan, and G.L. Dorsey. 1981. Effects of number of circular

plots on estimates of avian density and species richness. Pages 405-408 in C.J. Ralph

and J.M. Scott, editors. Estimating Numbers of Terrestrial Birds. Studies in Avian

Biology no. 6. Cooper Ornithological Society, Los Angeles, CA., USA.

Morrison, M.L. and E.C. Meslow. 1983. Bird community structure on early-growth

clearcuts in western Oregon. American Midland Naturalist 110: 129-137.

Morrison, M.L., I.C. Timossi, K.A. With, and P.N. Manley. 1985. Use of tree species by

forest birds during winter and summer. Journal of Wildlife Management 49:1098-

1102.

Morse, D.H. 1968. A quantitative study of foraging of male and female spruce-

woods warblers. Ecology. 49: 779-784.

Morse, D.H. 1976. Hostile encounters among spruce-woods warblers (Dendroica,

Parulidae). Animal Behaviour 24: 764-771.

Morse, D.H. 1971. The foraging of warblers isolated on small islands. Ecology.

52: 216-228.

Morse, D.H. 1977. The occupation of small islands by passerine birds. Condor 79: 399-

412.

Morse, D.H. 1978. Structure and foraging patterns of tit flocks in an English woodland.

Ibis 120: 298-312.

Morse, D.H. 1985. Habitat selection in North American Parulid warblers. Pages 131-158

in M.L. Cody editor. Habitat selection in birds. Academic Press. New York, NY.,

USA.

Morse, D.H. 1989. American Warblers. Harvard University Press. Cambridge, MA.

Morse, D.H. 1993. Black-throated green warbler (Dendroica virens). In The

birds of North America, No. 55 (A. Poole and F. Gill, eds.). The Academy of Natural

Sciences, Philadelphia, PA, and the American Ornithologists’ Union, Washington,

D.C., USA.

Mortberg, U.M. 2001. Resident bird species in urban forest remnants: landscape and

habitat perspectives. Landscape Ecology. 16: 193-203.

137

Norse, E.A., K.L. Rosenbaum, D.S. Wilcove, B.A. Wilcox, W.H. Romme, IM.W.

Johnston, and IM.L. Stout. 1986. Conserving biological diversity in our national

forests. The Wilderness Society, Washington, D.C., USA.

Orians, G.H. 1991. Supplement: Habitat Selection. American Naturalist.

137: 51-54.

Parrish, J. D. 1995a. Effects of needle architecture on warbler habitat selection in

a coastal spruce forest. Ecology. 76. 1813-1820.

Parrish, J.D. 1995b. Experimental evidence for intrinsic microhabitat preferences

in the black-throated green warbler. Condor. 97: 935-943.

Penhollow, M.E., and D.F. Stauffer. 2000. Large-scale habitat relationships of

neotropical migratory birds in Virginia. Journal of Wildlife Management. 64: 362-

373.

Pettersson, R.B., J.P. Ball, K-E Renhorn, P-A. Esseen, and K. Sjoberg. 1995. Invertebrate

communities in boreal forest canopies as influenced by forestry and lichens with

implications for passerine birds. Biological Conservation. 74: 56-63.

Porneluzi, P., J.C. Bednarz, L.J. Goodrich, N. Zawada, and J. Hoover. 1993.

Reproductive performance of territorial Ovenbirds occupying forest fragments and a

contiguous forest in Pennsylvania. Conservation Biology 7: 618-622.

Remsen, J. V., and S. K. Robinson. 1990. A classification scheme for foraging behaviour

of birds in terrestrial habitats. Pages 144-160 in M.L. Morrison, C.J. Ralph, J.

Verner, and J.M. Jehl Jr., editors. Avian Foraging: Theory, Methodology and

Applications. Studies in Avian Biology No. 13. Cooper Ornithological Society, Los

Angeles, CA., USA.

Rice, J., M.D. Ohmart, and D.W. Anderson. 1983. Habitat selection attributes of an avian

community: a discriminant analysis investigation. Ecological Monographs 53: 263-

290.

Robbins, C.S., D.K. Dawson, and B. Hoover. 1989. Habitat area requirements of

breeding forest birds of the middle Atlantic states. Wildlife Monographs 103: 1-34.

Robichaud, I., and M.-A. Villard. 1999. Do Black-throated Green Warblers prefer

conifers? Meso- and microhabitat use in a mixed wood forest. Condor 101: 262-271.

Robichaud, I., M-A Villard, and C.S. Machtans. 2002. Effects of forest regeneration on

songbird movements in a managed forest landscape of Alberta, Canada. Landscape

Ecology. 17: 247-262.

138

Robinson, S.K., and R.T. Holmes. 1982. Foraging behaviour of forest birds: the

relationships among search tactics, diet, and habitat structure. Ecology 63: 1918-

1931.

Robinson, S.K., and R.T. Holmes. 1984. Effects of plant species and foliage structure on

the foraging behaviour of forest birds. Auk 101: 672-684.

Robinson, S.K., M.M. Thompson III, T.M. Donovan, D.M. Whitehead, and J. Faaborg.

1995. Regional forest fragmentation and the nesting success of migratory birds.

Science 267: 1987-1990.

Roland, J., and P. P. Taylor. 1997. Insect parasitoid species response to forest structure at

different spatial scales. Nature 386: 710-713.

Rosenberg, K.V. and M.J. Cooper. 1990. Approaches to avian diet analysis. Pages 80-90

in M.L. Morrison, C.J. Ralph, J. Verner, and J.M. Jehl Jr., editors. Avian foraging:

theory, methodology, and applications. Studies in avian biology no. 13. Cooper

Ornithological Society, Los Angeles, CA., USA.

Sheskin, D.J. 1997. Handbook of parametric and nonparametric statistical procedures.

CRC Press, Boston, MA., USA.

Sherry, T.W., and MM.T. Holmes. 1985. Dispersion patterns and habitat responses of

birds in northern hardwoods forests. Pages 283-309 in M.L. Cody editor. Habitat

selection in birds. Academic Press. New York, NY., USA.

Sherry, T.W., and R.T. Holmes. 1997. American redstart (Setophaga ruticilla). In The

birds of North America, No. 277 (A. Poole and F. Gill, eds.). The Academy of Natural

Sciences, Philadelphia, PA, and the American Ornithologists’ Union, Washington,

D.C., USA.

Slater, P.J.B. 1978. Data collection. Pages 7-24 in P. Colgan, editor. Quantitative

ecology. John Wiley and Sons, New York, NY., USA.

South, A.B., S.P. Rushton, R. E. Kenward, and D.W. Macdonald. 2001.

Modelling vertebrate dispersal and demography in real landscapes: how does

uncertainty regarding dispersal behaviour influence predictions of spatial

population dynamics? Pages 327-349 in J.M. Bullock, R.E. Kenward, and

R.S. Hails, editors. Dispersal Ecology. The 42nd

Symposium of the British

Ecological Society, 2001. Blackwell Publishing, Oxford, UK.

139

Spies, T.A., W.J. Ripple, and G.A. Bradshaw. 1994. Dynamics and pattern of a managed

coniferous forest landscape in Oregon. Ecological Applications 4: 555-568.

Strong, A. 2000. Divergent foraging stategies of two neotropical migrant warblers:

implications for winter habitat use. The Auk. 117: 381-392.

Taylor, J.A., G.MM. Friend, and M.L. Dudzinski. 1984. Influence of sampling strategy

on the relationship between fauna and vegetation structure , plant life form and

floristics. Australian Journal of Ecology 9: 281-287.

Urban, D.L., H.H. Shugart Jr., D.L. DeAngelis, and M.V. O’Neil. 1988. Forest bird

demography in a landscape mosaic. Publication no. 2853. Oak Ridge National

Laboratory, Oak Ridge, TN., USA.

Villard, M.A., K. Freemark and G. Merriam. 1992. Metapopulation theory and

Neotropical migrant birds in temperate forests: an empirical investigation. Pages 474-

482 in J.M. Hagan III and D.W. Johnston, editors. Ecology and conservation of

Neotropical migrant landbirds. Smithsonian Institution Press, Washington, D.C.,

USA.

Villard, M.-A., P.M. Martin, and C.G. Drummond. 1993. Habitat fragmentation and pairing success in the Ovenbird (Seiurus aurocapillus). Auk 110: 759-768.

Virkkala, R. 1988. Foraging niches of foliage-gleaners in northern Finland. Ornis

Fennica. 65: 104-113.

Wahlter, B.A. 2002. Grounded ground birds and surfing canopy birds: variation of

foraging stratum breadth observed in neotropical forest birds and tested with

simulation models using boundary constraints. The Auk. 119: 658-675.

Whelan, C.J. 1989a. Avian structure preferences for foraging and the effect of

prey biomass. Animal Behaviour. 38: 839-846.

Whelan, C.J. 1989b. An experimental test of prey distribution learning in two paruline

warblers. Condor. 91: 113-119.

140

Whelan, C.J. 2001. Foliage structure influences foraging of insectivorous forest birds: an

experimental study. Ecology 82: 219-231.

Wiens, J.A. 1989. The Ecology of Bird Communities, Volume 1. Cambridge University

Press, New York, NY., USA.

Wiens, J.A. 1995. Landscape mosaics and ecological theory. Pages 1-26 in L. Hansson,

L. Fahrig, and G. Merriam, editors. Mosaic Landscapes and Ecological Processes.

Chapman and Hall, New York, NY., USA.

Wiens, J.A. 1996. Wildlife in patchy environments: Metapopulations, mosaics, and

management. Pages 53-84 in D.MM. McCullough, editor. Metapopulations and

wildlife conservation. Island Press, Washington, D.C., USA.

Wilcove, D.S. 1985. Nest predation on forest tracts and the decline of migratory

songbirds. Ecology 66: 12111-1214.

Willson, IM.F., and T.A. Comet. 1996. Bird communities of northern forests: patterns of

diversity and abundance. Condor 98: 337-349.

Willson, IM.F. 1974. Avian community organization and habitat structure. Ecology 55:

1017-1029.

Wolda, H. 1990. Pages 38-43 in M.L. Morrison, C.J. Ralph, J. Verner, and J.M. Jehl Jr.,

editors. Avian Foraging: Theory, Methodology and Applications. Studies in Avian

Biology No. 13. Cooper Ornithological Society, Los Angeles, CA., USA.

Wolter, P.T., and M.A. White. 2002. Recent cover type transitions and landscape

structural changes in northeast Minnesota, USA. Landscape Ecology. 17: 133-155.

Wu, J., and R. Hobbs. 2002. Key issues and research priorities in landscape ecology: An

idiosyncratic synthesis. Landscape Ecology. 17: 355-365.

Zannette, L., P. Doyle, and S. IM. Tremont. 2002. Food shortage in small fragments:

evidence from an area-sensitive passerine. Ecology 81: 1654-1666.

Zar, J. H. 1996. Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ.,USA.

141

Vita

Peter S. McKinley

Professional Experience

1999-Present (concurrent with Ph.D. program), Hancock Land Company, Casco, ME.,

USA. Vice President of Operations and Chief Ecologist.

1992-1995, Manomet Center for Conservation Sciences, Manomet, MA., and Brunswick,

ME., USA. Full time Project Manager and Biologist.

Summer 1991, Indiana University, Bloomington, IN., USA. Field Research Manager.

Summer of 1988, Concord Conservation Commission, Concord NH., USA. Conservation

Intern.

Education and School Related Experience

1996-Present, University of New Brunswick, Fredericton, NB, CA. Ph.D. student

in Department of Biology.

1988-1991, Indiana University, Bloomington, IN., USA.

M.S. Applied Ecology/Environmental Science.

1983-1987, Colby College, Waterville, ME., USA. B.A. Biology.

Publications

Hagan, J.M., M.W. Vander Haegen, and P.S. McKinley. 1996. The early development

of forest fragmentation effects on birds. Conserv. Biol. 10: 188-202.

Hagan, J.M., P.S. McKinley, A.L. Meehan, and S.L. Grove. 1997. Diversity and

abundance of birds in a north eastern industrial forest. J. Wildl. Manage. 61: 718-735.

McKinley, P.M. 1999. Songbird foraging success in two co-dominant hardwood tree

species in a managed mixed-wood forest. Proceedings of The Sustainable Forest

Management Network Conference. 69-73. Universtiy of Alberta, Edmunton, AB., CA.