Consequences of Rainforest Fragmentation for Frugivorous Vertebrates and Seed Dispersal

Author Moran, Catherine

Published 2007

Thesis Type Thesis (PhD Doctorate)

School Griffith School of Environment

DOI https://doi.org/10.25904/1912/54

Copyright Statement The author owns the copyright in this thesis, unless stated otherwise.

Downloaded from http://hdl.handle.net/10072/367385

Griffith Research Online https://research-repository.griffith.edu.au

Consequences of rainforest fragmentation for frugivorous

vertebrates and seed dispersal

Catherine Moran

B.Sc. (Hons.)

Griffith School of Environment

Faculty of Science, Engineering, Environment and Technology

Griffith University

Submitted in fulfilment of the requirements of the degree of Doctor of Philosophy

June, 2007

Abstract

Seed dispersal strongly influences patterns of plant regeneration. Frugivorous

(fruit eating) vertebrates disperse the seeds of between 70% and 90% of rainforest plant species. Forest fragmentation may affect the abundance and distribution of frugivore species. Consequently, patterns of seed dispersal and plant regeneration may vary between extensive forest and fragmented forest landscapes. This thesis assessed frugivorous vertebrates and seed dispersal in a rainforest landscape in subtropical

Australia. First, this study quantitatively compared the distribution and abundance of frugivorous bird and bat species between fragmented and extensive rainforest. Second, the roles of these frugivore species in seed dispersal were evaluated based on their functional attributes and the plant species that they had been recorded consuming.

Third, secondary consequences of forest fragmentation for seed dispersal were predicted from these results.

The field components of this study were conducted in the Sunshine Coast region of southern Queensland. Surveys of frugivorous bird and bat species were undertaken in a network of 48 study sites distributed throughout a 4 000 km2 area. Sites comprised 16 replicates of each of three site types: extensive forest (> 4 000 ha), rainforest remnants and patches of secondary regrowth. Extensive forest sites were stratified by altitude

(low (<200 m above sea level (a.s.l.), medium (200-500 m a.s.l.), and high (>500 m a.s.l.).

Birds were surveyed using 40 minute area searches within a one hectare plot during the early morning. Each site was surveyed for birds four times: twice during summer and twice in winter. Forty-two frugivorous bird species were identified during surveys. Twenty-six of these species occurred frequently enough to quantitatively assess their abundance pattern in remnant and regrowth sites relative to extensive forest. There

i were five species that were recorded in much lower numbers in remnants and/or regrowth than in extensive forest (‘decreasers’), seven that showed higher abundance in remnants and/or regrowth than in extensive forest (‘increasers’) and 14 whose abundance did not vary substantially between the three habitat types (‘tolerant’ species).

The decreasers included four rainforest pigeons (the wompoo, rose-crowned and superb fruit-doves Ptilinopus magnificus, P. regina and P. superbus and brown cuckoo-dove

Macropygia amboinensis) and the green catbird Ailuroedus crassirostris. There was no evidence for the complete seasonal movement of frugivorous bird species between high and low altitudes.

A lack of understanding of the functional roles of frugivorous species has previously limited our capacity to predict specific consequences for seed dispersal of frugivore declines. A major dimension of functional variation among frugivore species is the suite of plant species that they disperse, which depends initially on their patterns of consumption of plant species. In this thesis, frugivorous bird species that were expected to have similar patterns of plant species consumption were assembled into

‘functional groups’. These groupings were based on the bird species’ gape width, degree of frugivory and their methods of seed treatment. For example, it was proposed that species with wide gapes would be able to consume large fruits, whereas those with narrow gapes could only consume small fruits. It was also expected that species with fruit-dominated diets (‘major frugivores’) may consume a different suite of plant species than species with mixed diets or with diets dominated by non-fruit (‘minor frugivores’). Species that crushed seeds were expected to disperse few viable seeds.

Analyses showed that decreaser bird species were predominantly from functional groups that had the potential to disperse large-seeded plant species and may be the main dispersers of native laurels (Lauraceae). Consequently, it is likely that the dispersal of these plants may be reduced in fragmented forest.

ii Relationships between the functional attributes of frugivores and their actual patterns of plant species consumption were analysed using data on the plant species that each frugivore species was known to consume. Diet data were collated from 151 published sources as well as field observation and included records for 244 plant species. Major variation in patterns of plant species consumption corresponded with variation in frugivore species’ attributes. For example, the average size of fruits consumed by bird species increased with their gape width, although minor frugivores tended to consume fruits that were much smaller than their capacity. Statistical comparisons showed that highly frugivorous bird species consumed the highest number of plant species from the Lauraceae, whereas bird species with mixed diets consumed more arillate plant species from the Celastraceae, Sapindaceae, Mimosaceae and

Elaeocarpaceae than other frugivore groups. Bird species from a range of functional groups consumed figs and small-fruited plants from families such as Euphorbiaceae and

Solanaceae. Minor frugivores and a small number of major and mixed-diet bird species had species-poor diets that were dominated by these latter plant taxa.

In order to specifically assess the potential consequences of forest fragmentation for seed dispersal, patterns of plant species consumption were compared among decreaser, tolerant and increaser frugivore species. In particular, the potential for tolerant and increaser bird species to substitute for decreasers was evaluated. Analyses showed that dietary records for 12% of the 220 native plant species represented in the data set, including several from the Rubiaceae, were restricted to decreaser bird species.

In addition, analyses showed that few non-decreaser species consumed numbers of native plant species with fruits wider than 10 mm, or from the Lauraceae, Myrtaceae,

Meliaceae, Verbenaceae and Vitaceae that were comparable to decreaser bird species.

Consequently, it is predicted that there is limited potential for functional substitution by

iii other bird species for decreasers and, therefore, that the dispersal of these plant taxa may be substantially reduced in fragmented compared with extensive rainforest.

The potential for frugivorous bats to disperse seeds in fragmented forest was also assessed. Frugivorous bats were surveyed during summer in each of the 48 sites that had been surveyed for birds. Two observers conducted nocturnal, hour long searches along a 400-500 m transect. Two flying-fox species (grey-headed flying-fox

Pteropus poliocephalus and black flying-fox P. alecto) and the eastern tube-nosed bat

Nyctimene robinsoni were recorded during surveys. At the time of surveys, Pteropus spp. were most frequently recorded in regrowth, whereas N. robinsoni was detected more frequently in extensive forest and remnants than in regrowth. Decreaser bird species and N. robinsoni are rainforest and fruit specialists whereas tolerant and increaser bird species and Pteropus spp. have more generalist patterns of habitat and resource use. N. robinsoni has limited potential to substitute for decreaser bird species as a seed disperser in fragmented rainforest of the study region, because it is known to consume only a small number of plant species and because of its rarity in regrowth. In contrast, Pteropus spp. were widespread in fragmented forest and consumed approximately one-third of the plant species that were consumed by decreaser bird species. In fragmented landscapes, Pteropus spp. may potentially substitute for decreaser bird species as dispersers of large-fruited plant taxa and plants from the

Myrtaceae, although they appear unlikely to disperse seeds >9 mm more than short distances away from parent plants.

The results of this study show that fragmented remnant and regrowth patches of rainforest do not adequately conserve the full complement of frugivorous vertebrate species in the subtropics of eastern Australia. Although the number of frugivore species that showed sensitivity to rainforest fragmentation was relatively small, this may have substantial functional consequences. These consequences are likely because decreaser

iv species may be the sole or predominant dispersers of a substantial proportion of native plant species, which may consequently be susceptible to reduced dispersal away from parent plants in fragmented forest. Reduced dispersal may have a number of implications for plant regeneration. First, dispersal to recruitment sites within forest fragments is likely to be reduced, resulting in lower rates and clumped spatial patterns of recruitment. Second, dispersal of these species between rainforest fragments may be lower, leading to low rates of recolonisation following local extinctions. Third, short- distance dispersal to new habitats may be lower, resulting in low representation of susceptible plant species in regenerating forest on previously cleared land. Fourth, long distance dispersal of these plant taxa would be low, which would mean that they may have a limited capacity to shift their geographical range, for example in response to changing global climatic conditions.

Further clearing and fragmentation of rainforest would exacerbate the situation for decreaser frugivore species and may lead to the decline of additional frugivore species. It is recommended that remaining rainforest be protected from continued clearing. Restoration of forest areas based on the needs of decreaser frugivore species may help to re-establish them in fragmented landscapes. These actions could help to restore the regenerative capacity of many rainforest plant species and hence increase the long term integrity of fragmented rainforest ecosystems.

v Statement of originality

This thesis has not previously been submitted for a degree or diploma in any university.

To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where reference is made in the thesis itself.

…………………………………

C. Moran

Publications arising from this thesis

Slightly modified versions of Chapters Two and Three have previously been peer reviewed and published as a book chapter and journal article, respectively. I was responsible for conducting the research reported in those publications. The co-authors were listed in recognition of their contributions as my academic supervisors. These publications are listed below:

Chapter Two: Moran C, Catterall CP, Green RJ and Olsen MF (2004) Fates of feathered frugivores in fragmented forests. pp. 699-712 in Lunney D (Ed.) Conservation of Australia’s Forest Fauna. Second edition. Royal Zoological Society of NSW,

Mosman.

Chapter Three: Moran C, Catterall CP, Green RJ and Olsen MF (2004) Functional variation among frugivorous birds: implications for rainforest seed dispersal in a fragmented subtropical landscape. Oecologia 141, 584-595.

vi

Table of contents Page

ABSTRACT...... I

STATEMENT OF ORIGINALITY...... VI

PUBLICATIONS ARISING FROM THIS THESIS ...... VI

LIST OF FIGURES ...... XII

ACKNOWLEDGEMENTS ...... XIII

CHAPTER ONE: EFFECTS OF RAINFOREST FRAGMENTATION ON FRUGIVOROUS

VERTEBRATES AND THE POTENTIAL CONSEQUENCES FOR SEED DISPERSAL AND

PLANT REGENERATION ...... 1

1.1 RAINFOREST FRAGMENTATION, FOREST FAUNA AND SEED DISPERSAL...... 1

1.2 THE FUNCTIONAL ROLE OF FRUGIVORES IN SEED DISPERSAL ...... 2

1.3 THE ROLE OF SEED DISPERSAL IN PLANT REGENERATION...... 8

1.4 CONSEQUENCES OF RAINFOREST CLEARING AND FRAGMENTATION FOR FRUGIVORES ...... 11

1.5 RELATIONSHIPS BETWEEN FRUGIVORE SPECIES’ TRAITS AND THEIR SENSITIVITY TO

RAINFOREST FRAGMENTATION...... 16

1.6 CONSEQUENCES OF CHANGES IN THE COMPOSITION OF FRUGIVORE ASSEMBLAGES FOR SEED

DISPERSAL AND PATTERNS OF PLANT REGENERATION...... 20

1 .7 AIMS AND STRUCTURE OF THIS THESIS...... 26

1.8 RAINFOREST FRAGMENTATION, FRUGIVORES AND SEED DISPERSAL IN AUSTRALIA...... 31

CHAPTER TWO: CHANGES IN THE AVIAN FRUGIVORE ASSEMBLAGE IN

FRAGMENTED RAINFOREST COMPARED WITH EXTENSIVE FOREST IN SUBTROPICAL

AUSTRALIA ...... 34

2.1 INTRODUCTION ...... 34

2.2 METHODS ...... 35

2.2.1 Study region...... 35

2.2.2 Site network ...... 37

vii 2.2.3 Bird surveys ...... 42

2.2.4 Data treatment...... 42

2.2.5 Classification of frugivorous birds ...... 44

2.3 RESULTS ...... 45

2.3.1 Abundance of frugivorous bird species in extensive, remnant and regrowth sites ...... 45

2.3.2 Changes in the frugivorous bird assemblage in fragmented forest ...... 51

2.3.3 Seasonal patterns of frugivorous bird abundance ...... 54

2.3.4 Effects of altitude and season on frugivorous bird numbers...... 54

2.4 DISCUSSION ...... 57

2.4.1 Bird species showing a decreaser response to rainforest fragmentation ...... 57

2.4.2 Bird species showing an increaser response to fragmentation...... 60

2.4.3 Frugivore assemblage change in fragmented habitats ...... 61

2.4.4 Seasonal changes in frugivorous bird abundance ...... 62

2.4.5 Frugivorous birds and seed dispersal in remnant and regrowth rainforest:

conservation implications...... 63

CHAPTER THREE: SEED DISPERSAL POTENTIAL OF FRUGIVOROUS BIRD SPECIES IN

RELATION TO THEIR GAPE WIDTH, FRUGIVORY LEVEL AND SEED TREATMENT ...... 67

3.1 INTRODUCTION ...... 67

3.2 METHODS ...... 70

3.2.1 Assessment of the functional attributes of frugivorous bird species: gape width,

frugivory level and seed-crushing behaviour ...... 70

3.2.2 Data analyses ...... 72

3.3 RESULTS ...... 74

3.3.1 Variation in seed dispersal potential among species within the frugivorous bird

assemblage ...... 74

3.3.2 Functional group abundance in remnants and regrowth relative to extensive forest...... 78

3.4 DISCUSSION ...... 81

3.4.1 Characteristics of the frugivorous birds assemblage ...... 81

3.4.2 Functional characteristics of the frugivorous bird assemblage in fragmented

rainforest in subtropical Australia: assessment of potential consequences for seed

dispersal ...... 81 viii CHAPTER FOUR: VARIATION IN PATTERNS OF PLANT SPECIES CONSUMPTION BY

FRUGIVOROUS BIRD SPECIES IS RELATED TO GAPE WIDTH, DEGREE OF FRUGIVORY

AND SEED TREATMENT ...... 87

4.1 INTRODUCTION ...... 87

4.2 METHODS ...... 89

4.2.1 Diet composition of the frugivorous bird assemblage ...... 89

4.2.2 Functional attributes of bird species ...... 91

4.2.3 Data analyses ...... 92

4.3 RESULTS ...... 94

4.3.1 General patterns of plant consumption...... 94

4.3.2 The effect of gape width and frugivory level on diaspore size selection...... 97

4.3.3 Plant species richness of the diets of major, mixed and minor frugivores...... 99

4.3.4 Plant species diet composition in relation to frugivory level, gape width and seed

treatment...... 100

4.4. DISCUSSION ...... 109

4.4.1 Overlap and variation among frugivorous bird species in patterns of plant species

consumption...... 109

4.4.2 Frugivore gape width and patterns of fruit size consumption ...... 110

4.4.3 Frugivory level and patterns of plant species consumption...... 111

4.4.4 Variation among bird species within a frugivory level...... 113

4.4.5 Gape width and frugivory level as indicators of the functional potential of

frugivorous birds as seed dispersers ...... 114

CHAPTER FIVE: REDUCED DISPERSAL POTENTIAL OF NATIVE RAINFOREST PLANT

SPECIES IN FRAGMENTED RAINFOREST...... 116

5.1 INTRODUCTION ...... 116

5.2 METHODS ...... 119

5.2.1 Changes in the frugivorous bird assemblage in fragmented subtropical rainforest...... 119

5.2.2 Fruit consumption database ...... 119

5.2.3 Data analyses ...... 120

5.3 RESULTS ...... 121

5.3.1 Diet comparisons between frugivore response groups...... 121 ix 5.3.2 Specific substitution potential between frugivore taxa ...... 128

5.4 DISCUSSION ...... 137

5.4.1 Reduced dispersal of native rainforest plants as a consequence of rainforest

fragmentation...... 137

5.4.2 Potential for disperser substitution in fragmented forest ...... 139

5.4.3 Implications for conservation of regenerative potential in fragmented rainforest...... 141

CHAPTER SIX: THE DISTRIBUTION OF FRUGIVOROUS BATS AND THEIR POTENTIAL

TO DISPERSE SEEDS IN FRAGMENTED RAINFOREST...... 143

6.1 INTRODUCTION ...... 143

6.2 METHODS ...... 146

6.2.1 The study region and site network ...... 146

6.2.2 Surveys of frugivorous bat distribution...... 146

6.2.3 Frugivorous bird data...... 147

6.2.4 Information on the consumption of native plant species by frugivorous bat and bird

species...... 148

6.2.5 Data handling...... 149

6.3 RESULTS ...... 150

6.3.1 Distribution and abundance of frugivorous bats ...... 150

6.3.2 Association of bat distribution with environmental attributes...... 153

6.3.3 Comparison between frugivorous bat and bird species in their patterns of plant

species consumption ...... 154

6.3.4 Potential for frugivorous bat species to substitute for decreaser bird species as

dispersers in fragmented forest...... 158

6.4 DISCUSSION ...... 158

6.4.1 The distribution of flying-foxes in fragmented rainforest in the Sunshine Coast...... 158

6.4.2 The distribution of the eastern tube-nosed fruit-bat in fragmented rainforest in the

Sunshine Coast ...... 161

6.4.3 The potential for seed dispersal by frugivorous bats in remnants and regrowth:

comparison with frugivorous birds...... 162

CHAPTER 7: GENERAL DISCUSSION: CONSEQUENCES OF FOREST FRAGMENTATION

FOR FRUGIVORES AND IMPLCATIONS FOR SEED DISPERSAL...... 165 x 7.1 SUMMARY OF THE FINDINGS OF THIS THESIS...... 165

7.2 THE SENSITIVITY OF FRUGIVOROUS VERTEBRATE SPECIES TO RAINFOREST FRAGMENTATION

IN SUBTROPICAL AUSTRALIA ...... 168

Decreaser species...... 168

Tolerant species...... 171

7.3 CORRELATES OF FRUGIVORE SPECIES’ SENSITIVITY TO RAINFOREST FRAGMENTATION ...... 173

7.4 PATTERNS OF PLANT SPECIES CONSUMPTION ACROSS THE FRUGIVORE ASSEMBLAGE: AN

ALTERNATIVE MODEL ...... 175

7.5 POTENTIAL CONSEQUENCES OF RAINFOREST FRAGMENTATION FOR SEED DISPERSAL AND

PATTERNS OF PLANT REGENERATION ...... 179

7.6 CONSERVATION ISSUES...... 185

APPENDIX 1...... 187

APPENDIX 2...... 193

APPENDIX 3...... 198

REFERENCES ………………………………………………………………………………………. 203

xi List of figures Page

Figure 1.1 Conceptual representation of patterns of plant species consumption by frugivorous bird species………………………………………………………………………. 7 Figure 1.2 Potential seed dispersal trajectories in fragmented forest landscapes with respect to plant regeneration……………………………………………………………….. 22 Figure 1.3 Conceptual links between the chapters of this thesis…………………………… 30 Figure 2.1 Aerial view of part of the Sunshine Coast study region………………………… 37 Figure 2.2 Map of study region showing site locations……………………………………. 39 Figure 2.3 Examples of the seven patterns of abundance in remnants and regrowth compared with extensive forest……………………………………………………………. 50 Figure 2.4 Ordination of the 48 study sites based on the abundances of 39 frugivorous bird species ………………………………………………………………………….. 53 Figure 3.1 Inter-relationships between frugivorous bird attributes ……………………….. 77 Figure 3.2 Ordination of the 48 study sites based on numbers of birds from each functional group …………………………………………………………………………… 80 Figure 4.1 The average size of diaspores consumed compared with gape width …………. 98 Figure 4.2 The average proportion of diaspores close to the maximum handling capacity consumed by frugivores………………………………………………………… 99 Figure 4.3 The number of native plant species consumed by each frugivore ……………... 100 Figure 4.4 Overlap in the number of plant species consumed by frugivorous birds in relation to frugivory level………………………………………………………………….. 101 Figure 4.5 Classification of frugivore species based on Bray-Curtis similarity in patterns of consumption of native plant species……………………………………………. 106 Figure 4.6 Overlap in the number of plant species consumed by frugivorous birds in relation to gape width classes………………………………………………………………. 109 Figure 5.1 The proportion of native plant species with large (≥10 mm) diaspores that were consumed by decreaser, tolerant and increaser frugivore species………………. 125 Figure 5.2 Classification of frugivore species according to presence / absence of native plant species in the diet……………………………………………………..………… 127 Figure 5.3 The number of native plant species from selected plant families consumed by decreaser, tolerant and increaser frugivores…………………………………….. 130 Figure 5.4. Overlap in the number of native plant species consumed by frugivore species from the decreaser, tolerant and increaser response groups…………………………... 132 Figure 6.1. The abundance (mean ± SE) of flying-foxes recorded during a 60 minute search of extensive, remnant and regrowth forest sites…………………………………… 152 Figure 6.2 The proportion of native plant species with a median diaspore width ≥10 mm that were known to be consumed by decreaser, tolerant and increaser bird species, and by flying-foxes……………………………………………………………………… 156 Figure 6.3 Classification of frugivore species based on similarity of patterns of consumption of native plant species………………………………………………………………. 157 Figure 7.1 Map of Australia showing the approximate location of other studies in fragmented forest that have included frugivore species…………………………….………… 169 Figure 7.2 A model of variation in patterns of plant species consumption by frugivorous bird species in subtropical Australia………………………………………………….. 176 Figure 7.3 Relationship between bird species’ gape widths and their body mass………….. 179 Figure 7.4 The frugivore species that potentially disperse seeds along different dispersal trajectories in fragmented forest landscapes ………………………………… 182-3

xii Acknowledgements

I thank my principal supervisor, Associate Professor Carla Catterall, for her contributions to all stages of this project. This thesis and my PhD experience have greatly benefited from her dedicated attention and brilliant clear thinking. I am also grateful to my associate supervisors, Drs. Ronda Green and Mike Olsen, for their important contributions to this project, especially in its formative stages.

I acknowledge the traditional owners of the country in which the field components of this work were conducted, the Ka’bi or Gubbi Gubbi, Undumbi and

Badtjala people. I am grateful to the people who supported my field work on their land:

Caloundra City Council, Mim Coulstock, David and Bernie Daugaard, John and Joan

Dillon, Wally and Annalies Gogel, Barbara Hansa, Max and Chrissie Hendersen, Ken and Trish Long, Gillian and Neil MacLeod, Ted McCosker, Maroochy Shire Council,

Noosa Shire Council, John and Valerie Poulson, Arthur and Narelle Powter,

Queensland Environmental Protection Agency, Graham and Annabel Wearne and Greg and Charmaine Wightman.

I acknowledge and thank the people who have generously contributed data to this project. John Kanowski conducted bat surveys simultaneously with the author. Carl

Gosper, Damian Hackett and Stephen McKenna provided large amounts of unpublished data from their work on frugivory and Stephen McKenna contributed information on fruit attributes from his collection. Lyla, John and Francis Hansen, Val Jones, Valda

McLean and Shirley Rooke voluntarily conducted frugivory observations throughout the Sunshine Coast.

I am grateful to the people who have facilitated logistical aspects of this project.

Financial support was provided by an Australian Postgraduate Research Award, the

Rainforest Co-operative Research Centre, Griffith University, and the Norman

xiii Wettenhall Foundation. Dave Curmi assisted with field site set up. Heather Janetski arranged for access to bird specimens at the Queensland Museum and Chris Stansbury and John Kanowski helped to measure their gapes. Milton and Merle Rawson provided generous hospitality during components of my field work. Child care was provided by

Cath Cleary and Bev Moran and especially by Maureen and Peter Kanowski who made an enormous contribution to the care of Stella while I worked on my thesis.

Helpful advice on various technical aspects of the project was generously provided by the following people: Nick Clancy, Les Hall, Rachel King, Bill McDonald, and members of Wildlife Ecology Discussion Group at Griffith University (at various times including Carla Catterall, Paul Finn, Ronda Green, Peter Grimbacher, Clare

Hourigan, John Kanowski, Stephen McKenna, Aki Nakamura, Wendy Neilan, Scott

Piper, Terry Reis, Billie Roberts and Tang Yong).

During my time at Griffith University, Rachel King, Scott Piper, Naomi Doak and Sarah Boulter have provided encouragement, empathy, perspective, humour, and intellectual and musical adventure. These have been so important.

The kindness and encouragement of my friends, parents, siblings, grandparents and parents- and siblings-in-law has helped keep me going. My family, John, Stella and

Ruby, have been patient, considerate and positive, despite the toll this project has taken on our time together, among other things. In particular, I thank John for his many sacrifices and logistical and moral support.

xiv Chapter One

Effects of rainforest fragmentation on frugivorous vertebrates and the

potential consequences for seed dispersal and plant regeneration

1.1 Rainforest fragmentation, forest fauna and seed dispersal

Rainforest has been heavily cleared and fragmented worldwide (e.g., Myers,

1984; Turner and Corlett, 1996; Whitmore, 1997). The long-term survival of rainforest flora and fauna in the wild will therefore depend on their ability to persist in fragmented rainforest landscapes (Myers, 1984; Laurance, 1991; Daily et al., 2001; Sodhi et al.,

2004).

Rainforest fragmentation has negative consequences for the distribution and abundance of many forest biota (e.g., Turner, 1996; Laurance and Bierregaard, 1997;

Laurance and Peres, 2006). Because animals play important roles in rainforest dynamics, changes in the composition of rainforest fauna in fragmented forest landscapes may have secondary consequences for plant-animal interactions and ecosystem functions (Burkey, 1993; Didham et al., 1996; Daily et al., 2001; van Bael et al., 2003; Şekercioğlu et al., 2004; Hooper et al., 2005; Şekercioğlu, 2006). For example, between 70% and 90% of rainforest plant species are fleshy-fruited (Howe and Smallwood, 1982; Willson et al., 1989; Butler, 2003). Frugivorous (fruit eating) fauna are the main dispersers of the seeds of fleshy-fruited plants (van der Pijl, 1982;

Corlett, 1998). Therefore, changes in the frugivore assemblage in rainforest remnants could alter the dispersal of a large proportion of the rainforest flora in fragmented landscapes (Corlett, 1998; Silva and Tabarelli, 2000). The extent to which this occurs

1 will depend the level of variation among frugivore species in their function as seed dispersers.

1.2 The functional role of frugivores in seed dispersal

Seed dispersal is the movement of seed away from a parent plant (Howe and

Smallwood, 1982; Levin et al., 2003). Birds are an abundant and diverse element of the frugivorous fauna in forests worldwide (e.g., Corlett, 1998). Mammals and to a lesser extent, reptiles, amphibians, fish and invertebrates may also play a role in seed dispersal

(Corlett, 1998). Frugivorous animal species vary in their functional roles as seed dispersers as a consequence of several factors. Primary variation among frugivore species arises from differences in the suite of plant species that they consume and hence disperse (Crome, 1975; Snow, 1981; Herrera, 1984; Howe, 1986; Innis, 1989; Sun et al., 1997; Brown and Hopkins, 2002). Variation among frugivore species in the suite of plant species that they consume is affected by the interaction between the morphological, physiological and behavioural traits of the frugivore species and the morphological, chemical and nutritive traits of the fruits of plant species (Gautier-Hion et al., 1985; Moermond and Denslow, 1985; Corlett, 1996; Kitamura et al., 2002;

Poulsen et al., 2002; Silva et al., 2002).

There is broad variation in patterns of plant consumption among higher taxonomic groupings of frugivores (van der Pijl, 1982; Gautier-Hion et al., 1985;

Willson et al., 1989; Bollen et al., 2004). For example, because they have teeth, most mammals can consume large fruits with hard rinds or husks, whereas most birds, whose beaks limit their capacity to break into or swallow large fruits, cannot (van der Pijl,

1982; Gautier-Hion et al., 1985). This variation in patterns of consumption of plant taxa has been conceptualised in terms of taxonomic-based ‘dispersal syndromes’, which are described by suites of fruit characteristics, including size, colour, pulp characteristics

2 and location on a plant, and suites of related frugivore characteristics such as their perception of fruit colour or odour, digestive physiology and feeding behaviour (van der

Pijl, 1982; Bollen et al., 2004).

However, the broad categories of ‘bird’, ‘bat’ and ‘terrestrial mammal’ dispersal syndromes obscure the sometimes substantial variation in patterns of plant species consumption among faunal species within these categories (Willson et al., 1989; Stiles,

1993; Graham et al., 2002; Lord et al., 2002). For example, there is a maximum limit to the size of fruit that a given frugivore species can handle which results from its body mass and the size of its oral aperture (Herrera, 1981, 1984; Wheelwright, 1985; Mack,

1993). Consequently, the bird species in an assemblage may vary in substantially in their capacity to disperse large fruited species because, although they may consume the fleshy part of fruits piecemeal (Levey, 1987), only frugivore species with wide gapes or large body mass are able to transport large seeds (Wheelwright, 1985; Silva and

Tabarelli, 2000; Lord et al., 2002).

Beyond the intractable constraint on the maximum size of fruits that a frugivore species can handle, there is a lack of agreement regarding the factors that are important in determining major variation in patterns of fruit consumption within frugivore assemblages (Herrera, 1998, 2002; Levey and Martínez del Rio, 2001). However, the few studies that have examined interactions among multiple frugivore and plant species in rainforest (Pratt and Stiles, 1985; Hamann and Curio, 1999; Brown and Hopkins,

2002; Kitamura et al., 2002) have shown that there is additional, unexplained variation among frugivore species in their patterns of plant consumption beyond that related to fruit size. For example, Kitamura et al. (2002) considered patterns of consumption of

259 plant species by 25 frugivore species in north eastern Thailand and showed that certain frugivore species did not consume fruits from certain plant species, despite their size-related morphological capacity to handle them. Similarly, Pratt and Stiles (1985)

3 found that patterns of consumption of 20 plant species by 35 frugivorous bird species in

Papua New Guinea were related to the interaction between avian taxonomy and fruit structure in addition to fruit size: capsular fruits were consumed predominantly by birds of paradise (Paradisideae) whereas drupes and berries were mostly consumed by either small passerines (which took small fruits) or pigeons and bowerbirds (large fruits).

Several factors other than fruit size potentially influence patterns of plant species consumption by frugivore species. First, the chemical content of many fruits is dominated by carbohydrate and water, whereas a small proportion of fruits contain high levels of lipids or nutrients such as nitrogen (McKey, 1975; Izhaki and Safriel, 1989).

Highly frugivorous species may need to consume the fruits of particular nutrient-rich plant species in order to obtain a complete diet from fruit (Izhaki and Safriel, 1989;

Bairlein, 1996). Therefore, it is logical to expect that a frugivore species’ level of nutritional dependence on fruit may influence its patterns of plant species consumption

(Morton, 1973; McKey, 1975; Bairlein, 1996; Bosque and Calchi, 2003). For example, highly frugivorous species may actively select fruits with high lipid content (McKey,

1975), or fruits with high levels of protein (White, 1993) or other essential minerals and nutrients (Pulliam, 1975; Schaefer et al., 2003). In contrast, frugivores whose diets include non-fruit items such as animals or seeds may obtain substantial amounts of energy, minerals and nutrients from these sources (Izhaki and Safriel, 1989) and hence be less selective of particular plant species.

A frugivore species’ patterns of plant species consumption may also be strongly influenced by its digestive physiology. Variation among frugivore species in their digestive physiology affects their capacity to assimilate certain forms of carbohydrate

(Martínez del Rio et al., 1988; Martínez del Rio and Karasov, 1990) or lipids (Cipollini and Levey, 1997; Levey and Martínez del Rio, 2001). For example, frugivore species that are unable to digest sucrose would be expected to consume few of the plant species

4 that produce sucrose-rich fruits (Martínez del Rio and Restrepo, 1993). Furthermore, a frugivore species’ capacity to cope with secondary compounds is likely to have a strong influence over the plant species that it consumes (Sun et al., 1997; Izhaki et al., 2002;

Levey and Martínez del Rio, 2001). Because both the occurrence of secondary compounds in plants and an animal’s capacity to detoxify these compounds (at least in herbivores) has a strong phylogenetic basis (Bernays and Chapman, 1994), specialisation by frugivores on particular secondary compounds would be expected to result in an association between the diets of phylogenetically-related frugivores and phylogenetically-related plants.

Most discussions of the factors associated with variation in patterns of plant species consumption by frugivores, other than those associated with fruit size, relate to three different conceptual models: the lipid-carbohydrate dichotomy (McKey, 1975); specialisation on core plant taxa (Fleming, 1986); or a null model (Burns, 2006).

The lipid-carbohydrate dichotomy

Several studies in the Neotropics have suggested an association between highly frugivorous bird species and the consumption of plant species with lipid-rich fruits

(Snow BK, 1962; Snow DW, 1970, 1971; Howe and Primack, 1975; McKey, 1975;

Howe and Estabrook, 1977; Wheelwright, 1983). McKey (1975) proposed a model to synthesise these findings within a coevolutionary framework, suggesting that a small number of highly frugivorous species would consume ‘high quality’ fruits that have lipid-rich pulp, and would preferentially disperse seeds to good germination sites, whereas the larger group of opportunistic frugivores would consume ‘low quality’, carbohydrate-rich fruits. The predictions of this model, as they relate to variation among frugivore species in their patterns of plant consumption, are illustrated in Figure 1.1(a).

5 Empirical tests of the predictions of this model have advanced understanding of frugivore-plant interactions, although they have shown little support for either a dichotomous pattern of consumption based on lipid and carbohydrate content (e.g.,

Herrera, 1984; Fuentes, 1994; Corlett, 1996; Sun et al., 1997), or superior dispersal by highly frugivorous bird species (Wheelwright and Orians, 1982). Although highly frugivorous bird species may consume large proportions of lipid-rich fruits (Crome,

1975; Herrera, 1984; Stiles, 1993), they may also frequently consume fruits with relatively low lipid content (Fuentes, 1994; Sun et al., 1997). Furthermore, species that have mixed diets may also regularly consume lipid-rich fruits (Levey and Karasov,

1989; Howe, 1993; Martinez del Rio and Restrepo, 1993; Fuentes, 1994). It is likely that patterns of plant species consumption by frugivorous species are influenced by their need to balance intake of a variety of nutrients and minerals (Pulliam, 1975), or by chemical compounds (Izhaki et al., 2002), not only by their energetic requirements.

Specialisation on ‘core plant taxa’

Fleming (1986) developed a model of plant consumption for frugivorous phyllostomid bats, based on data collected in the Neotropics (Barro Colorado Island and

Costa Rica). This model described specialisation by particular bat species on certain

‘core plant taxa’, with the opportunistic addition of other plant species as their fruit became available (Figure 1.1(b)). The core plant taxa that Fleming identified were characterised by being available throughout the year; for one set of bat species (in the

Carollia and Sturnira genera) these plant taxa were ‘high quality’ fruits (Piper and

Solanum spp.) that occurred in low densities, while for bat species in the genus

Artibeus, the core taxon comprised high density, ‘low quality’ fruits (Ficus spp.). I am not aware of any subsequent tests of the generality of this model for frugivorous bats elsewhere, or for other frugivore taxa.

6 (a)

Specialist Large, lipid- Small, sugar Generalist frugivores rich fruits rich fruits frugivores

Piper spp. Ficus spp. Solanum spp. Artibeus (b) Carollia, Sturnira

Remaining fruits

(c) All frugivores

All fruits

Figure 1.1 Conceptual representations of patterns of plant species consumption by frugivorous bird species. The outer frames represent available plant species. Arrows indicate the consumption of plant taxa by frugivores. Three alternative models are represented: (a) the lipid-carbohydrate dichotomy proposed by McKey (1975) for Neotropical birds; (b) the ‘core plant taxa’ model proposed by Fleming (1986) for Neotropical bat genera; and (c) the neutral model proposed by Burns (2006) for temperate birds.

7 Null model of plant consumption

The frugivory literature has been dominated by ‘adaptive’ and ‘coevolutionary’ explanations of patterns of seed consumption and dispersal. Recently, Burns (2006) proposed a null model in which fruit preferences did not differ among frugivore species, but rather that frugivores consumed various fruits in proportion to their availability in the environment. Burns (2006) found some support for this hypothesis amongst a small number of bird and plant species (six and seven, respectively) in a northern hemisphere temperate forest. While a null model is valuable for focussing attention on patterns that can be explained by chance alone, many previous studies have reported strong deterministic patterns in frugivore feeding behaviour. For example, in a 12-year study of plant-frugivore interactions in Spanish scrubland, Herrera (1998) showed that plant species were not consumed in proportion to their availability in the environment. Even

Burns (2006, p.430) concluded that “…deterministic processes are not entirely unimportant in structuring pair-wise interactions between fruits and frugivores…” in his study system. The applicability of a neutral hypothesis to patterns of plant consumption by frugivore species in rainforest ecosystems has not been tested.

1.3 The role of seed dispersal in plant regeneration

Seed dispersal is one of several processes that determine the extent and patterns of plant regeneration (Wang and Smith, 2002). For example, following dispersal, the process of plant regeneration may be strongly influenced by seed predation, which in turn is related to seed predator abundance (Forget, 1993; Wright et al., 2000; Wright and Duber, 2001; Murray and Garcia, 2002; Babweteera et al., 2007). Nevertheless, seed dispersal establishes the critical template for plant regeneration (Herrera, 1985;

Nathan and Muller-Landau, 2000; Levin et al., 2003), and makes an important

8 contribution to individual plant reproductive success, plant population dynamics, and the ability of plant species to colonise new habitats (Howe and Smallwood, 1982).

Individual plant reproductive success

First, seed dispersal may increase the likelihood of successful reproduction by an individual plant by removing its seeds from the region of highest per capita seed and seedling mortality (Janzen, 1970; Connell, 1971; Howe and Smallwood, 1982; Harms et al., 2000). This concept forms the basis of the ‘Janzen-Connell’ hypothesis, which predicts that seed and/ or seedling mortality should be highest directly beneath parent plants as a result of density-dependent factors such as sibling competition, and the activity of fungal pathogens, seed predators and seedling herbivores (Janzen, 1970;

Connell, 1971). However, the benefits of escape (or consequences of not escaping) depend on how these agents of mortality vary in relation to the position of the parent tree and to seed and seedling density (Chapman and Chapman, 1995; Levin et al.,

2003). For example, seedlings of certain plant species may suffer very high mortality beneath parent plants (Howe et al., 1985; Schupp, 1988; Chapman and Chapman, 1995), whereas those of other plant species may not (Janzen and Martin, 1982; Chapman and

Chapman, 1995; Corlett and Turner, 1997; Baider and Florens, 2006). Nevertheless, there is increasing evidence that density-dependent mortality is a pervasive factor in structuring rainforest plant assemblages (Harms et al., 2000; Wright, 2002), and therefore that localised seed dispersal is an important functional process in rainforest dynamics (Terborgh et al., 2002).

Seed dispersal may also increase a plant’s reproductive success by delivering seeds to ‘microsites’ that contain combinations of abiotic conditions (e.g., soil fertility, moisture, light) and biotic factors (e.g., competitors, predators) that improve germination, survival and growth (Grubb, 1977; Hubbell, 1979). These may be sites of

9 limited spatial extent that occur in particular topographic positions (e.g., along watercourses) or that occur stochastically in dynamic forest systems (e.g., light gaps caused by tree falls) (Schupp, 1993). Hence, the probability of a seed reaching a suitable microsite and recruiting successfully is likely to increase with the number of seeds dispersed, and the spatial extent and temporal period of dispersal (Hurtt and Pacala,

1995). In addition, the seeds of many rainforest plant species may persist for only a short time in the seed bank (Hopkins and Graham, 1984; Alvarez-Bullya and Martínez-

Ramos, 1990). Ongoing dispersal would be required to maintain the chance that seeds of these plant species were present when a recruitment opportunity arose (Chesson and

Warner, 1981; Muller-Landau et al., 2002).

Plant population dynamics

Seed dispersal affects the demographic characteristics and dynamics of plant populations. For example, the size and rate of expansion of plant populations are products of the reproductive success of individual plants, which depends on the successful dispersal and establishment of propagules (Levin et al., 2003). Furthermore, seed dispersal to suitable microsites is a critical factor in the recovery of plant populations following localised extinctions, whether from stochastic environmental and demographic causes or from human activities (Cochrane et al., 1999).

Patterns of seed dispersal within and among populations may also affect gene flow and population genetic structure, which in turn may influence the susceptibility of populations to disturbances (Hamilton, 1999; Jordano and Godoy, 2002).

Colonisation of new habitats

Seed dispersal is fundamental to plant colonisation of new habitats. In fragmented forest landscapes, seed dispersal strongly influences patterns of plant

10 regeneration on cleared land (McDonnell and Stiles, 1983; Guevara et al., 1986; Silva et al., 1996; Holl et al., 2000; Zimmerman et al., 2000; Hooper et al., 2004; Laurence,

2004; Franklin and Rey, 2007).

Dispersal over long distances to new habitats determines the biogeographical distribution of plant species and the potential for species’ range expansions (Ridley,

1930; Levin et al., 2003). Long distance seed dispersal is likely to take on increasing importance given the changing climatic conditions associated with global warming

(Primack and Miao, 1992; Matlack, 1995; Westoby and Burgman, 2006; Weir and

Corlett, 2007).

1.4 Consequences of rainforest clearing and fragmentation for

frugivores

The composition of frugivore assemblages may change as a consequence of different species’ responses to forest clearing and fragmentation; throughout the world, some species have shown sensitivity to forest fragmentation, whereas others are more tolerant of forest fragmentation (Corlett, 1998; Silva and Tabarelli, 2000). Comparisons of historical bird species lists with contemporary surveys have revealed that certain frugivorous species are sensitive to forest fragmentation. For example, Castelletta et al.

(2000) reported that, within 20 years of widespread deforestation in Singapore, four of the ten frugivorous bird species had become locally extinct. In different regions of the

Colombian Andes, Kattan et al. (1994) documented the local extinction over an 80 year period of 36% (22 of 61 species) of frugivorous bird species, while Renjifo (1999) reported that 40% of frugivorous bird species (17 of 42 species) had become extinct following forest fragmentation. In the Brazilian Atlantic, Ribon et al. (2003) reported that 10 frugivorous bird species had become extinct and a further 11 were threatened

11 (i.e., approximately 66% of a total of 32 frugivorous bird species) following extensive forest clearing.

Other than these historical studies, research into the effects of forest fragmentation on frugivore species has generally involved comparisons of frugivore assemblages between continuous and fragmented forest, or evaluation of the effects of fragment size and isolation. I consider these in turn below.

Frugivore assemblage change in forest fragments compared with continuous forest

Two studies have compared frugivore species’ responses to fragmentation using systematic pre-fragmentation and post-fragmentation surveys. Working in the experimentally fragmented forests of the Biodynamics of Forest Fragmentation Project in Brazil (Bierregaard et al., 1992), Bierregaard and Stouffer (1997) compared average rates of capture of the 12 frugivorous bird species that were most common pre-clearing with their capture rates 2-3 years after the forest had been fragmented into 1 ha and 10 ha patches. Capture rates of these species declined significantly from 35 individuals/

1000 mist-net hours before clearing to approximately 20 individuals / 1000 net hours.

However, nine additional frugivorous bird species were recorded only after forest fragmentation. Cosson et al. (1999) compared the abundance of 14 frugivorous bat species in an area of forest in French Guiana before and after its fragmentation by flooding of the surrounding landscape to create a dam. Six bat species were not observed in any forest fragments following flooding, and the average abundance of seven of the eight remaining frugivorous bat species was lower in fragments (size range

5 - 40 ha) than in the mainland control site (0 – 65% of their abundance in the control site).

Several studies have compared frugivore assemblages in forest fragments and continuous forest ‘reference’ sites. In subtropical Australia, Date et al. (1996) surveyed

12 the incidence of nine frugivorous pigeon species in ten rainforest fragments (size range

1 ha to 80 ha) and 15 rainforest sites contiguous with large tracts of forest. Four of the pigeon species occurred more frequently in continuous than fragmented sites, five species were recorded in similar frequency between these two site types and one species was more common in fragments. In Uganda, the average number of frugivore species

(birds and monkeys) in a large tract of forest (8 500 ha in size) was 1.14 times the number of species in fragments (size range 130 ha – 1 400 ha), although this difference was not statistically significant (Farwig et al., 2006).

Other workers have compared the number of frugivore species visiting a focal tree species between continuous forest and rainforest fragments. For example, Graham et al. (2002) compared frugivorous bird assemblages at Dendropanax arboreus

(Araliaceae) and Bursera simaruba (Burseraceae) between fragments (mean 4.1 ha in size) and a large forest tract (650 ha) in Mexico. While the average numbers of bird species, visits to focal trees and fruits consumed were similar for both tree species between habitats, the species composition of frugivorous birds feeding at D. arboreus varied between continuous and fragmented forest, largely because two of the bird species from continuous forest did not visit trees in fragments. In the Atlantic forest of

Brazil, Pizo (1997) reported 35 bird species visiting fruiting Cabralea canjerana

(Meliaceae) trees in an extensive forest tract (49 000 ha) compared with 14 frugivorous bird species at the same tree species in a 250 ha rainforest remnant. The lower number of frugivorous bird species in the 250 ha remnant may have been a consequence of its reduced size and isolation, although the lower sampling effort in the smaller fragment

(45 hours of observation compared with 70 in the large forest tract) may also have contributed to this result. In Tanzania, five of the ten frugivorous bird species recorded in fruiting Leptonychia usambarensis (Sterculiaceae) in continuous forest (7 500 ha in size) were not recorded at trees of this species in three small fragments (2, 13 and 31 ha

13 in size), and mean visitation rates of two additional bird species were at least 75% lower in fragments than in continuous forest (Cordeiro and Howe, 2003). Although a higher number of individual trees were observed in continuous forest than in fragments (16 compared with 10) in this study, and this may have contributed to the difference in species’ totals, it should not have biased the data on average visitation rates per tree

(Cordeiro and Howe, 2003).

In summary, there is some evidence of reduced total frugivore species richness associated with forest fragmentation (Cordeiro and Howe, 2003; Farwig et al., 2006).

Studies that have evaluated the responses of individual species have reported declined abundance of one suite of frugivore species, maintained abundance of another group of species and, in some cases, increased numbers of a further suite of species (Date et al.,

1996; Bierregaard and Stouffer, 1997; Cosson et al., 1999).

The effect of fragment size on frugivore assemblages

Cordeiro and Howe (2001) conducted transect surveys in five forest patches in

Tanzania that varied in size from 0.5 ha to 3 500 ha. They detected the lowest numbers of frugivorous bird and primate species in the three smallest fragments (0.5, 9 and 30 ha) but similar species’ numbers between the 521 and 3 500 ha sites. The lower numbers of species detected in the smaller fragments may have been partly due to the lower survey effort in these habitats. Şekercioğlu et al. (2002) reported similar numbers of frugivorous bird species in a large (>200 ha) remnant and in small (approx. 5 ha) rainforest remnants in Costa Rica. Similarly, in French Guiana, Cosson et al. (1999) showed that three years after fragmentation, the patterns of reduced abundance of frugivorous bat species were similar between small (<5 ha) islands and a larger (40 ha) forest island. Date et al. (1996) reported no association between the abundance of frugivorous pigeon species and fragment size (with size ranging from 1 to 80 ha), even

14 for species that were less abundant in fragments than continuous forest overall (e.g.,

Ptilinopid fruit-dove species). In Brazil, Bierregaard and Stouffer (1997) reported that capture rates of only two of the six bird species tested varied between 1 ha and 10 ha fragments. In both cases, bird species’ abundances were significantly higher in one hectare than ten hectare fragments. The abundance of other species showed the opposite pattern, but data were not significant in statistical comparisons.

In summary, most studies have tended to show only a limited effect of fragment size on the abundance of frugivore species (Cosson et al., 1990; Date et al., 1996;

Şekercioğlu et al., 2002).

The effect of isolation on frugivore assemblages

In Costa Rica, Luck and Daily (2003) reported that the average number of frugivorous bird species declined from 21.5 at Micona spp. (Melastomaceae) trees that were located within 2 km of a large rainforest remnant in a low intensity agricultural matrix, to 14.1 at trees located 5-8 km from forest in areas of high agricultural intensity.

In subtropical Australia, Green (1993) compared visitation by frugivorous bird species at two species of fig (Ficus platypoda and F. superba), Ehretia acuminata

(Boraginaceae) and Diploglottis australis (Sapindaceae) in more-forested upland areas and less-forested valleys. This study found that fewer frugivorous bird species visited the fig trees in the valleys than in the mountain areas, whereas similar numbers of bird species were recorded at the other two plant species. In Kenya, Eshiamwata et al. (2006) reported a similar species richness of frugivorous birds at Ficus thonningii located within 200 m of forest compared with those over 1 km from forest, although the landscape they worked in may have contained a large amount of forest habitat.

In summary, lower numbers of frugivorous species have been reported visiting certain fruiting plant species in matrix habitats compared with relatively well-forested

15 areas (Green, 1993; Luck and Daily, 2003), however, there may be variation between regions or plant species.

1.5 Relationships between frugivore species’ traits and their sensitivity to rainforest fragmentation

Studies of frugivore species’ responses to forest fragmentation have shown that certain frugivore species are sensitive to forest fragmentation, whereas other frugivore species appeared to be relatively tolerant of, or even advantaged by, these changes (see

Section 1.4). Variation among species in their sensitivity to rainforest fragmentation may be due to differences in behavioural, ecological or demographic attributes (Lovejoy et al., 1986; Laurance, 1990; Stouffer and Bierregaard, 1995; Turner, 1996; Sieving and

Karr, 1997; Warburton, 1997; Corlett, 1998). Henle et al. (2004) reviewed empirical and theoretical evidence for the association of different plant and animal species’ attributes with their sensitivity to forest fragmentation. These authors identified certain aspects of demography (particularly population size and variability), and ecological traits (patterns of microhabitat and matrix use, rarity and biogeographical distribution) as being the most consistently related to differing fragmentation responses among species. Studies of the association between frugivore traits and their responses to forest fragmentation have focussed mainly on the effects of dispersal ability, degree of resources specialisation and body size. Among the studies evaluated by Henle et al.

(2004), several factors, including dispersal power, body size, and ecological specialisation had inconsistent associations with species’ fragmentation sensitivity.

Below, I review the findings of studies relating frugivore species’ attributes to their fragmentation responses.

16

Ability to disperse through the matrix

A species’ abundance in fragmented forest landscapes depends partly on its ability to disperse through matrix habitats (Wiens, 1994; Bierregard et al., 1992;

Warburton, 1997; Graham, 2001; Şekercioğlu et al., 2002). This ability would affect a species’ capacity to recolonise fragments after localised extinctions and to use networks of patches to satisfy resource requirements.

The natural dispersal ability of volant (flying) taxa is typically greater than that of terrestrial taxa. However, not all birds (Stouffer and Bierregaard, 1995) or bats

(Cosson et al., 1999) readily disperse through fragmented parts of the landscape. In practice, a species’ dispersal potential may be limited by resource availability, relative to cost. For example, Graham (2001) showed that although the keel-billed toucan

Ramphastos sulphuratus did fly among rainforest fragments in Mexico, its movements were limited to areas of the landscape that contained a minimum threshold amount of forest and fruit resources. This was interpreted as being a result of this species’ need to balance the cost of moving a certain distance with the energy gained from available resources (Graham, 2001).

Dispersal ability may also be related to a species’ scale of movement. For example, it has been proposed that migratory and nomadic species may have greater dispersal power than sedentary species and hence have a greater capacity to move through the modified matrix (reviewed in Henle et al., 2004). However, it has conversely been reasoned that sedentary species may be more likely to have smaller area requirements and hence be more capable of persisting in isolated fragments than species with large area needs (Henle et al., 2004). For example, nomadic frugivore species may move over large areas to find ripe fruit that is spatially and temporally patchy (Leighton and Leighton, 1983; Innis, 1989). Forest fragmentation may reduce

17 the capacity of these frugivore species to move among key fruit resources (Karr, 1976;

Leighton and Leighton, 1983; Terborgh, 1986; Wheelwright, 1986; Laurance and

Yensen, 1991). However, different frugivore species’ sensitivity to fragmentation has not been specifically correlated with variation in their movement patterns.

Degree of specialisation on resources

A species’ ability to traverse the matrix is not only associated with their capacity or willingness to move, but also with their use of matrix elements, such as isolated trees, copses of regrowth, windbreaks and agricultural crops (Estrada et al., 1993; Crome et al., 1994; Graham, 2001). Species with specialised patterns of forest resource use are arguably less likely to use habitat elements within the cleared matrix, and hence more likely to be adversely affected by fragmentation, than species with more generalised habitat requirements (Willis, 1974; Karr, 1976; Leck, 1979; Andrén, 1994; Christiansen and Pitter, 1994; Turner, 1996; Warburton, 1997; Gascon et al., 1999; Sigel et al.,

2006). For example, in a tropical rainforest landscape in Australia, frugivorous bird species that were dependent on rainforest were less likely to use matrix resources than species that used more open forest or a variety of forest types (Crome et al., 1994).

Studies conducted in the Neotropics (Christiansen and Pitter, 1997) and south east Asia

(Castelletta et al., 2000) have shown that frugivorous bird species that specialised on fruit were disproportionately sensitive to rainforest fragmentation. In contrast, species with diets that comprised more than one food type tended to be more resilient. In a study of natural fragments of monsoon rainforest in northern Australia, Price et al.

(1999) found that the use of rainforest patches by specialist frugivorous bird species was strongly affected by the cumulative amount of rainforest within a landscape, but that this did not clearly affect bird species that had the ability to switch from fruit to invertebrates or other dietary items.

18

Body size

Many studies have proposed that large body size may be associated with frugivore declines in fragmented forest (Kattan et al., 1994; Corlett, 1998, 2002;

Restrepo et al., 1997; Renjifo, 1999; Castelletta et al., 2000; McConkey and Drake,

2002). The distribution of large-bodied species in fragmented forest landscapes may be restricted as a result of their large area requirements (Leck, 1979; Pimm et al., 1988;

Turner, 1996; Sieving and Karr, 1997; Sodhi et al., 2004). Furthermore, increased hunting pressure in fragmented forest may disproportionately affect large-bodied species (Corlett, 2002; Sodhi et al., 2004; Terborgh and Nuñez-Iturri, 2006), especially large-bodied frugivores (Brash, 1987).

Restrepo et al. (1997) examined changes in the proportional distributions of body mass of frugivorous bird assemblages along a gradient from forest remnant to pasture in Colombia. Their results showed that larger-bodied species were consistently lost from avian frugivore assemblages in more disturbed situations, although small- bodied species were also lost from assemblages in one of the four landscapes surveyed.

However, Daily and Ehrich (1994) and Luck and Daily (2003) reported greater persistence of large-bodied than small-bodied avian frugivore species in agricultural landscapes in Costa Rica. This result was interpreted as a consequence of the superior position of larger birds in the foraging dominance hierarchy in this region (Daily and

Ehrlich, 1994). Similarly, Cosson et al. (1999) reported a clear positive relationship between the body size of bat species and their abundance in rainforest fragments in

French Guiana.

19 1.6 Consequences of changes in the composition of frugivore assemblages for seed dispersal and patterns of plant regeneration

This section develops a framework for the study of the consequences of forest fragmentation for frugivores, and the secondary consequences for seed dispersal and plant regeneration. There is variation among frugivore species in the plant species that they disperse (Section 1.2) and also in their responses to forest fragmentation (Section

1.4). Consequently, it would be expected that forest fragmentation would result in changes in the dispersal of frugivore-dispersed plant species, and that this in turn would be likely to affect patterns of plant regeneration (Section 1.3). This logic has been used to predict changes in seed dispersal and plant regeneration as a consequence of changes in the abundance of frugivore species in fragmented forest (e.g., Restrepo et al., 1997).

However, because of limited understanding of the specific roles of frugivore species in seed dispersal, the predicted changes have often been vague. Furthermore, there has been limited consideration of the potential for different spatial dimensions of seed dispersal (See Section 1.3) to be differentially affected by changes in the composition of frugivore assemblages.

Furthermore, for regenerating plants, the different processes of escape from density dependent mortality, recolonisation of microhabitats and colonisation of new habitats (described in Section 1.3) occur at different spatial scales. In Figure 1.2 and

Table 1.1, these are described as different seed dispersal trajectories for a given plant or plant species in fragmented forest landscapes. These seed dispersal trajectories can be used as a basis for systematically considering potential changes in qualitative aspects of the dispersal of different plant species that may result from changed composition of frugivore assemblages in fragmented forest. For example, variation among frugivore species in their patterns of movement may create different spatial patterns of seed deposition. First, dispersal along trajectory b (Figure 1.2) would be affected by 20 differences at the scale of patterns of microhabitat use by frugivore species (Reid, 1989;

Schupp, 1993; Wenny and Levey, 1998; Alcántara et al., 2000; Loiselle and Blake,

2002). Second, differences among frugivore species in their patterns of movement among fragments (e.g., Tewksbury et al., 2002) and into secondary regrowth (e.g., Silva et al., 1996) would affect dispersal along trajectories c, d and f. Finally, there is also variation among frugivore species in their propensity to move over large areas across the landscape (Holbrook et al., 2002; Dennis and Westcott, 2006), and hence to disperse seeds along trajectory e.

Table 1.1 Description of the potential trajectories of seed dispersal in fragmented forest landscapes (Figure 1.2) and their relationship to different aspects of the process of seed dispersal and plant regeneration.

Trajectory Description of seed movement path1 Processes affected a beyond the crown of the parent plant escape from density-dependent mortality b relatively short distance to regeneration recolonisation of microsites microsites within remnant c moderate distance across non-forest recolonisation following local extinction matrix between fragments d moderate distance into non-forest matrix colonisation of secondary regeneration e long distance across non-forest matrix biogeographical distribution and range expansion f moderate distance from non-forest matrix recolonisation of microsites/ colonisation of new into remnant habitat (e.g., introduced plant taxa) g moderate distance around non-forest recolonisation of microsites/ colonisation of new matrix habitat 1 relative distances involved in each trajectory refer to different scales of movement; ‘short’ is tens to a hundred metres; ‘moderate’ is hundreds of metres to a kilometre; ‘long’ is in order of kilometres.

21

g

b a

f c

d

e

Figure 1.2 Potential seed dispersal trajectories in fragmented forest landscapes with respect to plant regeneration. Forest fragments are shown in grey and the surrounding non-forest matrix is white. A focal plant individual is represented as a diamond. Arrows show paths of seed movement. Each trajectory is related to different aspects of the seed dispersal process in fragmented forest landscapes (Table 1.1). Trajectories (a) – (e) represent potential dispersal trajectories of a focal plant in a forest patch in relation to the following processes: escape from density-dependent mortality (a); recolonisation of regeneration microsites within a patch (b); recolonisation following local extinction in another forest patch (c); colonisation of new habitats in the non-forest matrix (d); and range expansion via colonisation over long distances (e). Trajectories (f) (re)colonisation of forest fragment from the non-forest matrix, and (g) (re)colonisation of new habitats in the non-forest matrix from the non-forest matrix, represent potential dispersal trajectories of plants in the non-forest matrix that may affect patterns of plant regeneration in fragments and the non-forest matrix. See Section 1.3 for further description.

Dispersal failure: potential consequences of the loss of disperser species for plant regeneration

The loss of all of the dispersers of a given plant species would result in dispersal failure for the plant species (e.g., Temple, 1977; Silva and Tabarelli, 2000; Terborgh and Nuñez-Iturri, 2006; Babweteera et al., 2007). Under this scenario, there would be no 22 dispersal along any of the trajectories shown in Figure 1.2. Dispersal failure, if combined with recruitment failure beneath parent plants, would eventually lead to a plant species’ extinction (Temple, 1977; Janzen and Vasquez-Yanez, 1991). However, since seedlings of many plant species may be able to recruit beneath parent plants

(Section 1.3), dispersal failure may not reduce reproduction to zero. Nevertheless, dispersal failure is likely to reduce plant reproductive success and lead to population decline over the longer term (Levin et al., 2003). For example, in Uganda, the tree

Balanites wilsoniana can only be dispersed by African elephants Loxodonta africana because of the very large size of its fruit (Babweteera et al., 2007). While seedlings of

B. wilsoniana recruited in forest fragments without elephants, the survival of these undispersed juveniles was substantially lower than that of juveniles that established away from the parent. Therefore, it would be expected that B. wilsoniana would have higher reproductive success in the forest where elephants were present (Babweteera et al., 2007). Other studies have also shown higher germination rates in dispersed than undispersed seeds (Asquith et al., 1999) and higher growth and survival of juvenile plants that have germinated beyond the crown of conspecific plants (Hubbell and

Foster, 1990; Bleher and Böhning-Gaese, 2001).

In addition to reducing per capita reproductive success, dispersal failure of a plant species would eliminate its ability to colonise microsites, either within a fragment

(Orrock et al., 2006; Figure 1.2, trajectory b), or between fragments (Poschlod et al.,

1996; McEuen and Curran, 2004; Figure 1.2, trajectory c). Consequently, the species would be unable to recolonise following localised extinctions, and its distribution would become more clumped. The population would consequently be more susceptible to local extinction if stochastic disturbances affected all individuals in the spatially-constrained population (Fahrig and Merriam, 1994; Cochrane et al., 1999). Furthermore, plant species that lacked dispersal would be unable to colonise new habitats, such as

23 regenerating vegetation (Figure 1.2, trajectory d). Finally, the failed dispersal of a plant species would mean it could not migrate over large distances. This may compromise the survival of that plant species in the longer term, for example if its existing range became climatically unsuitable as a result of changed global conditions (Primack and Miao,

1992; Westoby and Burgman, 2006; Weir and Corlett, 2007; Figure 1.2, trajectory e).

Potential consequences of reduced frugivore species richness or abundance for seed dispersal and plant regeneration

The majority of rainforest plant species are likely to be dispersed by multiple frugivore species (Wheelwright and Orians, 1982; Moermond and Denslow, 1985;

Bronstein and Hoffman, 1987). Consequently, the loss of all disperser species may be unlikely for most plant species. In most studies, lower numbers of disperser species and/or individuals have been reported, rather than the complete absence of dispersers

(e.g., Howe and Cordeiro, 2003; see Section 1.4).

Logically, reduced numbers of frugivore species may be associated with lower rates of visitation and fruit removal. This has been shown empirically in Tanzania

(Cordeiro and Howe, 2003), Madagascar (Bleher and Böhning-Gaese 2001, 2006) and

Brazil (Pizo, 1997). However, in Costa Rica, Luck and Daily (2003) reported substantially reduced frugivore species richness, but no change in rates of visitation to

Micona spp. (Melastomaceae). In Kenya, Farwig et al. (2006) reported slightly declined species richness of frugivores at fruiting Prunus africana (Rosaceae) in forest fragments compared with a large forest tract, but a concomitant increase in visitation and seed removal. Therefore, the number of frugivore species visiting a plant may not be directly related to the rate of dispersal of that plant species. Factors such as changed competitive interactions (e.g., Willson and Crome, 1989), and density or behavioural compensation

(Loiselle and Blake, 2002) may interact with changes in the composition of the

24 frugivore assemblage to influence seed removal rates in fragmented forest landscapes.

In addition, there is variation among frugivore species in the quantity of seeds that they disperse (Graham et al., 2002; Cordeiro and Howe, 2003), as well as the temporal period over which they disperse seeds (Greenberg et al., 1995). Furthermore, there may be variation among frugivore species in the proportion of seeds that they disperse to suitable germination microhabitats (e.g, Reid, 1989; Murray et al., 1993; Wenny and

Levey, 1998; Aukema and Martínez del Rio, 2002).

Consequently, the reduced abundance or loss of different frugivore species may have different impacts on the quantity and quality of seed dispersal. Therefore, the consequences for seed dispersal may be difficult to predict based on information about species richness or overall frugivore abundance. However, more detailed consequences for seed dispersal and plant regeneration of changes in the species composition or relative abundance of individual frugivore species may be predictable if information about frugivore species composition is combined with an understanding of functional variation among species and interpreted in the framework shown in Figure 1.2.

Limited knowledge regarding the disperser assemblage of most plant species constrains the capacity to predict changes in seed dispersal following changes in the frugivore assemblage. An exception may be large-fruited plant species; dispersal failure may be predicted from knowledge of the loss of frugivore species if all of the species with the morphological capacity to disperse large fruits have gone extinct, because small frugivores are unable to disperse large fruits (Herrera, 1984; Moermond and

Denslow, 1985; Wheelwright, 1985). Because large frugivores have declined in fragmented forest in many parts of the world (see Section 1.5), it has been predicted that dispersal of large-fruited plant species will consequently be reduced or fail (Corlett,

1996, 1998; Corlett and Turner, 1997; Silva and Tabarelli, 2000; McConkey and Drake,

2002; Kitamura et al., 2005). In Uganda, Chapman and Onderdonk (1998) found that

25 the abundance of seedlings, especially of large-seeded plant species, declined in fragments with reduced abundance of primates, compared with extensive forest.

1 .7 Aims and structure of this thesis

The aim of this thesis is to investigate changes in the frugivorous vertebrate assemblage as a consequence of rainforest fragmentation in the moist subtropics of

Australia, and to assess the potential for subsequent changes in seed dispersal. This broad aim is addressed first by investigating the effects of fragmentation on the frugivore species within a complete regional assemblage, and then by assessing the potential roles of the different frugivore species in seed dispersal through analyses of their morphology, behaviour and patterns of consumption of plant species. Finally, this information is synthesised to predict the consequences of fragmentation for seed dispersal by frugivores.

In his review of the biological effects of rainforest fragmentation, Turner (1996) recommended three key research directions to advance ecological understanding of the consequences of rainforest fragmentation: (1) study in older fragments, (2) the identification of susceptible groups of taxa, and (3) the assessment of higher order effects, including seed dispersal. Rainforest landscapes in subtropical Australia present an opportunity to undertake study of all three research areas. First, forests in this region have been fragmented for 70-150 years, sufficient time for assemblages to undergo some degree of ‘relaxation’ (Brooks et al., 1999). Second, for reasons detailed below, the vertebrate frugivore assemblage in subtropical Australia may be well suited to identifying variation among species in their responses to fragmentation and functional roles. Third, this approach may contribute to the development of a predictive understanding of the secondary consequences of rainforest fragmentation for seed dispersal. 26 Compared with many other regions of the world, the frugivore assemblage in

Australia is relatively simple (Crome, 1978; Dennis, 1997; Corlett and Primack, 2006).

The dominant frugivores are birds and bats (Green, 1993, 1995). There are no primates or other non-volant frugivorous mammals as in other regions (e.g., Corlett, 1998).

In addition to its relative simplicity, there is a substantial amount of information on the consumption of fleshy fruited plants by frugivore species in subtropical Australia

(e.g., Innis, 1989; Green, 1995; Recher et al., 1995; Church, 1997), and dietary information has been partly compiled for each species in the Handbook of Australian,

New Zealand and Antarctic Birds (HANZAB) series (Marchant & Higgins, 1993;

Higgins & Davies, 1996; Higgins, 1999; Higgins et al., 2001), although it has not been synthesised across any regional bird assemblage.

Furthermore, there is a sufficient number of species comprising the frugivore assemblage in subtropical Australia to detect strong patterns of variation among species, if they exist. Many previous community-wide studies of the of patterns of plant species consumption by frugivore species have been conducted in ecosystems with few frugivorous species, such as Mediterranean scrublands (18 bird species; Herrera

(1984)), littoral forest in Madagascar (6 bird species and 7 mammal species; Bollen et al. (2004)) and temperate rainforest in Canada (6 bird species; Burns (2006)). The small numbers of frugivore species in these studies may have made it difficult to detect statistically significant variation among species. Based on the number of frugivorous bird species recorded at fruiting trees, there are at least 32 frugivorous bird species in subtropical Australia (Green, 1993).

Specifically, this thesis addresses three sets of questions relating to:

1. the distribution and abundance of frugivorous bird and bat species in

fragmented forest relative to extensive rainforest in a specific study region

(the Sunshine Coast) of subtropical Australia;

27 2. variation among these frugivore species in their seed dispersal potential

(assessed using degree of frugivory, capacity for ingesting large seeds, and

other attributes), and their patterns of consumption of plant species and types

of fruits; and

3. the extent to which the seed dispersal potential of rainforest plant species is

likely to change as a consequence of changes in the frugivorous vertebrate

community in fragmented rainforest.

Figure 1.3 summarises the structure and conceptual links among the subsequent chapters of this thesis.

In Chapter Two, the specific composition of the frugivorous bird assemblage in the fragmented subtropical rainforest landscape of the Sunshine Coast, Australia is identified. This chapter also presents the results of field surveys aimed at assessing whether the abundance of frugivorous bird species is affected by forest fragmentation, using replicate sites of extensive, remnant and regrowth rainforest.

In Chapter Three, functional traits of frugivorous bird species that may influence their role as seed dispersers are assessed for the frugivorous bird assemblage.

Specifically, bird species’ gape width, degree of frugivory and seed treatment are analysed in relation to their responses to fragmentation. Information about the response of frugivorous bird species in extensive versus fragmented forest (Chapter Two) is used to assess the possibility of reduced dispersal potential of rainforest plant species in fragmented rainforest.

In Chapter Four, data on plant species consumption by frugivorous bird species are compiled from published literature and field records. Tests of association are conducted between the frugivore traits that were used in functional analyses conducted in Chapter Three, and frugivore species’ patterns of plant species consumption.

28 In Chapter Five, patterns of plant species consumption by frugivorous bird species are assessed using the same data as in Chapter Four in relation to the potential for reduced dispersal of certain plant species as a result of declined abundance of particular bird species in fragmented rainforest.

In Chapter Six, the species composition and responses to forest fragmentation of the frugivorous bat assemblage is identified for the same site network as the bird assemblage considered in Chapter Two. Information on patterns of plant species consumption by these bats is used to assess their potential to disperse similar plant species to frugivorous birds, particularly those bird species that decline in fragmented forest.

Chapter Seven provides a synthesis of the findings of the previous chapters in relation to understanding the functional roles of frugivore species in seed dispersal and predicting consequences of rainforest fragmentation for seed dispersal. This chapter makes specific predictions regarding seed dispersal and plant regeneration in fragmented rainforests of the study region and considers the implications for conservation.

29 Chapter Two: Changes in the avian frugivore assemblage in fragmented rainforest compared with extensive forest in subtropical Australia

Quantitative surveys of frugivorous bird species in extensive forest, remnant & regrowth.

Using attributes of frugivore species to Using patterns of plant species consumption by predict their functional roles in seed frugivore species to assess potential changes in dispersal seed dispersal and plant regeneration

Chapter Three: Seed dispersal potential of Chapter Five: Reduced dispersal potential of frugivorous bird species in relation to their native rainforest plant species in fragmented Chapter Six: The distribution of gape width, frugivory level and seed rainforest frugivorous bats and their potential treatment to disperse seeds in fragmented rainforest. Proposes key functional traits of frugivorous Uses information on patterns of plant species bird species. Predicts consequences of changes consumption to predict consequences of the in the frugivorous bird assemblage (Chapter 2) changes in the avian frugivore assemblage Surveys of frugivorous bat species in for seed dispersal using this approach. (Chapter 2) for seed dispersal. extensive forest, remnant and regrowth. Analyses patterns of plant species consumption, especially in relation to species consumed by frugivorous birds that decline in Chapter Four: Variation in patterns of plant fragmented forest (Chapter 5). species consumption by frugivorous bird Chapter Seven: General discussion: species is related to gape width, degree of Consequences of forest fragmentation frugivory and seed treatment for frugivorous vertebrates and rainforest

seed dispersal Tests predicted relationships between frugivore

traits and patterns of plant species consumption Synthesises the results from each chapter and (Chapter 3). makes recommendations for conservation and research.

Figure 1.3 Conceptual links between the chapters of this thesis. 30 1.8 Rainforest fragmentation, frugivores and seed dispersal in

Australia

There are three major areas of rainforest along the east coast of Australia. These are the tropical rainforests of north Queensland (approximately 15ºS - 19ºS), the subtropical rainforests of southern Queensland and northern New South Wales (26ºS - 30ºS), and the cooler temperate rainforests of Tasmania (40 - 44ºS) (Webb and Tracey, 1981).

Smaller patches of rainforest (including ‘dry rainforests’) occur across coastal and sub- coastal areas of northern and eastern Australia (Webb and Tracey, 1981; Bowman,

2000).

Australian rainforests are relicts of ancient rainforests that formerly covered extensive areas of the continent (Kershaw et al., 1991). Cool and dry climatic conditions during the late Tertiary, and especially during the Pleistocene, resulted in the retraction of rainforest to moist and protected refugial areas (Webb and Tracey, 1981; Adam,

1992; Goosem, 2000). During this period, much of the rainforest in Australia was replaced by open forests, woodland, savanna and grasslands, which were better suited to the changed climatic conditions and associated increase in fire (Martin, 1990; Kershaw et al., 1991). There has been some minor re-expansion of rainforest over the last few thousand years, although the distribution of rainforest remains disjunct, reflecting the

“archipelago of refugia” that were available during former climatic regimes (Webb and

Tracey, 1981: 609). The current distribution of rainforest in Australia is shown in the

National Land and Water Resources Audit (NLWRA) (2001).

Formerly, large tracts of continuous subtropical rainforest (tens of thousands of hectares) were associated with fertile soil on basalt lava flows on plateaux (e.g., the

Lamington and Maleny plateaux in southern Queensland, and the ‘Big Scrub’ of northern New South Wales). Rainforest also occurs in areas of less fertile soils. In the

Australian subtropics, rainforests on poor soil are restricted to areas that receive high 31 rainfall, are locally nutrient-enriched and moist (e.g., along watercourses), or are associated with topographic features that provide protection from fire (e.g., gullies)

(Webb and Tracey, 1981). Patches of subtropical rainforest in these situations are typically surrounded by more extensive, drier and fire-prone forest types (often dominated by dry-fruited Eucalyptus and related tree genera).

Large areas of subtropical rainforest have been cleared in Australia (Webb and

Tracey, 1981). Most of the clearing in these rainforest landscapes was for agriculture and occurred from the mid 1800s (Young and McDonald, 1987; Watson, 1989;

Frawley, 1991; Catterall and Kingston, 1993). The pre-European rainforest cover across the continent was estimated to have been four million ha; it has been estimated that approximately three-quarters of this remains (NLWRA, 2001). Subtropical rainforests have been heavily cleared from basalt plateaux, in the lowlands and along watercourses

(Catterall and Kingston, 1993). For example, less than 1% of the original rainforest cover remains of the Big Scrub, formerly the most extensive patch of lowland subtropical rainforest in Australia (Frith, 1952, 1976; Floyd, 1990). Large forest tracts are now mostly restricted to upland areas (Webb and Tracey, 1981; Catterall and

Kingston, 1993; Date et al., 1996).

Contemporary rainforest landscapes in Australia resemble those in many other regions of the world in comprising a mosaic of remnant forest patches, grazed land, agricultural cropland, tree crops, regrowth and suburban development (e.g., Guevara and Laborde, 1993; Benítez-Malvido and Martínez-Ramos, 2003). An important point of difference is that there is no shifting agriculture in Australia, whereas this is a feature of rainforests in some parts of Asia (e.g., Lawrence, 2004) and South America

(Tabarelli and Peres, 2002).

There is a lack of knowledge regarding ecological processes in fragmented rainforest landscapes in the Australian subtropics, including the dynamics of plant

32 regeneration (Adam, 1992; Green, 1995; Gilmore 1999; Hunter 1999). As a consequence, the long-term conservation values of the region’s remnant rainforests may be compromised, for example if regeneration trajectories are truncated by a lack of dispersal of certain plant taxa. This thesis provides new information relating to these issues.

33 Chapter Two

Changes in the avian frugivore assemblage in fragmented rainforest

compared with extensive forest in subtropical Australia

2.1 Introduction

Populations of many frugivorous bird species have declined following the fragmentation of tropical rainforests (Kattan et al. 1994; Date et al., 1996; Bierregaard and Stouffer, 1997; Pizo, 1997; Corlett, 1998; Renjifo, 1999; Castelletta et al., 2000;

Silva and Tabarelli 2000; Cordeiro and Howe, 2001, 2003; Ribon et al., 2003). It has been suggested that some frugivorous bird species from subtropical Australia may also be sensitive to rainforest fragmentation (Frith, 1952; Date et al., 1991; Date et al., 1996;

Price et al., 1999). Furthermore, it has been suggested that certain frugivorous bird species may migrate seasonally between upland and lowland areas in response to altitudinal differences in fruit availability (Innis, 1989; Date et al., 1991). Because rainforests have been heavily cleared in the lowlands of subtropical Australia (Catterall et al., 1997), seasonal dependence on these areas may limit populations of these frugivorous bird species (Date et al., 1991). However, species’ responses to forest fragmentation in subtropical Australia have not been assessed across the avian frugivore assemblage.

Understanding the use of fragmented forests by birds in the context of extensive and ongoing rainforest clearing may help develop management strategies appropriate for avian conservation (Saunders et al., 1991; Sodhi et al., 2004). Furthermore, frugivores disperse the seeds of a large proportion of rainforest plant species (Willson et al., 1989; Butler, 2003). Consequently, changed numbers of frugivores in fragmented rainforest may result in changed patterns of plant regeneration (Janzen and Vasquez-

Yanez, 1991; Harrington et al., 1997; Restrepo et al., 1997; Corlett, 1998; Silva and

34 Tabarelli, 2000). For example, the decline of particular frugivore species in fragmented landscapes may mean that certain plant species have lower dispersal potential in these areas. The declined abundance of a suite of frugivorous bird species in small forest fragments in Tanzania has been associated with reduced seed dispersal, lower levels of recruitment, and clumped spatial patterns of recruitment of certain plant species

(Cordeiro and Howe, 2001, 2003). In particular, large fruited plant species may be especially susceptible to reduced dispersal in fragmented landscapes because of the decline of large bodied frugivores (Corlett, 1996, 1998; Harrington et al., 1997; Silva and Tabarelli, 2000; Kitamura et al., 2002; McConkey and Drake, 2002; Meehan et al.,

2002).

In this chapter, frugivorous bird species’ abundances are assessed in fragmented rainforest in a subtropical Australian landscape. Specifically, differences in the abundance of frugivorous bird species between large tracts of forest, rainforest remnants and patches of rainforest regrowth are quantified. It is expected that the abundance of certain species may be lower in remnants and regrowth than in extensive forest, while that of other species may be higher. The effects of season and altitude on the birds’ use of rainforest habitats are also assessed. It is anticipated that certain species may occur in upland areas during summer and lowland areas during winter (Date et al., 1991; Recher et al. 1995). Potential implications of observed changes in the frugivorous bird assemblage for the dispersal of large fruited and other rainforest plants in fragmented landscapes are described.

2.2 Methods

2.2.1 Study region

The study was conducted in a 4 000 km2 subtropical rainforest landscape in the hinterland of the region known as the Sunshine Coast, approximately 100 km north of

35 the city of Brisbane in Southeast Queensland, Australia (152-154˚ E, 26-27˚ S).

Approximately two-thirds of the pre-European forest cover has been cleared throughout the region (Catterall et al., 1997), including extensive areas of rainforest (Meier and

Figgis, 1985; Young and McDonald, 1987). Extant forest fragments comprise a mosaic with cattle grazing land, agricultural cropland, plantation forests and suburban development (e.g., Figure 2.1).

Rainforest in coastal lowland areas of the study region had been almost totally cleared by the early 20th century (Frawley, 1991). Except for rainforest patches within open forests in the Cooloola area in the north, lowland rainforest has been reduced to scattered, small isolates behind coastal sand dunes or fringing watercourses. Rainforest in sub-coastal lowlands associated with the Mary River Valley, situated approximately

30 kilometres inland, have also been heavily cleared, mostly for cattle grazing. Much of the remainder of the study region comprises undulating terrain associated with the

Blackall and Conondale Ranges. A large expanse of continuous rainforest formerly occurred on the basaltic plateau of the Blackall Range, but this had been extensively cleared by the early twentieth century, firstly for timber and then for dairy farming

(Young and McDonald, 1987; Watson, 1989; Frawley, 1991), leaving rainforest remnants in gullies and along steeper slopes (e.g., Figure 2.1). Extensive eucalypt forest-rainforest mosaics extend from the northern and southern ends of the Blackall

Range and cover large areas of the Conondale Ranges. Unmanaged rainforest regrowth on previously cleared land makes an increasing contribution to regional forest cover.

Additionally, many small areas have been replanted by private landholders, community groups and local authorities over the past three decades (Catterall et al., 2004).

36

Figure 2.1 Aerial view of part of the Sunshine Coast study region showing remnant and regrowth forests interspersed with rural and residential land uses on the south eastern part of the Blackall Range (Source: Queensland Department of Natural Resources, 1997). Forest cover tends to be associated with undulating terrain or watercourses. The area seen in this view contains a moderate level of forest cover compared with other fragmented parts of the landscape.

2.2.2 Site network

Study sites were chosen to represent a range of situations in which rainforest remains or has re-established in the study region. Sixteen replicate sites within each of three different states of rainforest landscape context and condition were selected: (i) rainforest within extensive tracts of forest; (ii) remnant rainforest isolated from extensive forest by surrounding cleared and modified land; and (iii) regrowth, also isolated by cleared and modified land (Figure 2.2). Sites were identified using

37 vegetation mapping, aerial photography and on-ground assessment. As far as possible, replicate sites within each type were distributed throughout the study region. A one- hectare plot was marked within each of the study sites. The configuration of the plot was influenced by the shape and landform attributes of each site, but was usually either

200 m x 50 m or 100 m x 100 m.

Extensive forest sites were distributed along eastern slopes of the Conondale

Ranges, on the northern and southern ends of the Blackall Range, and on the Cooloola sand mass in the north of the study region (Figure 2.2). These sites were located within forest tracts greater that 4 000 ha in size and comprising at least 20% rainforest. Many of the extensive forest sites were located in patches of rainforest surrounded by eucalypt open forest and woodland, together with some large areas of forest timber plantations, usually the native hoop pine Araucaria cunninghamii. Extensive forest sites that were located in the smaller rainforest patches (several hectares) contained a more conspicuous eucalypt element, a lower diversity and abundance of rainforest plant species and were less structurally complex than sites located among larger rainforest patches (tens to hundreds of hectares).

The sixteen extensive forest sites were stratified by altitude: five were located in upland (>500 m a.s.l), six in mid-elevation (200-500 m a.s.l) and the remaining five in lowland (<200 m a.s.l.) forests (overall range 90-800 m a.s.l., mean 370 m, S.E. 53 m).

Remnants and regrowth were located at mid-elevations and in lowland areas (both ranging from 20-500 m a.s.l.: remnants, mean 206 m, S.E. 41 m; regrowth, mean 165 m, S.E. 41 m). It was not possible to locate high altitude replicate sites of remnant or regrowth rainforest.

38

Figure 2.2 Map of study region showing site locations in relation to the coast, watercourses (dark lines) and major water bodies (speckled areas). Inset: Location of study region in relation to the Australian continent.

Remnant sites were patches of native vegetation around which all or most of the original vegetation had been cleared. As far as possible, remnant sites were chosen to encompass the floristic and structural variation shown in extensive forest sites, so as to 39 concentrate on the influence of landscape situation rather than resource differences.

Eight were remnants of the formerly-extensive rainforest on the basaltic Blackall Range, seven had a sclerophyll forest component (e.g. trees from Eucalyptus and Lophostemon) and the remaining site was littoral rainforest. Remnant sites were often situated along watercourses, in gullies and on slopes that were too steep to be cleared. The interiors of remnant patches were generally intact, although some had been selectively logged. Sites were not currently grazed by cattle.

The following fleshy fruited plant taxa were characteristic of extensive and remnant sites: palms (e.g. Archontophoenix cunninghamiana and Livistona australis), figs Ficus spp., laurels (especially Cryptocarya spp. and Endiandra spp.),

Elaeocarpaceae (Elaeocarpus and Sloanea spp.), basswoods (especially Polyscias spp.),

Sapotaceae (e.g. Pouteria spp.), Sapindaceae (e.g. Diploglottis australis) and fleshy fruited Myrtaceae (e.g. Syzygium and Acmena spp.). Fleshy fruited vines, especially the native grapes (Cissus spp. family Vitaceae), whip vine Flagellaria indica and climbing pandans Freycinetia spp. (both genera from the Pandanaceae) were common throughout extensive and isolated remnant sites. Some remnant sites seemed to contain a greater proportion of pioneer species such as bleeding heart Homalanthus nutans and macaranga Macaranga tanarius (both Euphorbiaceae), blackwood wattle Acacia melanoxylon (Mimosaceae) than extensive sites. These and fleshy fruited weeds such as camphor laurel Cinnamomum camphora (Lauraceae), broad- and small-leaved privet

Ligustrum lucidum and L. sinense (Oleaceae), wild tobacco Solanum mauritianum

(Solanaceae) and lantana Lantana camara (Verbenaceae) were usually found in areas of ongoing disturbance, such as around walking tracks or near edges, especially in smaller remnants.

Regrowth sites were located mostly on former cattle pasture that had been regenerating for at least a decade. The floristic and structural composition of regrowth

40 sites differed from remnant and extensive forest sites. Patches with a developed tree species layer about 10 to 15 m in height were chosen. Regrowth sites generally contained a lower abundance and diversity of large-diameter trunks than remnant and extensive forest sites. Regrowth sites commonly contained the following plant taxa: sandpaper figs (Ficus coronata and F. fraseri Moraceae), jackwood Cryptocarya glaucescens, bleeding heart Homalanthus nutans, basswood Polyscias elegans

(Araliaceae), wild quince Guioa semiglauca (Sapindaceae), and piccabeen palms

Archontophoenix cunninghamiana. The suite of introduced woody weeds from the

Lauraceae, Oleaceae, Solanaceae and Verbenaceae that occurred in disturbed areas in remnants was common in most regrowth sites. Fleshy fruited vines were also common.

There were more patches of rainforest throughout extensive forest mosaics than in the landscapes surrounding remnant and regrowth sites. Individual remnant and regrowth sites varied in their isolation from other forest, with some sites located within relatively well forested (>50% forest cover) areas, many sites in moderately forested

(30-50%) parts of the landscape, and several in areas where over 70% of the forest had been cleared. Regrowth sites were often situated in more highly cleared parts of the landscape than remnant sites. Sites of the same type were separated by at least 2 km, and most were more than 5 km apart. Sites of different types were also usually well separated, although there were four cases where a remnant and a regrowth site were closely situated. Most remnant and regrowth sites were between five and 10 km from extensive forest, although some were located further away. Remnant sites ranged in size

(including interspersed eucalypt forest) from two to 100 ha (approximate mean 46.1 ha,

S.E. 9.4 ha) and regrowth sites were between approximately two and 10 ha in size

(mean 3.4 ha, S.E. 0.5 ha).

41 2.2.3 Bird surveys

The quantitative measure of bird abundance was the number of individuals of each frugivorous bird species seen or heard during a 40 minute visit to each 1 ha plot.

Bird counts were conducted within four hours of dawn and involved walking throughout the plot as many times as possible, following up on movements and sounds of falling fruit. A combination of visual detection and call recognition was used to identify the bird species. Most of the frugivorous bird species have loud, distinctive calls, making them equally detectible across site types. Small, canopy-dwelling species (e.g. mistletoebird) may have been under-recorded if they were not calling. The author surveyed all sites. Bird surveys were not conducted during strong wind or heavy rain.

Each plot was surveyed in this manner on four separate occasions; twice during

January-March (summer) and twice between July and September (winter) in 2001.

Consecutive surveys at any site were no less than three weeks apart. The total observation time at each site was 160 minutes; 80 minutes in both summer and winter.

The data on frugivorous bird distribution deriving from these bird surveys was used in subsequent chapters of this thesis (Chapters Three, Five and Six). The distribution of frugivorous bats was assessed using the same site network (reported in

Chapter Six).

2.2.4 Data treatment

The number of individuals of each frugivorous bird species was summed across the two visits made during a season. Data for species that were recorded in less than five sites in either season were not statistically analysed because their frequency was considered too low to determine a distribution pattern. A two-way split plot Analysis of

Variance (ANOVA) was used to test whether the abundance of birds that were recorded in at least five sites during both seasons varied between site types (three levels:

42 extensive, remnant, regrowth) and seasons (two levels: summer, winter). Season was used as the split, with site nested within site type (site:site type) as the error term when testing for effect of site type, and site:site type x season as the error term when testing for the effect of season or the interaction between season and site type. Where a species was recorded in at least five sites during one season only, a one-way ANOVA was conducted on the data from only that season to test for an effect of site type on abundance, and a paired t-test was used to test whether the difference in numbers between seasons was significant. A species was considered to show a substantial difference in numbers between summer and winter if the ANOVA result was significant and the abundance turnover exceeded 50% (after Catterall et al., 1998). The method used to calculate seasonal turnover:

(max. – min.) x 100 percent abundance turnover = max.

Where: max. is number of individuals recorded in the season in which the species was most

common; and

min. is the number in the season in which it was least common.

Where there was a significant effect of site type, Least Significant Difference

(LSD) comparisons were conducted to test for pair-wise differences. ANOVA procedures and LSD tests were conducted using the SAS statistical package (SAS Institute 1999).

Multidimensional scaling ordination using the semi-strong hybrid technique

(Faith et al., 1987) in WinPATN (Belbin et al., 2003) was used to describe differences among the 48 sites in terms of patterns of variation in frugivorous bird species composition. Data for 39 bird species were included in these analyses; three species that were detected at only one site (rock dove, blue-faced honeyeater and house sparrow)

43 were not included. Principal axis correlations were conducted to determine associations between the site ordination and the abundance of each bird species; associations found to be significant (at p<0.05) using a randomisation test (10 000 iterations) were displayed as species abundance vectors in the ordination space. Analysis of Similarity

(ANOSIM; Clarke and Green, 1988) was used to test for overall and pair-wise differences (using 10 000 iterations) among the three site types in their frugivorous bird species composition.

An interaction between site elevation (three levels: high (N=5), mid (N=6) or low (N=5)) and season (two levels: summer and winter) on selected frugivorous bird numbers in extensive forest was tested by way of a two-factor ANOVA using the PROC

GLM procedure in SAS (SAS Institute 1999). Analyses were conducted on pooled data for all frugivorous birds and separately on data for selected species (those nominated as being altitudinal migrants by Date et al. (1991)).

2.2.5 Classification of frugivorous birds

Literature searches revealed records of many bird species consuming fleshy fruit. As pointed out by Jones and Crome (1990), almost any rainforest-dwelling vertebrate will occasionally eat fleshy fruit, although some species do so very rarely.

Reference texts were used to systematically determine which of the bird species recorded during field surveys were frugivorous. A species was classified as frugivorous if frugivory was included in its dietary description in Blakers et al. (1984), or if it had been recorded consuming fruits from more than plant three genera in Barker and

Vestjens (1988, 1989) or in the Handbook of Australian, New Zealand and Antarctic

Birds (HANZAB) series (Marchant and Higgins, 1993; Higgins and Davies, 1996;

Higgins, 1999; Higgins et al., 2001)). Parrots, lorikeets, rosellas, cockatoos (Green,

1993) and some pigeons (Frith, 1982) grind or crush many, if not most of the seeds

44 from the fleshy fruits they consume. Although such birds may have relatively low potential as seed dispersers (Snow, 1981), they were included in the list of avian frugivores if they satisfied the previous criteria.

2.3 Results

2.3.1 Abundance of frugivorous bird species in extensive, remnant and regrowth sites

Using the criteria stated in Section 2.2.5, 42 bird species were classified as frugivores (Table 2.1). In total, 2768 individuals from these species were recorded during surveys. Other species that were known to eat fruit infrequently were recorded during surveys but these did not meet the stated criteria and are not considered further in this study. Of the 42 frugivorous bird species recorded during surveys, 26 were sufficiently common (present at five or more sites in at least one season) for statistical analyses (Table 2.1). Twelve of these 26 species showed a statistically significant

(P<0.05) difference in abundance among the three site types in one or both seasons. The patterns of abundance change between remnant and/or regrowth sites and extensive forest are indicated in Table 2.2. Eight frugivorous bird species showed a significant difference between seasons together with greater than 50% seasonal turnover in abundance. The rose-crowned fruit-dove Ptilinopus regina and scarlet honeyeater

Myzomela sanguinolenta showed a significant interaction between site type and season.

Patterns of statistically significant differences in abundance between site types (Table

2.2, Figure 2.3) grouped readily into three classes:

1. decreasers: species that showed lower numbers outside extensive forest in remnant and/or regrowth sites;

2. increasers: species that showed higher numbers outside extensive forest in remnant and/or regrowth sites; and

45 3. tolerant: no significant difference in numbers between remnants, regrowth and extensive forest.

There were five decreaser species (Tables 2.2, 2.3), three of which were fruit- doves (Ptilinopus spp.). The fruit-doves generally showed declining abundance from extensive forest through remnants to regrowth (Tables 2.2, Figure 2.3(i-ii)). The other two decreasers, the brown cuckoo-dove and green catbird, showed similar abundance in extensive and remnant forests but were less common in regrowth (Table 2.2, Figure

2.3(iii)). There were seven increaser species (Tables 2.2, 2.3, five of which (rainbow lorikeet, black-faced cuckoo-shrike, figbird, Torresian crow and silvereye) were significantly more abundant in regrowth than in either remnant or extensive forest

(Table 2.2, Figure 2.3(v)). The increaser bar-shouldered dove showed similar abundance between remnant and regrowth sites and was absent from extensive forest (Figure

2.3(iv)), while the Australian magpie was present, but least abundant in extensive forest, and most abundant in regrowth, with numbers in remnant forest intermediate (Table 2.2,

Figure 2.3(vi)). The remaining 14 species were classified as tolerant (Table 2.3), since their numbers did not differ significantly among site types (Table 2.2, Figure 2.3(vi)).

46 Table 2.1 Frugivorous bird species recorded in this study. Nomenclature and order follow Christidis and Boles (1994) († indicates seed grinder (likely to destroy seeds) and * indicates introduced species). ‘Number of sites’ indicates the number of sites (out of 48) in which the species was recorded in summer (two surveys), winter (two surveys) and across all surveys. Analyses (+) shows species that were analysed statistically. Number of sites: Common name Scientific name summer winter all Analyses surveys Australian brush-turkey† Alectura lathami 16 14 23 + rock dove†* Columba livia 0 1 1 white-headed pigeon† C. leucomela 16 13 22 + brown cuckoo-dove† Macropygia amboinensis 40 37 42 + emerald dove† Chalcophaps indica 12 8 15 + bar-shouldered dove† Geopelia humeralis 15 16 23 + wonga pigeon† Leucoscarcia melanoleuca 3 3 5 wompoo fruit-dove Ptilinopus magnificus 18 20 25 + superb fruit-dove P. superbus 12 1 13 + rose-crowned fruit-dove P. regina 36 6 36 + topknot pigeon Lopholaimus antarcticus 5 1 6 + galah† Cacatua roseicapilla 2 1 3 sulphur-crested cockatoo† C. galerita 8 14 16 + rainbow lorikeet† Trichoglossus haematodus 29 22 32 + scaly-breasted lorikeet† T. chlorolepidotus 4 3 7 Australian king-parrot† Alisterus scapularis 12 18 23 + crimson rosella † Platycercus elegans 3 1 4 pale-headed rosella† P. adscitus 4 10 14 + common koel Eudynamys scolopacea 17 0 17 + channel-billed cuckoo Scythrops novaehollandiae 7 0 7 + little wattlebird Anthochaera chrysoptera 0 6 6 + noisy friarbird Philemon corniculatus 0 3 3 blue-faced honeyeater Entomyzon cyanotis 0 1 1 noisy miner Manorina melanocephala 3 1 3 Lewin's honeyeater Meliphaga lewinii 48 48 48 + scarlet honeyeater Myzomela sanguinolenta 17 26 33 + black-faced cuckoo-shrike Coracina novaehollandiae 3 10 10 + barred cuckoo-shrike C. lineata 2 0 2 varied triller Lalage leucomela 2 2 4 olive-backed oriole Oriolus sagittatus 1 1 2 figbird Sphecotheres viridis 31 36 41 + grey butcherbird Cracticus torquatus 2 2 4 Australian magpie Gymnorhina tibicen 18 20 26 + pied currawong Strepera graculina 23 35 38 + paradise riflebird Ptiloris paradiseus 4 3 5 Torresian crow Corvus orru 20 34 35 + green catbird Ailuroedus crassirostris 32 28 35 + regent bowerbird Sericulus chrysocephalus 1 4 5 satin bowerbird Ptilonorhynchus violaceus 7 2 7 + house sparrow†* Passer domesticus 0 1 1 mistletoebird Dicaeum hirundinaceum 3 2 5 silvereye Zosterops lateralis 4 18 20 +

47 Table 2.2 Frugivorous bird species’ abundances in each of the three site types, during summer and winter. Nomenclature and order follow Christidis and Boles (1994). The mean abundance of individuals (No. individuals per hectare per hour; summed across two 40-minute surveys in 1 ha) is shown for all sites (Total), in Extensive forest (Ext, 16 sites), Remnants (Rem, n=16), and Regrowth (Reg, n=16). ANOVA p shows results of analyses testing for differences in abundance between site types (ST), seasons (S) and STxS. x indicates season for which effect of site type was not tested (species too infrequent). Letters next to means show LSD results (means with different letters are significantly different). Abund. pattern corresponds with Figure 2.2 (i to iii are “decreasers”, iv to vi “increasers”, and vii “tolerant”).

Mean abundance ANOVA p values Abund. Bird species Season Total Ext Rem Reg ST S ST x S pattern Australian brush-turkey s 0.63 0.44 0.82 0.63 0.35 0.17 0.96 vii w 0.38 0.13 0.56 0.44 white-headed pigeon s 0.63 0.38 0.88 0.63 0.10 0.55 0.11 vii

w 0.77 0.19 0.63 1.50 brown cuckoo-dove s 2.04 2.75a 2.63a 0.75b 0.02 0.20 0.06 iii w 2.46 2.06 3.63 1.69 emerald dove s 0.42 0.25 0.56 0.44 0.43 0.07 0.89 vii w 0.23 0.13 0.38 0.19 bar-shouldered dove s 0.52 0.00b 0.88a 0.69a 0.002 0.67 0.29 iv w 0.46 0.00 0.50 0.88 wompoo fruit-dove s 1.21 2.65a 1.00b 0.00c 0.0002 0.87 0.71 i w 1.25 2.65 0.82 0.31 superb fruit-dove s 0.29 0.56a 0.25ab 0.06b 0.031 0.00122 ii w 0.02 0.00 0.06 0.00 x rose-crowned fruit-dove s 1.88 2.81a 2.00b 0.81c 0.0021 0.0001 0.04 i w 0.23 0.56 0.13 0.00 0.071 topknot pigeon s 0.44 1.06 0.06 0.19 0.241 0.332 vii w 0.15 0.00 0.00 0.44 x sulphur-crested cockatoo s 0.35 0.56 0.31 0.19 0.07 0.54 0.08 vii w 0.42 0.94 0.13 0.19 rainbow lorikeet s 2.52 1.81b 1.19b 4.56a 0.01 0.08 0.16 v w 1.73 0.94 1.50 2.75 Australian king-parrot s 0.35 0.38 0.38 0.32 0.31 0.17 0.34 vii w 0.65 1.06 0.63 0.26 pale-headed rosella s 0.15 0.25 0.19 0.00 x w 0.69 0.00 0.75 1.31 0.101 0.362 vii common koel s 0.48 0.38 0.38 0.69 0.401 0.00012 vii w 0.00 0.00 0.00 0.00 x channel-billed cuckoo s 0.17 0.06 0.19 0.25 0.461 0.012 vii w 0.00 0.00 0.00 0.00 x little wattlebird s 0.00 0.00 0.00 0.00 x w 0.15 0.00 0.18 0.25 0.211 0.022 vii Lewin's honeyeater s 4.29 4.13 4.75 4.00 0.46 0.04 0.53 vii w 3.79 3.88 3.88 3.63 48 scarlet honeyeater s 0.60 1.00 0.56 0.25 0.061 0.12 0.03 vii w 0.92 0.56 1.13 1.06 0.271 black-faced cuckoo-shrike s 0.08 0.00 0.00 0.25 w 0.60 0.13b 0.06b 1.63a 0.0061 <0.00012 v figbird s 3.60 1.00b 2.56b 7.25a 0.0006 0.26 0.81 v w 4.96 1.56 4.94 8.38 Australian magpie s 1.40 0.00c 1.00b 3.19a <0.0001 0.31 0.17 vi w 1.08 0.13 1.06 2.06 pied currawong s 0.98 0.88 1.06 1.00 0.42 0.0004 0.40 vii w 2.31 1.88 3.06 2.00 Torresian crow s 1.04 0.25b 0.82b 2.06a 0.0001 0.02 0.81 v w 1.85 1.19 1.32 3.06 green catbird s 1.58 1.94a 2.44a 0.38b 0.0001 0.55 0.89 iii w 1.46 1.81 2.19 0.38 satin bowerbird s 0.19 0.13 0.19 0.25 0.751 0.452 vii w 0.10 0.06 0.00 0.25 x silvereye s 0.42 0.00 0.00 1.25 x w 2.59 1.06b 1.69b 5.00a 0.031 0.0022 v 1 p value from single-factor ANOVA testing site type effect within season 2 p value from paired t-test of difference between seasons; all other p values from two-way ANOVA

49

DECREASERS INCREASERS TOLERANT i) iv) vii) a a a a a a a a a 4 a a a 1 4 bb 2 cc b 2 b Abundance Abundance

0 0 0 a ii) a v) Ext Rem Reg a a 0.4

abab 8 b b 0.2 b b 4 bb

Abundance Abundance 0 0

iii) a a vi) a 3 4

2 b bb 2 c 1 Abundance Abundance

0 0

Ext Rem Reg Ext Rem Reg

Figure 2.3 Examples of the seven patterns of abundance in remnants and regrowth compared with extensive forest. Abundance (average of summer and winter data) shows mean and standard error in Ext = Extensive forest tracts; Rem = Remnant forest; and Reg = Regrowth patches; Patterns are exemplified using data from selected bird species: i) wompoo fruit-dove, ii) superb fruit-dove, iii) green catbird, iv) bar-shouldered dove, v) figbird, vi) Australian magpie, and vii) Lewin’s honeyeater. Means with different letters are significantly different (P <0.05 in LSD comparisons); see also Tables 2.1 and 2.2.

50

Table 2.3 Frugivorous bird species' responses to rainforest fragmentation, and their seasonality. Numerals (i-vii) show the pattern of abundance change among the three site types (see text, Fig. 2.3 and Table 2.2); Season shows the time of greater abundance if the effect of season was significant and turnover exceeded 50%. † indicates seed-crusher (likely to destroy seeds). Species Season i wompoo fruit-dove Decreasers rose-crowned fruit-dove summer ii superb fruit-dove summer iii brown cuckoo-dove† green catbird iv bar-shouldered dove† v rainbow lorikeet† Increasers black-faced cuckoo-shrike winter figbird Torresian crow silvereye winter vi Australian magpie vii Australian brush-turkey† white-headed pigeon† emerald dove† topknot pigeon Tolerant sulphur-crested cockatoo† Australian king-parrot† pale-headed rosella† common koel summer channel-billed cuckoo summer little wattlebird winter Lewin’s honeyeater scarlet honeyeater pied currawong winter satin bowerbird

2.3.2 Changes in the frugivorous bird assemblage in fragmented forest

Figure 2.4(i) displays an ordination of the 48 study sites based on the abundance of frugivorous bird species. The extensive forest sites are grouped towards one extreme of the ordination space, with regrowth sites at the other and remnant sites intermediate in terms of bird species composition. The composition of the frugivorous bird assemblage varied significantly among the three site types (ANOSIM global p<0.001) and between all site types in separate pair-wise comparisons (ANOSIM p = 0.001 for extensive versus remnant sites; p<0. 0001 for both other comparisons).

51 Figure 2.4(ii) shows that the bird species associated with the region of the ordination containing most of the extensive forest sites included the five decreaser species. In addition, the topknot pigeon and sulphur-crested cockatoo, both of which showed a decreasing trend (Table 2.2) and the paradise riflebird, which was only recorded in extensive forest (Table 2.4), were associated with this region of the ordination. The bird species associated with regrowth sites included the seven increaser species as well as two non rainforest species (noisy miner and grey butcherbird) and the pale headed rosella (Figure 2.4(ii)).

52

i)

ii) topknot pigeon (Tol)

paradise riflebird (U) noisy miner (U)

pale-headed rosella (Tol) sulphur-crested cockatoo (Tol) bar-shouldered dove (inc) wompoo fruit-dove (Dec) green catbird (Dec) grey butcherbird (U) figbird (Inc) brown cuckoo-dove (Dec) silvereye (Inc) rose-crowned fruit-dove (Dec) Australian magpie (Inc) black-faced cuckoo-shrike (Inc)

rainbow lorikeet (inc) superb fruit-dove (Dec) Torresian crow (inc) noisy friarbird (U)

Figure 2.4 (i) Ordination of the 48 study sites based on the abundances of 39 frugivorous bird species (Stress = 0.28). Extensive forest (filled square), remnants (open diamond), regrowth (filled triangle). (ii) Abundance vectors for bird species significantly (p <0.05) associated with the ordination. Fragmentation response patterns are shown in brackets: Dec decreaser, Tol tolerant, Inc increaser, U untested.

53 2.3.3 Seasonal patterns of frugivorous bird abundance

The common koel and channel-billed cuckoo were absent in winter, yet relatively common during summer, clearly the result of immigration. The rose-crowned and superb fruit-doves also showed large and significant summer increases, although they were present in low numbers during winter months, a pattern that is also consistent with immigration into the study region. The black-faced cuckoo-shrike, silvereye, little wattlebird, and pied currawong were recorded in substantially higher numbers during winter than summer. Numbers of the Lewin’s honeyeater and Torresian crow also differed between seasons, but their seasonal abundance turnover was less than 50%, and was probably due to factors such as reproduction or local movements rather than larger- scale migration. The remaining 18 species showed no significant difference in abundance between seasons. The decreasing response pattern detected for the rose- crowned fruit-dove was significant only during summer (Table 2.2). During winter, numbers of this species were similar across site types, although its abundance was very low. The significant interaction detected in the ANOVA for the scarlet honeyeater was not supported by LSD tests, although a tendency towards a decreasing response pattern was shown in summer, with a trend towards an increasing response pattern in winter (Table

2.2).

2.3.4 Effects of altitude and season on frugivorous bird numbers

The numbers of frugivorous birds (data for all species pooled) and of the wompoo and rose-crowned fruit-doves, white-headed and topknot pigeons and the brown cuckoo-dove in extensive forest at different elevations in summer and winter is shown in Table 2.5. No significant (p<0.05) interactions were detected between the two factors using ANOVA, indicating that the data on bird abundance patterns were not strongly influenced by altitudinal movements of rainforest pigeons. Note that the superb

54 fruit-dove was not recorded at any extensive forest sites, and very few other sites, in winter.

55 Table 2.5 Frugivorous bird abundance pattern in high, mid- and low elevation sites during summer (s) and winter (w). The mean (and standard error) number of individuals of all frugivores and selected rainforest pigeons is shown for each season (data from two surveys summed) in each elevation category; ANOVA p shows results of two-way ANOVA (E=elevation, S=season and ExS=interaction).

Elevation category (m a.s.l.) ANOVA p Bird species Seas. >500 (n=5) 200-500 (n=6) <200 (n=5) E S ExS all frugivorous birds s 30.20 (4.24) 23.33 (1.92) 20.80 (2.85) 0.68 0.40 0.39 w 20.80 (3.20) 21.17 (5.35) 23.40 (5.85) white-headed pigeon s 0.00 0.83 (0.39) 0.20 (0.20) 0.29 0.39 0.14 w 0.40 (0.41) 0.17 (0.16) 0.00 wompoo fruit-dove s 4.00 (0.96) 2.00 (0.57) 2.00 (0.85) 0.24 0.98 0.48 w 3.00 (1.44) 1.67 (0.74) 3.40 (1.31) rose-crowned fruit-dove s 2.00 (0.56) 3.50 (0.75) 2.80 (1.22) 0.51 0.0007 0.35 w 0.40 (0.25) 0.17 (0.16) 1.20 (0.81) brown cuckoo-dove s 3.40 (0.76) 2.83 (0.30) 2.00 (0.64) 0.15 0.16 0.56 w 2.00 (0.85) 2.67 (0.33) 1.40 (0.61)

56

2.4 Discussion

In this study, extensive forest sites were treated as a reference against which to quantify changes in numbers of frugivorous bird species in remnants and regrowth.

Since remnants and extensive forest sites were similar in fleshy fruited plant species composition, differences in frugivorous bird numbers between these two site types were most likely due to differences in site context rather than resource availability within the site. Differences in frugivorous bird numbers between remnants and extensive forest may reflect a response to several factors associated with the differing landscape context, including reduced total area of habitat, edge effects, or greater functional isolation. The patterns of bird abundance in regrowth sites reflect differences in both the availability of fleshy fruit resources and the landscape context.

2.4.1 Bird species showing a decreaser response to rainforest fragmentation

A number of studies in different parts of the world have documented bird declines and local extinctions in fragmented rainforest (e.g., Kattan et al., 1994; Corlett,

1998; Renjifo, 1999; Castelletta et al., 2000; Ribon et al., 2003). Consistent with descriptions provided by Frith (1952), the wompoo, rose-crowned and superb fruit- doves were generally less abundant in remnants and regrowth than extensive forest.

Despite being known to fly across cleared land (Frith, 1952; Howe et al., 1981; Date et al., 1991, 1996; Gosper and Holmes, 2002), fruit-doves used remnant and regrowth rainforest habitats in the Sunshine Coast much less frequently than extensive forest areas. The Australian fruit-doves are obligate frugivores and the plants that characterise their diets are typical of mature rainforest (Crome, 1975, 1990; Innis, 1989).

Consequently, their low numbers in fragmented forest may be due to low availability of

57 the required plant species in remnants and in the local landscapes surrounding remnant sites.

Although fruit-doves showed overall decreasing abundance in remnants compared with extensive forest, their abundance in certain individual remnants resembled that of extensive rainforest sites. These remnant sites, plotted among extensive forest sites in the ordination space, were mostly located in relatively well forested parts of the landscape. These remnants may function as part of a network of patches (Howe et al., 1981; Date et al., 1991, 1996; Price et al., 1999) that, although discontinuous, are sufficiently close that the energetic costs and perceived predation risk associated with movement between patches are not too high. In Mexico, Graham (2001) described a similar situation for the strong-flying keel-billed toucan Ramphastos sulphuratus, which moved between forest fragments, but only in parts of the landscape that contained at least a minimum threshold amount of forest and fruit resources.

The brown cuckoo-dove and green catbird showed much lower abundance in regrowth than in remnants and extensive forest. The brown cuckoo-dove is noted for its conspicuousness in regrowth vegetation at forest edges and consumes a range of plants that are common in rainforest regrowth (Frith, 1952; Crome, 1975; Willson and Crome,

1989). However, this species roosts in well-developed forest (Frith, 1982) and may therefore be limited in its use of isolated regrowth patches, unless these are located close to mature forest. Furthermore, the brown cuckoo-dove and green catbird are sedentary or only locally nomadic in subtropical rainforests (Frith, 1982; Blakers et al.,

1984; Innis and McEvoy, 1992; Date et al., 1996). These species may require a larger area of contiguous vegetation than is provided within or in local landscapes surrounding most of the regrowth patches surveyed in this study. In general, these areas occur as narrow strips or small patches with low amounts of surrounding forest cover compared with many remnant sites. The data presented in this chapter may indicate area-

58 sensitivity of these species at very small patch sizes. Data presented by Howe et al.

(1981) also suggest that brown cuckoo-doves may be unable to persist in very small remnants; they were ‘common’ in patches of 1 ha to 2.5 ha, but ‘rare’ in those under 1 ha. While both the brown cuckoo-dove and green catbird are highly frugivorous, the former grinds ingested seeds in a muscular gizzard (Frith, 1982; Dennis and Westcott,

2006) and the later eats flowers, invertebrates, and the eggs and nestlings of other birds in addition to fruit (Blakers et al., 1984). The consumption of alternative food sources that are high in protein relative to the pulp of most fruits (Morton, 1973) may mean that these species are less constrained in their fruit preferences than the fruit-specialists that rely only on the nutritional quality of fruit pulp (Snow, 1981). They may consequently be able to satisfy their energetic and nutritional needs in the smaller area of forest available in remnants than the obligate frugivore species.

Species declines in fragmented habitats may also result from changed biotic interactions (Terborgh and Winter, 1980; Karr, 1982; Doak et al., 1992). For example, the aggressive, non-rainforest noisy miner invaded regrowth patches from the surrounding matrix. This species has been shown to exclude other bird species in open eucalypt forest (Piper and Catterall, 2003). Rates of nest predation may also increase in fragmented rainforest (Sieving and Karr, 1997). However, it is not clear whether the species classified as decreasers in the present work would be disproportionately affected by these changes.

The differences in bird species’ abundances between habitat types in the present study were clearly the result of habitat choice by birds within individual species. For example, four regrowth sites were located adjacent to, or within tens of metres of, remnant sites (Figure 2.2). However, these remnant and regrowth sites contained different bird assemblages and were plotted in separate regions of the ordination based on bird species’ abundance. At one of these site pairs, several bird species were present

59 in the remnant site but never recorded in the regrowth (e.g. the wompoo and rose- crowned fruit-doves, and green catbird), or were recorded in much lower numbers in the regrowth than adjacent remnant rainforest (e.g. brown cuckoo-dove). These results suggest greater habitat preferences of frugivorous birds in subtropical Australia than shown by Laurance et al. (1996) in their comparison of adjoining extensive and regrowth rainforest patches in the Atherton Tableland area of North Queensland.

2.4.2 Bird species showing an increaser response to fragmentation

As well as decreasers, the present work detected several species that increased in abundance in remnant and regrowth rainforest compared with extensive forest.

Consistent with observations made during the 1950s in subtropical rainforest remnants

(Frith, 1952), the seed-crushing bar-shouldered dove was absent from extensive forest but invaded some rainforest remnant and regrowth patches. This may reflect greater availability of grasses or other food within and surrounding remnants and regrowth patches (Frith, 1952).

The black-faced cuckoo-shrike, figbird, rainbow lorikeet, Torresian crow and silvereye were found in similar abundance in remnants and extensive rainforest but were most abundant in regrowth. These species commonly use non-rainforest habitats, in contrast with the rainforest-dependent decreaser species (Blakers et al., 1984; Catterall et al., 1998). Furthermore, with the exception of the highly frugivorous figbird, these increaser frugivores regularly eat a variety of other food types (Blakers, et al. 1984). As suggested for the bar-shouldered dove, high numbers of these species in regrowth may reflect ready use of the types of resources occurring within and surrounding regrowth patches. These may include non-fruit food types such as the high nectar availability in ornamental garden plantings, or resources like fleshy fruited weeds that would boost the

60 availability of food at various times for frugivores that may not be more opportunistic than fruit-specialists.

Fleshy fruited weeds, in particular camphor laurel, have been identified as potentially important food sources for the topknot and white-headed pigeons (Frith,

1982; Innis, 1989; Date et al., 1996). It has been suggested that the spread of camphor laurel has contributed to an increase in numbers of these species throughout certain

Australian subtropical rainforest landscapes following dramatic declines during the early half of 1900s (Frith, 1952, 1982; Date et al., 1996). Importantly, camphor laurel bears fruits at a time when the availability of fruiting native rainforest plants is low

(Innis, 1989; Date et al., 1991; Scanlon et al., 2000). Unlike the fruit-doves, neither the topknot nor white-headed pigeon, both fruit-specialists, showed significantly different numbers in rainforest remnants or regrowth compared with extensive forests in the present study, although topknot pigeon numbers showed a decreasing trend.

2.4.3 Frugivore assemblage change in fragmented habitats

The abundance of over half of the bird species (14 of 26) evaluated in the present work was similar among extensive forest, remnants and regrowth. It has been proposed that rainforest contraction in regions such as Africa and Australia during the

Pleistocene may have imposed an ‘extinction filter’ on rainforest fauna (Howe et al.,

1981; Balmford, 1996; Corlett and Primack, 2006). As a result, the species that have persisted may be those that are unspecialised and with the capacity for movement between disjunct rainforest patches. These attributes may be associated with relatively high tolerance of anthropogenic forest fragmentation (Howe et al., 1981; Balmford,

1996; Corlett and Primack, 2006).

The changes in the frugivorous bird assemblage in fragmented forest in the study region resemble those described by a ‘cut-and-paste’ model (Woinarski 1993; Crome et

61 al., 1994). That is, the avian frugivore communities in remnant and regrowth sites comprised species from diverse habitats, formed by the decline of some species and concurrent increase of others, in response to changes in habitat quality, area and/ or landscape forest cover. This model better describes the assemblage changes documented between extensive, remnant and especially regrowth sites in the present study than those that emphasise declines (e.g. ‘nested subsets’; Patterson, 1987).

2.4.4 Seasonal changes in frugivorous bird abundance

A greater abundance and diversity of native fleshy fruits are generally available in subtropical Australian rainforests during summer than winter (Innis, 1989; Church,

1997). During winter, fruit availability seems to be highest in lowland areas (Innis,

1989; Date et al., 1991). The seasonal differences in fruit availability may influence the abundance of frugivorous birds in rainforest habitats within the study region. Numbers of the common koel, channel-billed cuckoo, superb and rose-crowned fruit-doves increased substantially during summer and all are regular summer immigrants to the study region. The first two species are total migrants (Higgins, 1999). The rose-crowned fruit-dove is a partial migrant, with some individuals over-wintering in forests within the study region while the majority of individuals appear to return to tropical forests in northern Australia or Papua New Guinea (Blakers et al., 1984). The superb fruit-dove is considered vagrant in subtropical Australia.

Numbers of the little wattlebird, black-faced cuckoo-shrike, pied currawong and silvereye increased substantially during winter. The higher winter numbers of silvereyes reflect an influx of individuals of this species from the south to the study region

(Blakers et al., 1984). The silvereye and the black-faced cuckoo-shrike were classed as increasers in the present study, while the other two winter-abundant species (little wattlebird and pied currawong) were classified as tolerant. This contrasted with the bird

62 species that were more common in summer than winter in that they were either decreasers (the fruit-doves) or classed as tolerant (common koel and channel-billed cuckoo). All four winter-abundant bird species make use of remnant and especially regrowth rainforests, and their increased winter abundance may indicate a response to winter fruit availability in regrowth habitat, potentially including the winter-fruiting weeds. Indeed, silvereyes of a subtropical island population were found to increase their intake of fruit during winter (Catterall, 1985), and pied currawongs have been reported to move from eucalypt open forests into rainforest during winter (Lindsey,

1995), concurrent with a dietary shift from insects to fruit (Blakers et al., 1984). There is some evidence of a winter influx of the little wattlebird into eastern Queensland

(Blakers et al., 1984). This species usually occupies coastal eucalypt forests and heathlands rather than rainforests (Blakers et al., 1984), but the occurrence of this species in coastal remnant and regrowth rainforest in the Sunshine Coast, possibly reflects increased fruit intake during winter.

Date et al. (1991, 1996) suggested that there may be seasonal altitudinal migration in some species of rainforest pigeon, and proposed the general scenario of movement into upland forests during summer and lowland forests during winter. The surveys conducted for the present study may have been too infrequent to detect detailed seasonal movement patterns, but would have captured substantial turnover between altitudes. The results do not show a seasonal exchange of frugivorous birds between extensive forest sites at different elevations.

2.4.5 Frugivorous birds and seed dispersal in remnant and regrowth rainforest:

conservation implications

The wompoo fruit-dove suffered population declines and localised extinctions from southern parts of its range (the southern limits of subtropical rainforest in

63 Australia) during the early part of the 20th century (Recher et al., 1995) and appeared to be declining in other parts of subtropical Australia from the late 1920s, following widespread rainforest clearing and fragmentation. This prompted Frith (1952) to predict that this species was …”doomed to early extinction” (pp.91-92). Frith (1952) also forecast ongoing decline in superb fruit-dove populations as a result of rainforest loss but suggested the nomadic behaviour of rose-crowned fruit-doves would give them greater resilience to habitat destruction and fragmentation (Frith, 1982). Patterns of frugivorous bird abundance in Sunshine Coast habitats suggest that neither regrowth nor remnant rainforest patches provide suitable habitat for significant numbers of these three fruit-dove species.

Changed dispersal of rainforest plants has been predicted as a consequence of frugivore declines in fragmented landscapes throughout the world (Howe, 1984; Crome

1990; Janzen and Vasquez-Yanez, 1991; Corlett, 1998; Sodhi et al., 2004; Terborgh and

Nuñez-Iturri, 2006). Silva and Tabarelli (2000) predicted that around 30% of native plant species could be lost from forest fragments in Brazil, based on an assessment of the patterns of frugivore decline in that region and the potential for frugivore species to disperse large seeds and certain plant families. While the present study classified a group of decreaser frugivore species, it also showed the numerical replacement of these by a suite of increaser species. Increasers also potentially disperse rainforest seeds, but it is unclear whether these species may substitute in fragmented forests by dispersing the same suite of plant species as birds from the decreaser group. In subtropical

Australia, the fruit-doves may swallow and disperse larger fruits and seeds than most other frugivorous species (Green 1993). If increaser or tolerant species do not disperse the same large fruited plant species as the decreasers, seedlings of such plants may not recruit in many rainforest regrowth or remnant patches. Even if non-decreaser frugivores disperse similar plant species to decreasers, rates of seed dispersal for these

64 plant species will be lower, unless tolerant or increaser frugivores show density compensation (Renjifo, 1999; Loiselle and Blake, 2002) or increased rates of fruit consumption. Higher-order interactions involving seed or seedling predators may also change in fragmented forest landscapes, and may exacerbate or offset the effects of changed seed dispersal (Harrington et al. 1997; Murray and Garcia, 2002; Wright et al.

2002). Following seed dispersal, factors that influence plant establishment, growth and survival determine regeneration outcomes (Wang and Smith, 2002).

The seeds of plants dispersed by increaser birds are likely to be moved into and around fragmented forest landscapes at greater rates than in extensive forest. It has been suggested that fruits consumed by mixed diet, opportunistic frugivores, such as characterise the increaser species of the present study, may be mostly sugary, watery and small seeded (McKey, 1975). Many fleshy fruited weeds fit this description

(Richardson et al., 2000) and their increased dispersal and recruitment in remnants and regrowth can be expected as a result of the regular use of these habitats by the increaser bird species. This may lead to positive feedback cycles between the fleshy fruited weeds and the fragmentation-tolerant opportunistic frugivores in regrowth areas of highly disturbed rainforest landscapes.

Qualitative aspects of seed dispersal may also change in fragmented forests. For example, the abundance of two seed-crusher species increased (bar-shouldered dove and rainbow lorikeet), while only one (the brown cuckoo-dove) decreased in fragmented forest habitats. This may mean that a greater proportion of the seeds of the plant species consumed in fragmented forest are destroyed than are dispersed in viable condition, although neither of the increaser species seem to consume large amounts of fleshy fruit.

Furthermore, the lump-lined stomach of the decliner fruit-doves (Crome, 1975) potentially influences germination success of seeds. Differences among frugivore species in their use of particular habitat elements may also change spatial patterns of

65 seed dispersal in fragmented parts of the landscape (Schupp, 1993; Silva et al., 1996;

Alcántara et al., 2000; Jordano and Schupp, 2000; Loiselle and Blake, 2002; Dennis and

Westcott, 2006).

66

Chapter Three

Seed dispersal potential of frugivorous bird species in relation to their

gape width, frugivory level and seed treatment

3.1 Introduction

Fruit-eating birds may disperse the seeds of up to 70% of plant species in subtropical Australian rainforests (Willson et al., 1989; Green, 1995; Butler, 2003).

There may be variation among bird species in the plant species that they disperse because of different patterns of plant species consumption (Crome, 1975; Snow, 1981;

Herrera, 1984; Howe, 1986; Innis, 1989; Sun et al., 1997; Brown and Hopkins, 2002).

However, there is little agreement regarding the factors that influence patterns of plant species consumption by frugivore species (Herrera, 1998, 2002). Consequently, in the absence of detailed dietary information, which is typically available for only a small proportion of the species in a frugivore assemblage (e.g., Crome, 1975), there is only a limited basis for predicting patterns of plant consumption by frugivore species. Major differences among frugivore species in their patterns of plant species consumption may be related to certain morphological and behavioural attributes. If so, it may be possible to use these attributes to describe groups of frugivore species with similar combinations of attributes (functional groups) that potentially disperse similar plant species to one another.

The role that a bird species plays in the dispersal of seeds of a particular plant species depends first on whether it feeds on fruit from that plant species and second on whether it disperses viable seeds or destroys seeds. Birds may crush and destroy seeds either in the bill while feeding (parrots: Crome and Shields 1992) or by grinding during

67 digestion (some pigeons and doves: Frith 1982, and megapodes). Seed-crushing birds would make a small contribution to seed dispersal.

Patterns of plant species consumption reflect variation among frugivorous bird species in their gape sizes. Gape width imposes an intractable limit on the size of fruits and/or seeds able to be ingested by a bird species (Herrera, 1981; Moermond and

Denslow, 1985; Wheelwright, 1985). Consequently, bird species with narrow gapes are physically constrained to swallowing only small fruits, whereas birds with wider gapes are capable of consuming plant species with larger fruits (Wheelwright, 1985).

The plant species consumed by a bird species may also be influenced by the extent to which it specialises on fruit rather than other food sources. For example, it was predicted that birds with fruit-dominated diets would consume fruits with greater protein and lipid content rather than those with high sugar content (Snow, 1971; Crome,

1975; McKey, 1975; Howe, 1977; Howe and Estabrook, 1977; Stiles, 1993).

Subsequent studies have shown patterns that are inconsistent with this prediction

(Fuentes, 1994; Sun et al., 1997; Witmer and Van Soest, 1998). This may be due to interspecific variation in factors that influence the net nutritional value of fruit, such as frugivore digestive adaptations (Martínez del Rio and Restrepo, 1993) and plant secondary compounds (Cipollini and Levey, 1997). However, it may be reasonable to expect that highly frugivorous species, and particularly obligate frugivores such as the

Australian fruit-doves (Crome, 1975; Frith 1982; Recher et al., 1995), would need to consume fruits with certain energetic or nutritional values to satisfy their requirements

(White, 1993; Morton, 1973; Izhaki and Safriel, 1989; Bairlein, 1996). This may generate variation among frugivore species in the plant species that they consume. For example, highly frugivorous bird species are associated with the consumption of plants from the Lauraceae, a family noted for its high lipid content (Snow, 1971, 1981; Crome,

1975; Wheelwright, 1983; Stiles, 1993). Frugivore species that have mixed diets can

68 obtain nutrition from other food sources, most of which are rich in protein and other nutrients compared with fruit (Morton, 1973; Izhaki and Safriel, 1989). Consequently, the plant species composition of frugivore diets may vary among frugivore species depending on their degree of dependence on fruit. Furthermore, the species richness and volume of fruits consumed by a bird species may be positively related to its degree of frugivory (Wheelwright et al., 1984), although Pratt and Stiles (1985) proposed that mixed-diet species were less selective and therefore would consume a wider variety of plant species.

The identification of groups of functionally distinct frugivore species may make it possible to forecast consequences of changes in frugivore assemblage composition for seed dispersal. For example, it has been predicted that declined abundance of certain frugivorous bird species in fragmented rainforest may result in declining plant populations in fragmented landscapes (Howe, 1984; Crome 1990; Janzen and Vasquez-

Yanez, 1991; Corlett, 1998; Sodhi et al., 2004; Terborgh and Nuñez-Iturri, 2006). In particular, it has been proposed that the dispersal of large-seeded plant species will be limited in fragmented rainforest regions throughout the world (Corlett, 1996, 1998;

Harrington et al,. 1997; Silva and Tabarelli, 2000; Kitamura et al., 2002; McConkey and

Drake, 2002).

Chapter Two described the changes in the frugivorous bird assemblage resulting from either declined or increased numbers of individual species in fragmented compared with extensive rainforest in a subtropical region of Australia. The present chapter asseses the potential functional role of different frugivorous bird species in the dispersal of seeds from rainforest plants. The approach employed considers variation among frugivore species in their dietary preferences to be the primary dimension of functional variation among disperser species.

69 Bird species are categorised into functional groups using information about their gape width, frugivory level and seed-crushing behaviour. It is proposed that this approach assembles groups of bird species that potentially disperse a similar suite of plant species. Data from the field study of bird distribution in the Sunshine Coast region

(Chapter Two) is used to compare between these functional groups in terms of their numbers in rainforest remnants and regrowth, relative to extensive forest. The extent to which seed dispersal may vary between forested and fragmented parts of the landscape as a consequence of functional variation among frugivorous bird species is considered.

3.2 Methods

Chapter Two describes the study region (Section 2.2.1) and site network (2.2.2), field surveys of bird species abundance (2.2.3), and the classification of the 42 frugivorous bird species (2.2.5). The different abundance patterns that were detected among frugivorous bird species are detailed in Section 2.3.1 in Chapter Two.

Note that the frugivore attribute data used in this chapter are used in analyses in

Chapter Four.

3.2.1 Assessment of the functional attributes of frugivorous bird species: gape width,

frugivory level and seed-crushing behaviour

Vernier callipers were used to measure (to 0.01 mm) the gape width (width of the bill at the junction of the upper and lower mandibles) of skin specimens kept in the

Queensland Museum. There may have been some shrinkage of specimens, but this is should be proportional across species. Measurements were taken from 10 individuals of each species, except for the Australian king-parrot Alisterus scapulatus for which only nine specimens were available. Specimens collected from the vicinity of the study area

(i.e. southeast Queensland) were measured where possible. Where there was probably

70 sexual dimorphism in gape width, measurements were taken from five male and five female specimens. Fruit-doves and the topknot pigeon have a peculiar gape morphology which enables gape distension (Crome and Shields, 1992). To quantify the extent of this distension, the closed gape width was measured (as for dried specimens) from two thawed specimens of the wompoo fruit-dove and one of the rose-crowned fruit-dove that had been frozen fresh (hence the inter-mandibular flesh was intact) and then compared with the maximum width of a plasticine ball that could be swallowed by the specimens. The distension was similar for both species (25%) and this value was used to augment the average width of measurements taken from skins for the fruit-doves and topknot pigeon.

Each bird species was allocated to one of three diet groups, reflecting the relative dominance of fleshy fruit in their diets: major (fruit-dominated diet), mixed

(diet comprising two to several main food types, one of which was fruit), and minor

(diet dominated by foods other than fruit, but occasionally including fruit). These categories were determined from descriptions in the literature (Blakers et al. (1984) and the Handbook of Australian, New Zealand and Antarctic Birds (HANZAB) series

(Marchant and Higgins, 1993; Higgins and Davies, 1996; Higgins, 1999, Higgins et al.,

2001)) and in discussion with experts. Dietary descriptions in the literature were usually qualitative but sufficiently consistent to enable the allocation of species to a relative level of frugivory.

Bird species that potentially destroy seeds were identified from references in the literature to the destruction of seeds during feeding, grinding-based digestion, or detection of crushed seeds in faecal samples (Blakers et al., 1984; Crome and Shields,

1992; and the HANZAB series). Seed-crushing bird species were considered to have low potential as seed dispersers relative to other frugivorous birds, although some seeds, particularly if very small or hard, may pass through the bird’s gut intact (Snow, 1981).

71 Bird species for which seed-crushing behaviour was not mentioned were considered to generally disperse viable seed, although it is recognised that even birds that do not crush seeds may not always function as effective dispersers, for example if they consume the flesh of a fruit but not the seed (“fruit-thieving”) (Green, 1993).

3.2.2 Data analyses

To establish whether a bird species’ frugivory level, gape size and seed treatment were confounded with one another, pair-wise and three-way interactions were tested using a log linear model in the statistical package SPSS (2001). The factors were frugivory level (three levels, major (1), mixed (2) and minor (3)), gape width classes

(three levels, small (<10 mm), medium (10 – 15 mm) or large (>15 mm)) and seed treatment (two levels, seed-disperser and seed-crusher). The data were the number of bird species (total 42) within each cell of the three-way table of factors.

Each of the 42 bird species recorded during the field study was classed into one of the following seven functional groups based on combinations of the measured attributes of seed treatment, gape width and frugivory level. All seed-crushing bird species were grouped together (as group 8) since the seed dispersal potential of these birds is likely to be similarly low, irrespective of gape width and frugivory level. Seed- dispersing birds (species that do not crush seeds) were allocated to the following groups:

1. large-gaped (>15 mm) with fruit-dominated diet;

2. medium-gaped (10-15 mm) with fruit-dominated diet;

3. large-gaped with mixed diet;

4. medium-gaped with mixed diet;

5. small-gaped (<10 mm) with fruit-dominated diet;

6. small-gaped with mixed diet; and

72 7. fruit a minor dietary component (all gape widths pooled).

The abundance of birds within each functional group was calculated by summing the data for all species comprising each group, including those species that were too uncommon for individual analysis (recorded in less than five sites during summer and winter surveys; Chapter Two, Section 2.3.1). One-way ANOVA, together with Least Significant Difference (LSD) tests, were conducted using SPSS (2001) to test for differences in the abundance of each functional group, between extensive, remnant and regrowth sites.

The eight functional groups described above were further combined to form three groups that are proposed to differ substantially in their influence on overall seed dispersal dynamics; functional groups 1- 3 (likely to have high influence), 4 - 6

(medium), and 7 - 8 (low). An association between these three groups and sensitivity to forest fragmentation, as indicated by decreasing (1), tolerant (2) or increasing (3) abundance patterns, was tested using Spearman’s rank correlations in SPSS (2001).

Multidimensional scaling ordination using the semi-strong hybrid technique

(Faith et al. 1987) in WinPATN (Belbin et al., 2003) was used to describe differences among the 48 sites in terms of patterns of variation in functional group composition.

Data for 39 bird species were included in these analyses; three species that were detected at only one site (rock dove, blue-faced honeyeater and house sparrow) were not included. Principal axis correlations were conducted to determine associations between the site ordination and the abundance of each functional group; associations found to be significant (at p< 0.05) using a randomisation test (MCAO, 10 000 iterations) were displayed as functional group abundance vectors in the ordination space.

73 3.3 Results

3.3.1 Variation in seed dispersal potential among species within the frugivorous bird

assemblage

The 42 frugivorous bird species that occur in rainforest habitats in the Sunshine

Coast, along with their gape width, frugivory level and seed-crushing behaviour, are shown in Table 3.1. Just over one-third (15) of these species destroy seeds by crushing or grinding them. Species’ gape widths ranged from 5.2 mm (scarlet honeyeater) to 32.8 mm (channel-billed cuckoo) (Table 3.1). There were nine species with small (<10 mm) gapes, 15 with medium (10 – 15 mm) gapes, and 18 with large (>15 mm) gapes. Eleven bird species had fruit-dominated diets (major frugivores), 15 had mixed diets and 16 had diets dominated by food types other than fruit (minor frugivores). None of the interactions between frugivory level, gape width and seed treatment were significant in the Loglinear model (loglinear model likelihood ratio (L.R.) χ2 = 6.38, p = 0.90 for three-way interaction; pair-wise chi-squared tests showed no interaction between gape width and frugivory level (χ2 = 0.38, p = 0.98), frugivory level and seed treatment (χ2 =

0.82, p = 0.66), nor gape width and seed treatment (χ2 = 2.51, p = 0.29) (Figure 3.1).

74 Table 3.1 Characteristics and response pattern of frugivorous bird species recorded in the field study. Seed treat. indicates seed treatment (C = crusher, D = disperser). The abundance of individuals recorded in 1 ha (averaged across four surveys) is shown for extensive (N=16), remnant (N=16) and regrowth (N=16) sites. No. sites shows the number of sites (maximum 48) at which the species was recorded (summer plus winter surveys). Response pattern shows the abundance pattern in relation to forest fragmentation (based on analyses in Chapter Two: D = decreaser, I = increaser, T = tolerant, and U = uncertain (i.e. too rare for statistical analysis).

1 1 Seed Gape width Gape Frugivory No. Response Scientific name Common name 2 3 4 Extensive Remnant Regrowth treat. (mm) class level sites pattern Alectura lathami Australian brush-turkey C 18.3 (0.5) L Mixed 0.56 1.38 1.06 23 T Columba livia rock dove * + C 10.2 (0.4) M Minor 0.00 0.00 0.13 1 U C. leucomela white-headed pigeon C 11.8 (0.4) M Major 0.56 1.50 2.13 22 T Macropygia amboinensis brown cuckoo-dove C 10.1 (0.3) M Major 4.81 6.31 2.44 42 D Chalcophaps indica emerald dove C 8.5 (0.2) S Major 0.38 0.94 0.63 15 T Geopelia humeralis bar-shouldered dove C 6.6 (0.4) S Minor 0.00 1.38 1.56 23 I Leucosarcia melanoleuca wonga pigeon C 9.3 (0.3) S Mixed 0.50 0.06 0.00 5 U Ptilinopus magnificus wompoo fruit-dove D 19.0 (0.6) L Major 5.25 1.81 0.31 25 D P. superbus superb fruit-dove D 12.6 (0.3) M Major 0.56 0.31 0.06 13 D P. regina rose-crowned fruit-dove D 11.5 (0.3) M Major 3.38 2.23 0.81 36 D Lopholaimus antarcticus topknot pigeon D 17.5 (0.5) L Major 1.06 0.06 0.63 6 T Cacatua roseicapilla galah C 15.5 (0.2) L Minor 0.00 0.13 0.25 3 U C. galerita sulphur-crested cockatoo C 22.9 (0.4) L Minor 1.50 0.44 0.38 16 T Trichoglossus haematodus rainbow lorikeet C 12.2 (0.1) M Minor 2.75 2.69 7.31 32 I T. chlorolepidotus scaly-breasted lorikeet C 11.3 (0.2) M Minor 0.13 0.69 0.56 7 U Alisterus scapularis Australian king-parrot C 17.4 (0.3) § L Mixed 0.56 0.26 0.32 23 T Platycercus elegans crimson rosella C 14.4 (0.2) M Mixed 0.31 0.06 0.00 4 U P. adscitus pale-headed rosella C 12.1 (0.2) M Mixed 0.25 0.94 1.31 14 T Eudynamys scolopacea common koel D 18.2 (0.2) § L Major 0.38 0.38 0.69 17 T Scythrops novaehollandiae channel-billed cuckoo D 32.8 (0.6) L Major 0.06 0.19 0.25 7 T Anthochaera chrysoptera little wattlebird D 9.9 (0.1) S Minor 0.00 0.19 0.25 6 T Philemon corniculatus noisy friarbird D 11.5 (0.4) M Minor 0.06 0.00 0.44 3 U Entomyzon cyanotis blue-faced honeyeater + D 13.1 (0.4) M Minor 0.00 0.13 0.00 1 U Manorina melanocephala noisy miner D 10.2 (0.3) M Minor 0.00 0.00 0.38 3 U Meliphaga lewinii lewin’s honeyeater D 10.5 (0.2) M Mixed 8.00 8.75 7.63 48 T Myzomela sanguinolenta scarlet honeyeater D 5.2 (0.1) S Minor 1.56 1.69 1.25 32 T Coracina novaehollandiae black-faced cuckoo-shrike D 17.4 (0.2) L Minor 0.13 0.06 1.88 10 I C. lineata barred cuckoo-shrike + D 13.5 (0.2) M Mixed 0.00 0.00 0.14 2 U

75

Table 3.1 (cont.) 1 1 Seed Gape width Gape Frugivory No. Response Scientific name Common name 2 3 4 Extensive Remnant Regrowth treat. (cm) class level sites pattern Lalage leucomela varied triller D 9.1 (0.2) S Mixed 0.06 0.13 0.06 4 U Oriolus sagittatus olive-backed oriole D 15.8 (0.3) L Mixed 0.00 0.00 0.13 2 U Sphecotheres viridis figbird D 17.8 (0.3) L Major 2.56 7.50 15.63 41 I Cracticus torquatus grey butcherbird D 15.3 (0.2) L Minor 0.00 0.00 0.38 4 U Gymnorhina tibicen Australian magpie D 18.5 (0.2) L Minor 0.13 2.06 5.25 26 I Strepera graculina pied currawong D 20.1 (0.4) L Mixed 2.75 4.25 3.00 38 T Ptiloris paradiseus paradise riflebird D 16.7 (0.3) § L Mixed 0.50 0.13 0.00 5 U Corvus orru Torresian crow D 19.5 (0.2) L Minor 1.44 2.25 5.13 35 I Ailuroedus crassirostris green catbird D 19.5 (0.2) L Mixed 3.75 4.63 0.75 35 D Sericulus chrysocephalus regent bowerbird D 13.9 (0.4) § M Mixed 0.13 0.06 0.13 5 U Ptilonorhynchus violaceus satin bowerbird D 18.5 (0.2) § L Mixed 0.19 0.19 0.50 7 T Passer domesticus house sparrow * + C 8.6 (0.1) S Minor 0.00 0.00 0.19 1 U Dicaeum hirundinaceum mistletoebird D 6.9 (0.1) S Major 0.19 0.00 0.13 5 U Zosterops lateralis silvereye D 6.0 (0.2) S Mixed 1.06 1.69 6.25 20 I 1 Nomenculature and bird species order after Christidis and Boles (1994). * next to common name denotes introduced species; + shows species excluded from multivariate analyses (present in only 1 site) 2 Mean (and standard error), N=10 except Australian King Parrot (N=9); includes adjustment for fruit-doves and Topknot Pigeon (see text). § shows species that may have sexually dimorphic gape widths. 3 S = Small (<10 mm), M = Medium (10-15 mm), L = Large (>15 mm). 4 Major indicates fruit-dominated diet, Mixed is a mixed diet that includes fruit, Minor indicates that fruit is a relatively minor dietary component.

76

.

16 Disperser Crusher 12

8

4

No. frugivore species frugivore No. 0 Major Mixed Minor Frugivory level

20 Disperser Crusher 16

12 8

4

No. frugivore species frugivore No. 0 Small Medium Large Gape width class

40

30

20

10 Gape widthGape (mm)

0 Major Mixed Minor

Frugivory

Figure 3.1 Inter-relationships between frugivorous bird attributes; i) the number of seed-crushing and seed-dispersing bird species from each frugivory level (major n=11, mixed-diet n=15, and minor n=12); ii) the number of seed-crushing and seed-dispersing bird species from each gape width class (small n=6, medium n=15, and large n=17); and iii) the actual gape width of each bird species in major, mixed-diet and minor frugivory levels

77 3.3.2 Functional group abundance in remnants and regrowth relative to extensive

forest

The abundance of two functional groups (medium-gaped birds with fruit- dominated diets and large-gaped birds with mixed diets) showed a decreasing response to fragmentation and two groups (small-gaped, mixed-diet birds and minor frugivores) showed an increaser abundance pattern (Table 3.3). The remaining four functional groups showed a tolerant abundance pattern, although both seed-crushers and the large- gaped, major frugivore group showed an increasing trend that was not statistically significant.

Table 3.2 The mean (and S.E.) abundance of frugivorous bird species within each of the eight functional groups in extensive (n=16), remnant (n=16) and regrowth (n=16) forest in southeast Queensland.

1 2 Abund. Functional group Extensive Remnant Regrowth ANOVA p pattern3 1. Large gape, fruit-dominated diet 9.31 (1.69) 9.94 (1.60) 17.50 (3.84) 0.06 T (5) 2. Medium gape, fruit-dominated 3.94 (0.62)a 2.44 (0.35)b 0.88 (0.30)c <0.0001 D diet (2) 3. Large gape, mixed diet (5) 7.19 (0.88)a 9.19 (1.18)a 4.38 (0.85)b 0.005 D 4. Medium gape, mixed diet (3) 8.13 (0.50) 8.81 (0.63) 7.94 (0.60) 0.54 T 5. Small gape, fruit-dominated diet 0.19 (0.10) 0.00 0.13 (0.09) 0.22 T (1) 6. Small gape, mixed diet (2) 1.13 (0.49)a 1.81 (0.97)a 6.31 (1.70)b 0.005 I 7. Fruit a minor dietary component 3.31 (0.67)a 6.38 (1.13)a 14.94 (1.66)b <0.0001 I (all gape width classes) (9) 8. Seed-crushers (all gape width 0.07 13.19 (1.29) 17.50 (1.76) 18.50 (1.87) T classes and frugivory levels) (15) 1 The number of species in each group is shown in brackets next to group name. The species comprising each group are shown in Table 3.2. 2 Results of ANOVA between site types (d.f. = 2, 47). Groups differing significantly in abundance (p<0.05) are given different letters. 3 The response to fragmentation for each group (D=decreaser, T=tolerant, I=increaser, see Chapter Two for description of patterns).

There was a positive association between a species’ seed dispersal potential and its sensitivity to forest fragmentation and disturbance (Spearman’s r = 0.39, p = 0.049).

Bird species making a relatively high potential contribution to seed dispersal tended to show decreaser or tolerant abundance patterns, while bird species considered to have

78 relatively low dispersal potential mostly had tolerant or increaser abundance patterns

(Table 3.3).

Table 3.3 Relationship between frugivorous bird species’ abundance pattern and their relative seed dispersal potential. Only bird species that were sufficiently frequent during field surveys to determine an abundance response pattern (26 out of the 42 species) are included. ‘ Disp.pot.’ is relative influence over seed dispersal dynamics.

Number of species showing abundance pattern1 Disp. pot. Functional groups2 Decreaser Tolerant Increaser High 1, 2 and 3 4 5 1 Medium 4 and 6 0 1 1 Low 7 and 8 1 8 5 1 see Table 3.1 for species’ abundance pattern data and functional attributes 2 see Table 3.2 for explanation of functional group attributes

Ordination of the study sites based on frugivore abundance at the functional group level showed extensive forest and regrowth sites tending to cluster at opposite extremes of the ordination space (Figure 3.2). The three site types differed significantly in terms of functional group abundance (ANOSIM global p<0.01). Pair-wise tests showed that regrowth sites differed from both extensive and remnant sites (ANOSIM p<0.01 for both), but that extensive and remnant sites were not substantially different

(p=0.06). The two functional groups significantly associated with the region of the ordination containing extensive and some remnant forest sites (Figure 3.2) were those that that showed decreasing abundance patterns with fragmentation (Table 3.2). The four groups associated with regrowth sites comprised two ‘increaser’ and two ‘tolerant’ functional groups.

79

Large-gaped, mixed- diet frugivores

Large-gaped, major frugivores

Small-gaped, mixed- diet frugivores

Crushers Medium-gaped, major frugivores Minor frugivores

Figure 3.2 Ordination of the 48 study sites based on numbers of birds from each functional group (Stress = 0.22). is extensive forest, is remnants, is regrowth. The lower panel shows abundance vectors for those functional groups significantly (p<0.05) associated with the ordination.

80 3.4 Discussion

3.4.1 Characteristics of the frugivorous birds assemblage

Approximately one-third of the 42 frugivorous bird species recorded in these surveys crush seeds. This is a similar proportion of the frugivorous bird assemblage as that recorded in the Neotropics (Terborgh 1986). Seed-crushing species included several species of parrot, plus certain doves and pigeons, such as the white-headed pigeon and brown cuckoo-dove. Birds that crush seeds probably contribute relatively little to seed dispersal, although a small proportion of ingested seeds may remain intact following passage through the gut of these species (Snow 1981; Corlett, 1998; Dennis and

Westcott, 2006).

Of the 27 seed-dispersing bird species identified, seven species may also contribute comparatively little to broad patterns of seed dispersal, since fleshy fruits are only a minor component of their diets. The 20 remaining species do not crush seeds and have either fruit-dominated or mixed diets in which fruit is a conspicuous component.

These birds would have most influence over seed dispersal dynamics in the study region, although some species were recorded in low numbers during surveys.

In the frugivore assemblage studied, there was no consistent association between any of the measured functional attributes. The positive association between gape size and dietary dominance of fruit that had been reported elsewhere (Fuentes 1994) was not apparent in this assemblage.

3.4.2 Functional characteristics of the frugivorous bird assemblage in fragmented

rainforest in subtropical Australia: assessment of potential consequences for

seed dispersal

The frugivore attributes of seed crushing behaviour, gape width and frugivory level were used to describe groups of species that may have similar functional roles in

81 terms of their capacity to disperse the seeds of similar plants, especially plants with similar-sized diaspores. This dimension of functional variation among frugivore species is useful in assessing whether changes in the species composition of a frugivore assemblage may result in the loss of potential dispersal agents for certain plant species

(e.g., Silva and Tabarelli, 2000). In addition to dietary composition, variation among frugivore species in factors such as their abundance in a habitat, feeding rates, dispersal distances and patterns of microhabitat use is likely to lead to variation in their role as seed dispersers (Schupp, 1993; Jordano and Schupp, 2000; Loiselle and Blake, 2002).

The functional groupings were not assembled using specific dietary information, but the attributes used to form the groups have been linked with dietary composition

(Crome, 1975, 1978; Moermond and Denslow, 1983, 1985; Wheelwright, 1985;

Moermond et al., 1986; Whelan and Willson, 1994). All but one of the groups formed using these attributes were multispecific, indicating functional overlap and the potential for substitution among frugivore species (Loiselle and Blake 2002). The only functional group with a single member (the mistletoebird) was the small-gaped, fruit-dominated diet group. This species is likely to have a different diet to most other frugivore species in that it is largely restricted to fruits from mistletoe plants (Loranthaceae). However, other bird species may also be efficient and effective dispersers of seeds from these plants (Reid, 1989), and the importance of mistletoe as food for the mistletoebird may be disproportionate to the importance of the mistletoebird as a disperser of mistletoe, as has been described for other frugivore-plant interactions (Herrera, 1984; Jordano, 1987;

Silva et al., 2002).

In the context of changes in the frugivore assemblage in fragmented landscapes, functional overlap may mean that the decline of one frugivore species may be offset by increased numbers of a functionally similar bird species. Dispersers of large fruited plants are vulnerable to decline in tropical fragmented landscapes throughout the world

82 (Corlett, 1996, 1998; Harrington et al,. 1997; Silva and Tabarelli, 2000; Kitamura et al.,

2002; McConkey and Drake, 2002). Foraging observations at large-seeded plant species suggest that this may also be the case in subtropical rainforests of southeast Queensland

(Green, 1993). Approximately 30% of the species comprising the avian frugivore assemblage of subtropical Australia (14 out of 42 of the species considered in this study) have wide gapes (>15 mm). Four of these species are minor frugivores and may make relatively little contribution to the dispersal of these plants. A similar percentage

(24%) of the frugivorous bird species had gapes wider than 15 mm in the Brazilian

Atlantic (Silva and Tabarelli, 2000).

However, the overall abundance of the functional group that may have the greatest potential to disperse large-seeded plants (large-gaped, major frugivores) did not decline in fragmented rainforest in the Sunshine Coast, and in fact showed a strong tendency towards increasing in abundance in regrowth. This increasing tendency was entirely due to the greatly increased abundance of the figbird in remnant, and particularly in regrowth, sites, compared with extensive forest. Numbers of other bird species from this functional group were either low and similar among site types (e.g. common koel) or much lower in remnants (e.g. wompoo fruit-dove). Although fruit- doves have been nominated as having the greatest potential among avian frugivores to disperse large-seeded plants in subtropical Australian rainforests (Frith, 1982; Green,

1995), the present work has shown that figbirds are also morphologically capable of dispersing large fruits. Consequently, the high abundance of figbirds in remnants and regrowth habitats of the study region potentially maintains the dispersal of large-seeded plant species in these parts of the landscape. Furthermore, other bird species with the potential to disperse large-seeded plants (the topknot pigeon, pied currawong and satin bowerbird) were also present in these habitats. Hence, in contrast to tropical landscapes elsewhere, the loss or decline of bird species such as the wompoo fruit-dove from

83 fragmented subtropical Australian rainforest landscapes may not result in reduced dispersal of large fruited plant species.

However, the functional group approach used here may overstate potential similarity among frugivore species. For example, although the channel-billed cuckoo has a large gape and fruit-dominated diet, available information suggests that their diet may be dominated by figs (Blakers et al., 1984). Consequently, this species may not contribute to the dispersal of large fruited plants, despite their very large gape. Second, movement patterns and gut passage rates may mean that certain frugivore species do not disperse seeds among fragmented habitats (Silva et al., 1996; Alcántara et al., 2000). In the case of figbirds, radiotracking in northern Australia has shown that birds of this species regularly fly distances of several kilometres between rainforest patches, and also visit isolated trees (Price, 1999). Furthermore, aviary tests showed that the gut passage times of figbirds resembled those of fruit pigeons (Price 1999). Therefore, the role of figbirds in dispersing seeds within and among remnant and regrowth forests would probably not be limited by restricted movement patterns nor by very rapid gut passage times. Third, patterns of plant species consumption are likely to be influenced by a frugivore species’ digestive adaptations (Levey and Grajal, 1991; Martínez del Rio and

Restrepo, 1993; Cipollini and Levey, 1997) and ability to handle secondary compounds

(Levey and Martínez del Rio, 2001). The relationship between these factors and frugivory level is not known, although may be likely to be associated with taxonomy.

The two functional groups whose overall abundance decreased in remnants and/or regrowth were large-gaped birds with mixed diets and medium-gaped birds with fruit-dominated diets. An overall decrease in the dispersal of plants eaten by birds from these two groups would be expected to result from declines of these functional groups.

The medium-gaped birds with fruit-dominated diets (superb and rose-crowned fruit- doves) small fruit-doves may consume a suite of rainforest plants that is distinct from

84 other birds, even the closely-related wompoo fruit-dove. In a detailed dietary study of these species in north Queensland, the superb and rose-crowned fruit-doves predominantly consumed species from the family Araliaceae and from the lauraceous genera Endiandra and Litsea, whereas the wompoo fruit-dove fed more on plants from the family Elaeocarpaceae and the lauraceous genus Cryptocarya (Crome, 1975). All three species were associated with extensive forest rather than remnant or regrowth sites in the present study. It seems likely that the greatly reduced numbers of these birds in fragmented and disturbed habitats would cause changes in the composition and rate of seed dispersal in these habitats, even though the figbird replaces them numerically.

Numbers of the small-gaped, mixed diet functional group increased in regrowth compared with remnant and extensive forest. This group comprised the varied triller and the silvereye, although the increasing pattern was driven by numbers of the silvereye.

The plant species dispersed by these frugivores would be expected to increase in fragmented rainforest landscapes of subtropical Australia. Consequently, the regeneration trajectory of rainforest regrowth in fragmented parts of the landscape may be strongly influenced by the plant species consumption patterns of these bird species.

If, as may be expected from their small size and mixed diet, these birds consume large numbers of introduced weed species (Richardson et al., 2000), fragmented rainforest may be overwhelmed by the input of fleshy-fruited weeds.

There was an overall positive association between a species’ seed dispersal potential and the extent to which their abundance was negatively impacted in remnants and regrowth. It was considered that birds with fruit-dominated diets and large or medium gapes were likely to have a disproportionately high influence over general seed dispersal patterns. Because these species are nutritionally dependent on fruit, they may be likely to consistently eat a large volume of fruit and to need to feed on a range of plant species to cope with the temporal variation in availability of any given species

85 (Leighton and Leighton, 1983). Since having a larger gape gives birds access to a greater range of available fruits (Wheelwright 1985), it was also expected that large- gaped birds with mixed diets may have relatively high seed dispersal potential. All except one of the species (figbird) within these three functional groups showed either a decreasing or tolerant abundance response to forest fragmentation and disturbance.

Consequently, it is predicted that the overall rate of seed dispersal may be lower, and that fewer plant taxa plant taxa would be dispersed in remnants and regrowth than in extensive forest. The plants that are predicted to be most likely to be affected by these changes are large fruited plants and Lauraceae. Beyond this, predictive power is limited by a lack of information regarding associations between frugivore traits and patterns of consumption of plant species.

The approach presented here provides a means of systematically assessing dispersal potential in relation to frugivore assemblage composition. As this has not previously been done, specific dietary information is required to assess whether the functional attributes selected do reflect major variation among frugivorous bird species in their diet composition.

86 Chapter Four

Variation in patterns of plant species consumption by frugivorous bird species is related to gape width, degree of frugivory and seed treatment

4.1 Introduction

A primary determinant of functional variation among frugivore species is the suite of plant species that they consume and disperse (Gautier-Hion et al., 1985; Corlett,

1998; Hamann and Curio, 1999; Kitamura et al., 2002; Poulsen et al., 2002; Silva et al.,

2002). However, information on plant species consumption is extremely time- consuming to collect and only limited information is available for most of the frugivore species in an assemblage (Hamann and Curio, 1999; Kitamura et al., 2002; Silva et al.,

2002). In contrast, information regarding the morphology and behaviour of frugivore species may be more easily gained (e.g, Dunning, 1993). These factors may influence patterns of plant species consumption by interacting with variation in the traits of plant species (e.g., fruit size) (van der Pijl, 1982; Gautier-Hion et al., 1985). Potentially, information about the relevant traits of frugivore species may be used to assess their role in seed dispersal when detailed dietary information is not available.

Functional classifications of frugivore species, which relate types of frugivores to the types of plants they consume, are a step in the development of a predictive understanding of seed dispersal interactions. Most studies of frugivore diets have documented variation among only a subset of the frugivore or plant species in an assemblage (e.g, Crome, 1975; Innis, 1989; Sun et al., 1997; Brown and Hopkins, 2002;

Kitamura et al., 2002). Consequently, our understanding of patterns of fruit-frugivore interactions across frugivore communities is limited (Herrera, 1998, 2001). The

87 traditional approach of comparing among taxonomic groups (e.g., birds and mammals) has yielded knowledge of broad patterns of dietary variation across a frugivore community (Van der Pijl, 1982; Gautier-Hion et al., 1985; Bollen et al., 2004).

However, this approach obscures the variation among species within taxonomic groups

(Willson et al., 1989; Stiles, 1993; Graham et al., 2002; Lord et al., 2002). For example, the gape width of frugivorous bird species imposes an intractable upper threshold on the size of fruits that they can swallow (Herrera, 1981; Wheelwright, 1985). Consequently, bird species with wider gapes are capable of consuming plant species with larger fruits while bird species with narrow gapes are physically constrained to swallowing only small fruits (Wheelwright, 1985; Silva and Tabarelli, 2000; McConkey and Drake,

2002; Kitamura et al., 2002). In Costa Rica, Wheelwright (1985) showed that the maximum and mean size of fruits that were consumed by frugivorous bird species were positively correlated with their gape widths.

Variation in patterns of plant species consumption by different frugivore species may also be explained by variation in the energetic or chemical content of fruits

(Herrera, 1982, 1987; Stiles, 1993; Jordano, 1995; Cipollini and Levey, 1997).

Frugivore species vary in their capacity to digest plant toxins, lipids and sugars (Izhaki and Safriel, 1989; Martínez del Rio and Restrepo, 1993; Cipollini and Levey, 1997).

Frugivore species also vary in their degree of nutritional dependence on fruit (Snow,

1971; McKey, 1975). Most frugivorous species eat fruit as part of a varied diet that also includes other items, such as nectar or invertebrates. The tissues of non-fruit foods, especially insects, are relatively rich in protein and other nutrients (Morton, 1973;

Herrera, 1987; Izhaki and Safriel, 1989). Frugivore species that only consume fruits, but crush and digest seeds would obtain nutrition from the seeds as well as the fruit pulp

(Snow, 1981; Innis, 1989; Jones and Crome, 1990). However, fruit is the sole source of nutrition for a small number of species. It has been proposed that these dietary

88 specialists may be likely to select fruits with relatively high energy and protein content

(Morton, 1973; McKey, 1975; Bairlein, 1996; Bosque and Calchi, 2003). For example, it has been reported that highly frugivorous bird species prefer to consume lipid-rich fruits (Snow, 1971; Crome, 1975; McKey, 1975; Howe and Estabrook, 1977; Howe,

1981; Wheelwright, 1983; Stiles, 1993). This may be because lipids provide more energy than carbohydrates (Johnson et al., 1985; Witmer and Van Soest, 1998), or because lipids tend to co-occur with high nitrogen levels (Sun et al., 1997).

In Chapter Three, it was proposed that gape width, frugivory level and seed treatment were important functional attributes of frugivorous bird species because they are likely to determine variation in the composition of plant species consumed by bird species. In the present Chapter, actual patterns of plant species consumption by frugivorous species are examined across the subtropical bird assemblage of eastern

Australia, in relation to the nominated attributes. This is made possible by a relatively large published record of foraging information for the bird species of this region.

Specifically, tests are conducted for relationships between frugivore species’ gape width, frugivory level and seed treatment and their patterns of fruit size consumption, the dietary composition of plant species, and the frequency of specific plant families in the diet. The results are used to assess whether the selected functional attributes are useful indicators of differences among frugivorous bird species in their roles as seed dispersers.

4.2 Methods

4.2.1 Diet composition of the frugivorous bird assemblage

In Chapter Two, 42 frugivorous bird species were identified in subtropical rainforests of the Sunshine Coast, eastern Australia. Records of the consumption of native plant species by these birds were obtained from 151 published sources (Appendix 89 1), together with several unpublished data sets (See Acknowledgements). Most records had been obtained from opportunistic field observation of fruit consumption, although some were from targeted surveys of particular plant or frugivore species. A small number of records were obtained from bird gut contents. There was large variation among bird species in the number of individual frugivory records available (i.e., the total number of sources that had reported observations of frugivory). These ranged from

2 records for the galah to 228 for the satin bowerbird (including repeat observations of the consumption of a given plant species by a bird species).

Foraging records included native and exotic species of tree, shrub, vine and herb

(Appendix 2). Because of the large geographic range of many of the bird species from subtropical Australian rainforests, frugivory records may have been collected from a region extending from temperate southern Australia (e.g., French, 1990) to tropical

Papua New Guinea (Frith et al., 1976). Records obtained from observation outside the eastern Australian subtropics were included in analyses only if the plant species also occurred within this region. The data set included records accumulated during a period of more than 140 years, with the earliest account published by Gould (1865). For a given frugivore species, the data potentially included foraging records from multiple years, seasons and geographic locations. These records were compiled into a binary data matrix containing the presence or absence of each fleshy-fruited plant species in the diet of each of the frugivorous bird species.

Records were rejected if it appeared that the bird had not been observed actually consuming the fruit (e.g., if it was simply observed in the fruiting plant), or if it was judged from accompanying information that the interaction was an instance of fruit theft

(consumption of the fruit pulp without ingestion of the seed). Diaspore size (see below) was used to further screen the records for likely cases of fruit theft; if the size of the fruit greatly exceeded the gape width of the bird species, it was assumed that it would

90 only be capable of consuming part of the fruit and had probably acted as a fruit thief

(Howe and Estabrook, 1977). Cases were excluded if the median diaspore size was more than twice the gape width of the frugivore species. This approach accounts for potentially substantial intraspecific variation in diaspore size (Edwards, 2005). Each record in the screened data set was treated as evidence of the potential for the bird species to consume and disperse viable seed from the plant species.

For each plant species, diaspore size (the functional dispersal unit) was recorded.

For most plant species, the measurement was the shorter axis (usually diameter) of the whole fruit. However, seed diameter was used for soft fruits with multiple small seeds

(e.g. many species in the Solanaceae and Moraceae) that can be dispersed as a result of piecemeal consumption (Corlett, 1998; Kitamura et al., 2002). Similarly, for dehiscent, arillate fruits such as Alectryon spp. (family Sapindaceae), the diameter of an individual seed plus the fleshy aril was used. Diaspore size data were collected from published literature (Williams et al., 1984; Floyd, 1989; Cooper and Cooper, 1994; Hauser and

Blok, 1998; Butler, 2003), supplemented with data from field collections (S. McKenna,

C. Moran) and biological web sites. In most cases, a range of diaspore size values was reported and the median of these was used in analyses.

Diaspores with a median width greater than 75% of the gape width of a bird species were considered to be close to the bird species’ morphological handling capacity. This was used to calculate, for each bird species, the percentage of plant species consumed that had diaspores close in size to the bird’s handling capacity.

4.2.2 Functional attributes of bird species

Four of the 42 bird species (scarlet honeyeater Myzomela sanguinolenta, house sparrow Passer domesticus, rock dove Columba livia and Australian magpie

Gymnorhina tibicen) were rarely recorded consuming fruit of native plant species (less

91 than two native plant species recorded) and are therefore not considered further. The remaining 38 bird species were categorised based on three attributes considered to influence their patterns of fruit consumption in terms of plant species (see Chapter

Three). Gape width was measured on museum specimens of each bird species and ranged from 6.0 mm (silvereye Zosterops lateralis) to 32.8 mm (channel-billed cuckoo

Scythrops novaehollandiae). For some analyses these data were used to group bird species into three gape width categories: small (<10 mm), medium (10 – 15 mm), or large (>15 mm). Each species’ degree of frugivory was classified into one of three categories: 11 bird species had a fruit-dominated diet (‘major frugivores’), 15 had a diet comprising more than one main food type, one of which was fruit (‘mixed-diet frugivores’) and 12 had a diet dominated by food other than fruit, but which occasionally included fruit (‘minor frugivores’). The relative level of dependence of each bird species on fruit was determined primarily from qualitative descriptions in the literature (Blakers et al., 1984; the Handbook of Australian, New Zealand and Antarctic

Birds (HANZAB) series: Marchant and Higgins, 1993; Higgins and Davies, 1996;

Higgins, 1999; Higgins et al., 2001). Each bird species was also categorised as being either a seed-disperser or seed-crusher, based on information contained in the HANZAB series about the destruction of seeds during either feeding (parrots, cockatoos) or digestion (certain pigeon and dove species, Australian brush turkey). The measurement and classification of frugivore attributes is explained in more detail in Chapter Three

(Section 3.2.1).

4.2.3 Data analyses

Tests to establish whether a bird species’ frugivory level, gape size and seed treatment were independent of one another were described in Chapter Three (Section

3.2.2).

92 One-way analyses of covariance (ANCOVA) were used to test the effect of the independent variable frugivory level (three levels: major (n=10 (the channel-billed cuckoo was excluded from analyses)), mixed-diet (n=15) and minor (n=12)) on the following dependent variables: i) the size of diaspores consumed; ii) the percentage of diaspores consumed that were close to a bird species’ handling capacity (i.e., wider than

75% of its gape width); and iii) the total number of plant species consumed. Species’ gape widths (actual width in mm) were used as the covariate in analyses. Homogeneity of regression slopes was tested using the interaction between frugivory level and gape width.

Differences between major and mixed-diet frugivores in the number of species consumed from plant families with at least three species represented in the data set were tested using t-tests. A parametric procedure was used for families with >15 plant species in the data set (three families), while randomisation (1 000 iterations) was used for the

25 plant families with 4 - 15 species represented (using the Pop-tools add-in in MS

Excel (Hood, 2003).

A classification tree for frugivore species, based on their dietary composition, was generated using the UPGMA algorithm (Manly, 1994) and Bray-Curtis dissimilarity metric in PRIMER (Clarke and Warwick, 2001). Plant species that had been recorded in the diet of less than three frugivore species, and frugivores that had been recorded consuming less than three plant species were excluded from all multivariate analyses to reduce variation in the data. Multivariate analyses were subsequently conducted on a data matrix containing information for 35 bird species and

151 plant species. Similarity percentages (SIMPER; Clarke, 1993) analyses were used to identify the plant species that contributed most to the dissimilarity between these groups.

93 4.3 Results

Frugivory level, gape width and seed treatment varied independently across the

38 bird species (Chapter Three, Section 3.3.1, Figure 3.1). Therefore, relationships between each attribute and patterns of plant species consumption were tested independently.

4.3.1 General patterns of plant consumption

The functional attributes and patterns of native plant species consumption of frugivorous bird species are shown in Table 4.1. The data set relating bird species to plant species contained information for 244 native plant species that had been consumed by at least one of the 38 frugivorous bird species. There was considerable variation between plant species in the suites of frugivore species that were known to consume them. An average of 5.2 (S.E. 0.30) frugivorous bird species was recorded consuming each native plant species (range 1 – 26 bird species).

94 Table 4.1 The functional attributes and patterns of native plant species consumption of 38 frugivorous bird species in Australian subtropical rainforests. ‘O.’ order, ‘F.’ family. ‘Frug. lev.’ frugivory level; ‘Gape’ gape width; ‘Seed treat.’ seed treatment (C seed-crusher, D seed-disperser), ‘No. plant spp.’ the number of native plant species recorded for each bird species; ‘Av. dias. size.’ average size (mm) of diaspores consumed; ‘Dias. size range’ shows the minimum and maximum sizes of fruits consumed and the number of plant species with diaspores < 10 mm and ≥ 10 mm consumed; ‘Perc. ≥ 75% gape’ is the percentage of the plant species recorded in the diet of each bird species with diaspores wider than 75% of the gape width; and ‘Perc. Ficus spp.’ is the percentage of native plant species consumed that were from the genus Ficus. Diaspore size range2 Av. No. dias. No. Perc ≥ Perc Frug. Gape Seed plant size. No. ≥ 75% Ficus Bird order, family1 Bird species1 Common name1 lev. (mm) treat. spp. (mm) min max <10 10 gape spp. O. Galliformes F.Megapodiiae Alectura lathami Australian brush turkey Mixed 18.3 C 9 9.44 1 27.5 6 3 33 11 O. Columbiformes F.Columbidae Columba lecuomela white-headed pigeon Major 11.8 C 39 9.29 1 18.5 19 20 59 8 Macropygia amboinensis brown cuckoo-dove Major 10.1 C 63 5.48 1 12.5 56 7 25 10 Chalcophaps indica emerald dove Major 8.5 C 9 6.39 1 14.5 6 3 44 33 Geopelia humeralis bar-shouldered dove Minor 6.6 C 3 2.83 1 4.0 3 0 0 33 Leucosarcia melanoleuca wonga pigeon Mixed 9.3 C 9 6.65 1 14.0 5 4 44 33 Ptilinops magnificus wompoo fruit-dove Major 19.0 D 81 10.01 1 27.5 34 47 15 10 P. superbus superb fruit-dove Major 12.6 D 26 9.08 1 22.5 13 13 54 12 P. regina rose-crowned fruit dove Major 11.5 D 74 8.75 1 22.5 41 33 49 12 Lopholaimus antarcticus topknot pigeon Major 17.5 D 73 10.25 1 27.5 31 42 23 10 O. Psittaciformes F.Cacatuidae Cacatua roseicapilla galah Minor 15.5 C 2 5.00 1 9.0 2 0 0 50 sulphur-crested C. galerita cockatoo Minor 22.9 C 11 7.45 1 27.5 9 2 9 10 F.Psittacidae Trichoglossus haematodus rainbow lorikeet Minor 12.2 C 12 5.25 1 14.0 10 2 17 33 T. chlorolepidotus scaly-breasted lorikeet Minor 11.3 C 8 4.00 1 7.5 8 0 0 38 Alisterus scapularis Australian king parrot Mixed 17.4 C 59 7.82 1 20.0 37 22 8 5 Platycercus elegans crimson rosella Mixed 14.4 C 42 7.20 1 27.5 31 11 24 10 P. adscitus pale-headed rosella Mixed 12.1 C 5 5.00 2 8.5 5 0 0 0

95

Table 4.1 (cont.) Diaspore size range2 Av. No. dias. No. Perc ≥ Perc Frug. Gape Seed plant size. No. ≥ 75% Ficus Bird order, family1 Bird species1 Common name1 lev. (mm) treat. spp. (mm) min max <10 10 gape spp.

O. Cuculiformes F.Cuculidae Eudynamys scolopacea common koel Major 18.2 D 25 7.04 1 20.0 19 6 8 24 Scythrops novaehollandiae channel-billed cuckoo Major 32.8 D 6 2.00 1 7.0 6 0 0 83 O. Passeriformes F.Meliphagidae Anthochaera chrysoptera little wattlebird Minor 9.9 D 3 3.50 2 6.0 3 0 0 0 Philemon corniculatus noisy friarbird Minor 11.5 D 2 7.25 6 8.5 2 0 0 0 Entomyzon cyanotis blue-faced honeyeater Minor 13.1 D 2 1.00 1 1.0 2 0 0 100 Manorina melanocephala noisy miner Minor 10.2 D 7 4.29 1 9.0 7 0 29 43 Meliphaga lewinii Lewin’s honeyeater Mixed 10.5 D 106 6.65 1 20.0 82 24 32 9 black-faced cuckoo- F.Campephagidae Coracina novaehollandiae shrike Minor 17.4 D 10 3.45 1 8.0 10 0 0 50 C. lineata barred cuckoo-shrike Mixed 13.5 D 10 2.05 1 6.5 10 0 0 70 Lalage leucomela varied triller Mixed 9.1 D 10 4.50 1 8.5 10 0 20 30 F.Artamidae Cracticus torquatus grey butcherbird Minor 15.3 D 3 4.83 4 6.0 3 0 0 0 Strepera graculina pied currawong Mixed 20.1 D 50 8.20 1 27.5 32 18 6 20 F.Paradisidae Ptiloris paradiseus paradise riflebird Mixed 16.7 D 33 6.57 1 17.5 26 7 9 12 F.Oriolidae Oriolus sagittatus olive-backed oriole Mixed 15.8 D 30 5.11 1 12.5 26 4 7 20 Sphecotheres viridis figbird Major 17.8 D 74 7.11 1 18.5 52 22 7 14 F.Corvidae Corvus orru Torresian crow Minor 19.5 D 10 4.15 1 9.0 9 1 0 50 F.Ptilonorhynchidae Ailuroedus crassirostris green catbird Mixed 19.5 D 104 9.08 1 35.0 60 44 9 9 Sericulus chrysocephalus regent bowerbird Mixed 13.9 D 108 7.62 1 27.5 73 35 28 9 Ptilonorhynchus violaceus satin bowerbird Mixed 18.5 D 106 8.81 1 35.0 58 48 10 8 F.Dicaeidae Dicaeum hirundinaceum mistletoebird Major 6.9 D 6 4.33 1.5 7.0 6 0 33 0 F.Zosteropidae Zosterops lateralis silvereye Mixed 6.0 D 37 4.45 1 9.0 37 0 59 19 1 Taxonomy and nomenclature follow Christidis and Boles (1994). 2 Median diaspore size shown. Note that analyses assumed intra-specific variation in diaspore size to a maximum of 50% larger or smaller than median size.

96

4.3.2 The effect of gape width and frugivory level on diaspore size selection

The channel-billed cuckoo was removed from analyses relating to patterns of diaspore size consumption, since it showed a strongly different pattern from other major frugivore species; it predominantly consumed plants with very small diaspores (average

2 mm), despite its very large gape (32.8 mm; Table 4.1). This was a consequence of the dominance of Ficus spp. in its diet (83%; Table 4.1). ANCOVA showed that the average size of diaspores consumed varied significantly among the three frugivory levels. Inspection of the data (Figure 4.1) showed that major and mixed-diet frugivores consumed larger diaspores than minor frugivores (Figure 4.1; Table 4.2). The average width (mm) of diaspores that was consumed by a bird species increased with its gape width (mm) at a ratio of approximately 1:2 (Table 4.2; Figure 4.1). Frugivory level and gape width together explained 74% of the variation in diaspore size consumption (eta squared 0.43 and 0.31, respectively).

The consumption of plant species with diaspores that were close to the limit of a bird species’ handling capacity was also influenced by frugivory level; minor frugivores consumed a low dietary proportion of plant species with diaspores that were close to the limit of their handling capacity (Figure 4.2, Table 4.2). The dietary proportion of these fruits decreased with gape width (Table 4.2, Figure 4.2). Together, these two factors explained 65% of the variation in consumption of diaspores that were close in size to bird species’ handling capacity (eta squared 0.39 and 0.26 for frugivory level and gape width, respectively).

97

Table 4.2 Results of ANCOVA tests for effects of gape width (G.w.) and frugivory level (F.l.) on i) the average size of diaspores consumed (Dias. size); ii) the dietary proportion of diaspores that were close to the limit of a bird species’ handling capacity (Perc. >75% gape); and iii) the number of native plant species consumed (No. plant spp.). ANCOVA showed homogeneity of regressions for each comparison (no significant interaction between factors (F.l. x G.w.). β is the slope of the relationship between gape width and the variable tested. The mean, number of cases, and strength of regression (r2 ) is shown separately for each frugivory level in Figures 4.2 to 4.4. Frugivory level Gape width (G.w.) (F.l.) G.w. x F.l. Factor F p r2 β F p F p i) Dias. size 14.04 0.001 0.50 0.46 12.36 <0.0001 0.57 0.57 ii) Prop. ≥75% gape 1.48 0.002 0.50 -0.42 10.49 <0.0001 2.7 0.08 iii) No. plant spp. 4.36 0.045 0.40 0.28 8.53 0.001 0.83 0.44

12

10

8

6

4 Minor Average diaspore size Average 2 Mixed

0 Major 0 5 10 15 20 25 30

Gape width (mm)

Figure 4.1 The average size of diaspores consumed compared with gape width for major (n = 10, av. = 7.77, r2 = 0.36), mixed (n = 15, av. = 6.61, r2 = 0.42) and minor (n = 12, av. = 4.42, r2 = 0.16) frugivores.

98 0.6

0.5

0.4

0.3 ≥

0.2 Minor 0.1 Mixed Prop. species gape width 75% 00.0 Major 0 5 10 15 20 25 30

Gape width (mm)

Figure 4.2 The average proportion of diaspores close to the maximum handling capacity (≥ 75% of gape width) consumed by major (n = 10, av. = 0.31, r2 = 0.44), mixed-diet (n = 15, av. = 0.20, r2 = 0.39) and minor (n = 12, av. = 0.04, r2 = 0.01) frugivores.

4.3.3 Plant species richness of the diets of major, mixed and minor frugivores

The minimum number of native plant species recorded in the diet of a frugivore species was two (galah, noisy friarbird and blue-faced honeyeater, all minor frugivores), compared with a maximum of 128 (the mixed-diet Lewin’s honeyeater) (Table 4.1).

Major and mixed-diet frugivores consumed a higher average number of plant species than minor frugivores (Table 4.2, Figure 4.3; eta squared 0.34). The relationship between the number of native plant species consumed by a frugivore and its gape width

(approximately 3:1) was marginally significant (Table 4.2; eta squared 0.12). There was large variation in the number of native plant species consumed by different major and mixed-diet frugivore species (Figure 4.3). The frugivore species with low dietary plant species richness tended to consume plants from a narrow range of plant taxa; for example, the major frugivore mistletoebird mostly consumed plant species from the

99 Loranthaceae and the mixed-diet barred cuckoo-shrike consumed a high dietary proportion of Ficus spp. (70%; Table 4.1).

120

100

80

60

40 Minor

No. native plant species 20 Mixed

0 Major 0 5 10 15 20 25 30

Gape width (mm)

Figure 4.3 The number of native plant species consumed by each frugivore in major (n = 11, av. = 47.00, r2 = 0.35), mixed (n = 15, av. 47.87, r2 = 0.09) and minor (n = 12, av. = 6.08, r2 = 0.24) frugivory levels.

4.3.4 Plant species diet composition in relation to frugivory level, gape width and seed

treatment

Almost half (104) of the 244 plant species in the data set were recorded in the diet of both major and mixed-diet frugivores (but were not known to be consumed by minor frugivores); a further 34 plant species were consumed by at least one frugivore species from each of the three frugivory levels (Figure 4.4). Minor frugivores consumed a subset of the plant species consumed by major and mixed-diet frugivores, except for a single plant species (Melicope vitiflora Rutaceae). There were 32 and 65 plant species known only from the diets of major or mixed-diet frugivores, respectively (Figure 4.4).

In most cases, these plant species were from families that were also known from the 100 diets of bird species from other frugivory levels, suggesting that these taxa may actually be consumed by both major and mixed-diet bird species. However, all three of the plant species in the data set that were from Agavaceae (all from the genus Cordyline) were only known to be consumed by the mixed-diet regent bowerbird. The four species from

Celastraceae were also only known to be consumed by mixed-diet frugivores.

32

104 Major 65

Mixed 34 1 1 7 Minor

Figure 4.4 Overlap in the number of plant species consumed by frugivorous birds in relation to frugivory level (major (n = 11, including channel-billed cuckoo), mixed-diet (n = 15) and minor (n = 12) frugivores)

Only four plant species, all figs, were consumed by at least half of the 38 bird species: Ficus macrophylla (consumed by 26 bird species), F. platypoda (25), F. obliqua (23) and F. fraseri (21). At least two-thirds of the bird species had been recorded consuming plant species from the families Moraceae, Euphorbiaceae,

Sapindaceae, Myrtaceae and Elaeocarpaceae, although few minor frugivore species were known to consume these last two families (Table 4.3).

Major frugivores consumed higher numbers of native plant species from the

Lauraceae than mixed-diet species (Table 4.4). Mixed-diet frugivores consumed a higher number of native plant species from the families Celastraceae, Mimosaceae,

Sapindaceae, Smilacaceae and Urticaceae.

101 Table 4.3 The proportion of frugivorous bird species in each frugivory level that had been recorded consuming native plant species from 40 of the plant families represented in the data set1. Plant families that were consumed by at least half of the bird species in a frugivory level are shown in bold. The total number of bird species that was known to consume plant species from each family is also shown.

Proportion of frugivore species Total No. Mixed- Major Minor no. Plant family plant diet (n=11) (n=12) frug. spp. (n=15) spp. Agavaceae 3 0.00 0.13 0.00 2 Philesiaceae 2 0.00 0.20 0.00 3 Thymelaecae 2 0.09 0.20 0.00 4 Apocynaceae 2 0.00 0.27 0.11 5 Araceae 2 0.18 0.20 0.00 5 Celastraceae 4 0.00 0.33 0.00 5 Cucurbitaceae 3 0.18 0.20 0.00 5 Epacridaceae 3 0.09 0.20 0.11 5 Eupomatiaceae 2 0.18 0.20 0.00 5 Smilacaceae 3 0.09 0.27 0.00 5 Sterculiaceae 2 0.00 0.33 0.00 5 Menispermaceae 4 0.27 0.20 0.00 6 Myrsinaceae 3 0.18 0.27 0.00 6 Santalaceae 2 0.00 0.27 0.22 6 Sapotaceae 4 0.27 0.20 0.00 6 Icacinaceae 2 0.45 0.13 0.00 7 Symplocaceae 2 0.36 0.20 0.00 7 Rosaceae 4 0.27 0.33 0.00 8 Mimosaceae 3 0.09 0.53 0.11 10 Solanaceae 3 0.27 0.40 0.11 10 Pittosporaceae 4 0.36 0.47 0.00 11 Verbenaceae 5 0.36 0.40 0.00 11 Ebenaceae 4 0.64 0.33 0.00 12 Oleaceae 6 0.64 0.33 0.11 13 Rubiaceae 10 0.45 0.53 0.00 13 Urticaceae 3 0.18 0.53 0.22 13 Anacardiaceae 2 0.45 0.60 0.00 14 Vitaceae 6 0.64 0.47 0.00 14 Lauraceae 21 0.73 0.53 0.00 16 Ulmaceae 2 0.18 0.60 0.56 16 Rutaceae 11 0.64 0.60 0.11 17 Arecaceae 5 0.64 0.60 0.11 18 Araliaceae 4 0.73 0.67 0.11 19 Meliaceae 7 0.64 0.67 0.22 19 Rhamnaceaae 4 0.45 0.80 0.22 19 Elaeocarpaceae 6 0.73 0.73 0.11 20 Myrtaceae 20 0.91 0.53 0.22 21 Sapindaceae 17 0.64 0.80 0.67 26 Euphorbiaceae 11 0.82 0.80 0.67 28 Moraceae 13 0.91 0.87 0.78 31 1 The remaining 28 plant families represented in the data set had only been recorded in the diet of one frugivorous bird species

102

Table 4.4 The average number of native plant species from selected plant families that were consumed by major and mixed-diet frugivores. Plant families with three or more plant species in the data set were included. ‘No spp.’ shows the number of native plant species from each family represented in the data set. ‘p’ shows the results of t-tests comparing the number of plant species between major and mixed-diet frugivores; significant (p<0.05) results are shown in bold. Statistical significance was determined from randomisation for all plant families except Lauraceae, Myrtaceae and Sapindaceae (see text).

Average no. plant spp. Plant family No. Major1 Mixed1 p spp. n = 8 n = 10 Araliaceae 4 1.63 1.60 0.38 Arecaceae 5 1.88 1.60 0.25 Celastraceae 4 0.00 0.60 0.005 Curcurbitaceae 3 0.25 0.40 0.58 Ebenaceae 4 1.25 1.20 0.85 Elaeocarpaceae 6 2.75 3.20 0.40 Epacridaceae 3 0.25 0.30 0.94 Euphorbiaceae 11 2.38 3.20 0.81 Lauraceae 21 9.88 4.80 0.04 Meliaceae 7 1.88 3.20 0.16 Menispermaceae 4 0.38 0.40 0.66 Mimosaceae 3 0.25 1.40 0.03 Moraceae 13 7.13 7.40 0.49 Myrsinaceae 3 0.38 0.40 0.58 Myrtaceae 20 4.13 5.50 0.26 Oleaceae 6 1.75 1.20 0.24 Pittosporaceae 4 0.50 0.80 0.17 Rhamnaceae 4 0.88 1.70 0.08 Rosaceae 4 0.63 0.90 0.57 Rubiaceae 10 1.13 1.70 0.45 Rutaceae 11 2.13 2.10 0.90 Sapindaceae 17 2.13 5.00 0.003 Sapotaceae 4 0.38 0.30 0.65 Smilacaceae 3 0.13 0.40 0.01 Solanaceae 3 0.63 0.70 0.93 Urticaceae 3 0.13 1.10 <0.001 Verbenaceae 5 0.63 0.90 0.47 Vitaceae 6 2.63 2.50 0.83 1 Three major frugivore species (emerald dove, channel-billed cuckoo and mistletoebird) and five mixed- diet frugivore species (Australian brush turkey, wonga pigeon, pale-headed rosella, barred cuckoo-shrike and varied triller) were not included in these analyses because of the low plant species richness of their diets (Table 4.1, Figure 4.4). Minor frugivores were not included in analyses because of the low number of plant species that they consumed from most families.

The UPGMA classification assembled frugivore species into four groups that broadly corresponded with frugivory level (Figure 4.5). Most major frugivores were classified together in Group 1, most mixed-diet frugivores were in Group 2 and most 103 minor frugivores were in Group 3 (Figure 4.5). Group 4 comprised species from all three frugivory levels; the two major frugivore species in Group 4 were small gaped, and the three mixed-diet species were either small gaped or seed crushers.

Table 4.5 shows the percentage dissimilarity between the different groupings of bird species in the classification and the plant species that contributed most to these differences. There was a high level of dissimilarity among all groups: the two groups of bird species that had the least dissimilar patterns of plant species consumption were

Groups 1 and 2 (Table 4.5). The plant species listed for each set of pair-wise comparisons were consumed by many of the bird species in one of the groups and relatively few of the bird species in the other group. In many cases, the same plant species may have contributed to differences among more than two groups. For example,

Guioa semiglauca (Sapindaceae) was more common in the diets of bird species in

Group 4 than 3 but more common among Group 2 birds than those in Group 4. Bird species in Groups 3 and 4 generally consumed low numbers of plant species.

Consequently, the plant species that contributed most to the dissimilarity between the groups were all more frequent in the diets of species in Groups 1 and 2 than 3 or 4

(Table 4.5). In terms of differences in diet composition, bird species in Groups 3 and 4 consumed few of the species consumed by Groups 1 and 2. In contrast, bird species in

Groups 1 and 2 consumed most of the plant species consumed by the other birds.

The plant taxa that distinguished the bird species in Group 1 from Groups 2, 3 and 4 consistently included several from the Lauraceae (Table 4.5). This was consistent with the results of univariate comparisons of major and mixed-diet bird species (Table

4.4). In addition, consumption of the large fruited Eleaocarpus grandis (22.5 mm median diameter) and individual species from Burseraceae, Ebenaceae and Vitaceae distinguished this group of bird species from all other groups (Table 4.5). The bird species in Group 2 were distinguished from birds in Groups 1, 3 and 4 by their

104 consumption of arillate species from the Elaeocarpaceae, Mimosaceae and Sapindaceae

(Table 4.5); five of the six plant species that distinguished between Groups 2 and 1, and which were more abundant in the diets of species in Group 2, were arillate species.

Shared consumption of species from Araliaceae and Arecaceae distinguished birds in

Group 1 and 2 from those in Groups 3 and 4. Bird species in Group 3 were distinguished from Group 4 by their consumption of Ficus spp. Bird species in Group 4 were distinguished by Euphorbiaceae, Solanaceae and certain species in Sapindaceae and Rhamnaceae.

105 ST G FL

D S ■ C L ◊ C S ▲ C S ■ 4 D S ◊ D L ▲ C S ◊ D S ▲ D M ◊ D L ▲ D L ■ D L ■ D M ▲ 3 D L ▲ C S ◊ C M ▲ C M ▲ C L ▲ C M ◊ C L ◊ D L ◊ D M ◊ 2 C M ■ D S ◊ D L ◊ D M ◊ D L ◊ D L ◊ D L ■ D L ■ 1 D M ■ D L ◊ D L ■ D M ■ C M ■

Figure 4.5 Classification of frugivore species based on Bray-Curtis similarity in patterns of consumption of native plant species. ‘ST.’ is seed treatment (D seed-disperser, C seed-crusher); ‘G’ is gape width class ( S small (< 10 mm), M medium (10 - 15 mm) , L large (> 15 mm)); and ‘FL’ is frugivory level ( ■ Major, ◊ Mixed, ▲ Minor). 106

Table 4.5 The ten plant species that contributed most to the dissimilarity between each pair of groups formed in the classification. ‘Pair-wise comparison’ shows the two groups being compared, ‘Average dissimilarity’is the percentage dissimilarity between the two groups based on their patterns of plant species consumption (100 is total dissimilarity). The numbers in columns show the group in which the corresponding plant species was consumed by the higher number of bird species. Note that in comparisons of Groups 1 (n=10) and 2 (n=7) with Groups 3 (n=10) and 4 (n=8), distinguishing plant species were always consumed by the highest number of bird species from Groups 1 and 2. Pair-wise comparison 1 v 2 1 v 3 1 v 4 2 v 3 2 v 4 3 v 4

Average dissimilarity 63.58 83.65 92.40 79.55 88.52 90.25 Plant species family Euroschinus falcata Anacardiaceae 2 2 Polyscias elegans Araliaceae 1 1 2 2 P. murrayi Araliaceae 1 1 2 2 Archontophoenix Arecaceae 1 1 2 2 cunninghamiana Canarium australasicum Burseraceae 1 1 1 Diospyros pentamera Ebenaceae 1 1 1 Elaeocarpus grandis Elaeocarpaceae 1 1 1 E. kirtonii Elaeocarpaceae 1 1 E. obovatus Elaeocarpaceae 2 2 Sloanea australis Elaeocarpaceae 2 2 2 Glochidion ferdinandi Euphorbiaceae 2 4 Macaranga tanarius Euphorbiaceae 3 Omalanthus nutans Euphorbiaceae 1 4 Beilschmedia obtusifolia Lauraceae 1 B. elliptica Lauraceae 1 1 1 Cinnamomum oliveri Lauraceae 1 1 1 Cryptocarya Lauraceae 1 glaucescens C. obovata Lauraceae 1 1 1 C. triplinervis Lauraceae 1 1 Litsea australis Lauraceae 1 1 1 L.reticulata Lauraceae 1 1 1 Neolitsea dealbata Lauraceae 1 1 1 Melia azedarach Meliaceae 1 3 Acacia maidenii Mimosaceae 2 2 2 A. melanoxylon Mimosaceae 2 2 2 Ficus coronata Moraceae 2 2 F. fraseri Moraceae 3 F.macrophylla. Moraceae 2 3 F. obliqua Moraceae 1 2 3 F. platypoda Moraceae 2 3 F. rubiginosa Moraceae 1 3 F. superba Moraceae 3 F. virens Moraceae 3 F. watkinsiana Moraceae 1 4 Streblus brunonianus Moraceae 2 2 Acmena smithii Myrtaceae 1 1 3 Olea.paniculata Oleaceae 1 Piper novae-hollandiae Piperaceae 1 1 Alphitonia excelsa Rhamnaceae 2 2 4

107 Table 4.5 (cont.) Pair-wise comparison 1 v 2 1 v 3 1 v 4 2 v 3 2 v 4 3 v 4

Average dissimilarity 63.58 83.65 92.40 79.55 88.52 90.25 Plant species family A. petreii Rhamnaceae 2 2 Coprosma quadrifida Rubiaceae 2 Acronychia oblongifolia Rutaceae 2 1 1 Melicope micrococca Rutaceae 2 2 2 Exocarpus Santalaceae 3 cupressifolius Diploglottis australis Sapindaceae 2 2 3 Elattostachys xylocarpa Sapindaceae 2 2 2 Guioa semiglauca Sapindaceae 2 2 4 Jagera pseudorhus Sapindaceae 4 Sarcopteryx stipitata Sapindaceae 2 Solanum aviculare Solanaceae . 4 Aphananthe philipensis Ulmaceae 2 3 Dendrocnide excelsa Urticaceae 2 Cissus sterculifolia Vitaceae 1 1 1

Large- and medium-gaped bird species are interspersed in the classification (Figure 4.5).

Frugivores with small (<10 mm) gapes mostly consumed a subset of the plant species that were consumed by frugivores with medium or large gapes (Figure 4.6). There was considerable overlap in the plant species consumed by birds from all gape widths. Of the 48 plant species only consumed by large-gaped frugivores, a large number (27; 56%) had large (≥10 mm) diaspores, compared with seven (23%) of the 30 species for medium-gaped birds and one (14%) of the seven plant species only consumed by small-gaped birds (Chi square 3x2 contingency table comparing the number of plant species with diaspores ≥10 mm and <10 mm among the three gape width classes, χ2 = 11.05, p

= 0.004).

108

30 108

Medium 48

46 2 Large

7 3

Small

Figure 4.6 Overlap in the number of plant species consumed by frugivorous birds in relation to gape width classes (small (< 10 mm, n = 6), medium (10 – 15 mm, n = 15) and large (> 15 mm, n = 17) gape widths).

4.4. Discussion

4.4.1 Overlap and variation among frugivorous bird species in patterns of plant species

consumption

This study shows that many of the rainforest plant species in subtropical Australia are consumed by multiple frugivorous bird species. Frequently, the bird species that consume fruits from a given plant species may vary in their degree of frugivory, gape size and seed treatment. As in other regions, figs are consumed by most frugivore species, irrespective of their degree of frugivory, gape size or other attributes (Ridley, 1930; Janzen, 1979; Snow, 1981; Wheelwright et al., 1984; Shanahan et al., 2001). Many of the plant species in the Euphorbiaceae and Sapindaceae are also consumed by many frugivorous bird species in the present study, as in other regions (Snow,

1981; Wheelwright et al., 1984; Silva et al., 2002). However, the present work also shows that there is substantial variation among sets of frugivore species in their consumption of certain plant species.

This variation is related to bird species’ gape width, degree of frugivory and seed treatment. For example, species in the Araliaceae and Arecaceae were recorded in the diets of a large number of major and mixed-diet bird species, particularly seed dispersing species with gapes > 10 mm, but were consumed by few minor frugivores. 109 The present study documents less overlap among frugivorous bird species in their patterns of plant species consumption than has previously been implied in functional classifications that have grouped most bird species together as a functional unit (e.g., van der Pijl, 1982; Gautier-Hion et al.,

1985; Bollen et al., 2004). It is possible that increased data on fruit-frugivore interactions in the study region would show greater overlap among bird species in their patterns of consumption of plant species. For example, interactions involving rarer plant and/or bird taxa are likely to be under- recorded in field observations of frugivory (Silva et al., 2002). However, the data used in the present work were collected over large geographical and temporal scales. This suggests that the major differences among frugivore species in their pattern of plant species consumption may be consistent over space and time.

4.4.2 Frugivore gape width and patterns of fruit size consumption

There was a positive association between a bird species’ gape width and the average size of fruits that it consumed. The average size of fruits consumed by minor frugivore species was substantially smaller than their handling capacity. This contrasted with major and mixed-diet frugivores that consumed relatively high proportions of fruits that were close to their maximum handling capacity. The patterns of plant consumption shown by most major and mixed-diet frugivorous bird species are consistent with the gape-limited patterns of fruit size consumption reported by Wheelwright (1985) for frugivorous birds in Costa Rica. It has been proposed that net energy yield would be higher for fruits that are close to a species’ fruit size handling capacity because there may be additional searching and handling costs associated with consuming small fruits (Martin, 1985; Herrera, 1987; Sallabanks and Courtney, 1993). Furthermore, because larger seeds may be regurgitated more readily than smaller seeds, and hence eliminated more rapidly than small seeds can be defecated, net rates of pulp intake may be higher when consuming larger fruits

(Murray et al., 1993).

110 However, larger-gaped birds also consumed small fruits, as has been documented in Costa

Rica (Wheelwright, 1985) and Thailand (Kitamura et al., 2002). Even the exclusively frugivorous bird species with larger gapes (the Ptilinopus fruit doves and topknot pigeon) consumed some small fruits (Table 4.1). Consequently, rather than consuming a distinctly different set of plant species, small-gaped frugivore species consumed the small-fruited subset of the plant species that were collectively consumed by larger-gaped bird species. With the exception of the largest fruits, which were only consumed by some of the bird species with gapes wider than 15 mm, there was also considerable overlap in patterns of plant species consumption between birds with medium and large gapes. Although there may be energetic or nutritional advantages to consuming larger fruits, patterns of fruit size consumption may be complicated by variation among plant species in factors such as their fruit pulp to seed ratios (Howe and van der Kerckhove, 1980; Herrera, 1987) and pulp chemistry (Martínez del Rio and Restrepo, 1993; Cipollini and Levey, 1997).

4.4.3 Frugivory level and patterns of plant species consumption

McKey (1975) proposed that patterns of plant consumption would vary between highly frugivorous species and those with mixed diets. Specifically, it was reasoned that highly frugivorous species would specialise on lipid-rich fruits, while mixed-diet frugivores would consume fruits from carbohydrate-rich plant species (McKey, 1975; Howe and Estabrook, 1977;

Snow, 1981). In subtropical Australia, there is substantial overlap among major and mixed-diet frugivores in their patterns of plant species consumption, contrary to the predicted dichotomy.

Major frugivores did consume the highest number of plant species from the family Lauraceae, a family known for the high lipid content of fruit pulp (Snow, 1971, 1981; Crome, 1975; Stiles 1993).

However, as in other regions (e.g., Howe, 1981; Herrera, 1984; Fuentes, 1994; Sun et al., 1997), major frugivores were not the only consumers of plant species that may contain high lipid content.

There is little information regarding nutrient, mineral or chemical content of Australian fruits.

However, the plant taxa that characterised the diets of mixed-diet frugivores included families that

111 bear fruits with lipid-rich pulp in other regions, including Celastraceae (Corlett, 1996) and

Sapindaceae (Snow, 1981). These data suggest that lipids may be energetically important for both major and mixed-diet frugivores in subtropical Australia.

However, neither major nor mixed-diet species specialised on lipid-rich fruits, and both groups consumed many plant taxa that are associated with high carbohydrate content in other regions (e.g., Moraceae). This may reflect the need to consume a variety of nutrients and minerals

(Pulliam, 1975; Schaefer et al., 2003), or to avoid consuming toxic amounts of fats (Bairlein, 1998) or secondary chemicals that may be associated with particular plant taxa (Bairlein, 1996; Cipollini and Levey, 1997). In addition, lipid-specialisation may not be possible because plants that bear lipid-rich fruits may not always be fruiting (Leighton and Leighton, 1983; Wheelwright, 1986;

Innis, 1989).

There was a consistent difference between major and mixed-diet frugivores in their consumption of plant species with arillate fruits. It is difficult to provide a mechanistic explanation for the disproportionate consumption of arillate species by mixed-diet frugivores. It was reported that birds of paradise (family Paradisidae) were the sole consumers of certain arillate fruits in Papua

New Guinea, interpreted to be a function of their relatively long and narrow bills (Pratt and Stiles,

1985). However, a more recent study showed that other frugivore groups, including fruit pigeons, also consumed arillate fruits in Papua New Guinea (Brown and Hopkins, 2002).

In the present study, a set of bird species that were classified as mixed-diet frugivores showed patterns of plant species consumption that closely resembled several major frugivores

(Group 1 in the classification). The major frugivore bird species in this group included the

Ptilinopus fruit-doves and topknot pigeon, which are among the most highly frugivorous bird species in the world (Crome, 1975). These species shared a substantial proportion of plant species with the mixed-diet green catbird, satin and regent bowerbirds (all in the family Ptilonorhynchidae) and pied currawong. Similar patterns of plant species consumption were shown for fruit-dove and bowerbird species from these same genera in Papua New Guinea (Pratt and Stiles, 1985; Brown and

112 Hopkins, 2002). Similarity in patterns of plant species consumption between these bird species may be related to a substantial increase in the degree of frugivory of the bowerbirds, catbird and pied currawong during the non-breeding season (Blakers et al. 1984; Innis and McEvoy, 1992; Frith et al., 2004). This seasonal switch to a fruit-dominated diet may necessitate the consumption of plant taxa with particular nutritional attributes (Bairlein, 1996). Hence, frugivore species that have fruit- dominated diets during part of the year may show overall patterns of plant consumption similar to species that have fruit-dominated diets throughout the year.

4.4.4 Variation among bird species within a frugivory level

Although the structure of fruit-frugivore interactions in the assemblage studied is related to degree of frugivory, there is also substantial variation in patterns of plant consumption within the major and mixed-diet frugivore groups. Minor frugivore species were similar to one another; they consumed a low number of native plant species, predominantly Ficus spp., and were mostly classified together in Group 3. However, there were several major and mixed-diet frugivore species

(in Groups 3 and 4 in the classification) that consumed only a subset of the plant species consumed by other major and mixed-diet frugivores. Based on data collected in Papua New Guinea, Brown and Hopkins (2002) suggested that some relatively frugivorous species, including the barred cuckoo-shrike (“yellow-eyed cuckoo-shrike” in their study), may specialise on figs. It appears that this is the case for this and certain other major and mixed-diet species that had high dietary proportions of figs (e.g., in the present study, channel-billed cuckoo, common koel, wonga pigeon).

Overall, seed-crushing species were not distinguished from seed-dispersing species based on dietary composition, although there were insufficient data to test for an effect of seed treatment within frugivory levels. For example, it may be reasonable to expect that seed-crushing major frugivore species show patterns of plant species consumption similar to mixed-diet frugivores, since they derive nutrition from seed as well as fruit pulp (Snow, 1981; Innis, 1989; Jones and Crome,

1990). In this study, two of the three seed-crushing species that were classed as major frugivores

113 (emerald dove and brown cuckoo-dove) had diets that resembled mixed-diet species more than major frugivores. The diet of the seed-crushing white-headed pigeon was relatively similar to the other major frugivore species.

Classification based on functional attributes may overestimate the functional similarity between certain species at finer scales. For example, in the present study, both the figbird and wompoo fruit-dove have large gapes and fruit-dominated diets and have relatively similar dietary composition in the context of the entire avian frugivore assemblage. However, in a pair-wise comparison of plant species dietary composition, the figbird was only known to consume 44% of the plant species consumed by the wompoo fruit-dove (see Chapter Five). Consideration of taxonomic relatedness among species may help elucidate some of the additional variation within the groups formed using functional attributes. In the present study, close taxonomic relatives with similar functional traits tended to have the most similar diets (e.g., the wompoo and rose-crowned fruit-doves and topknot pigeon; bowerbirds and catbird). However, some sets of close relatives with different functional traits also had similar diets (e.g., black-faced and barred cuckoo-shrike), whereas others did not (e.g., figbird and olive-backed oriole). Secondary chemical compounds may generate similar patterns of plant species consumption by taxonomic relatives.

4.4.5 Gape width and frugivory level as indicators of the functional potential of frugivorous birds

as seed dispersers

There is a worldwide concern that large fruited plant species may not be dispersed in disturbed rainforest (Corlett, 1996, 1998; Corlett and Turner, 1997; Harrington et al,. 1997; Silva and Tabarelli, 2000; McConkey and Drake, 2002). Studies investigating the potential for a frugivorous bird species to disperse large-fruited plants have often considered its gape width as an indication of its functional capacity (e.g., Silva and Tabarelli, 2000; McConkey and Drake, 2002). It has recently been argued that gape size is an unreliable measure of fruit size handling capacity

(Dennis and Westcott, 2006). However, with the notable exception of the channel-billed cuckoo, the

114 present study has shown a strong association between gape width and patterns of fruit size consumption among the avian frugivore assemblage of subtropical Australia, provided that gape distensibility and frugivory level are accounted for. For example, the consumption of large fruits by

Ptilinopus fruit-doves could not be predicted on the basis of hard-tissue bill dimensions (Dennis and

Westcott, 2006). However, their capacity to handle larger fruits is evident if their gape distensibility is incorporated into the measurement of gape width (as described in Chapter Three, Section 3.2.1).

The present work has also shown the importance of incorporating a measure of frugivory level into analyses of seed dispersal potential to avoid over-estimating the ability of minor frugivore species to disperse large seeds. For example, based on the results presented here, it would be predicted that, despite its wide gape, the black-faced cuckoo-shrike would only consume small fruits because of its low degree of frugivory. In most studies relating patterns of fruit size selection to frugivore attributes (e.g., Silva and Tabarelli, 2000), frugivory level has not been explicitly considered.

In Chapter Three, it was proposed that classification of frugivorous bird species based on their gape width, frugivory level and seed crushing behaviour should yield functionally similar groups of species. The results of the present chapter show that groups formed using these frugivore attributes are associated with major differences in dietary composition among bird species in subtropical Australia. The classification of frugivorous species using readily available functional attributes provides a framework for predicting substantial differences among frugivore species in their roles as seed dispersers. In application to conservation, this approach could be used to forecast and manage the consequences of frugivore declines in fragmented forests for seed dispersal (Silva and Tabarelli, 2000; Kitamura et al., 2002). Ideally, classification based on the chosen attributes would not entirely substitute for detailed, species-specific dietary information. However, the collection of such information is extremely time-consuming. Furthermore, scientists are required to inform management decisions in the absence of this information. The approach demonstrated here provides a systematic means of identifying the plant and frugivore species that may be a priority for conservation and for research into patterns of plant-frugivore interactions.

115

Chapter Five

Reduced dispersal potential of native rainforest plant species in fragmented

rainforest

5.1 Introduction

Seed dispersal enhances the reproductive success of plants by removing seeds from competition, predation and other causes of seed and seedling mortality that are most intense directly beneath the parent (Janzen, 1970; Connell, 1971; Harms et al., 2000). Seed dispersal is also the agent of plant mobility, enabling colonisation of suitable germination microsites that become available following local disturbances within forest (Grubb, 1977; Hubbell, 1979). Frugivorous vertebrates disperse the seeds of most rainforest plants (Howe and Smallwood, 1982: Willson et al.,

1989). Therefore, declines in the abundance of frugivores following rainforest clearing and fragmentation may alter the rates or patterns of seed dispersal and plant regeneration (Corlett, 1998,

2002; Bleher and Böhning-Gaese, 2001, 2006; Cordeiro and Howe, 2001, 2003). For example, the complete absence of dispersers for a particular plant species would mean that recruitment could only occur beneath the crown of the parent plant, and may result in reduced recruitment (Bleher and

Böhning-Gaese, 2001; Cordeiro and Howe, 2003; Babweteera et al., 2007). Dispersal failure in fragmented forests would prevent the plant species from recolonising forest remnants from which it had gone extinct, and would mean it was unable to colonise cleared land during secondary succession (Poschlod et al., 1996; Duncan and Chapman, 2002). Consequently, plant species that experience dispersal failure would have low recruitment rates and restricted spatial distribution and be vulnerable to stochastic extinction (Fahrig and Merriam, 1994; Cochrane et al., 1999).

In the situation where dispersers are present but their abundance is greatly reduced, dispersal may not fail but would potentially be reduced. The consequences of substantially reduced dispersal

116 of a plant species for plant regeneration may resemble those described for dispersal failure, although the extent to which recruitment would be spatially and quantitatively limited would depend on the feeding rates and patterns of intra- and inter-habitat movements by remaining dispersers (Loiselle and Blake, 2002; Schupp et al., 2002; Dennis and Westcott, 2006).

The decline of certain frugivore species in fragmented landscapes may result in dispersal failure or reduction for plant species, but this depends on the dietary composition of other frugivore species in the regional frugivore assemblage. For example, it has been predicted that large-seeded plant species are unlikely to be dispersed in fragmented tropical rainforest regions worldwide as a result of the decline of the entire suite of frugivore species that are capable of dispersing large seeds

(Chapman and Chapman, 1995; Corlett, 1998, 2002; Silva and Tabarelli, 2000; Kitamura et al.,

2002; McConkey and Drake, 2002). On the other hand, most fleshy-fruited plant species are eaten and dispersed by multiple frugivore species (Howe, 1977; Howe and Smallwood, 1982;

Wheelwright and Orians, 1982; Brown and Hopkins, 2002), and many frugivore species do occur in fragmented landscapes (Estrada et al., 1993; Corlett, 1998; Renjifo, 1999; Chapters Two and Six of this thesis). Therefore, the consequences of the decline of one frugivore species for plant dispersal may be offset by increases in the density or consumption rates of other, functionally similar frugivore species (Corlett, 1998; Renjifo, 1999; Nathan and Muller-Landau, 2000; Loiselle and

Blake, 2002).

The potential for functional substitution among frugivore species can be examined by identifying the attributes of frugivores that reflect their role as seed dispersers (Silva and Tabarelli,

2000; Dennis and Westcott, 2006). For example, in Chapter Three, I showed that, while the bird species that declined in fragmented parts of subtropical Australia had large gapes and fruit- dominated diets, these attributes were shared by some of the frugivorous bird species that persisted or increased in abundance in fragmented parts of the landscape. I proposed that the plant species dispersed by decreaser species may not necessarily experience dispersal failure, although dispersal may be reduced, depending on whether the abundance and behaviour of substitute disperser species

117 fully compensated for the declines. The use of functional attributes of frugivores may be a useful means of identifying broad sets of plant taxa that they consume (Chapter Four), and hence the plants that may be vulnerable to frugivore declines (e.g., Silva and Tabarelli, 2000; Kitamura et al.,

2002; Chapter Three of this thesis). However, it may be necessary to compare actual dietary composition across the frugivore assemblage to determine the specific plant taxa that would be affected by changes in the abundance of individual frugivore species (Galetti, 2001). Among the studies that have considered the specific plant species consumed by individual frugivore species, most attention has been paid to large frugivore species and large-fruited plant species (e.g.,

Kitamura et al., 2002; McConkey and Drake, 2002; Babweteera et al., 2007).

The present study considers patterns of plant species consumption across an entire regional avian frugivore assemblage to assess the likelihood that the effects of fragmentation-related decreases in some bird species could be offset by the presence of other frugivore species with similar dietary composition. This study does not explicitly evaluate the potential for the changed abundance of certain frugivore species to result in other changes to seed dispersal, such as the dispersal of fewer seeds, or seed input to fewer or different microsites (Schupp, 1993; Jordano and

Schupp, 2000; Loiselle and Blake, 2002). Frugivore species were previously identified as showing decreased, increased or similar abundance in fragmented compared with intact forest in a rainforest landscape of subtropical Australia (Chapter Two). Here, records of the consumption of plant species by frugivore species are used to assess dietary similarity among frugivore species and among the groups of species that showed different abundance responses to fragmentation. Diet composition is assessed in terms of plant species, genus, family, and fruit size. This approach identifies plant taxa that are vulnerable to reduced dispersal, for example because they are known only from the diets of frugivores with decreased abundance in fragmented landscapes. This chapter will also examine potential changes in the dispersal of introduced plant species. The potential implications for the maintenance of plant regenerative potential in fragmented rainforest are considered.

118 5.2 Methods

5.2.1 Changes in the frugivorous bird assemblage in fragmented subtropical rainforest

The distribution and clearing history of subtropical rainforest in Australia were described in

Chapter One (Section 1.4.1). Chapter Two described the study region and site network used in the present work (Section 2.2.1), and field and analytical methods (Section 2.2.2).

The responses of avian frugivores to rainforest fragmentation in a region of subtropical

Australia were assessed in Chapter Two (Section 2.3.1). Three response patterns were identified among 26 frugivorous bird species by comparing species’ abundances between rainforest remnants and areas of regrowth relative to extensive forest. These three patterns were “decreaser” species

(n=5), which had lower abundance in remnants and / or regrowth than in extensive forest;

“increaser” species (n=7), which had higher abundance in remnants and / or regrowth than extensive forest; and “tolerant” species (n=14), which showed no clear difference in abundance in either remnant or regrowth habitats, compared with extensive forest. Other frugivorous bird species

(n=16) in the region were too uncommon for statistical analyses, and may therefore be likely to make a relatively small contribution to seed dispersal because of their low abundance. Nine of the

26 bird species grind or crush seeds (Chapter Three, Section 3.2.1) and are not considered further in the present chapter as they probably contribute relatively little to the dispersal of viable seed. The non-seed-crushing scarlet honeyeater Myzomela sanguinolenta was also excluded, due to the low number of observations of fruit consumption recorded to the level of plant species. Therefore, 16 frugivorous bird species (four decreaser, five increaser, and seven tolerant species) are considered here.

5.2.2 Fruit consumption database

Data on the consumption of plant species by the 16 frugivore species were derived from 100 published sources (Appendix 1), together with several unpublished data sets. The data used in the present study of 16 bird species in relation to their fragmentation-related abundance responses

119 comprised a subset of the database described in Chapter Four (Section 4.2.1); this chapter deals with 16 of the 38 frugivorous bird species that were considered in Chapter Four. The plant species included in analyses are listed in Appendix 2.

5.2.3 Data analyses

Spearman rank correlations were used to test for an association between the frugivore species’ sensitivity to fragmentation (scored as increaser (low sensitivity) = 1, tolerant = 2, decreaser (high sensitivity) = 3) and the total number of native plant species, genera and families that they consumed. Spearman rank correlation was also used to test the association between sensitivity to fragmentation and the number of native plant species consumed from the plant families with at least five species in the data set. The dietary proportions of exotic plant species and of native plant species with large (≥10 mm diameter) diaspores were compared among decreaser, tolerant and increaser frugivores, using Spearman’s rank correlations and with analysis of variance

(ANOVA), using frugivore species as replicates within each fragmentation response group. Pair- wise differences were tested using least significant difference (LSD) comparisons.

To examine similarities among the 16 frugivore species in their dietary composition, a classification tree was generated using the UPGMA algorithm (Manly, 1994) in PRIMER (5.2.9)

(Clarke and Warwick, 2001), with the Bray-Curtis dissimilarity metric. The statistical significance of overall dietary differences between frugivore response groups was tested using analysis of similarity, with 9 999 iterations (ANOSIM; Clarke and Green, 1988), also in PRIMER. Plant species with less than three consumer species, and frugivore species that consumed less than three native plant species were excluded. For analyses based on patterns of consumption of plant species, the Bray-Curtis dissimilarity was based on the presence of native plant species in the diet of frugivore species. For analyses at higher taxonomic levels, counts of the number of native species consumed from each plant genus or family were used. Genera or families with only one plant species in the data set were excluded.

120 Potential redundancy between pairs of frugivore species was quantified as the percentage of plant species in the diet of each decreaser frugivore that was also consumed by the other frugivore species. The redundancy between individual decreaser species and particular combinations of other species was also similarly quantified. The magnitude of potential dispersal reduction that would result from the absence of each individual frugivore species was assessed by calculating the number of plant species recorded solely in the diet of each frugivore species, as well as the number of plant species known only from the collective diet of groups of certain frugivore species (for example, all decreaser frugivores). The attributes of the plant species that were recorded only in the diet of decreaser frugivores were identified (higher taxonomic association, growth form and diaspore size) and compared with those of the remaining plant species in the data set. This comparison was made using chi-squared tests on cross-tabulations of species’ frequencies within attribute classes in SPSS

(2001).

5.3 Results

The data matrix comprised information for 254 plant species from 164 genera and 67 families, including 31 plant species introduced to subtropical eastern Australia from other continents, and three introduced from tropical Australia (collectively referred to as “exotic species”)

(Appendix 2). The data on plant species' presence in the diet of the 16 frugivore species yielded records of 912 different combinations of plant and frugivore species. Most of the 220 native plants

(70%) were recorded in the diet of more than one frugivore species.

5.3.1 Diet comparisons between frugivore response groups

There was considerable variation among frugivore species in the numbers of plant taxa they consumed, with numbers of native plant species ranging from one to 106 (Table 5.1). All decreaser frugivores, together with several tolerant or increaser species, consumed relatively high numbers of plant species, genera and families (Table 5.1). There was no statistically significant correlation

121 between sensitivity to fragmentation and the number of native plant species (Rs = 0.45, p = 0.08, n

= 16), genera (Rs = 0.36, p = 0.17) or families (Rs = 0.37, p = 0.16) consumed, although all associations were positive. Exotic plants comprised a larger average percentage of the diet of increasers (41%) than of tolerant (25%) or decreaser (9%) frugivores (Table 5.1) (Rs = 0.70, p =

0.002; ANOVA F = 4.24, p = 0.04).

122 Table 5.1 Numbers of plant taxa (N native, E exotic) consumed by each frugivore species. Bird species with gapes <10 mm are asterisked. ‘Gen.’ and

‘Fam.’ are genera and families, respectively.

Numbers of plant taxa Response Common name2 Code Genus and species Family Species Gen. Fam. Species Species pattern1 <10 mm ≥10 mm N E N E N E Decreaser wompoo fruit-dove womp Ptilinopus magnificus Columbidae 81 4 56 37 34 4 47 0 superb fruit-dove sfd P. superbus Columbidae 25 4 31 22 11 3 14 1 rose-crowned fruit-dove rcfd P. regina Columbidae 74 7 54 32 41 6 33 1 green catbird gcat Ailurioedus crassirostris Ptilonorhynchidae 104 7 77 40 60 7 44 0

Tolerant topknot pigeon topk Lopholaimus antarcticus Columbidae 73 4 46 28 30 3 43 1 common koel koel Eudynamys scolopacea Cuculidae 25 8 26 19 19 4 7 4 channel-billed cuckoo chan Scythrops novaehollandiae Cuculidae 6 0 2 3 6 0 0 0 little wattlebird * lwat Anthochaera chrysoptera Meliphagidae 3 3 7 7 3 3 0 0 Lewin’s honeyeater Lewhe Meliphaga lewinii Meliphagidae 104 21 89 47 77 17 27 4 pied currawong pcurr Strepera graculina Artamidae 50 10 39 25 32 7 18 3 satin bowerbird satbb Ptilonorhynchus violaceus Ptilonorhynchidae 106 12 89 52 55 9 51 3

Increaser black-faced cuckoo-shrike blfcs Coracina novaehollandiae Campephagidae 10 5 12 11 10 5 0 0 figbird figb Sphecotheres viridis Oriolidae 74 17 60 33 49 13 25 4 Torresian crow Tcrow Corvus orru Corvidae 10 6 16 15 9 5 1 1 silvereye * seye Zosterops lateralis Zosteropidae 36 17 39 30 35 15 1 2 Australian magpie Amag Gymnorhina tibicen Artamidae 1 5 6 6 1 3 0 2

Mean 48.9 8.1 40.6 25.4 29.5 6.5 19.4 1.6 SE 9.7 1.5 7.2 3.7 5.7 1.2 4.8 0.4 Total 220 34 164 67 130 22 90 12 1 From Chapter Two (Section 2.3.1); comparisons of abundance in extensive forest (E), remnants (M) and regrowth (G); Decreasers' abundance pattern is E>M>G except for green catbird (E=M>G); Tolerant pattern is E=M=G; Increasers' pattern is E=M

123

The average dietary proportion of native plant species with large diaspores (≥10 mm diameter) was significantly greater for decreasers (0.49) and tolerant frugivores (0.31) than increasers (0.09), (Rs = 0.72, p =0.004; ANOVA F = 6.09, p = 0.02, Figure 5.1). There was substantial variation among individual species within the tolerant response group; only two species

(topknot pigeon and satin bowerbird) consumed dietary proportions of native plants with large diaspores within the range shown by decreaser frugivores (Figure 5.1).

ANOSIM showed that the overall native plant species composition of the diets of decreaser frugivores was not significantly different from that of either tolerant or increaser frugivores (global

R = 0.137, p = 0.12). This was consistent both at the level of plant genus (R = 0.115, p = 0.15) and family (R = 0.093, p = 0.18). Most decreaser frugivore species consumed a broadly similar suite of plant species to one another (Figure 5.2), and were also similar to two tolerant frugivores: topknot pigeon (which resembled wompoo and rose-crowned fruit-doves), and satin bowerbird (similar to green catbird). The increaser figbird, tolerant Lewin’s honeyeater and tolerant pied currawong had the next most similar dietary composition to the group containing most of the decreaser species

(Figure 5.2). These patterns were similar when classification was conducted on both plant genus and family data. The superb fruit-dove’s diet did not closely resemble that of the other decreaser species, probably due to the low abundance of this mostly tropical bird species in subtropical

Australia (Innis, 1989; Date et al., 1996; Gosper and Holmes, 2002), and consequently the low number of subtropical plant species known in its diet (Table 5.1).

124 a .7 0 topknot pigeon ab satin bowerbird Lewin’s honeyeater .6 0

.5 0 b

.4 0

.3 0

.2 0

.1 0

10mm Proportion native species ≥ 10mm 0. 0 0

Dec Tol Inc

Response group Figure 5.1 The proportion of native plant species with large (≥10 mm) diaspores that were consumed by decreaser (Dec), tolerant (Tol) and increaser (Inc) frugivore species. Only species with gape widths >10 mm are included. The horizontal lines show the mean values. Letters above the scatter plot for each group indicate results of LSD comparisons; groups with different letters had significantly different means (p<0.05).

Spearman rank correlation showed that there was a positive association between increasing sensitivity to fragmentation and the number of plant species consumed from six of the 13 plant families that had more than five plant species represented in the data set (Table 5.2). The decreaser wompoo and rose-crowned fruit doves and green catbird generally consumed high numbers of native plant species from Lauraceae, Meliaceae, Myrtaceae, Rubiaceae, Verbenaceae and Vitaceae

(Figure 5.3). Two tolerant species, satin bowerbird and Lewin’s honeyeater, consumed numbers of plant species from the Meliaceae, Myrtaceae, Rubiaceae, Verbenaceae and Vitaceae, within the range shown by decreaser frugivore species (Figure 5.3). The tolerant topknot pigeon consumed a high number of native plant species from the Lauraceae and Vitaceae, similar to decreaser species.

The only increaser frugivore known to consume comparable numbers of native plant species from

125 these families was the figbird, which consumed relatively high numbers of species from Lauraceae and Meliaceae (Figure 5.3).

126 Decreaser Tolerant

Lewin's honeyeater Increaser figbird pied currawong wompoo fruit-dove topknot pigeon rose-crowned fruit-dove satin bowerbird green catbird superb fruit-dove common koel Torresian crow black-faced cuckoo-shrike channel-billed cuckoo silvereye 100 80 60 40 20 0

Dissimilarity

Figure 5.2 Classification of frugivore species (based on Bray-Curtis dissimilarity matrix and UPGMA sorting) according to presence / absence of native plant species in the diet. Symbols next to names show the response group for each frugivore species (see text). The little wattlebird and Australian magpie were not included because they consumed only three and one native species, respectively.

127

Table 5.2 The average number of native plant species from specified families consumed by decreaser (Dec n = 4), tolerant (Tol n = 7) and increaser (Inc n = 5) frugivores. The total number of native plant species from these families that are represented in the data set is shown. Significant (p <0.05) results are shown in bold.

Number of plant species No. in Mean no. consumed Spearman rank data set correlation Plant family Dec Tol Inc Rs p Arecaceae 5 1.8 1.1 0.6 0.38 0.08 Elaeocarpaceae 6 3.3 2.6 1.4 0.32 0.11 Euphorbiaceae 8 0.8 2.0 1.8 -0.23 0.20 Lauraceae 21 12.8 5.4 2.0 0.63 0.004 Meliaceae 7 2.8 2.1 0.6 0.44 0.04 Moraceae 13 8.0 7.4 5.8 0.21 0.21 Myrtaceae 19 6.5 4.4 1.0 0.59 0.008 Oleaceae 5 1.5 0.9 0.4 0.39 0.07 Rubiaceae 10 3.0 0.6 0.2 0.64 0.004 Rutaceae 10 2.8 1.7 0.8 0.41 0.06 Sapindaceae 15 2.5 3.1 2.8 0.11 0.68 Verbenaceae 5 1.3 0.7 0.0 0.58 0.01 Vitaceae 6 4.3 2.3 0.2 0.71 0.001

5.3.2 Specific substitution potential between frugivore taxa

Among tolerant and increaser frugivores, two tolerant species, topknot pigeon and satin bowerbird, consumed the greatest percentage of plant species that were consumed by individual decreaser species (Table 5.3; 56-73% and 52-66% respectively). A moderate percentage of the plant species recorded in the diets of individual decreaser frugivore species was consumed by the tolerant

Lewin’s honeyeater (38-49%), tolerant pied currawong (35-48%) and increaser figbird (40-53%), while other individual frugivore species consumed only a small percentage of the plant species recorded in the diets of decreasers (Table 5.3). In combination, the topknot pigeon and satin bowerbird consumed 72-81% of the plants recorded in the diets of individual decreaser frugivore species. The cumulative effect of remaining tolerant frugivores increased the percentage of shared plant species to 80-86%, while the addition of increaser species did not increase this further (80-

88%) (Table 5.3).

128

Lauraceae Meliaceae Myrtaceae

20 6 15 15 4 10 10 2 5 5

0 0 0

Rubiaceae Verbenaceae Vitaceae

8 4 8

6 3 6 No. of native species species No. of native 4 2 4

2 1 2 0 0 0

Dec Tol Inc Dec Tol Inc Dec Tol Inc

Response group

Figure 5.3 The number of native plant species consumed by decreaser (Dec), tolerant (Tol) and increaser (Inc) frugivores, for plant families where there was a significant (p < 0.05) association between sensitivity to fragmentation and the number of plant species consumed by different response groups. topknot pigeon, satin bowerbird, Lewin’s honeyeater. Most increaser species were not known to consume any plant species from these families.

130

Table 5.3 The percentages of native plant species that were recorded in the diets of each decreaser frugivore species (column head) and also consumed by each other frugivore (bird) species (row head)1, and by particular groups of species (‘Bird groups’). The number of native plant species consumed by each bird species is shown in Table 5.1.

Decreaser bird species womp sfd rcfd gcat Decreaser birds wompoo fruit-dove (womp) 80 70 55 superb fruit-dove (sfd) 25 23 15 rose-crowned fruit-dove (rcfd) 64 68 51 green catbird (gcat) 70 64 72 Tolerant birds topknot pigeon (topk) 73 68 68 56 common koel 19 16 22 16 chanel-billed cuckoo 7 8 7 5 little wattlebird 0 0 0 1 Lewin’s honeyeater 38 40 49 46 pied currawong 41 48 42 35 satin bowerbird (satbb) 58 52 61 66 Increaser birds black-faced cuckoo-shrike 6 8 7 8 figbird 44 52 53 40 Torresian crow 9 8 9 9 silvereye 12 24 18 18 Australian magpie 1 0 1 1 Bird groups topk & satbb 80 72 78 81 tolerant spp. (excl. topk & satbb) 59 64 66 58 tolerant & increaser spp. (excl. topk & satbb) 64 68 70 63 all tolerant spp. 86 80 86 86 all increaser spp. 46 56 55 48 all tolerant & increaser spp. 86 80 86 88 e.g., the cells in the top left of the table show that the wompoo fruit-dove is known to eat 80% of the plants recorded as eaten by the superb fruit-dove, but that the superb fruit-dove has only been recorded eating 25% of the plants eaten by the wompoo fruit-dove.

Twenty-seven native plant species were recorded only in the diet of decreaser frugivores (Figure 5.4, Table 5.4). These varied widely in their taxonomy, growth form and diaspore size, although plant species from the Rubiaceae comprised a much greater percentage of the 27 species (26%), than they did in the remainder of the data set (2%)

(χ2 = 27.1, p<0.0001). Tree species comprised a smaller proportion of plants consumed only by decreaser frugivores (29%), compared with the proportion of trees among the remaining species in the data set (52%) (χ2 = 4.02, p = 0.045). There was a trend for 131 vines to be more common among the plant species only known from the diets of decreasers than among remaining species (29%, 15%, χ2 = 2.93, p = 0.086). Among the plant species that were only known to be consumed by decreasers, there was no significant difference in the number of species that were shrubs (41%, 33%, χ2 = 0.32, p

= 0.57), had large (≥10 mm) diaspores (48%, 40%; χ2 = 0.37, p = 0.54), or were from the Rutaceae, Lauraceae or Myrtaceae (11%, 7%, 11%, compared with 4%, 10%, 8%; p

= 0.21, 0.96, 0.90). Eight of the 27 plant species that were only known to be consumed by decreaser frugivores belong to genera that were known to be consumed by tolerant or increaser frugivores, and all but one of the plant species were from families that were known to be consumed by non-decreaser bird species (Group 1, Table 5.5).

Figure 5.4 Overlap in the number of native plant species consumed by frugivore

13 3 Increaser 27 Decreaser 61 18

54

44 Tolerant species from the decreaser, tolerant and increaser response groups.

The satin bowerbird and topknot pigeon consumed a relatively high number of native plant species that were otherwise only consumed by decreaser frugivores (11 and

9, respectively; Table 5.4). The magnitude of dispersal reduction in fragmented parts of the landscape would be substantially higher if, in addition to decreasers, either the satin bowerbird or topknot pigeon were absent from these areas (23% and 17%, respectively;

132 Table 5.4). Apart from these two tolerant species and decreaser species, there are no known additional dispersers among the species analysed here for 32% of native plant species. Species with diaspores ≥10 mm diameter were much more common among the plant species that were only consumed by the bird species group comprising the topknot pigeon, satin bowerbird and decreasers, than in the remainder of the data set (60% compared with 32%; χ2 =14.34, p = 0.0002). In addition, species from the Rubiaceae

(11%, 1%; χ2 = 9.00, p = 0.003) were more frequent among the plant species that were only consumed by members of this group. There was no significant difference in the number of species from the Lauraceae (13% , 8%; χ2 = 0.80, p = 0.37), Myrtaceae (both

9%; χ2 = 0.05, p = 0.81), or Rutaceae (7%, 3%; χ2 0.84, p = 0.36), or in the number of species that were trees (44%, 52%; χ2 = 0.85, p = 0.36), shrubs (33%, 35%; χ2 = 0.01, p

= 0.91) or vines (23%, 13%; χ2 = 2.51, p = 0.11). Plant species that were only consumed by members of this group included both of the species from the Icacinaceae, three of the four native Verbenaceae and two of the three native Sapotaceae that were represented in the data set.

The Lewin’s honeyeater, satin bowerbird and figbird were the unique consumers of a relatively high number of native plant species (20, 12 and 8, respectively, Table

5.4). Therefore, the loss of any of these species may also result in substantially reduced dispersal of a noteworthy percentage of native rainforest plant species. In particular, the declined abundance of the Lewin’s honeyeater may cause a substantial reduction in the dispersal of 23% of the native plant species in the data set. Most other non-decreaser frugivore species had few native plant species for which they were the only recorded consumer (Table 5.4).

133 Table 5.4 For each frugivore (frug.), the number of native plant species that it consumed that were not consumed by another frugivore species (sp.), or were only also consumed by a decreaser (Dec.) frugivore(s) is shown. The magnitude of potential dispersal failure that would result from the absence of each individual frugivore species together with decreaser frugivores, if there were no gaps in the data set, is also shown. Bird codes are explained in Table 5.1. No native plant species Dispersal failure in consumed: absence of decreasers plus each other species1: not by Only by no. of plant % of plant other frug. subject sp. species species plus any Dec (n=220) Decreaser species womp 6 5 sfd 1 4 rcfd 5 5 gcat 8 4 All decreaser 27 27 27 12 Tolerant species topk 1 9 37 17 koel 0 0 27 12 chan 0 0 27 12 lwat 0 0 27 12 Lewhe 20 4 51 23 pcurr 1 0 28 13 satbb 12 11 50 23 Increaser species bfcs 1 1 29 13 figb 8 0 35 16 Tcrow 0 2 29 13 seye 3 0 30 13 Amag 0 0 27 12 Species groups topk and satbb 13 30 70 32 all tolerant 44 54 125 57 all tolerant and inc 75 118 220 100 1 calculated by adding the 27 plant species recorded only in the diet of decreasers, the number of plant species recorded only in the diet of the subject non-decreaser, and those plant species only recorded in their diet plus that of decreasers.

134 Table 5.5 Plant species that were only recorded in the diet of decreaser frugivores, and tolerant or increaser species that may substitute for them. Bird species codes are explained in Table 5.1. Med. dias. size = median diaspore size (for explanation see Section 4.2.1). Group Plant species Family Med. Growth Decreaser Tolerant and increaser spp. dias. form3 frugivore known to consume congeneric size species (or con-familial) plants4 (mm)

Group 11 Rubus moorei Rosaceae 1.0 V womp Lewhe, satbb, figb, Tcrow, seye (pcurr, Amag) Acronychia wilcoxiana Rutaceae 14.0 S womp topk, koel, Lewhe, pcurr, satbb, figb (seye) Jasminium didyum Oleaceae 6.0 V womp, rcfd satbb (topk, koel, Lewhe, pcurr, bfcs, figb, seye) Syzygium francisii Myrtaceae 15.0 T womp topk, koel, Lewhe, satbb, figb (pcurr) Syzygium johnsonii Myrtaceae 12.5 T sfd topk, koel, Lewhe, satbb, figb (pcurr) Cryptocarya rigida Lauraceae 13.0 T rcfd topk, Lewhe, pcurr, satbb, figb (koel, bfcs, Tcrow, seye) Endiandra muelleri Lauraceae 14.0 T womp, rcfd, sfd topk, koel (Lewhe, pcurr, satbb, figb, bfcs, Tcrow, seye) Symplocos stawellii Symplocaceae 3.0 T womp, rcfd, sfd topk, satbb

Group 22 Smilax australis Smilacaceae 7.5 V womp (satbb) Smilax glyciphylla Smilacaceae 10.0 V gcat (satbb) Calamus muelleri Arecaceae 10.5 V womp (topk, koel, Lewhe, pcurr, satbb, figb, Tcrow) Elaeagnus triflora Elaeagnaceae 12.0 V womp - Legnephora moorei Menispermaceae 13.0 V rcfd, gcat (Lewhe) Owenia cepiodora Meliaceae 17.5 T rcfd (topk, koel, Lewhe, pcurr, satbb, figb) Embelia australiana Myrsinaceae 6.5 V sfd, gcat (koel, Lewhe, satbb, seye) Micromelum minutum Rutaceae 6.0 S rcfd (topk, koel, Lewhe, pcurr, satbb, figb, seye) Sarcomelicope Rutaceae 12.5 T gcat (topk, koel, Lewhe, pcurr, satbb, figb, simplicifolia seye) Archirhodomyrtus beckleri Myrtaceae 8.5 S gcat (topk, koel, Lewhe, pcurr, satbb, figb) Phaleria chermsideana Thymelaceae 9.0 S gcat (Lewhe, seye)

135 Table 5.5 (cont.)

Plant species Family Median Growth Decreaser Tolerant and increaser spp. diaspor form3 frugivore known to consume congeneric e size species (or confamilial only) plants4 (mm)

Group 2 Rhodosphaera Anacardiaceae 9.5 T gcat (topk, koel, Lewhe, satbb, figb, seye) (cont.) rhodanthema Hodgkinsonia ovatiflora Rubiaceae 4.0 S gcat (Lewhe, satbb, bfcs, Tcrow, seye) Psychotria loniceroides Rubiaceae 5.5 S gcat (Lewhe, satbb, bfcs, Tcrow, seye) Aidia racemosa Rubiaceae 7.5 S rcfd (Lewhe, satbb, bfcs, Tcrow, seye) Ixora beckleri Rubiaceae 8.0 S womp, rcfd, gcat (Lewhe, satbb, bfcs, Tcrow, seye) Canthium coprosmoides Rubiaceae 11.0 S womp, sfd, gcat (Lewhe, satbb, bfcs, Tcrow, seye) Canthium odoratum Rubiaceae 6.5 S rcfd (Lewhe, satbb, bfcs, Tcrow, seye) Randia benthamiana Rubiaceae 17.5 S gcat (Lewhe, satbb, bfcs, Tcrow, seye) 1 plant species that are in genera consumed by tolerant or increaser species. 2 plant species that are in genera not consumed by tolerant or increaser species. 3 Source: Butler, 2003; S = shrubs and small trees, T = trees, C = tall climbers and understorey climbers (combined in analyses). 4 including records of consumption of plant genus only (i.e. unidentified species) and from exotic plant species.

136

5.4 Discussion

5.4.1 Reduced dispersal of native rainforest plants as a consequence of rainforest

fragmentation

Based on patterns of consumption of plant species by frugivorous bird species, substantially reduced dispersal is likely for 27 native rainforest plant species in the absence of four frugivorous bird species from fragmented rainforest. The analyses underpinning this result are based on data for approximately half of the fleshy-fruited plant species that were recorded in subtropical Australian rainforest in a comprehensive inventory by Butler (2003). If these data reflect patterns among the general fleshy- fruited flora of the region, 12% of native rainforest plant species may have severely reduced regenerative potential in fragmented forest because of the absence of known disperser species. If these plant species are unable to germinate under parent plants, for example due to high levels of fungal attack or seed predation, these species will fail to regenerate and may become extinct in rainforest fragments. Similar predictions have been made elsewhere where numbers of disperser species have declined dramatically

(Janzen and Vasquez-Yanez, 1991; Chapman and Chapman, 1995; Terborgh and

Nuñez-Iturri, 2006).

However, many plant species do regenerate in the absence of dispersers (Janzen and Martin, 1982; Chapman and Chapman, 1995; Corlett and Turner, 1997), although their recruits may be less abundant and more spatially aggregated than in forest with an intact disperser assemblage (Hubbell and Foster, 1990; Bleher and Böhning-Gaese,

2001; Schupp et al., 2002; Cordeiro and Howe, 2003). Plant species without dispersers would also be unable to colonise rainforest fragments following local extinction

(McEuan and Curran, 2004) or secondary regrowth on cleared land (Duncan and

Chapman, 2002; Ingle, 2003). The low regenerative potential of plant species may result

137 in high vulnerability to extinction in fragmented rainforest (Poschlod et al., 1996;

Cochrane et al., 1999).

The plant family Rubiaceae is well represented among the species that were identified as being vulnerable to substantially reduced dispersal in the fragmented rainforest landscapes of subtropical Australia. Substantially reduced dispersal in forest fragments may result in reduced numbers and a clumped spatial distribution of recruits

(e.g., Cordeiro and Howe, 2003; Babweteera et al., 2006). Recruitment of plants in the

Rubiaceae has been shown to be relatively low in isolated forest patches in both Brazil

(Tabarelli et al., 1999) and Singapore (Turner et al., 1996). This has been attributed to unsuitable germination conditions in fragments (Turner et al., 1996; Tabarelli et al.,

1999). If the patterns detected in the present study reflect the situation in other regions, low recruitment of plants from the Rubiaceae may be a consequence of frugivore declines, and subsequently reduced dispersal in fragmented forest.

However, it is also possible that plant species from the Rubiaceae are consumed by additional frugivore species in subtropical Australia than available data show. It is thought that many frugivore species disperse Rubiaceae in tropical regions (Snow,

1981; Silva et al., 2002), although patterns of consumption of this family are not well understood (Wheelwright et al., 1984). As in other regions (e.g., Corlett, 1996), plant species within the Rubiaceae in subtropical Australia are typically shrubs or small trees that bear medium-sized (average diameter 8.04 mm, n = 22) drupes or berries that are mostly black or orange in colour (Butler, 2003). These fruit characteristics are typically associated with consumption by many bird species (Gautier-Hion et al., 1985;

Wheelwright, 1985; Brown and Hopkins, 2002; Bollen et al., 2004). On the other hand, factors such as secondary metabolites that require specific digestive adaptation may limit the suite of frugivores that consume a plant taxon (Martínez del Rio and Restrepo,

1993; Izhaki et al., 2002). It is possible that chemistry is a factor limiting the

138 consumption of Rubiaceae by many bird species, since Izhaki et al. (2002) have shown that anthraquinones, a particular class of secondary compound, were especially common in the species of Rubiaceae that they studied, and that these deterred consumption by some bird species.

5.4.2 Potential for disperser substitution in fragmented forest

In general, increaser species in subtropical Australia have low potential to substitute for decreaser species as seed dispersers, since they consume a low number of native plant taxa, a high dietary proportion of exotic plants and a low dietary proportion of plant species with large fruits. In contrast, the overall diet consumption patterns of tolerant frugivores resembled those of the decreaser species, suggesting considerable potential for disperser substitution by tolerant frugivores in fragmented rainforest.

Hence, many native rainforest plants should retain some potential for dispersal within and between remnants and into regrowth, although this would depend on whether increaser species such as the figbird, and tolerant frugivores, particularly the topknot pigeon, satin bowerbird and Lewin’s honeyeater actually moved and transported seeds across habitat boundaries. However, this assessment of bird species’ potential to substitute for one another as seed dispersers is based on the presence or absence of plant species in their diets and does not account for variation in other factors that may influence plant regeneration outcomes, including the numbers of fruits consumed, or the microsites to which seeds are dispersed (Schupp, 1993).

Where the movement of frugivore species in fragmented forest has been studied in other parts of the world, few frugivore species have moved between forest fragments or into cleared areas (Duncan and Chapman, 2002; Silva et al., 1996; McEuen and

Curran, 2004). Patterns of native plant recruitment in weed-dominated regrowth in subtropical eastern Australia indicate that frugivorous birds do disperse seeds from a

139 variety of native plant species across cleared land (Neilan et al., 2006). In particular figbirds and large flocks of topknot pigeons regularly travel many kilometres over cleared land (Frith, 1957; Price 1999) and consequently may disperse seeds among fragmented rainforest patches.

However, most individual tolerant species consumed a relatively low percentage of the plant species that were consumed by decreaser frugivores. Consequently, the continued dispersal in remnant and regrowth patches of 30 of the plant species dispersed by decreasers (an additional 14% of plant species in the data set) may depend on only two tolerant frugivore species; the topknot pigeon and satin bowerbird. The reduced number of disperser species for these plant species, combined with a relatively low abundance of both topknot pigeon and satin bowerbird during site-based surveys in remnants and regrowth (Chapter Two, Table 2.2), suggests that dispersal of seeds of these plant species would be reduced in fragmented rainforest.

In addition to the Rubiaceae, this study identifies the plant families Lauraceae,

Meliaceae, Myrtaceae, Verbenaceae and Vitaceae, as well as those species with diaspores ≥ 10 mm wide, as being susceptible to reduced dispersal in fragmented forest in subtropical Australia. As with Rubiaceae, restricted recruitment of plants from the families Lauraceae, Myrtaceae and Meliaceae in forest fragments in other regions has been attributed to germination limitation (Turner et al., 1996; Tabarelli et al., 1999). As suggested by the results of the present study for subtropical Australia, reduced dispersal may also be important in these regions, especially since many plant species from these families are consumed by only a subset of the frugivore assemblage (Snow, 1981;

Wheelwright et al., 1984; Silva and Tabarelli, 2000).

In subtropical Australia, the Lewin’s honeyeater is known to consume several plant species from the families identified as being susceptible to reduced dispersal in fragmented forest. Populations of these bird species appear to be stable (Blakers et al.,

140 1984; Higgins et al., 2001), and numbers of this species were consistently high in the surveys conducted for this study (Chapter Two, Table 2.2). Therefore, in addition to the topknot pigeon and satin bowerbird, the Lewin’s honeyeater may help maintain the regenerative potential of some plant families in fragmented forest, although it may contribute little to the dispersal of large-seeded plant species from any family. In the study region, frugivorous Pteropid bats also use remnants and regrowth (Chapter Six).

These bats consume a range of plant species from the Myrtaceae (Eby, 1995; Chapter

Six of this thesis) and can transport large fruits in their teeth or claws (Ratcliffe, 1932).

However the seeds that they move beyond about 100-200 m from source trees are likely to be those that they swallow or carry in cheek pouches; smaller than 9 mm diameter

(Eby, 1995; McConkey and Drake, 2002; Meehan et al., 2005). Therefore, frugivorous bats may play an important role in dispersing seeds of these plants within remnants and possibly into adjacent cleared areas (Galindo-González and Sosa, 2003), but may contribute little to their dispersal among more widely separated fragments.

5.4.3 Implications for conservation of regenerative potential in fragmented rainforest

Data presented here and in Chapter Four suggest that most native Australian rainforest plant species are probably dispersed by multiple frugivore species. However, it is predicted that fragmentation-related changes in the frugivore assemblage of subtropical rainforests may result in substantially reduced dispersal of a suite of plant species. The size of this suite of plants is strongly dependent on the responses of two

“tolerant” frugivore species. Additional understanding of the factors affecting the abundance of both the topknot pigeon and satin bowerbird in response to landscape change is required because of the disproportionate effect that losing these species would have over seed dispersal. The topknot pigeon is widespread in certain fragmented rainforest landscapes of subtropical Australia (Date et al., 1996; Gosper and Holmes,

141 2002; Neilan et al., 2006). However, the distribution of the topknot pigeon was shown to be restricted in small fragments compared with extensive forest in another area of subtropical Australia (Howe et al., 1981). While the abundance pattern of topknot pigeons suggested fragmentation tolerance, there was a trend toward a decreaser abundance pattern, and numbers of this species were relatively low and highly variable in the region of the present study (Chapter Two). This species has previously undergone dramatic population declines following rapid rainforest clearing by European settlers in the late nineteenth century (Frith, 1952, 1957; Date et al., 1996). Similarly, the assessment of fragmentation tolerance of the satin bowerbird was based on relatively low occurrence during the surveys conducted for the present study, although this species was reported by Howe et al. (1981) to be relatively common in fragmented subtropical rainforest. If, in addition to identified decreaser species, the satin bowerbird and topknot pigeon declined in fragmented parts of the landscape, the present analyses suggest that the dispersal of one-third of native rainforest plant species may be substantially reduced in these areas. A comparable magnitude of dispersal reduction has been predicted from

Brazil (Silva and Tabarelli, 2000) and Thailand (Kitamura et al., 2002), regions in which a large proportion of rainforest has also been cleared.

142 Chapter Six

The distribution of frugivorous bats and their potential to disperse

seeds in fragmented rainforest.

6.1 Introduction

Seed dispersal by frugivorous fauna plays several important roles in the maintenance of biodiversity in fragmented landscapes. The dispersal of seeds within and between large forest tracts, remnant forest patches and other habitats helps maintain species and genetic diversity, initiates recolonisation after local extinction, and is crucial for natural regeneration of rainforest on cleared land (Howe and Smallwood, 1982;

Guevara et al., 1986; Young et al., 1996; Galindo-González et al., 2000; Wright, 2002).

Frugivore species differ in their capacity to disperse plant species in fragmented parts of the landscape, depending first on whether they use fragmented habitats, second, on the suite of plant species they consume and third on their patterns of movement within and between different habitat types. Hence, the species composition of the frugivore assemblage in remnant and regrowth forest will influence patterns of seed dispersal and forest regeneration. Dispersal of certain plant species may be reduced in these areas if they are not dispersed by the frugivore species that occur in fragmented parts of the landscape (Hamann and Curio, 1999; Silva and Tabarelli, 2000; Corlett,

1998, 2002; Cordeiro and Howe, 2001, 2003). For example, relatively few frugivores have the capacity to disperse fruits with large seeds and those that do are often vulnerable to the effects of fragmentation (Wheelwright, 1985; Chapman and Chapman,

1995; Corlett, 1998, 2002; Silva and Tabarelli, 2000; Kitamura et al., 2002; McConkey and Drake, 2002; Walker, 2006).

143 Frugivorous birds and bats are the main seed dispersers in Australian subtropical rainforests (Green, 1995), where approximately 70% of plant species are fleshy-fruited

(Willson et al., 1989; Butler 2003). Surveys in remnants and regrowth patches in a fragmented subtropical rainforest landscape in Australia have found that frugivorous bird species showed one of three patterns of abundance relative to extensive forest: (i) lower numbers in remnant and/or regrowth rainforest patches compared with extensive forest (‘decreaser’ pattern); (ii) higher numbers in remnant and/or regrowth rainforest patches compared with extensive forest (‘increaser’ pattern); or (iii), no substantial difference in numbers between the three site types (‘tolerant’ pattern: Chapter Two).

The impact of the decreased abundance of certain bird species on seed dispersal in remnant and regrowth rainforest patches depends on whether remaining frugivore species perform similar seed dispersal roles in these habitats.

In Australia, three species of frugivorous bat species occur regularly in subtropical rainforests: the grey-headed and black flying-foxes (Pteropus poliocephalus and P. alecto) and eastern tube-nosed fruit-bat (Nyctimene robinsoni). Two additional species, the little red flying-fox (P. scapulatus) and Queensland blossom bat

(Syconycteris australis), occasionally feed on fruit but are not common in rainforest

(Ratcliffe, 1932; Law and Spencer, 1995). The consequences of the extensive loss and fragmentation of subtropical Australian rainforest for the distribution of frugivorous bats are not well understood. It is widely held that populations of P. poliocephalus have suffered dramatic declines since European settlement (Eby et al., 1999; Eby and

Lunney, 2002; Dickman and Fleming, 2002). Elsewhere throughout the Old World tropics, pteropid populations have undergone large declines (Cox et al., 1991; Corlett,

1998; McConkey and Drake, 2002), in some cases associated with restricted distribution in fragmented habitats (Mildenstein et al., 2005). In Australia, both P. poliocephalus

(Eby 1991a, 1998; McDonald-Madden et al., 2005) and P. alecto (Markus and Hall,

144 2004) are known to use forest resources in fragmented parts of the landscape, but their distribution has not been systematically compared between fragmented and intact forest.

The distribution and habits of N. robinsoni are poorly known, especially in the southern parts of its range.

Chapter Five assessed the potential for tolerant and increaser bird species to substitute for decreaser bird species as seed dispersers, based on a comparison of their patterns of plant species consumption. The results of analyses showed that certain plant taxa may be solely or predominantly dispersed by decreaser bird species, and therefore may be susceptible to substantially reduced dispersal in fragmented rainforest.

However, Chapter Five also showed that a large proportion of the plant species dispersed by decreaser bird species are potentially dispersed by certain tolerant bird species in fragmented forest. The ability of frugivorous bats to substitute for decreaser bird species as seed dispersers in fragmented forest landscapes of subtropical Australia has not been evaluated. If frugivorous bats do not consume a similar suite of plant species to decreaser frugivores, or if their distribution is restricted to extensive forest, they would have low potential to substitute for decreaser bird species as seed dispersers.

Studies conducted in north Queensland, the Philippines, Madagascar and Peru have reported little overlap between the diets of frugivorous bird and bat species (Gorchov et al., 1995; Hamann and Curio, 1999; Bollen et al., 2004; Richards, 1990). In contrast,

Eby (1998) found that the plant species consumed by P. poliocephalus in subtropical

Australian rainforests comprised a subset of those that were collectively consumed by the sympatric assemblage of frugivorous birds. However, more detailed dietary comparisons between individual species are needed to show whether frugivorous bats have the potential to substitute as seed dispersers for the frugivorous bird species that decline in fragmented subtropical rainforest landscapes.

145 This chapter tests the overall hypothesis that the distribution of frugivorous bats is restricted in a fragmented rainforest landscape of subtropical Australia. The presence and abundance of foraging flying-foxes and N. robinsoni are compared between extensive forest, and patches of remnant and regrowth patches that have been isolated by clearing (16 sites of each type). The effect of site altitude and presence of a watercourse are also evaluated. Information about the native plant species known to be consumed by frugivorous bats is compiled and used, in conjunction with information about their use of remnant and regrowth habitats, to assess their potential to disperse seeds in fragmented parts of the study region. In particular, the potential for frugivorous bat species to substitute for decreaser bird species as seed dispersers in fragmented habitats is assessed, especially in relation to the dispersal of plant species with large seeds and other plant taxa that have been identified as vulnerable to reduced dispersal in fragmented forest (Chapter Five).

6.2 Methods

6.2.1 The study region and site network

Bat surveys were conducted in the same network of 48 sites as the bird surveys.

A description of the site network and study region is provided in Chapter Two (Section

2.2.1). The distribution and clearing history of subtropical rainforest in Australia were described in Chapter One (Section 1.4.1).

6.2.2 Surveys of frugivorous bat distribution

The occurrence of frugivorous bats was assessed using a single, hour-long nocturnal search at each site during summer (January-February) 2003. Surveys were timed to occur during the period of maximum fruit abundance in rainforest in subtropical Australia (Innis, 1989; Church, 1997). Searches were conducted between

146 one hour after sunset (usually around 8 pm) and 2 am. Two observers searched for bats

(the author and J. Kanowski), each using a spotlight (30 W) and walking slowly for approximately 400-500 m, usually along a watercourse or path. Bats were located through movement, calls and foraging sounds, using binoculars for identification when necessary. Most flying-fox records involved sighting of individuals, although records were occasionally made from calls. Abundance estimates may consequently have been biased towards remnant and especially regrowth sites where visibility was greater.

However, presence-absence information would be a reliable measure of site use, since flying-foxes were usually heard when they alighted in vegetation or when they dropped or dislodged fruit. Where possible, flying-foxes were identified to species; otherwise the record was made as ‘Pteropus sp.’ Only large flying-foxes were involved in the instances of undetermined species, so it was assumed that these were either P. poliocephalus or P. alecto, since P. scapulatus are noticeably smaller (Hall and

Richards, 2000; personal observation). Some N. robinsoni individuals were both seen and heard but this species was usually detected by the distinctive, squeaky call which it emits while flying. Except in the unlikely circumstance that this species has different calling behaviour between habitat types, presence-absence information for this species would not be biased toward any particular habitat type.

6.2.3 Frugivorous bird data

The patterns of abundance of frugivorous birds in the same site network were determined from 40 minute searches of a 1 ha plot at each site, conducted twice in summer and twice in winter, 2001 by the author (described in Section 2.2.3). Of the 26 bird species that had been recorded frequently enough to assign a fragmentation response pattern, 14 are known to either destroy seeds (i.e., ‘seed crushers’; e.g., white- headed pigeon Columba leucomela, Australian king-parrot Alisterus scapulatus) or to

147 consume fleshy fruits only infrequently (e.g., black-faced cuckoo-shrike Coracina novaehollandiae, Torresian crow Corvus orru (i.e., ‘minor frugivores’)) (Chapter

Three). These species potentially make relatively little contribution to seed dispersal, either because they do not disperse viable seed or because they consume a small number of plant taxa (Chapter Four). Therefore, only the remaining 12 bird species that usually disperse intact seeds and have fruit-dominated or mixed diets are considered in the present chapter.

6.2.4 Information on the consumption of native plant species by frugivorous bat and

bird species

Information about the fleshy-fruited plant species consumed by the 12 frugivorous bird species, P. poliocephalus, P. alecto and N. robinsoni were obtained from 130 published sources (Appendix 1) and several unpublished data sets. The data set containing records of plant species consumption by frugivorous bird species was described in Chapter Four (Section 4.2.1). The majority of the foraging records for flying-foxes came from data published by Eby (1995, 1998). Most of the published foraging records for both birds and bats were obtained from direct field observation although a relatively small proportion of records were obtained from gut contents, scats, or regurgitated seeds. There was large variation among frugivore species in the amount of foraging information available. Except in the case of targeted surveys of particular frugivore or plant species, records were typically accompanied by minimal information about the observed interaction, such as details of fruit handling. Records were rejected if it appeared that the frugivore had not been observed actually consuming the fruit (i.e., it was only observed in the fruiting plant), or if it was judged from accompanying information that the interaction was likely to be an instance of fruit theft (consumption of the flesh without ingesting the seed). Diaspore size (the average width of the

148 functional dispersal unit; see Section 4.2.1) was used to evaluate the likelihood that flying-foxes would transport the seed away from source plants. Although the size of fruits that flying-foxes are able to consume is not constrained by their gape width

(Ratcliffe, 1932), their ability to transport seeds is size-limited, since only small seeds can pass through their gut (c.a. 4 mm for P. poliocephalus (Eby, 1991b)), or be carried in cheek pouches (c.a. 9 mm (Eby, 1995)). Flying-foxes may also carry larger fruits in their jaws, but are only likely to transport these over short distances (in the order of metres) (Ratcliffe, 1932). For the purpose of this work, it was considered that flying- foxes were potentially able to carry in their cheek pouches diaspores with a maximum median width of 18 mm, based on the possibility of large intraspecific variation in fruit size (Edwards, 2005).

Because of the wide geographical range of many of the frugivore species that occur in subtropical Australia, frugivory records may have been collected from an area extending from temperate southern Australia to tropical Papua New Guinea, but the analyses presented here only considered records of the consumption of plant species that were native to the study region (based on published accounts of plant distribution and expert advice). Plant taxa included in analyses are listed in Appendix 2. For a given frugivore species, the data potentially included foraging records from multiple years, seasons and geographic locations. The data were compiled into a binary matrix showing whether or not each fleshy-fruited plant species had been recorded in the diet of each of the frugivore species.

6.2.5 Data handling

Bat distribution in extensive, remnant and regrowth sites

The presence of frugivorous bats (and the abundance of flying-foxes) was (i) compared across habitat types; and (ii) analysed in relation to environmental attributes

149 of sites (altitude and the presence of a watercourse), which may be related to the foraging distribution of frugivorous bats Australia (Palmer and Woinarski 1999; Palmer et al. 2000). I did not attempt to quantify abundance of N. robinsoni as most records were from calls. For flying-foxes, the abundance measure was the number of individuals recorded during the hour survey. Log-transformation normalised abundance data for ‘all flying-foxes’ (i.e., positively identified P. poliocephalus and P. alecto plus unidentified large flying-foxes), but not for P. poliocephalus alone. Non-parametric statistical tests were used on raw abundance data for P. poliocephalus.

Patterns of plant species consumption

The number of native plant species, genera and families, the proportion of plant species with a median diaspore size ≥10 mm, and the average diaspore size of plant species consumed were calculated for each frugivore species under consideration in the present work. A dendrogram showing multivariate similarities among the diets of frugivores (flying-foxes and the 12 frugivorous bird species) was generated using the

UPGMA algorithm and Bray-Curtis similarity metric (Manly 1994) in the statistical program PRIMER, based on the presence or absence of native plant species in the frugivores’ diets. Plant species known to be eaten by less than three frugivores were excluded from this analysis.

6.3 Results

6.3.1 Distribution and abundance of frugivorous bats

Frugivorous bats were recorded in most of the sites surveyed (Table 6.1). P. scapulatus and S. australis were recorded only once each (both in the same coastal remnant site); data for these two species are not considered further. While it was often not possible to distinguish between P. poliocephalus and P. alecto during surveys, many 150 more individuals of P. poliocephalus were positively identified than P. alecto. During the survey period, P. poliocephalus was recorded in significantly more sites in remnant and regrowth forest than extensive forest (Table 6.1). The abundance of P. poliocephalus also varied significantly between site types (Kruskal-Wallis H = 11.17, d.f .= 2, P = 0.004), being higher in regrowth and remnants than extensive forest (Figure

6.1). The occurrence of ‘all flying-foxes’ (P. poliocephalus, P. alecto and unidentified large flying-foxes) was not statistically different between site types, although there was a similar trend to that shown when data for definitely identified P. poliocephalus was analysed separately (Table 6.1). The abundance of ‘all flying-foxes’ was higher in both remnant and regrowth forest than in extensive forest (ANOVA F2, 47 =8.99, P=0.001)

(Figure 6.1). N. robinsoni was detected in more extensive forest and remnant sites than in regrowth (Table 6.1).

Table 6.1 Distribution of frugivorous bats in surveys of extensive, remnant and regrowth rainforest in the Sunshine Coast, Queensland, Australia. The table shows the number of sites in which each species was recorded. Species Total Extensive Remnant Regrowth p (n = 48) (n = 16) (n = 16) (n = 16) P. poliocephalus 25 4 9 12 0.044 P. alecto 3 1 1 1 - Unidentified large flying-foxes 19 7 8 4 - All large flying-foxes1 39 10 15 14 0.064 N. robinsoni2 13 6 6 1 0.025 Any fruit-eating bat3 41 12 15 14 0.314 1 includes unidentified individuals considered to be P. poliocephalus or P. alecto. 2 extensive and remnant sites pooled for statistical test. 3 includes all flying-foxes and N. robinsoni. 4 χ2 test of independence for distribution in different forest types, d.f.=2 5 Fisher’s exact test, regrowth versus extensive and remnant sites

151

i) P. poliocephalus

b 8 b 6

4 a 2 Abundance

0

ii) all flying-foxes

b 10

8 b

6

4 a

Abundance 2

0 Ext Rem Reg

Site type

Figure 6.1 The abundance (mean ± SE) of flying-foxes recorded during a 60 minute search of extensive, remnant and regrowth forest sites within the Sunshine Coast, Queensland, Australia. i) P. poliocephalus; ii) all flying-foxes (grey-headed, black and unidentified large flying-foxes). Ext = extensive forest, Rem = remnant forest and Reg = regrowth (n = 16 for each site type). Means with different letters were significantly different (P<0.05) using: i) Fisher’s exact tests for P. poliocephalus data, and ii) LSD comparisons for all flying-foxes.

152 6.3.2 Association of bat distribution with environmental attributes

N. robinsoni was recorded at nine of the 34 sites that contained a watercourse and four of the 14 sites without, showing no clear association with watercourses (χ2 =

0.02). In contrast, the presence of flying-foxes was strongly associated with watercourses; P. poliocephalus was present at 24 of the 34 sites with watercourses and at only one of the 14 sites without (Fisher’s exact test, p<0.001), while ‘all large flying- foxes’ (P. poliocephalus, P. alecto and unidentified large flying-foxes) were present in

32 of the sites with watercourses and in seven of the sites without (Fisher’s exact test, p

= 0.001). The mean abundance of P. poliocephalus was also higher in sites with a watercourse (average of 4.1 bats per survey) than in sites without (0.1 bats per survey;

Wilcoxon rank test z = -3.83, p = 0.0001). There was a similar result for ‘all large flying-foxes’ (5.3 bats per survey in sites with watercourse, 1.5 bats per survey in sites without; t-test = 3.38, p = 0.001, d.f. = 46).

For N. robinsoni, there was no clear association between altitude and occurrence within extensive and remnant forest (logistic regression R = -0.13, N = 32, p = 0.09).

This species was recorded at 7 of 13 sites below 200 m, 5 of 13 sites located between

200 and 500 m, and none of six sites above 500 m a.s.l. Similarly, no altitudinal trend was detected in the distribution of flying foxes within the habitats in which they were most abundant (remnant and regrowth sites), either in terms of abundance (Pearson’s correlation coefficient R = -0.15, N = 32, p = 0.21), or presence/ absence (logistic regression R<0.0001, N = 32, p = 0.28). Flying-foxes were recorded at 16 of the 17 remnant and regrowth sites below 200 m and 13 of 15 sites above 200 m.

153 6.3.3 Comparison between frugivorous bat and bird species in their patterns of plant

species consumption

There were a total of 811 foraging records for birds and bats, from 221 native species of trees, shrubs, vines and herbs. N. robinsoni had only been positively recorded feeding on the fruits of four native plant species in subtropical Australia (Table 6.2),

Elaeocarpus grandis (Elaeocarpaceae), Ficus watkinsiana (Moraceae), Endiandra discolor (Lauraceae) and Melodorum leichardtii (Annonaceae). Several additional plant species, especially figs, are likely to be consumed by this bat, but as observations had been recorded only to genus level (e.g., ‘Ficus sp.’), these were not quantifiable in the present data set. Flying-foxes were known to consume 48 species, 31 genera and 29 families of native plants; values which are moderate in comparison with the ranges shown by bird species (6 – 106 species, 2 – 89 genera, 3 – 52 families) (Table 6.2).

Flying-foxes were known to consume ten of the 20 plant species in the dataset from the

Myrtaceae, and ten of the 13 species from the Moraceae. For the remaining 27 plant families known to be consumed by flying-foxes, only one or two plant species had been recorded.

154 Table 6.2 Patterns of native plant consumption for frugivorous bats and the most important frugivorous bird species in the study region (see Section 6.2.3). ‘Av. dias. size’ is the average size (mm) of diaspores consumed. ‘Prop cons f-fox’ is the proportion of the plants species consumed by each bird species that were also known to be consumed by flying-foxes. Frugivore species Abund. Number of plant taxa Av. Prop pattern1 dias. cons size2 f-fox Species Genera Families ≥10 mm4 Bats Nyctimene robinsoni 4 4 4 3 12.1 flying-foxes Pteropus spp. 48 31 29 22 9.5 (44) (30) (28) (18) (7.8) Birds1 wompoo fruit-dove Ptilinopus magnificus Dec 81 56 37 47 10.1 0.36 superb fruit-dove P. superbus Dec 26 31 22 13 9.1 0.27 rose-crowned fruit-dove P. regina Dec 74 54 32 33 8.8 0.31 green catbird Ailuroedus crassirostris Dec 104 77 40 44 9.1 0.34 topknot pigeon Lopholaimus antarcticus Tol 73 46 28 42 10.3 0.36 common koel Eudynamys scolopacea Tol 25 26 19 6 7.0 0.52 channel-billed cuckoo Scythrops novaehollandiae Tol 6 2 3 0 2.0 0.83 Lewin’s honeyeater Meliphaga lewinii Tol 106 89 47 24 6.7 0.26 pied currawong Strepera graculina Tol 50 39 25 18 8.2 0.50 satin bowerbird Ptilonorhynchus violaceus Tol 106 89 52 48 8.8 0.31 figbird Sphecotheres viridis Inc 74 60 33 22 7.1 0.32 silvereye Zosterops lateralis Inc 37 39 30 0 4.5 0.35 1 Abund. pattern is the abundance pattern detected for each bird species; Dec decreaser, Tol tolerant, Inc increaser (Chapter Two). 2 The number of native plant species that had diaspores with a median size of ≥10 mm that were known to be consumed by each frugivore species. For flying-foxes, the second number (in brackets) shows results when fruits that are too large to be transported internally are excluded. It was considered that plant species with a median diaspore size up to 18 mm may be transported internally due to intraspecific variation in fruit dimension.

155

The proportion of native plant species with a median diaspore size ≥10 mm that was consumed by flying-foxes was similar to that consumed by decreaser bird species

(Figure 6.2), although many of these diaspores could only be transported externally by flying-foxes. The only other non-decreaser frugivore taxa that were known to consume a similar proportion of plant species with diaspores ≥10 mm were two tolerant bird species, topknot pigeon L. antarcticus and the satin bowerbird Ptilonorhynchus violaceus (Table 6.2). Similar patterns were evident when the data were analysed in terms of: (i) the average size of diaspores consumed by flying-foxes, relative to birds, and (ii) the number of diaspores with a median width ≥10 mm consumed by flying- foxes, relative to birds (Table 6.2).

0.6

0.4 ≥

0.2

Prop. diaspores 10 mm 10 Prop. diaspores 0 Dec Tol Inc F-foxes

Frugivore group

Figure 6.2 The proportion of native plant species with a median diaspore width ≥ 10 mm that were known to be consumed by decreaser (Dec, n=4), tolerant (Tol, n=6) and increaser (Inc, n=2) bird species, and by flying-foxes (F-foxes). The open circle shows the proportion of plant species with a median diaspore ≥10 mm that could be transported internally by flying-foxes, allowing for up to 50% intraspecific variation in diaspore width.

156

superb fruit-dove fruit-dove topknottopknot pigeon pigeon wompoo fruit-dove fruit-dove rose-crownedrose-crowned fruit-dove fruit-dove satin bowerbird bowerbird green catbird catbird Lewin's honeyeater honeyeater figbirdfigbird Pteropusflying-foxes pied currawong currawong channel-billed cukoo cuckoo common koel koel silversilvereye eye 100100 8080 6060 4040 2020 00

Dissimilarity Dissimilarity

Figure 6.3 Classification of frugivore species based on Bray-Curtis dissimilarity metric using patterns of consumption of native plant species.

Frugivorous bats did not consume a different set of plant species to frugivorous birds overall; for all of the native plant species known to be consumed by either flying- foxes or N. robinsoni, at least one of the 12 bird species considered here was also known to be a consumer. In the multivariate analysis, flying-foxes were not strongly separated from frugivorous birds based on dietary composition (Figure 6.3). The plant species that comprised flying-fox diets represented around one-third of the plant species consumed by most of the frugivorous bird species considered here, including decreaser bird species (Table 6.3).

157

6.3.4 Potential for frugivorous bat species to substitute for decreaser bird species as

dispersers in fragmented forest

The eastern tube-nosed fruit-bat consumed a very low percentage of the plant species that were consumed by decreaser bird species. The percentage of plant species consumed by each decreaser bird species that was known to be consumed by flying- foxes is shown in Table 6.2; flying-foxes consumed around one-third of the plant species known from the diets of each decreaser bird species. Neither flying-foxes nor the eastern tube-nosed bat were known to consume any of the plant species that were only known from the diets of decreaser species among frugivorous birds (Chapter Five,

Table 5.5). In relation to the plant families that were most frequent in the diets of decreaser species (and hence predicted to be vulnerable to reduced dispersal in fragmented rainforest), flying-foxes consumed half of the plant species from Myrtaceae

(10 out of 20). The average number of species from the Myrtaceae consumed by decreaser bird species was 6.5 (Table 5.2). There were no records of flying-foxes consuming plants from the Verbenaceae, and they were only known to consume one species from each Meliaceae (Melia azedarach; out of a total of seven species) and

Lauraceae (Cryptocarya obovata; out of 21 species). Flying-foxes had been recorded consuming one-third (two out of six) of the plant species from the Vitaceae in the data set.

6.4 Discussion

6.4.1 The distribution of flying-foxes in fragmented rainforest in the Sunshine Coast

Pteropus poliocephalus and P. alecto are similar in many respects, including size and reproductive characteristics, communal roosting behaviour (Ratcliffe 1932;

Hall and Richards 2000) and the consumption of both fruit and nectar (Richards and

158 Hall 1998). Because it was not always possible to identify these flying-foxes to species during surveys for the present work, the two species are henceforth considered collectively, although it is acknowledged that all comments may not apply equally to both species. P. scapulatus is not included in subsequent uses of the term ‘flying-foxes’ in the context of the present work.

During surveys for this work, flying-foxes were recorded foraging in most of the

48 sites across all three habitat types. They were not restricted to extensive forest, and used remnants and regrowth, despite their relatively small size and isolation and any floristic and structural differences between the site types. The ability of flying-foxes to use fragmented habitats may be due, in part, to their mobility over large geographic areas. P. poliocephalus has been recorded travelling tens of kilometres from daytime roosts to forage in multiple feeding areas (Eby, 1991b; Spencer et al,. 1991). This is in contrast to the Philippines were large flying-foxes have been reported to have comparatively restricted foraging ranges (0.4 to 12 km) and apparently avoid disturbed habitats in agricultural areas (Mildenstein et al., 2005). The capacity of Australian flying-foxes to routinely traverse large distances would readily enable movement between most forest patches in the Sunshine Coast, including those that have been isolated by clearing. Australian flying-foxes also forage in a range of forest types, including rainforest, eucalypt forests, paperbark and mangrove forests (Ratcliffe, 1932;

Eby, 1995). Their diets are fairly broad and comprise nectar, pollen and fruit from a range of flowering and fruiting plant species, genera and families (Parry-Jones and

Augee, 1991; Eby, 1998; Southerton et al., 2004). Similarly, generalist patterns of forest and food resource use by Neotropical phyllostomid bats in Brazil are associated with higher abundance in fragmented and modified habitats, compared with specialist bat taxa (Marinho-Filho and Sazima, 1998).

159 During the period of the surveys conducted for the present study, flying-foxes were recorded in higher frequency and abundance in remnant and especially regrowth rainforest sites compared with extensively forested areas. Since the distribution of flying-foxes is known to correspond with localised food availability (Eby, 1991a; Parry-

Jones and Augee, 1992; Palmer et al., 2000), the relatively high numbers observed in remnants and regrowth probably reflected fruit availability in these sites at this time.

During surveys, most observations of foraging flying-foxes in all three site types were at native sandpaper figs, especially Ficus coronata. These figs appeared to be especially common in regrowth sites, and are associated with early stages of rainforest regrowth in subtropical Australia (Kooyman, 1996).

The capacity to infer general patterns of flying-fox distribution from the survey conducted for this work is limited. This is due to the combination of a temporally restricted survey effort, and the potential for the geographic distribution of nomadic flying-foxes in eastern Australia to vary considerably with the availability of ephemeral food resources (Nelson, 1965; Parry-Jones and Augee, 1992; Eby, 1995). Many daytime roost sites are located in small remnants and regrowth vegetation in extensively-cleared parts of the region (Roberts, 2005) and flying-foxes forage and roost in vegetation in urban landscapes elsewhere in Australia (Parry-Jones and Augee, 1991; Markus, 2004;

McDonald-Madden et al., 2005)

However, while flying-foxes may make use of fragmented and disturbed rainforest habitats in the Sunshine Coast, they apparently have not benefited at a population level from such changes to the landscape. As is the case for Pteropid populations throughout the Old World tropics (Fujita and Tuttle, 1991), flying-fox numbers are declining in subtropical Australia (Lunney and Moon, 1997; Eby et al.,

1999; Eby and Lunney, 2002; Dickman and Fleming, 2002). Declines were reported by the early part of the 20th century as a result of habitat loss and persecution (Ratcliffe,

160 1932). Although they use rainforest, including disturbed habitats, flying-foxes

(particularly P. poliocephalus) appear to depend on nectar resources for at least part of the year (Ratcliffe, 1932; Eby et al., 1999; Southerton et al., 2004). Hence, clearing of large tracts of nectar-producing open forests in south-east Queensland (Catterall et al.,

1997) would have removed resources that may be critical for maintaining flying-fox populations (Eby et al., 1999). Consequently, even a scenario of increased extent of rainforest regrowth would be unlikely to compensate for the loss of seasonally- important nectar resources in terms of maintaining flying-fox populations.

Flying-foxes were more abundant in sites associated with a watercourse than in

‘dry’ sites. This may reflect their use of Ficus coronata as a major food resource at the time of surveys, since these plants tend to occur most abundantly close to water (Floyd,

1989). However, watercourses are also strongly associated with the location of flying- fox colonial day roosts (‘camps’) in south-east Queensland (Roberts, 2005), and may be used by flying-foxes to navigate through the landscape (Palmer and Woinarski, 1999).

Hence, flying-fox foraging activity in the Australian subtropics may generally be concentrated in vegetation along drainage lines, as has been described for P. alecto in the monsoonal forests of northern Australia (Palmer et al., 2000), for Pteropus spp. in the Philippines (Mildenstein et al., 2005) and for frugivorous bats in Mexico (Galindo-

González and Sosa, 2003). Although flying-fox camps in south-east Queensland are mostly restricted to low altitudes (below 120 m a.s.l., Roberts, 2005), foraging is not confined to lowland rainforest in the Sunshine Coast.

6.4.2 The distribution of the eastern tube-nosed fruit-bat in fragmented rainforest in

the Sunshine Coast

Nyctimene robinsoni was recorded in 13 of the 48 subtropical rainforest sites in the study region. Previous reports have suggested that the geographical range of N.

161 robinsoni in Australia is mostly tropical, based on infrequent records of this species from subtropical rainforests (Hall and Richards, 1979; Milledge, 1987; Hall et al.,

1995). However, the surveys conducted for the present study indicate that it may be more common in subtropical rainforest than previously thought, at least in the study region.

N. robinsoni was recorded more frequently in extensive and remnant rainforest than in regrowth sites. In contrast with flying-foxes and the frugivorous bird species that showed tolerant or increaser abundance patterns (Chapter Two), N. robinsoni depends on rainforest, mostly consumes fruit, and makes only limited foraging movements

(Spencer and Fleming, 1989; Hall and Richards, 2000). These attributes were also typical of the frugivorous bird species that were less common in remnants and regrowth than in extensive forest in the study region. These characteristics were also associated with the bat species that had restricted distribution in fragmented forests in the neo- tropics (Cosson et al., 1999). The physical separation of remnants did not prevent their use by N. robinsoni. The low frequency of this species in regrowth patches may have been a result of insufficient availability of food plants and the small size of these sites

(see Chapter Two, Section 2.2.2 for description of characteristic floristics of regrowth in relation to remnants and extensive forest). Given the habitat specialisation of N. robinsoni, coupled with its apparent preference for mature rainforest rather than regrowth, rainforest clearing must have substantially reduced the extent of suitable habitat and is likely to have led to reduced populations of this species.

6.4.3 The potential for seed dispersal by frugivorous bats in remnants and regrowth:

comparison with frugivorous birds

Both N. robinsoni and flying-foxes potentially disperse seeds in fragmented subtropical rainforest. N. robinsoni probably only disperses a small number of plant

162 species, and may disperse few seeds into or within patches of regrowth. In contrast, flying-foxes are known to consume at least one-third of the native plant species consumed by decreaser bird species. Since opportunistic observation of the foraging patterns of night-feeding flying-foxes would be lower than for diurnal birds, the actual proportion may be higher than reported here. Hence, flying-foxes have the potential to at least partially compensate for decreaser bird species with respect to the dispersal of rainforest plant species. For example, only a few frugivorous bird species, predominantly decreasers, consume high numbers of native plant species from the

Myrtaceae in subtropical Australia (Chapter Five). However, flying-foxes are also known to consume many plant species from the Myrtaceae (Eby, 1995; this study) and may help maintain dispersal of these plants in fragmented parts of the landscape.

In addition to consuming the fruits of a particular plant, the potential of a frugivore to disperse seeds between fragmented habitats is influenced by its foraging and ranging behaviour, combined with its gut passage rate (Schupp, 1993; Wenny and

Levey, 1998; Loiselle and Blake, 2002). Flying-foxes may consume fruits within the source plant and drop the seeds beneath the crown of the parent. Alternatively, flying- foxes may transport seeds relatively short distances away from parent plants by flying to a nearby tree to consume harvested fruits, behaviour which may be particularly common when other foraging flying-foxes are present (Richards, 1990). In addition, P. poliocephalus may move continually between successive feeding trees, sometimes over several kilometres, including across cleared land (Eby, 1991b). Hence, flying-foxes potentially disperse seeds both within and between extensive forest, remnant and regrowth patches. Many frugivorous bird species also routinely travel across cleared land between forest areas in subtropical Australia, and may play a similar role in seed dispersal in fragmented parts of the landscape. However, birds tend to eliminate seeds while perched (McDonell and Stiles, 1983). In contrast, frugivorous bats may also

163 defecate in flight (Charles-Dominique, 1986), and hence potentially disperse seeds to cleared areas as well as areas with standing vegetation. Work in West Africa (Thomas,

1982) and Mexico (Medellín and Gaona, 1999; Galindo-González and Sosa, 2003) has shown that frugivorous bats are responsible for the majority of seed input to cleared land, while birds make little contribution to seed input in these areas, although the reverse has been found in the Philippines (Ingle, 2003).

The frugivorous bird species that decline in remnants and regrowth in the

Sunshine Coast consume many large-fruited plant species. Nevertheless, flying-foxes, along with two tolerant bird species topknot and satin bowerbird, may disperse large- seeded plant species in fragmented parts of subtropical Australia. However, while flying-foxes can carry fruits as large as mangoes in their jaws or claws (Ratcliffe, 1932), they spit out or drop most large seeds close to the parent tree (van der Pijl, 1982; Eby,

1995; McConkey and Drake, 2002; Meehan et al., 2005). Longer distance dispersal may be restricted to small seeds that can be transported internally in the gut, or possibly cheek pouches. Indeed, Pteropid fruit bats may potentially disperse some very small seeds over hundreds of kilometres (Shilton et al., 1999). Hence, although flying-foxes potentially disperse large seeds a short distance away from the source tree, they may contribute relatively little to the dispersal of large seeds between distant forest patches or regenerating areas.

Although flying-foxes do not consume different plant species to frugivorous birds in subtropical Australia (c.f., Fleming et al., 1987; Richards, 1990; Gorchov et al.,

1995; Hamann and Curio, 1999; Bollen et al., 2004), they potentially have a distinctive role as dispersers of rainforest plants in deforested parts of the landscape. This is because of their use of fragmented rainforest, mobility over long distances, ability to defecate seeds in treeless areas, and consumption of a large number of plant species from the Myrtaceae.

164 Chapter 7

General discussion: Consequences of forest fragmentation for

frugivores and implcations for seed dispersal

7.1 Summary of the findings of this thesis

This thesis has evaluated consequences of forest fragmentation for fauna and the potential for impacts on higher order interactions. Specifically, this work has assessed changes in the abundance and distribution of frugivorous vertebrate species in a fragmented rainforest landscape and evaluated the functional roles of frugivore species to make predictions regarding potential changes in the process of seed dispersal in fragmented rainforest.

The salient findings of each chapter of this thesis are summarised in Table 7.1.

Frugivorous bird and bat species showed varied responses to rainforest fragmentation

(Chapters 2 and 6). The abundance of a subset of frugivorous species was lower in fragmented compared with extensive rainforest. Frugivorous bird species that declined in fragmented rainforest habitats shared similar functional traits (Chapter 3). There was a strong association between functional attributes of frugivores and their actual patterns of plant species consumption (Chapter 4). A substantial proportion of native rainforest plant species may be dispersed solely or predominantly by the frugivore species that declined in fragmented forest landscapes (Chapters 5 and 6). It was predicted that dispersal of these plant species in fragmented habitats would depend on only a small subset of frugivore species, and consequently that regeneration of these plants would be reduced in fragmented forest.

165 Table 7.1 Summary of the scope and major findings of the component studies reported in the chapters of this thesis.

Focus Key findings Chapter 2 Assessment of Forty-two subtropical Australian bird species are at least partly frugivorous. Field surveys documented three general the abundance of frugivorous abundance patterns among the 26 most common frugivorous bird species: decreaser (abundance lower in remnant bird species in fragmented and/or regrowth than in extensive forest (five species), increaser (abundance higher in remnant and/or regrowth than and extensive forest. extensive forest (seven spp.), and tolerant (abundance similar across site types (14 spp.)). Response patterns were generally consistent between seasons. Furthermore, the abundance of frugivorous bird species in extensive forest at different altitudes (low (<200 m a.s.l.), medium (200-500 m a.s.l.) and high (>500 m a.s.l.)) did not vary between seasons.

Chapter 3 Assessment of The bird traits proposed to strongly influence a species’ seed dispersal potential were gape width (small, medium, the functional roles of bird large), frugivory level (major, mixed-diet and minor frugivores) and seed treatment (seed-dispersers and seed- species and potential crushers). Tests showed that decreaser bird species tended to be seed-dispersing major frugivores with large or changes in seed dispersal in medium gapes. It was predicted that this would result in reduced dispersal of plant species with large fruits and from fragmented forest. Lauraceae.

Chapter 4 Test of the The average size of fruits consumed by a bird species increased with gape width, except for minor frugivores. Minor association between patterns frugivores consumed small fruits, irrespective of their gape width. Major and mixed-diet frugivores consumed a higher of plant species consumption number of plant species than minor frugivores. The diets of minor frugivores and bird species with small gapes mostly and the traits identified in comprised plant species from the Moraceae and Euphorbiaceae. Among the species with gapes wider than 10 mm, Chapter 3. major frugivores consumed the highest number of native plant species from the family Lauraceae, whereas mixed- diet frugivores consumed the highest number of species from the Celastraceae, Mimosaceae, Sapindaceae and Smilacaceae.

Chapter 5 Evaluation of the There was considerable overlap in the diet of frugivore species that showed different responses to fragmentation. potential for changed seed However, there were no known tolerant or increaser dispersers for 12% of the native plant species in the data set. dispersal in fragmented This included several species from the Rubiaceae. It was predicted that dispersal of these plant species would be rainforest landscapes, using severely reduced in fragmented rainforest. In addition, decreasers consumed the highest number of plant species information on patterns of with fruits ≥10 mm in diameter, and of plant species from the Lauraceae, Meliaceae, Myrtaceae, Verbenaceae and plant species consumption Vitaceae. Only a small number of other bird species consumed similar numbers of these plant taxa. It was predicted by frugivore species, that dispersal of these plants would be substantially reduced in fragmented rainforest. Increaser frugivores had limited together with data on their potential to substitute for decreasers as seed dispersers in fragmented forest. response to fragmentation (Chapter 2)

166 Focus Key findings

Chapter 6 Assessment of Three frugivorous bat species occur in rainforests of subtropical Australia. Flying-foxes used all three habitats studied the distribution of frugivorous (intact, remnant and regrowth rainforest), whereas the eastern tube-nosed bat was largely restricted to extensive and bat species in fragmented remnant forest. The distribution of flying-foxes, but not the eastern tube-nosed bat, was positively associated with and extensive forest. watercourses. Frugivorous bats used rainforest at all altitudes. The plant species consumed by frugivorous bats were Evaluation of their potential also consumed by frugivorous bird species. Flying-foxes potentially substitute for decreasers as dispersers of plant to disperse the plant species species with large fruits and from the Myrtaceae in fragmented parts of the landscape, although they may not consumed by ‘decreaser’ bird disperse fruits >9 mm very far from parent trees. species.

167 7.2 The sensitivity of frugivorous vertebrate species to rainforest fragmentation in subtropical Australia

There have been no previous assessments of effects of forest fragmentation across a regional frugivore assemblage in Australia. However, some of the frugivore species that were evaluated in the present work had been included in other studies of faunal change in fragmented landscapes. These studies provide an opportunity to asses whether the species’ response patterns detected in the present study may be consistent among regions or at different times. The locations of the studies that have been conducted in Australia are shown in Figure 7.1.

Decreaser species

In the present study, the abundance of five bird species, and the occurrence of the eastern tube-nosed bat, was lower in fragmented than in extensive rainforest. Three bird species, all seed-dispersing Columbidae from the genus Ptilinopus (wompoo, superb and rose-crowned fruit-doves), showed a clear reduction in abundance in rainforest fragments compared with extensive forest. They were also absent from most of the isolated regrowth patches surveyed for this study. Similarly, Howe et al. (1981) only once recorded the wompoo fruit-dove during surveys of small rainforest remnants in the Dorrigo region, while Warburton (1997) reported that both the wompoo and superb fruit-doves were relatively uncommon in small remnants in the Wet Tropics. In contrast, it has been reported that Ptilinopus species are widespread in fragmented rainforest landscapes in the Big Scrub region (Gosper and Holmes, 2002). It has been postulated that extensive patches of advanced regrowth dominated by the fleshy-fruited, introduced species camphor laurel Cinnamomum camphora may facilitate the use of fragmented rainforest landscapes in the Big Scrub region by these species (Date et al.,

1996; Gosper and Holmes, 2002; Neilan et al., 2006). The present work showed that, in 168 the Sunshine Coast region, Ptilinopus species made limited use of regrowth, including patches dominated by C. camphora. However, patches of C. camphora are far more extensive in the Big Scrub region than in other parts of Australia (Scanlon et al., 2000) and provide a high degree of forest cover in a formerly highly cleared landscape.

Crome et al., (1994) Warburton (1997)

Wet Tropics

Sunshine Coast Eby, 1991a Big Scrub Date et al. (1996) Dorrigo Gosper and Holmes (2002) Neilan et al. 2006) Howe et al. (1981)

McDonald-Madden et al. (2005)

Figure 7.1 Map of Australia showing the approximate location of other studies in fragmented forest that have included frugivore species. Rainforest is shown in red. (Source of base map: National Land and Water Resources Audit, 2001). The star shows the location of the present study.

Outside Australia, Ptilinopus species have been reported to be sensitive to forest clearing on Pacific Islands (Steadman and Freifeld, 1998; McConkey and Drake, 2002) and in parts of south-east Asia (Hamann and Curio, 1999; Kitamura et al., 2002), although hunting has also been implicated in these declines. Species in the Columbidae were detected in relatively low numbers in rainforest fragments during surveys in

Tanzania (genus Columba; Cordeiro and Howe, 2001), but at least some Columbid species persist in fragmented rainforest landscapes in the Indo-Malaysian region

(Corlett, 1998). 169 In the present study, three frugivore species, the brown cuckoo-dove

(Columbidae), green catbird (Ptilonorhynchidae) and eastern tube-nosed bat

(Pteropidae) had similar numbers in extensive forest and remnants, but much lower abundance or frequency of occurrence in regrowth. To my knowledge, there have been no other studies of the eastern tube-nosed fruit-bat in fragmented landscapes. Consistent with the results of the present study, the frequency of the brown-cuckoo-dove was similar between remnants and extensive forest in studies in the Big Scrub (Date et al.,

1996) and Wet Tropics (Warburton, 1997). Neilan et al. (2006) reported that the brown cuckoo-dove was common in regrowth patches adjacent to extensive forest in the Big

Scrub region, but that this species was uncommon in regrowth sites distanced from extensive forest.

The two other studies of the distribution of the endemic green catbird in fragmented forest landscapes have reported divergent patterns from those documented in the present study. The first (Howe et al., 1981) reported very low abundance in remnants in the Dorrigo region (i.e., greater sensitivity to fragmentation than reported here), while the second (Neilan et al., 2006) reported high abundance in regrowth patches in the Big Scrub region (i.e., greater tolerance of fragmentation than detected in the present study). In the case of the remnants studied by Howe et al. (1981), many sites may have been too small (0.1-2.5 ha) for this species to maintain a territory (average 2.1 ha; Innis and McEvoy, 1992). The greater resilience of the green catbird in the Big

Scrub region mirrors the trend already described for other decreaser bird species, and may at least partly be attributed to the more extensive occurrence of advanced regrowth in that region.

170 Tolerant species

Fifteen vertebrate frugivore taxa (14 bird species and Pteropid flying foxes) showed similar abundances across extensive, remnant and regrowth in the study region.

Among these tolerant species was the topknot pigeon, the only seed-dispersing

Columbid species that did not decrease in fragmented rainforest in the study region. The topknot pigeon is also widespread in fragmented landscapes in the Big Scrub region

(Date et al., 1996; Neilan et al., 2006), but its numbers were much lower in small remnants than in nearby extensive forest in the Dorrigo area (Howe et al., 1981). In the

Wet Tropics, this species occurred only infrequently in remnants smaller than 30 ha

(Warburton, 1997).

In agreement with the classification of the satin bowerbird as tolerant in the present study, this species occurred in most of the small remnants surveyed in the

Dorrigo region by Howe et al. (1981). However, in the Wet Tropics, Warburton (1997) reported strongly decreased abundance of the satin bowerbird in remnants smaller than around 660 ha. However, this species occupies highland rainforest in north Queensland

(Nix and Switzer 1991), and its low incidence in the remnants surveyed by Warburton may have been the result of the lower altitude of smaller remnants, rather than their decreased size. The satin bowerbird was uncommon in the regrowth sites surveyed in the Big Scrub region by Neilan et al. (2006), whereas it may have been expected to be more widespread if it were tolerant of fragmentation.

Consistent with the present study, the Lewin’s honeyeater and pied currawong were recorded in all (or almost all) of the remnant and regrowth sites surveyed respectively by Warburton (1997) and Neilan et al. (2006). The white-headed pigeon, emerald dove and Australian king-parrot occurred in most of the rainforest remnants surveyed in the Big Scrub by Date et al. (1996) and/ or in the Wet Tropics by

Warburton (1997). These species, as well as the Australian brush turkey were also

171 frequently recorded in regrowth patches in the Big Scrub region by Neilan et al. (1996).

There have been no other studies of the fragmentation sensitivity of the common koel, but a congeneric species, Eudynamys cyanocephala, also uses fragmented forest landscapes in south east Asia (Corlett and Ko, 1995).

Australian flying-foxes are known to range widely in fragmented forest landscapes (Eby, 1991a; McDonald-Madden et al., 2005). Outside Australia, the distribution of flying-foxes appears to be limited in fragmented forest landscapes

(Mildenstein et al., 2005), although increased hunting and persecution of flying-foxes have typically accompanied forest loss and fragmentation in these areas (Cox et al.,

1991; Corlett, 1998; McConkey and Drake, 2002). Use of modified parts of the

Sunshine Coast landscape by flying-foxes may reflect reduced levels of persecution since 1995 (Eby and Lunney, 2002).

Increaser species

In the present study, the bar-shouldered dove only occurred in remnants and regrowth sites. Similarly, Date et al. (1996) detected this species in remnants but not in extensive forest in the Big Scrub. While the figbird was moderately common in the extensive and remnant forests surveyed for the present study, it was very abundant in the regrowth sites surveyed. Similarly, this species was common in regrowth in the Big

Scrub region (Neilan et al., 2006), and occurred in the fragments and planted windbreaks assessed in the Wet Tropics by Crome et al. (1994). In contrast, the incidence of the figbird was relatively low in the smallest fragments surveyed by

Warburton (1997) in the Wet Tropics. As in the present study, the Torresian crow and silvereye were common in regrowth in the Big Scrub (Neilan et al., 2006), and the silvereye was present in rainforest fragments in the Wet Tropics (Crome et al., 1994).

172 White-eyes (Zosteropidae) also appear to be tolerant of forest fragmentation in Asia

(Corlett, 1998).

In summary, frugivore species that were classified as increasers in the present study have generally also been reported to be widespread in fragmented rainforest landscapes in other parts of Australia. However, for decreasers and certain tolerant species, there was considerable variation among studies in the reported consequences of forest fragmentation. In particular, decreaser species may be more widespread in the Big

Scrub region than in the Sunshine Coast, possibly due to the extent of advanced regrowth in the former. Although classified as tolerant species in the present study, the topknot pigeon, satin bowerbird and flying-foxes showed sensitivity to forest fragmentation in certain studies in other regions.

7.3 Correlates of frugivore species’ sensitivity to rainforest fragmentation

This thesis has shown large interspecific variation in sensitivity to forest fragmentation within the broadly defined frugivore guild in subtropical Australia. There is also variation among frugivore species in their responses to forest fragmentation in other regions (e.g., Restrepo et al., 1997; Corlett, 1998; Luck and Daily, 2003).

Understanding species’ characteristics associated with different responses to fragmentation may improve our predictive capacity in relation to the biological consequences of forest clearing and fragmentation (Henle et al., 2004). Because most studies of the consequences of forest fragmentation for frugivores have not evaluated species-specific responses, there is limited understanding of the general profile of a fragmentation-sensitive frugivore species. Although the present study was not explicitly deigned to test correlates of susceptibility to fragmentation, the community-wide data 173 set provides an opportunity to assess whether certain factors were clearly correlated with variation in species’ responses to forest fragmentation.

As proposed by Henle et al. (2004) and reviewed in Chapter One of this thesis

(Section 1.5), a species’ sensitivity to forest fragmentation is likely to be influenced by combinations of demographic traits (particularly population size and variability, dispersal power and generation time), and ecological traits (specialised patterns of microhabitat and matrix use and biogeographical distribution). Specific demographic data were unavailable for most of the species in the present study, but information on their ecological traits was generally accessible. Thus, information on body size, biogeographical position, rarity and patterns of resource use of frugivore species (see

Appendix 3) were evaluated with respect to their sensitivity to rainforest fragmentation in subtropical Australia.

In summary, of the factors examined, specialised patterns of resource use were most closely associated with sensitivity to fragmentation (see analyses in Appendix 3).

Specifically, frugivorous bird and bat species that were sensitive to forest fragmentation were rainforest-dependent fruit-specialists. In contrast, species that used open eucalypt forest or a variety of forest types, or that consumed fruit as well as other food types were more likely to be tolerant or to show increased abundance in fragmented rainforest. This is consistent with results from other regions (Kattan et al., 1994;

Castelletta et al., 2000). For example, bird species that consumed one type of food went extinct more rapidly in Singapore than species that consumed multiple food types (e.g.,

Castelletta et al., 2000). Rainforest-dependent bird species were also more sensitive to rainforest fragmentation than species that used a variety of forest types in the Australian

Wet Tropics (Crome et al., 1994; Warburton, 1997). Contrary to findings in other regions (e.g., Restrepo et al., 1997; Renjifo, 1999), large body size was not associated with sensitivity to forest fragmentation.

174 An exception to the general profile of fragmentation-sensitive species presented above may be the topknot pigeon. The topknot pigeon is a rainforest and fruit-specialist that was classified as tolerant in the present study, although this species has shown fragmentation sensitivity in some other regions (Howe et al., 1981; Warburton, 1997).

The topknot pigeon differs from the rainforest- and fruit-specialist decreaser species in its frequent aggregation into large flocks (50 to hundreds of individuals), whereas the decreaser pigeon species mostly forage alone or in pairs (Frith, 1982; Westcott and

Dennis, 2006). Gregarious foraging behaviour may reduce risk of predation (Pulliam,

1973; Howe, 1979; Watson et al., 2007), and hence increase a species’ willingness to traverse the non-forest matrix. Thus, the profile of a fragmentation-sensitive frugivore species may include non-gregarious foraging behaviour, at least in subtropical

Australia.

7.4 Patterns of plant species consumption across the frugivore

assemblage: an alternative model

Frugivory level, gape width and seed treatment were associated with variation among frugivorous bird species in their patterns of plant species consumption (Chapter

4). In Figure 7.2, I propose a model of the relationships between bird attributes and major dimensions of dietary variation in subtropical Australian rainforest. To my knowledge, this has been the first assessment of patterns of plant species consumption based on information for individual frugivore species across an entire assemblage of frugivorous rainforest birds. This is also the first community-wide evaluation of associations among frugivore species’ attributes and their patterns of consumption of a large number of plant species.

175

GROUP A GROUP B Lauraceae, Mostly arillate fruits Burseraceae, from Sapindaceae, Ebenaceae, Celastraceae Vitaceae Mimosaceae, Elaeocarpaceae Mixed-diet frugivores Major frugivores* GROUP C (>10 mm gapes) (>10 mm gapes) Larger fruits (>10 mm),

Araliaceae, Arecaceae

GROUP D Small fruits (<10 mm) e.g., Euphorbiaceae, Solanaceae, Small-gaped and/or certain Sapindaceae seed-crushing major and mixed-diet

Increasing number of frugivore species of frugivore species Increasing number frugivores Most minor frugivores, a GROUP E small number of major Moraceae and mixed-diet frugivores

Figure 7.2 A model of variation in patterns of plant species consumption by frugivorous bird species in subtropical Australia. The outer frame represents all available plant species and the inner frames represent the division of these plant resources by frugivore species. Arrows show the groups of plant taxa consumed by each frugivore group. The vertical arrow shows that an increasing number of frugivore species consumes plant taxa in lower sections of the frame. *The major frugivore classification may include species that are seasonally highly frugivorous but have mixed diets during other times of the year.

The conceptual model presented in Figure 7.2 synthesises the variation among frugivore species that has been shown in the present study (see especially Tables 4.3,

4.4 and 4.5 and Figure 4.5). Plant taxa that contributed to the distinctions between bird groups classified in Chapter Four (Figure 4.5, Table 4.5) have been used to illustrate the structure of frugivore-plant interactions in subtropical Australia. For example, Group A comprises the Lauraceae, Burseraceae, Ebenaceae and Vitaceae, which were mostly consumed by highly frugivorous bird species with gapes wider than 10 mm. Group B includes plant taxa that distinguished mixed-diet frugivores (gapes >10 mm) from other 176 bird species; these are mostly arillate fruited species from Sapindaceae, Mimosaceae and Celastraceae. Plant taxa represented as Group C (Araliaceae, Arecaceae) in Figure

7.2 were consumed by both of these groups of bird species, but generally not by minor frugivores, bird species with small gapes or by seed-crushing species. Although most frugivorous bird species consumed Moraceae (Chapter Four, Table 4.3), bird species in

Group D may also consume small-fruited pioneer taxa (e.g., Euphorbiaceae), whereas the minor and other frugivore species in Group E may have had fig-dominated diets

(Figure 7.2).

Although patterns of plant species consumption across this frugivore community did not conform neatly to any of the existing models of plant species consumption

(Section 1.2, Figure 1.2), elements of all of these models may describe components of the model presented in Figure 7.2. For example, certain small fruited plant species

(Group D) and figs (Group E) were consumed by most frugivorous birds, despite variation among species in their gape width, frugivory level or seed treatment (Figure

7.2). A neutral model of plant-frugivore interactions, as proposed by Burns (2006), may describe patterns of consumption of these plant species. However, the dichotomous consumption of plants in Groups A and B by major and mixed-diet frugivores respectively may more closely resemble the pattern proposed by McKey (1975).

Contrary to initial predictions (McKey, 1975), the differences are not based on strict lipid or carbohydrate preferences. Furthermore, this dietary distinction was only evident in a portion of the diets of major and mixed-diet frugivores; species from both groups consumed several other plant taxa (e.g., Group C in Figure 7.2), despite their different degrees of frugivory.

It is possible that, for major and mixed-diet frugivores (with gapes wider than 10 mm), plant species in Groups D and E resemble ‘core plant taxa’, while those in Groups

A, B and C are added as they become available, in the manner proposed in the ‘core

177 plant taxa’ model by Fleming (1986). A major characteristic of the ‘core plant taxa’ that distinguished between groups of Neotropical bat genera was their extended fruit availability (Fleming, 1986). An analysis of fruiting phenology in subtropical Australian rainforest showed that several plant in Groups D and E (e.g., Moraceae, Solanaceae and

Euphorbiaceae) fruit over extended periods (Innis, 1989). However, in contrast to

Fleming’s model, major and mixed-diet frugivore species in the present study were distinguished by their consumption of intermittently-available species (i.e., Groups A and B), rather than by their consumption of plant species that had longer periods of availability.

The extent to which the model presented in Figure 7.2 may be consistent in other regions is not known, although it is based on a much larger amount of data than existing models of frugivore-plant interactions. However, a functional attribute approach, such as that used in Chapter Three, could be used to assess potential functional similarity and variation among frugivore species in many situations where diet information is lacking.

Data on bird species’ body mass may be more accessible than gape width measurements for species in other regions (e.g., Dunning, 1993). For the bird species included in the present study, gape width is strongly positively associated with body mass (Figure 7.3).

The gape width of seven bird species was smaller than expected from their body size; six of these species (encircled) were the seed-crushing Columbidae in the assemblage

(feral pigeon, emerald dove, bar-shouldered dove, brown cuckoo-dove, wonga pigeon and white-headed pigeon). The seventh species was the Australian brush-turkey, also a seed-crusher. Seed-crushers may therefore have a larger ratio of body mass to gape width. The capacity of a bird species to fly with seeds is limited by its body size (Mack,

1993) and non-crushing species may therefore rapidly regurgitate or defecate indigestible seed ballast (Murray et al., 1993). In contrast, seed-crushing species may retain seeds for a relatively long time in order to digest them. Therefore, seed-crushers,

178 especially those species that have a diet dominated by the seeds of fleshy-fruited plants

(e.g., brown cuckoo-dove, white-headed pigeon), may require a large body mass. The channel-billed-cuckoo had a much larger gape than expected from its body size (Figure

7.3). In general, body mass may be useful as a surrogate measure of relative seed size handling capacity for most bird species, but would over estimate this capacity for seed crushers. This would not affect predictions if seed crushers were classified in a separate functional group with limited potential to disperse viable seed, as they were in the present study.

40

30 channel-billed cuckoo

20

Gape width(mm) 10 Australian seed-crushing brush-turkey 0 Columbidae 2 76543 8

Log body mass (g)

Figure 7.3 Relationship between bird species’ gape widths and their (ln+1) body mass (R2 = 0.48, p <0.00001, d.f. = 40; with outliers removed, R2 = 0.83, p <0.00001, d.f. = 32).

7.5 Potential consequences of rainforest fragmentation for seed dispersal and patterns of plant regeneration

Chapter One introduced the concept of seed dispersal trajectories to highlight the consequences of the multiple spatial dimensions of seed dispersal for plant regeneration outcomes (Section 1.6, Figure 1.2, Table 1.1). Incorporating the knowledge developed in this thesis regarding frugivore species’ responses to forest fragmentation and

179 information regarding their patterns of movement in the landscape, may be used to asses: (1) the potential for variation among frugivore species in their contribution to the movement of seeds along certain trajectories; and therefore, (2) whether particular aspects of plant regeneration may be disproportionately affected by changes in the frugivore assemblage in fragmented forest.

Dennis and Westcott (2006) developed a classification of frugivorous vertebrate species in tropical Australia based partly on differences in their spatial scale and rate of movements and gut passage rates. Table 7.1 combines information from the present study with that presented by Dennis and Westcott (2006) to examine consequences of forest fragmentation for seed dispersal in the study region. For example, species that range widely across the landscape, travel fast and have a slow gut passage rate and use remnant and regrowth habitats (e.g., topknot pigeon, channel-billed cuckoo, barred cuckoo-shrike; Table 7.1), may disperse seeds along all trajectories, except over short distances to microsites within a forest fragment (trajectory b). In contrast, the black- faced cuckoo-shrike, which predominantly uses regrowth rainforest (Chapter Two), would be expected to disperse relatively few seeds from or into remnants (trajectories c, d, e and f). Because decreaser fruit-doves were in very low numbers in fragmented forest (Chapter Two), they may not disperse seeds along any of the dispersal trajectories in fragmented forest (Table 7.1). The decreaser green catbird and eastern tube-nosed bat were considered to have the potential to disperse seeds only along trajectory b (Table

7.1) because, although these species occurred in fragments during the present work

(Chapters Two and Six), the eastern tube-nosed bat makes limited foraging movements

(Spencer and Fleming 1989) and the green catbird is territorial (Innis and McEvoy,

1992). The satin bowerbird potentially disperses seeds along trajectories c, d, f and g if the distances are short.

180 Table 7.1 Characteristics of frugivore species that may influence their role in dispersing seed along different dispersal trajectories in fragmented rainforest in subtropical Australia. ‘Move. scale’ is scale of movement, ‘Move rate’ is rate of movement, and ‘Gut pass.’ is gut passage rate. ‘Frag. resp.’ shows species’ fragmentation responses. This information was combined to asses the potential for each species to contribute to dispersal along individual trajectories (Refer to Figure 7.4 and Section 1.6). Note that, although it was included in the original (Figure 1.2), dispersal away from the crown of the parent plant (trajectory a) is not shown as it is implied by the movement of seed along any of the other trajectories.

Bird species1 Characteristics2 Frag. Potential dispersal resp.3 trajectories Move. scale Move. Gut b c d e f g rate pass. wompoo fruit-dove short slow long Dec superb fruit-dove short slow long Dec rose-crowned fruit-dove short slow long Dec topknot pigeon wide-very wide fast long Tol x x x x x common koel moderate-wide fast long Tol x x x x x channel-billed cuckoo wide-very wide fast long Tol x x x x x Lewin's honeyeater short slow long Tol x x x x x black-faced cuckoo-shrike wide-very wide fast long Inc x x barred cuckoo-shrike wide-very wide fast long Undet x x x x x varied triller short slow short Undet x olive-backed oriole short slow long Undet x x x x x figbird wide-very wide fast short Inc x x x x x pied currawong moderate-wide fast long Tol x x x x x green catbird - - - Dec x satin bowerbird short slow short Tol x mistletoebird short slow long Undet x x x x x silvereye short slow long Inc x x x x x eastern tube-nosed bat wide-very wide fast long Dec x flying foxes wide-very wide fast long Tol x xS xS xS xS xS

1 Frugivore species shown were those considered in both the Wet Tropics study of Dennis and Westcott (2006) and the present study, except the green catbird which is endemic to subtropical Australia but for which sufficient information was available to characterise its potential to contribute to dispersal along different trajectories. Where species’ fragmentation responses in the study region had not been determined (e.g., barred cuckoo-shrike, varied triller, olive-backed oriole and mistletoebird), potential dispersal trajectories shown may be affected by unmeasured differences in their patterns of use of remnants and regrowth. Details for seed crushing bird species and minor frugivores were not provided by Dennis and Westcott (2006). 2 Information on species’ characteristics after Dennis and Westcott (2006). Movement scale during the average gut passage time of a seed (short <100 m, moderate = 100-200 m, wide = 200-800 m, very wide >800 m). Movement rate was determined by the slope of the relationship between distance moved and time, measured using radio-tracked individuals; slow <2, fast >2. Gut passage was short (<30 min.), long (>30 min). 3 Fragmentation response is the response pattern detected in the present work (Chapters Two and Six) ; Dec is Decreaser, Tol is Tolerant, Inc is Increaser and Undet is Undetermined. S only small (ca. < 9mm) seeds would be likely to be moved along these trajectories by flying-foxes.

181 i) Plant taxa consumed by most frugivore species

topknot pigeon figbird common koel common koel pied currawong Lewin’s honeyeater channel-billed cuckoo mistletoebird varied triller Lewin’s honeyeater silvereye S olive-backed oriole barred cuckoo-shrike flying-foxes figbird olive-backed oriole pied currawong topknot pigeon green catbird common koel channel-billed cuckoo satin bowerbird g mistletoebird Lewin’s honeyeater silvereye black-faced cuckoo-shrike eastern tube-nosed bat b barred cuckoo-shrike flying-foxes olive-backed oriole figbird pied currawong topknot pigeon f mistletoebird common koel silvereye channel-billed cuckoo flying-foxesS Lewin’s honeyeater c barred cuckoo-shrike olive-backed oriole d figbird topknot pigeon pied currawong e channel-billed cuckoo mistletoebird barred cuckoo-shrike silvereye flying-foxesS flyin g-foxesS

ii) Plant taxa consumed predominantly by decreaser frugivore species

topknot pigeon Lewin’s honeyeater flying-foxesS

Lewin’s honeyeater green catbird g satin bowerbird eastern tube-nosed bat b topknot pigeon flying-foxes Lewin’s honeyeater flying-foxesS

topknot pigeon f Lewin’s honeyeater flying-foxes S c

d e topknot pigeon flying-foxesS

ii) Plant taxa consumed only by decreaser frugivore species

green catbird g eastern tube-nosed bat b

f c d e

182 Figure 7.4 (on previous page)The frugivore species that potentially disperse seeds along different dispersal trajectories in fragmented forest landscapes for plant taxa that are i) consumed by most frugivore species; ii) predominantly consumed by decreaser species; and iii) only consumed by decreaser species. The grey patches in this figure represent forest fragments and the white area represents the cleared or modified matrix. ‘Decreaser’ frugivore species that occurred in low frequency in rainforest remnants in this study (all fruit-doves; Chapter 2), are not shown in this figure because they are likely to disperse few seeds in fragmented forest. Seed-crushing frugivore species are not shown in this figure because they are likely to disperse few viable seeds. The white diamond represents an individual plant. Arrows show the trajectories of seed movement. Table 7.1 shows the characteristics associated with bird species’ potential to disperse seeds along different trajectories. S flying-foxes are likely to disperse only small seeds (<9 mm) along trajectories involving movement beyond tens of metres.

Many plant species, including Moraceae, and Euphorbiaceae, are consumed by most frugivore species (Chapters 4 and 5). Consequently, seeds of these plant taxa are potentially dispersed by multiple frugivore species along most dispersal trajectories

(Figure 7.4i). However, a relatively small number of frugivore species had the potential to disperse plant taxa over long distances to new habitats (trajectory e). With the exception of the topknot pigeon and flying-foxes, these frugivore species had fig- dominated diets (Chapter Four). Flying-foxes may only disperse small-seeded plant species along this trajectory (Eby, 1991b, 1995; Shilton et al., 1999). Consequently, the topknot pigeon may be the main agent of colonisation of new habitats for most plant taxa. Plant taxa that are not dispersed to new habitats may have a limited ability to cope with climate change (Primack and Miao, 1992; Matlack, 1995; Westoby and Burgman,

2006; Weir and Corlett, 2007).

There was a suite of plant taxa, including Lauraceae, Myrtaceae, Meliaceae,

Verbenaceae and Vitaceae and plant species with large fruits (>10 mm), that were predominantly consumed by decreaser frugivores. Analyses of the potential of tolerant and increaser species to disperse these plants have shown that only a small subset of 183 frugivore species (topknot pigeon, satin bowerbird, Lewin’s honeyeater or flying-foxes) may be likely to substitute for decreaser frugivore species as dispersers of these plant taxa in fragmented rainforest (Chapters Five and Six). The decreaser green catbird and eastern tube-nosed bat potentially disperse seeds within fragments, although the eastern tube-nosed bat may only consume a small number of these plant species (Chapter Six).

Figure 7.4 (ii) shows that a small number of frugivore species potentially disperse these plant taxa along each trajectory. Unless these frugivore species increase their feeding rates on these plant taxa concurrent with the decline of decreaser species (i.e.,

‘behavioural compensation’ (Loiselle and Blake, (2002)), dispersal rates of these plants would be likely to be reduced in fragmented forest.

In the case of plant taxa that were only known to be consumed by decreaser frugivore species, which is likely to include several species from the Rubiaceae

(Chapter Five), dispersal along any of the trajectories shown would be severely limited in fragmented rainforest (Figure 7.4 (iii)). Consequently, regeneration of these plant taxa would be likely to be substantially reduced in fragmented forest landscapes.

The predicted reductions in dispersal of certain plant taxa may be tested by comparing observations of fruit removal between fragmented and extensive forest landscapes. Predictions relating to the consequences of frugivore declines for seed dispersal along certain trajectories in fragmented rainforest, for example, reductions in dispersal to recruitment microsites, may be assessed using seed trapping (e.g., Harvey,

2000), or seed tracking techniques (e.g., Levey and Sargent, 2000; Tewksbury et al.,

2002). Dispersal over long distances to new habitats (Figure 7.4, trajectory e) may be infrequent for most plant species (Nathan and Muller-Landau, 2000), but may be severely limited in fragmented landscapes for certain plant taxa (e.g., those not consumed by the topknot pigeon). Differences among plant species in their potential for

184 dispersal over long distances may be inferred by analysing changes in species’ distributions, for example in response to global climatic changes.

Many frugivore species may contribute to the dispersal of seeds from the non- forest matrix into fragments and within the non-forest matrix (trajectories f and g,

Figure 1.6; Table 7.1). Consequently, high rates of fruit removal from plants in the non- forest matrix, and seed input from these plants into fragments and secondary regrowth are expected. Introduced plant species are ubiquitous in most fragmented landscapes

(Buckley et al., 2006) and, depending on the patterns of consumption of plant species by frugivore species, may be among the plant species with high dispersal rates in fragmented forest landscapes. Based on their consumption of a relatively high dietary proportion of introduced plant taxa (Chapter Five, Table 5.1), it is predicted that increaser bird species would disperse the seeds of introduced plant species. Studies of fruit removal and seed input (e.g., Guevara and Laborde, 1993; Medellín and Gaona,

1999), and of plant recruitment within fragments (e.g., Janzen, 1983) or in regrowth

(e.g., Neilan et al., 2006) may asses these predictions.

7.6 Conservation issues

This thesis has shown that existing fragmented rainforest in the study region is likely to have lost a component of the vertebrate frugivorous fauna and, as a consequence, that the regenerative potential of a substantial proportion of native plant species may be reduced. Continued clearing and fragmentation of rainforest would be likely to exacerbate this situation by further disadvantaging decreaser species, and potentially leading to the decline of additional frugivore species.

All of the decreaser frugivore species identified in this work used at least a subset of the rainforest remnants surveyed. The specific patterns of use of fragmented rainforest habitats by these frugivore species may be related to the size and

185 configuration of fragments (Doak et al., 1992; Andrén, 1994; Wiens, 1994; With et al.,

1997; Price et al., 1999; Graham, 2001; Graham and Blake, 2001; Develey and

Metzger, 2006). Identification of factors that influence the distribution of decreaser frugivore species in fragmented parts of the landscape, such as a threshold fragment size or degree of connectivity, would highlight the areas of highest conservation value for these susceptible vertebrate taxa. This understanding could also inform rainforest restoration actions in terms of the landscape attributes that are necessary to reinstate populations of decreaser frugivore species in fragmented landscapes. Information provided in this thesis may be used to identify specific plant resources that would be useful for decreaser frugivore species. For example, Lauraceae may be important for fruit-pigeons (Crome, 1975; Innis, 1989; Recher et al., 1995), while strangling figs (e.g.,

Ficus watkinsiana, F. macrophylla) and Elaeocarpus grandis may be important for the eastern tube-nosed bat.

Based on the secondary consequences of frugivore declines that have been predicted in this thesis, rainforest restoration would also be necessary to return the regenerative capacity of several plant taxa in fragmented parts of the landscape. In the short-term, replanting or direct-seeding of the plant taxa that are vulnerable to reduced dispersal in fragmented forest could be incorporated into ecological management and restoration works. In the longer term, ecological restoration that enables the movement of frugivore species across modified parts of the landscape may facilitate seed dispersal by frugivores in these habitats (Tucker and Murphy, 1997; Tewksbury et al., 2002;

Martínez-Garza and Howe, 2003; Jansen, 2005).

186 Appendix 1

Original published sources of records of plant species consumption by frugivore species. The thesis chapters in which data from each source was included are shown following each reference.

Barker, R.D., Vestjens, W.J.M., 1989. The food of Australian birds: Non-Passerines. C.S.I.R.O., Melbourne. 4, 5, 6 Barker, R.D., Vestjens, W.J.M., 1990. The food of Australian birds: Passerines. C.S.I.R.O., Melbourne. 4, 5, 6 Bedggood, G.W., 1970. Bird notes from East Gippsland. Australian Bird Watcher 3, 252-265. 4, 5, 6 Béland, P., 1977. Mimicry in Orioles of south-eastern Queensland. Emu 77, 215-218. 4, 5, 6 Binns, G. 1954. The camp out at Lake Barrine, Atherton Tableland, North Queensland. Emu 54, 29-46. 4 Bourke, P.A., 1949. The breeding population of a thirty-five acre ‘Timber Paddock’. Emu 49, 73-83. 4, 5, 6 Bravery, J.A., 1970. The birds of Atherton Shire, Queensland. Emu 66, 267-271.4, 5, 6 Brookes, G.B. 1919. Report on investigation in regard to the spread of prickly-pear by the Scrub Turkey. Q. Ag. J. 11, new series, 26-28. 4 Burton, A.C.G., Morris A.K., 1993. New South Wales Bird Report. Aust. Birds 26, 89- 133. 4, 5, 6 Campbell, A.J., Barnard, H.G., 1917. Birds of the Rockingham Bay district, North Queensland. Emu 17, 1-12. 4, 5, 6 Carter, T., 1924. Birds of the Broome Hill district. Part III. Emu 23, 306-318. 4, 5, 6 Chisholm, A.H., 1938. The birds of Barellan, N.S.W. with botanical and other notes. Emu 37, 301-313. 4 Chisholm, A.H., 1944. An interesting old notebook. Emu 43, 281-288. 4, 5, 6 Church R., 1997. Avian frugivory in a subtropical rainforest: Eleven years of observations in Lamington National Park. The Sunbird 27, 85-97. 4, 5, 6 Cleland, J.B., 1911. Examination of contents of stomachs and crops of Australian birds. Emu 11, 79-95. 4 Cleland, J.B. Cited in Higgins, P.J. (Ed.), 1999. Handbook of Australian, New Zealand and Antarctic Birds: V4 Parrots to Dollarbirds. Oxford University Press, Melbourne. 4, 5, 6 Cleland, J.B., Maiden, J.H., Froggatt, W.W., Ferguson, E.W., Musson, C.T., 1918. The food of Australian birds. N.S.W. Department of Agricultural Science Bulletin No. 15. 4, 5, 6 Cooper, R.P., 1962. A revision of the distribution of the Brown Pigeon. Emu 61, 266- 269. 4 Cooper, R.M., Knight, B., 1989. New South Wales Bird Report for 1985 Aust. Birds 22, 1-45. Cooper W., Cooper, W.T., 1994. Fruits of the rain forest: A guide to fruits in Australian tropical rain forests. Geo Production, Chatswood. 4, 5, 6 Crome, F.H.J., 1975a. The ecology of fruit pigeons in tropical northern Queensland. Aust. Wildl. Res. 2, 155-185. 4, 5, 6

187 Crome, F.H.J., 1975b. Breeding, feeding and status of the Torres Strait Pigeon at Low Isles, north Queensland. Emu 75,189-98. 4, 5, 6 Crome, F.H.J., 1978. Foraging ecology of an assemblage of birds in lowland rainforest in northern Queensland. Aust. J. Ecol. 3, 195-212. 4, 5, 6 Crome F.H.J., Shields J., 1992. Parrots and Pigeons of Australia. Angus & Robertson, Pymble. 4, 5, 6 Crouther, M.M., 1985. Some breeding records of the common koel Eudynamis scolopacea. Australian Bird Watcher 11, 49-56. 4, 5, 6 Date, E.M, Recher, H.F., 1990. Ecology and Management of Rainforest Pigeons in NSW – An interim report. NSW NPWS, Sydney. 4, 5, 6 Denny, T., Dudman, D., 1979. Koel behaviour. Sunbird 10,78.4, 5, 6 De Warren, J.J., 1928. The avifauna of the upper reaches of the Macleay River, NSW. Emu 28,11-120. 4, 5, 6 Drew, R.A.I., 1987. Reduction in fruit fly (Tephritidae: Dacinae) populations in their endemic rainforest habitat by frugivorous vertebrates. Aust. J. Zool. 35, 283-8. 4, 5, 6 Eby, P. ,1991 “Finger-winged night workers”: managing forests to conserve the role of Grey-headed Flying Foxes as pollinators and seed dispersers. pp. 91-100 in (D. Lunney (ed.)) Conservation of Australia’s Forest Fauna. Royal Zoological Society of NSW, Mosman. 6 Eby, P., 1995. The biology and management of flying foxes in New South Wales. Species Management Report No. 18. New South Wales National Parks and Wildlife Service, Hurstville. 6 Eby, P. 1998. An analysis of diet specialisation in frugivorous Pteropus poliocephalus (Megachiroptera) in Australian subtropical rainforest. Aust. J. Ecol. 23, 443-456. 6 Firth, D.J., 1979. Ecology of Cinnamomum camphora (l.) Nees and Eberm (camphor laurel) in the Richmond-Tweed region of north-eastern New South Wales. Unpublished thesis, Department of Botany, University of New England, Armidale. 4, 5 Floyd, A.G., 1989. Rainforest trees of mainland south-eastern Australia. Forestry Commission of New South Wales, Sydney. 4, 5, 6 Floyd, A.G., 1990. Australian Rainforests in New South Wales. Volume 1. Surrey Beatty & Sons, Chipping Norton. 4, 5, 6 Forde, N., 1986. Relationships between birds and fruits in temperate Australia. in Ford, H.A., Paton, D.C. (Eds.), The dynamic partnership: Birds and plants in southern Australia. D.J. Woolman, Government Printers, Adelaide, pp. 42-58. 4, 5, 6 Forshaw, J.M., 1969. Australian Parrots. Landsdowne Press, Melbourne. 4 Forshaw, J.M., Muller, K.A., 1978. Annotated list of birds observed at Iron Range, Cape York Peninsula, Queensland, during October 1974. Aust. Bird Watcher 7, 171-194. 4, 5, 6 Forshaw, J.M., Cooper, W.T. (1981) Australian Parrots. Second edition. Landsdowne Editions, Melbourne. 4 French, K., 1990. Evidence for frugivory by birds in montane and lowland forests in South-east Australia. Emu 90,185-189. 4, 5, 6 Frith, H.J., 1952. Notes on the pigeons of the Richmond River. Emu 52,89-99. 4, 5, 6 Frith H.J., 1957. Food habits of the Topknot Pigeon. Emu 57, 341-345. 4, 5, 6 Frith H.J., 1982. Pigeons and doves of Australia. Rigby Publishers, Adelaide. 4, 5, 6 Frith, H.J., Crome, F.H.J., Wolfe, T.O., 1976. Food of fruit-pigeons in New Guinea. Emu 76, 49-58. 4, 5, 6 Frith, C.B., Frith, D.W., 2004. Bird Families of the World: The Bowerbirds Ptilonorhynchidae. Oxford University Press: Oxford. 4, 5, 6 188 Gannon, G.R., 1936. Plants spread by the Silvereye. Emu 35, 314-316. 4, 5, 6 Gibson, J.D., 1977. The birds of the County of Camden (including the Illawarra district). Aust. Birds 11, 41-80. 4, 5, 6 Gilbert, P.A., 1936. The Topknot Pigeon. Emu 35, 301-312. 4, 5, 6 Gosper, C.R., 1994. Comparison of the avifauna of rainforest remnants with regrowth dominated by the exotic tree camphor laurel Cinnamomum camphora. Unpublished thesis, University of New England, Armidale. 4, 5, 6 Gosper, C.R., 1999. Plant food resources of birds in coastal dune communities in New South Wales. Corella 2, 53-62. 4, 5, 6 Gosper, D.G., cited in Higgins, P.J. (Ed.), 1999. Handbook of Australian, New Zealand and Antarctic Birds: V4 Parrots to Dollarbirds. Oxford University Press, Melbourne. 4, 5, 6 Gosper, D.G., 1962. Breeding records of the koel. Aust. Bird Watcher 1, 226-228. 4, 5, 6 Gould, J., 1865. Handbook to the Birds of Australia. 2 volumes. J. Gould, London. 4, 5, 6 Green R.J., 1993. Avian seed dispersal in and near subtropical rainforests. Wildl. Res. 20, 535-557. 4, 5, 6 Hackett, D., 1996. Frugivory and Ligustrum lucidum in north-eastern NSW: Implications for seed dispersal and avifauna conservation. Integrated project, Southern Cross University, Lismore. 4, 5, 6 Hall, L.S., Richards, G. 2000. Flying foxes: Fruit and blossom bats of Australia. UNSW Press: Sydney. 6 Halse, S.A., 1978. Feeding habits of six species of honeyeaters in south-western Australia. Emu 78, 145-148. 4, 5, 6 Hindwood, K.A., 1959. The Purple-crowned Pigeon in southeastern Australia. Emu 59, 219-20. 4, 5, 6 Hindwood, K.A., 1970. The Regent Bowerbird near Sydney. Birds (Sydney) 5, 21-24. 4, 5, 6 Holland, L., 1964. Pigeons of the Woolgoolga District, N.S.W. Aust. Bird Watcher 2, 61-69. 4, 5, 6 Holmes, G., 1987. Avifauna of the Big Scrub region. Report prepared for the Australian National Parks and Wildlife Service & NSW National Parks and Wildlife Service, Sydney. 4, 5, 6 Holmes, G., 1990. Biology and Ecology of Coxen's Fig-parrot. Royal Australasian Ornithologists Union, Report 65. 4, 5, 6 Holmes, G., cited in Higgins, P.J. (Ed.), 1999. Handbook of Australian, New Zealand and Antarctic Birds: V4 Parrots to Dollarbirds. Oxford University Press, Melbourne. 4, 5, 6 Hoskin, E., 1991. Birds of Sydney 1770-1989. Surrey Beatty and Sons: Sydney. 4, 5, 6 Howe, F.E., 1928. Notes on some Victorian birds. Emu 27, 252-265. 4, 5, 6 Innis, G.J., 1989. Feeding ecology of fruit pigeons in subtropical rainforests of south- eastern Queensland. Aust. Wildl. Res. 16, 365-94. 4, 5, 6 Innis, G.J., McEvoy, J., 1992. Feeding ecology of green catbirds (Ailuroedus crassirostris) in subtropical rainforests of south-eastern Queensland. Wildl. Res. 19, 317-29. 4, 5, 6 Jarvis, H., 1929. Feeding habits of some Queensland birds. Q. Agric. J. 32, 8-12. 4, 5, 6 Jenkins, C.F.H., 1968. Notes on the feeding habits of some West Australian birds. W. A. Nat. 11, 52-55. 4 Johncock, C.F., 1903. Notes on Loranthus exocarpi. Trans. R. Soc. S. Aust. 27, 253- 255. 4, 5

189 Johnstone, R.E., Burbidge, A.H., 1991. The avifauna of Kimberley rainforests, in McKenzie, N.L., Johnstone, R.B., Kendrick, P.G. (Eds.) Kimberley Rainforests in Australia. Surrey Beatty, Sydney, pp.361-395. 4, 5, 6 Keast, A., 1958. The influence of ecology on variation in the Mistletoebird (Dicaeum hirundinaceum). Emu 58, 195-206. 4 Keast, A., 1968. Seasonal movements in Australian honeyeaters (Meliphagidae) and their ecological significance. Emu 67, 159-209. 4, 5 Knight, B.J., 1970. A popular native. Bird Observer 462, 4-5. 4 Lea, A.M. and Gray, J.T. (1935b). The food of Australian Birds. Part II. Emu 35, 63-98. Lea, A.M., Gray, J.T., 1936. The food of Australian Birds. Part IV. Emu 35, 251-280. 4, 5, 6 Leach, G.J., Hines, H.B., 1987. Birds of the Marburg District, South-East Queensland. Sunbird 17, 65-95. 4, 5 Lenz, N.H.G., 1993. Behavioural and reproductive biology of the regent bowerbird Sericulus chrysocephalus (Lewin, 1808). PhD Thesis, Griffith University, Nathan. 4, 5, 6 Lepschi, B.J., 1997. Food of some birds in southern Australia: Additions to Barker and Vestjens, Part 2. Emu 97, 84-87. 4, 6 Liddy, J., 1985. A note on the associations of birds and Lantana near Beerburrum, south-eastern Queensland. Corella 9, 125-126. 4, 5 Lindsey, T.R., 1984. New South Wales Bird Report for 1982. Aust. Birds 18, 37-69. 4, 5, 6 Lindsey, T.R., 1985. New South Wales Bird Report for 1983. Aust. Birds 19, 65-100. 4, 5 Lord, E.A.R., 1943. Migratory notes (1941 – 1942). Emu 43, 18-23. 4, 5, 6 Lord, E.A.R., 1956. Birds of the Murphy’s Creek district, Queensland. Emu 56, 100- 128. 4, 5, 6 Makin, D., 1969. Birds of Sandy Cape, Fraser Island. Q. Nat. 19, 31-42. 4, 5 Male, T.D. Roberts, G.E., 2002. Defence of fruiting trees by birds in an Australian forest. Biotropica 34, 172-6. 4, 5, 6 Mannes, E., 1976. Brown Pigeon observations. N.S.W. Field Ornithology Club Newsletter 16, 3. 4, 5, 6 Marshall, A.J., 1935. On the birds of the McPherson Ranges, Mt. Warning and contiguous lowlands. Emu 35, 36-48. 4, 5 Marshall, A.J., 1964. Southern Figbird near Bass Strait. Emu 63, 339. 4, 5, 6 Mees, G.F., 1969. A systematic review of the Indo-Australian Zosteropidae. III. Zoologische Verhandelingen 102, 1-390. 4, 5, 6 Morris, J.G., 1969. The control of feral pigeons and sparrows associated with intensive animal production. Aust. J. Sc. 32, 9-15. Morris, A.K., 1975. The birds of Gosford, Wyong, and Newcastle (County of Cumberland). Aust. Birds 9, 37-76. 4, 5, 6 North cited in Higgins, P.J., (Ed.), 1999. Handbook of Australian, New Zealand and Antarctic Birds: V4 Parrots to Dollarbirds. Oxford University Press, Melbourne. 4, 5, 6 North cited in Marchant, S., Higgins, P.J., (Eds.), 1993. Handbook of Australian, New Zealand and Antarctic Birds: V2 Raptors to Lapwings. Oxford University Press, Melbourne. 4, 5, 6 Norton, J., 1897. Magpies (Black and Grey). Agric. Gazette of N.S.W. 8, 535-537. 4, 5, 6

190 Panetta, F.D., Sparkes, E.C., 2001. Reinvasion of a riparian forest community by an animal-dispersed tree weed following control measures. Biological Invasions 3:75-88. 4, 5, 6 Parry-Jones, K., Augee, M.L. 1991. Food selection by Grey-headed Flying-foxes (Pteropus poliocephalus) occupying a summer colony site near Gosford, New South Wales. Aust. Wildl. Res. 18,111-124. 6 Parry-Jones, K., Augee, M.L., 2001. Factors affecting the occupation of a colony site in Sydney, New South Wales by the Grey-headed Flying-Fox Pteropus poliocephalus (Pteropidae). Aust. Ecol. 26, 47-55. 6 Paton, D.C., 1986. Honeyeaters and their plants in southeastern Australia, in Ford, H.A., Paton, D.C., (Eds.), The dynamic partnership: Birds and plants in southern Australia. D.J. Woolman, Government Printers, Adelaide, pp. 9-19. 4, 5 Perkins, D.L., 1973. Some observations from Fraser Island. Sunbird 4, 40-41.4, 5, 6 Ratcliffe, F., 1932. Notes on the fruit bats (Pteropus spp.) of Australia. J. An. Ecol. 1, 32-57. 6 Recher, H.F., Date, E.M., Ford, H.A., 1995. The biology and management of rainforest pigeons in N.S.W.. N.S.W. National Parks and Wildlife Species Management Report No. 16. 4, 5, 6 Reid, N., 1986. Pollination and seed dispersal of mistletoes (Loranthaceae) by birds in southern Australia in Ford, H.A., Paton, D.C., (Eds.), The dynamic partnership: Birds and plants in southern Australia. D.J. Woolman, Government Printers, Adelaide, pp. 64-84. 4, 5, 6 Richards, G.C. 1990. The Spectacled flying-fox, Pteropus conspicillatus, in North Queensland: Part 2, Diet, feeding ecology and seed dispersal. Aust. Mamm. 13:25-31. 6 Roberts, G.J., Ingram, G.J., 1976. An annotated list of the land birds of Cooloola. Sunbird 7, 1-20. 4, 5, 6 Rogers, J., 1998. A coastal walk. Bird Observer 789, 10-11. 4 Rose, A.B., 1973. Food of some Australian birds. Emu 73, 177-183. 4, 5, 6 Rose, A.B., 1997. Notes on the diet of cuckoos in New South Wales. Australian Bird Watcher 17: 130-133. Aust. Bird Watcher 17, 134-137. 4, 5, 6 Rose, A.B., 1999. Notes on the non-nectar foods of some honeyeaters in eastern New South Wales. Aust. Bird Watcher 18, 26-34. 4, 5, 6 Rowley, I., Vestjens, W.J.M., 1969. The comparative ecology of Australian corvids. V. Food. C.S.I.R.O. Wildl. Res. 18, 131-155. 4, 5, 6 Saunders, T., cited in Higgins, P.J., Davies, S.J.J.F., (Eds.), 1996. Handbook of Australian, New Zealand and Antarctic Birds: V3 Snipe to Pigeons. Oxford University Press, Melbourne. 4, 5, 6 Scanlon, T., the Camphor Laurel Taskforce, 2001. Camphor Laurel Kit: Everything you need to know about camphor laurel and its control. North Coast Weed Advisory Committee, Grafton, NSW http://www.fncw.nsw.gov.au/camphor_kit.html 4, 5, 6 Shanks, D.. 1949. Observations from the Upper King River district, Victoria. Emu 49, 132-141. 4 Smith, L., 1984. Garden plants attractive to birds. Aust. Birds 18, 17-32. 4, 5, 6 Spencer, H.J., Fleming, T.H. 1989. Roosting and Foraging Behaviour of the Queensland Tube-nosed Bat, Nyctimene robinsoni (Pteropidae): Preliminary Radio-tracking Observations. Aust. Wildl. Res. 16:413-420. 6 Stansbury, C.D., Vivian-Smith, G., 2003. Interactions between frugivorous birds and weeds in Queensland as determined from a survey of birders. Plant Prot. Quart. 18, 157-165. 4, 5, 6

191 Storr, G.M., 1953. Birds of the Cooktown and Laura districts, north Queensland. Emu 53, 225-248. 4, 5, 6 Vestjens, W.J.M., Carrick, R., 1974. Food of the Black-backed Magpie, Gymnorhina tibicen, at Canberra. Aust. Wildl. Res. 1, 71-83. 4, 5 Waterhouse cited in Higgins, P.J., Davies, S.J.J.F., (Eds.), 1996. Handbook of Australian, New Zealand and Antarctic Birds: V3 Snipe to Pigeons. Oxford University Press, Melbourne. 4, 5, 6 Waterhouse, R.D., 1995. Observations on the diet of the Lewin's Honeyeater Meliphaga lewinii in the Illawarra Rainforest, New South Wales. Corella 19, 102-105. 4, 5, 6 Wheeler, W.R., 1967. The birds of Cairns, Cooktown and the Atherton Tablelands. Aust. Bird Watcher 3, 55-76. 4, 5, 6 Wheeler, W.R., 1972. Bird notes 1970-71. Bird Obs. 484, 4-8. 4, 5, 6

192 Appendix 2

List of plant species used in the data sets analysed in Chapters Four, Five and

Six and the total number of frugivorous bird and bat species (data for Pteropus alecto and P. poliocephalus combined) known to consume each plant species.

Thesis Chapter Family Genus Species 4 5 6 Total number frugivore species Agavaceae Cordyline petiolaris 4 5 6 2 Agavaceae Cordyline rubra 5 6 1 Agavaceae Cordyline stricta 5 6 1 Akaniaceae Akania bidwillii 4 5 6 4 Alangiaceae Alangium villosum 4 5 6 5 Anacardiaceae Euroschinus falcata 4 5 6 14 Anacardiaceae Rhodosphaera rhodanthema 4 5 6 1 Annonaceae Melodorum leichhardtii 4 5 6 2 Annonaceae Polyalthia nitidissima 4 5 6 1 Apocynaceae Carissa ovata 4 5 6 2 Apocynaceae Melodinus australis 4 5 6 4 Araceae Alocasia brisbanensis 4 5 6 1 Araceae Pothos longipes 4 5 6 5 Araliaceae Cephalaralia cephalobotrys 5 6 1 Araliaceae Polyscias elegans 4 5 6 18 Araliaceae Polyscias murrayi 4 5 6 15 Araliaceae Polyscias sambucifolia 4 5 6 1 Arecaceae Archontophoenix cunninghamia 4 5 6 18 Arecaceae Calamus muelleri 4 5 6 3 Arecaceae Linospadix monostachya 4 5 6 3 Arecaceae Livistonia australis 5 6 7 Arecaceae Ptychosperma elegans 4 5 6 1 Boraginaceae Ehretia acuminata 4 5 6 15 Burseraceae Canarium australasicum 4 5 6 9 Caprifoliaceae Sambucus australasica 4 5 6 2 Celastraceae Celastrus subspicata 4 5 6 4 Celastraceae Maytenus bilocularis 4 5 6 1 Celastraceae Maytenus cunninghamii 4 5 6 1 Celastraceae Siphonodon australis 5 6 2 Chenopodiaceae Enchylaena tomentosa 4 5 6 1 Cucurbitaceae Diplocyclos palmatus 5 6 4 Cucurbitaceae Sicyos australis 5 6 1 Cucurbitaceae Zehneria cunninghamii 4 5 6 2 Cunoniaceae Schizomeria ovata 4 5 6 7 Dilleniaceae Hibbertia scandens 4 5 6 6 Ebenaceae Diospyros australis 4 5 6 3 Ebenaceae Diospyros fasciculosa 4 5 6 3 Ebenaceae Diospyros geminata 4 5 6 6 Ebenaceae Diospyros pentamera 4 5 6 12 Elaeagnaceae Elaeagnus triflora 4 5 6 1 Elaeocarpaceae Elaeocarpus grandis 4 5 6 8 Elaeocarpaceae Elaeocarpus kirtonii 4 5 6 9 Elaeocarpaceae Elaeocarpus obovatus 4 5 6 15 Elaeocarpaceae Elaeocarpus reticulatus 4 5 6 11 Elaeocarpaceae Sloanea australis 4 5 6 8 Elaeocarpaceae Sloanea woollsii 4 5 6 11 Epacridaceae Leucopogon parviflorus 4 5 6 1 Epacridaceae Monotoca elliptica 4 5 6 2 Epacridaceae Trochocarpa laurina 4 5 6 3 Euphorbiaceae Actephila lindleyi 5 6 1 Euphorbiaceae Breynia oblongifolia 4 5 6 4 Euphorbiaceae Claoxylon australe 4 5 6 3 Euphorbiaceae Cleistanthus cunninghamii 5 6 3 Euphorbiaceae Drypetes deplanchei 4 5 6 7

193 Thesis Chapter Family Genus Species 4 5 6 Total number frugivore species Euphorbiaceae Flueggea leucopyrus 5 6 1 Euphorbiaceae Glochidion ferdinandi 4 5 6 12 Euphorbiaceae Glochidion sumatranum 4 5 6 4 Euphorbiaceae Macaranga tanarius 4 5 6 7 Euphorbiaceae Mallotus discolor 4 5 6 7 Euphorbiaceae Omalanthus nutans 4 5 6 15 Eupomatiaceae Eupomatia laurina 4 5 6 1 Eupomatiaceae Galbulimima belgraveana 4 5 6 4 Flacourtiaceae Caesaria multinervosa 4 5 6 2 Flacourtiaceae Scolopia braunii 4 5 6 6 Flacourticaeae Berberidopsis beckleri 5 6 1 Flagellariaceae Flagellaria indica 4 5 6 6 Grossulariaceae Polyosma cunninghamii 4 5 6 4 Icacinaceae Citronella moorei 4 5 6 4 Icacinaceae Pennantia cunninghamii 4 5 6 7 Lauraceae Beilschmedia elliptica 4 5 6 7 Lauraceae Beilschmedia obtusifolia 4 5 6 8 Lauraceae Cinnamomum oliveri 4 5 6 10 Lauraceae Cinnamomum virens 4 5 6 6 Lauraceae Cryptocarya bidwillii 4 5 6 6 Lauraceae Cryptocarya erythroxylon 4 5 6 4 Lauraceae Cryptocarya foetida 4 5 6 1 Lauraceae Cryptocarya glaucescens 4 5 6 6 Lauraceae Cryptocarya hypospodia 4 5 6 3 Lauraceae Cryptocarya macdonaldii 4 5 6 4 Lauraceae Cryptocarya microneura 4 5 6 4 Lauraceae Cryptocarya obovata 4 5 6 9 Lauraceae Cryptocarya rigida 4 5 6 1 Lauraceae Cryptocarya triplinervis 4 5 6 11 Lauraceae Endiandra discolor 4 5 6 5 Lauraceae Endiandra muelleri 4 5 6 4 Lauraceae Endiandra sieberi 4 5 6 2 Lauraceae Litsea australis 4 5 6 9 Lauraceae Litsea reticulata 4 5 6 11 Lauraceae Neolitsea australiensis 4 5 6 7 Lauraceae Neolitsea dealbata 4 5 6 10 Liliaceae Dianella caerulea 4 5 6 5 Loganiaceae Strychnos psilosperma 4 5 6 4 Meliaceae Anthocarapa nitidula 4 5 6 7 Meliaceae Dysoxylum fraserianum 4 5 6 7 Meliaceae Dysoxylum mollissimum 4 5 6 8 Meliaceae Dysoxylum rufum 4 5 6 3 Meliaceae Melia azedarach 4 5 6 18 Meliaceae Owenia cepiodora 4 5 6 2 Meliaceae Synoum glandulosum 4 5 6 5 Menispermaceae Hypserpa laurina 5 6 1 Menispermaceae Legnephora moorei 4 5 6 2 Menispermaceae Stephania japonica 4 5 6 2 Menispermaceae Tinospora smilacina 5 6 1 Mimosaceae Acacia aulacocarpa 5 6 2 Mimosaceae Acacia maidenii 4 5 6 7 Mimosaceae Acacia melanoxylon 4 5 6 8 Monimaceae Palmeria scandens 4 5 6 2 Monimiaceae Hedycarya angustifolia 4 5 6 4 Moraceae Ficus coronata 4 5 6 13 Moraceae Ficus fraseri 4 5 6 22 Moraceae Ficus macrophylla 4 5 6 27 Moraceae Ficus microcarpa 4 5 6 6 Moraceae Ficus obliqua 4 5 6 24 Moraceae Ficus platypoda 4 5 6 25 Moraceae Ficus rubiginosa 4 5 6 15 Moraceae Ficus superba 4 5 6 17 Moraceae Ficus virens 4 5 6 12 Moraceae Ficus watkinsiana 4 5 6 17 Moraceae Maclura cochinchinensis 4 5 6 7 Moraceae Streblus brunonianus 4 5 6 10 Moraceae Trophis scandens 4 5 6 3 Myoporaceae Myoporum insulare 4 5 6 2 Myrsinaceae Embelia australiana 4 5 6 2

194

Thesis Chapter Family Genus Species 4 5 6 Total number frugivore species Myrsinaceae Rapanea howittiana 4 5 6 4 Myrsinaceae Rapanea variabilis 4 5 6 2 Myrtaceae Acmena hemilampra 4 5 6 6 Myrtaceae Acmena ingens 4 5 6 10 Myrtaceae Acmena smithii 4 5 6 15 Myrtaceae Archirhodomyrtus beckleri 4 5 6 2 Myrtaceae Austromyrtus bidwillii 4 5 6 3 Myrtaceae Austromyrtus dulcis 5 6 1 Myrtaceae Austromyrtus hillii 4 5 6 2 Myrtaceae Decaspermum humile 4 5 6 7 Myrtaceae Pilidiostigma rhytispermum 4 5 6 1 Myrtaceae Rhodamnia argentea 4 5 6 8 Myrtaceae Rhodamnia rubescens 4 5 6 7 Myrtaceae Rhodomyrtus psidioides 4 5 6 6 Myrtaceae Syzgium luehmannii 4 5 6 2 Myrtaceae Syzygium australe 4 5 6 8 Myrtaceae Syzygium corynanthum 4 5 6 8 Myrtaceae Syzygium crebrinerve 4 5 6 6 Myrtaceae Syzygium francisii 4 5 6 1 Myrtaceae Syzygium johnsonii 4 5 6 1 Myrtaceae Syzygium oleosum 4 5 6 2 Myrtaceae Syzygium paniculatum 4 5 6 7 Oleaceae Chionanthus ramiflora 4 5 6 4 Oleaceae Jasminum dallachii 5 6 2 Oleaceae Jasminum didymum 4 5 6 4 Oleaceae Jasminum simplicifolium 4 5 6 1 Oleaceae Notelaea longifolia 4 5 6 4 Oleaceae Olea paniculata 4 5 6 11 Pandanaceae Freycinetia excelsa 5 6 1 Pandanaceae Freycinetia scandens 5 6 1 Philesiaceae Eustrephus latifolius 4 5 6 2 Philesiaceae Geitonoplesium cymosum 4 5 6 2 Piperaceae Piper novae-hollandiae 4 5 6 12 Pittosporaceae Citriobatus pauciflorus 4 5 6 1 Pittosporaceae Pittosporum rhombifolium 4 5 6 2 Pittosporaceae Pittosporum undulatum 4 5 6 10 Pittosporaceae Pittosporum venulosum 4 5 6 1 Podocarpaceae Podocarpus elatus 4 5 6 5 Polygonaceae Muehlenbeckia gracillima 4 5 6 1 Rhamnaceaae Alphitonia excelsa 4 5 6 18 Rhamnaceaae Alphitonia petriei 4 5 6 10 Rhamnaceae Emmenosperma alphitonioides 4 5 6 4 Rhamnanceae Rhamnella vitiensis 5 6 1 Rosaceae Rubus moluccanus 4 5 6 4 Rosaceae Rubus moorei 4 5 6 3 Rosaceae Rubus parvifolius 4 5 6 1 Rosaceae Rubus rosifolius 4 5 6 6 Rubiaceae Aidia racemosa 4 5 6 1 Rubiaceae Canthium coprosmoides 4 5 6 4 Rubiaceae Canthium odoratum 4 5 6 1 Rubiaceae Coprosma quadrifida 4 5 6 4 Rubiaceae Hodgkinsonia ovatiflora 4 5 6 4 Rubiaceae Ixora beckleri 4 5 6 3 Rubiaceae Morinda jasminoides 4 5 6 7 Rubiaceae Morinda umbellata 4 5 6 1 Rubiaceae Psychotria loniceroides 4 5 6 1 Rubiaceae Randia benthamianus 4 5 6 1 Rutaceae Acronychia laevis 4 5 6 5 Rutaceae Acronychia oblongifolia 4 5 6 11 Rutaceae Acronychia suberosa 4 5 6 2 Rutaceae Acronychia wilcoxiana 4 5 6 1 Rutaceae Halfordia kendack 4 5 6 5 Rutaceae Melicope elleryana 4 5 6 5 Rutaceae Melicope micrococca 4 5 6 6 Rutaceae Melicope vitiflora 5 6 1 Rutaceae Micromelum minutum 4 5 6 1 Rutaceae Sarcomelicope simplicifolia 4 5 6 2 Rutaceae Zanthoxylum brachyacanthum 4 5 6 1 Santalaceae Exocarpos cupressiformis 4 5 6 7

195

Thesis Chapter Family Genus Species 4 5 6 Total number frugivore species Santalaceae Santalum lanceolatum 5 6 1 Sapindaceae Alectryon coriaceus 4 5 6 2 Sapindaceae Alectryon subcinereus 4 5 6 2 Sapindaceae Alectryon tomentosus 4 5 6 2 Sapindaceae Arytera distylis 4 5 6 4 Sapindaceae Cupaniopsis anacardioides 4 5 6 5 Sapindaceae Cupaniopsis baileyana 5 6 1 Sapindaceae Cupaniopsis flagelliformus 5 6 1 Sapindaceae Cupaniopsis parvifolia 4 5 6 1 Sapindaceae Diploglottis australis 4 5 6 19 Sapindaceae Elattostachys xylocarpa 4 5 6 7 Sapindaceae Guioa acutifolia 4 5 6 1 Sapindaceae Guioa semiglauca 4 5 6 15 Sapindaceae Jagera pseudorhus 4 5 6 11 Sapindaceae Mischarytera lautereriana 4 5 6 4 Sapindaceae Mischocarpus anodontus 4 5 6 3 Sapindaceae Mischocarpus pyriformis 4 5 6 2 Sapindaceae Sarcopteryx stipata 4 5 6 7 Sapotaceae Planchonella queenslandica 4 5 6 2 Sapotaceae Pouteria australis 4 5 6 3 Sapotaceae Pouteria chartacea 4 5 6 1 Sapotaceae Pouteria myrsinoides 5 6 1 Smilacaceae Ripogonum album 4 5 6 2 Smilacaceae Smilax australis 4 5 6 2 Smilacaceae Smilax glycophylla 4 5 6 1 Solanaceae Duboisia myoporoides 4 5 6 5 Solanaceae Solanum aviculare 4 5 6 8 Solanaceae Solanum stelligerum 5 6 1 Sterculiaceae Brachychiton acerifolius 4 5 6 4 Sterculiaceae Brachychiton discolor 4 5 6 3 Surianaceae Guilfoylia monostylis 4 5 6 2 Symplocaceae Symplocos stawellii 4 5 6 3 Symplocaceae Symplocos thwaitesii 4 5 6 6 Thymelaecae Wikstroemia indica 4 5 6 4 Thymelaeceae Phaleria chermsideana 4 5 6 2 Ulmaceae Aphananthe philippinensis 4 5 6 15 Ulmaceae Trema tomentosa 4 5 6 7 Urticaceae Dendrocnide excelsa 4 5 6 10 Urticaceae Dendrocnide photinophylla 4 5 6 3 Urticaceae Pipturus argenteus 4 5 6 9 Verbenaceae Callicarpa pedunculata 4 5 6 3 Verbenaceae Clerodendrum floribundum 4 5 6 2 Verbenaceae Clerodendrum tomentosum 4 5 6 1 Verbenaceae Gmelina leichhardtii 4 5 6 2 Verbenaceae Vitex lignum-vitae 4 5 6 7 Vitaceae Cayratia clematidea 4 5 6 3 Vitaceae Cayratia eurynema 4 5 6 8 Vitaceae Cissus antarctica 4 5 6 11 Vitaceae Cissus hypoglauca 4 5 6 8 Vitaceae Cissus sterculiifolia 4 5 6 10 Vitaceae Tetrastigma nitens 4 5 6 8 Zingiberaceae Alpinia caerulea 4 5 6 4

Introduced from outside Australian subtropics Araliaceae Schefflera actinophylla 4 5 6 11 Arecaceae Archontophoenix alexandrae 4 5 6 5 Boraginaceae Cordia dichotoma 4 5 6 1

Introduced from outside Australia Anacardiaceae Schinus terebinthifolia 4 5 6 7 Arecaceae Syagrus romanzoffianum 5 6 1 Caesalpiniaceae Tamarindus indica 4 5 6 3 Lauraceae Cinnamomum camphora 4 5 6 26 Liliaceae Asparagus africanus 4 5 6 2 Liliaceae Asparagus densiflorus 5 6 2

196

Thesis Chapter Family Genus Species 4 5 6 Total number frugivore species Liliaceae Asparagus plumosus 4 5 6 2 Moraceae Ficus benjamima 4 5 6 11 Myrsinaceae Ardissia crenata 4 5 6 1 Myrtaceae Eugenia uniflora 4 5 6 2 Myrtaceae Psidium guajava 5 6 6 Ochnaceae Ochna serrulata 4 5 6 11 Oleaceae Ligustrum lucidum 4 5 6 15 Oleaceae Ligustrum sinense 4 5 6 13 Oleaceae Olea europea 4 5 6 3 Passifloraceae Passiflora suberosa 4 5 6 4 Phytolaccaceae Phytolacca americana 4 5 6 2 Phytolaccaceae Phytolacca octandra 4 5 6 18 Phytolaccaceae Rivina humilis 4 5 6 2 Rosaceae Duchesnea indica 4 5 6 4 Rosaceae Rosa rubiginosa 4 5 6 2 Rosaceae Rubus fructosus 4 5 6 8 Rubiaceae Coffea aribica 4 5 6 1 Rutaceae Murraya paniculata 4 5 6 2 Solanaceae Lycium ferocissimum 4 5 6 9 Solanaceae Physalis peruviana 4 5 6 5 Solanaceae Solanum americanum 4 5 6 9 Solanaceae Solanum capsicoides 5 6 1 Solanaceae Solanum erianthum 5 6 2 Solanaceae Solanum hispidum 4 5 6 2 Solanaceae Solanum mauritianum 4 5 6 21 Solanaceae Solanum nigrum 4 5 6 5 Solanaceae Solanum pseudocapsicum 5 6 1 Solanaceae Solanum seaforthianum 4 5 6 6 Ulmaceae Celtis sinensis 4 5 6 10 Verbenaceae Lantana camara 4 5 6 21 Vitaceae Vitis vinifera 4 5 6 11

197 Appendix 3

Associations between frugivore species’ responses to forest fragmentation as detected in this thesis and their ecological attributes are assessed (Table A3.1). This topic was introduced in Section 1.4.2 and the results are discussed in Section 7.1.2 in the main body of this thesis.

Body mass

Body mass among frugivorous species in subtropical Australia ranges from 9 g

(mistletoebird) to 2 300 g (Australian brush turkey) (average 263 g S.E. 59 g). There is no association between body size and fragmentation sensitivity (sensitivity scored as

Increaser = 1, Tolerant = 2, Decreaser = 3; Spearman’s Rank correlation coefficient =

0.01, p = 0.96). The body mass of the species that were sensitive to forest fragmentation ranged from 48 g to 485 g (average 199 g). Of the nine large-bodied species (> 400 g), seven were tolerant of fragmentation, one was a decreaser and one was an increaser (Table A3.1).

Biogeographical distribution

There is no clear association between species that are endemic to Australia and sensitivity to forest fragmentation. Only six of the frugivore species that were common enough in survey data to analyse for response to fragmentation are endemic to

Australia; one was a decreaser, one an increaser and four were tolerant (Table A3.1). Of the six frugivore species that were decreasers, only one (the green catbird) is endemic to

Australia; the remaining decreaser species had large geographical distributions.

198 Rarity

In the present study, a species’ relative abundance (or rarity) was indicated by its average abundance in extensive forest sites during surveys (Table A3.1). There is an association between a species’ abundance and its sensitivity to fragmentation

(sensitivity scored as Increaser = 1, Tolerant = 2, Decreaser = 3; Spearman’s Rank correlation coefficient = 0.37, p = 0.05). However, contrary to the expectation that less common species may be more susceptible to fragmentation (Henle et al., 2004), this analysis shows that decreaser species were typically common in extensive forest.

Patterns of resource specialisation

A frugivore species’ degree of habitat and dietary specialisation was determined from information on their degree of dependence on rainforest and fruit, respectively.

Consistent with the expectation that resource specialists are more fragmentation sensitive (see Section 1.4.2.2), decreaser species tended to be both rainforest and fruit specialists. Five out of six (83%) decreaser frugivore species were rainforest specialists, compared with (4 of 14) 29% of tolerant frugivore species and none of the five increaser species were rainforest specialists (Fisher’s exact test (decreaser versus tolerant and increaser species) χ2 = 8.68, p =0.008; Table 7.2). Similarly, 83% of decreaser frugivore species were fruit specialists, compared with 36% of tolerant frugivores and 14% of increasers (Fisher’s exact test (decreaser versus tolerant and increaser species) χ2 = 5.80, p =0.03; Table 7.2)

199 Table A3.1 Ecological attributes of vertebrate frugivore species in subtropical Australia in relation to their response to fragmentation (Frag. Resp.). Distribution shows the countries (other than Australia) in which the species occurs. Av. abundance shows the average abundance of each species during surveys for the present study. Forest type shows the characteristic habitat of each species. Diet shows the main food items. No. genera is the number of plant genera (including introduced taxa) known to be consumed by each species. Frag. Resp.1 Common name Body Distribution outside Australia3 Av. Forest Diet6 mass abund. type5 (g)2 Decreaser eastern tube-nosed fruit-bat 48 PNG 0.38 RF F brown cuckoo-dove† 240 Phil., Born., Suma., Moluc., Sulaw., PNG 4.80 MF F, S wompoo fruit-dove 465 PNG 5.25 RF F superb fruit-dove 110 Indo., Sulaw., Moluc., PNG 0.56 RF F rose-crowned fruit-dove 125 Indo. 3.38 RF F green catbird 207 Endemic (subtrop) 3.75 RF F, I, V Tolerant grey-headed flying-fox 700 Endemic MF N, F 0.50 black flying-fox 674 Endemic MF F, N Australian brush-turkey† 2300 Endemic 0.56 MF S, F, I, V white-headed pigeon† 420 Endemic 0.56 RF F, S emerald dove† 135 India, China, Indo-Malaya, Phil. PNG 0.38 MF S, F, I topknot pigeon 475 Endemic 1.06 RF F sulphur-crested cockatoo† 860 PNG 1.5 MF S, F, i scaly-breasted lorikeet† 75 Endemic 0.13 MF N, F, S, I Australian king-parrot† 243 Endemic 1.44 RF S, F, N pale-headed rosella† 110 Endemic 0.25 OF S, F, N, I common koel 245 Iran, Pakistan, India, China, Phil., Indo. 0.38 MF F channel-billed cuckoo 748 Indo., PNG 0.06 MF F, I, V little wattlebird 65 Endemic 0.00 OF N, F, I Lewin's honeyeater 34 Endemic 8.00 RF F, N, I pied currawong 398 Endemic 2.75 MF F, I, V satin bowerbird 201 Endemic 0.19 MF F, I, P

200 Frag. Resp.1 Common name Body Distribution outside Australia3 Av. Forest Diet6 mass abund. type5 (g)2 Increaser rock dove†* 308 all continents except Antarctic 0.00 MOD S, I bar-shouldered dove† 130 PNG 0.00 MF S rainbow lorikeet† 125 Indo., PNG 2.75 MF N, S, F, I black-faced cuckoo-shrike 134 South-east Asia, PNG, India 0.13 MF I, S, F figbird 128 Endemic 2.56 Mf, Mod F Australian magpie 299 PNG (Introd in NZ) 0.13 OF, Mod. I, S Torresian crow 499 PNG 1.44 OF, Mod V, I, P, F silvereye 11 Pacific Islands, NZ 1.06 MF N, I, F Undetermined wonga pigeon† 415 Endemic 0.50 MF S, F, I paradise riflebird 104* Endemic (subtrop.) 0.50 RF F, I regent bowerbird 102 Endemic (subtrop.) 0.13 RF F, I crimson rosella † 135 Endemic 0.31 MF S, F, N, I noisy friarbird 110 PNG 0.06 OF N, F, I, V blue-faced honeyeater 105 PNG 0.00 OF I, N, F barred cuckoo-shrike 70 PNG 0.00 MF F, I varied triller 34 PNG 0.06 RF I, F, S olive-backed oriole 96 PNG 0.00 MF I, F grey butcherbird 91 Endemic 0.00 OF, Mod I, V, P, S, F mistletoebird 9 Pacific Islands 0.19 MF F, I galah† 330 Endemic 0.00 OF S, F, I noisy miner 75 Endemic 0.00 OF I, N, F, V 1 Fragmentation response is the response pattern (Decreaser, Tolerant or Increaser) shown in this study (Chapter Two). Undetermined is species that were detected too infrequently during surveys to assign a fragmentation response. Bat responses were determined from a single summer survey. 2 Mass data was obtained from (Churchill, 1998) for bats and Baker et al. (1997) for birds, except the barred cuckoo-shrike (Dunning, 1993). * data for the paradise riflebird from one individual. and bats. 3 Distribution information was from the HANZAB series for birds and (Churchill, 1998) for bats ; ‘subtrop.’ shown in brackets after Endemic species indicates that the species is endemic to subtropical Australia; remaining endemic species are endemic to the Australian continent. Born. is Borneo, Indo. is Indonesia, Moluc. is Moluccas, PNG is Papua New Guinea, Phil. is Philippines, Suma. is Sumatra, Sulaw. is Sulawesi. 4 Bird species’ average abundance was determined from four surveys in each of the 16 sites. Bat abundance was surveyed once in each site.

201 5 Forest type(s): RF is rainforest, MF is rainforest as well as open forest and/or woodland, OF is open forest, G is grassland, MOD is modified. 6 Food is: F (fruit) S (seed) I (invertebrates) N (nectar) P (non-fruit/seed plant material) V (vertebrates). Listed in approximate order of relative dietary proportion. Food and forest type information from Blakers et al (1984), HANZAB series or Catterall et al. (2004) for birds and from Eby (1995), Richards and Hall (1998), Hall and Richards (2000) for bats.

202

References

Adam P (1992) Australian Rainforests. Oxford University Press, New York

Alcántara JM, Rey PJ, Valera F and Sánchez-Lafuente AM (2000) Factors shaping the

seedfall pattern of a bird-dispersed plant. Ecology 81, 1937-1950.

Alvarez-Bullya ER and Martínez-Ramos M (1990) Seed bank versus seed rain in the

regeneration of a tropical pioneer tree. Oecologia 84, 314-325.

Andrén H (1994) Effects of habitat fragmentation on birds and mammals in landscapes

with different proportions of suitable habitat. Oikos 71, 355-366.

Asquith NM, Terborgh J, Arnold AE and Riveros CM (1999) The fruits the agouti ate:

Hymenaea courabil seed fate when its disperser is absent. Journal of Tropical

Ecology 15, 229-235.

Aukema JE and Martínez del Rio C (2002) Where does a fruit-eating bird deposit

mistletoe seeds? Seed deposition patterns and an experiment. Ecology 83, 3489-

3496.

Babweteera F, Savill P and Brown N (2007) Balanites wilsoniana: regeneration with

and without elephants. Biological Conservation 134, 40-47.

Baider C and Florens FBV (2006) Current decline of the “Dodo tree”: A case of broken-

down interactions with extinct species or the result of new interactions with alien

invaders? pp. 199-214 in Laurance WF and Peres CA (eds) Emerging Threats to

Tropical Forests. University of Chicago Press, Chicago.

Bairlein F (1996) Fruit-eating in birds and its nutritional consequences. Comparative

Biochemistry and Physiology 113, 215-224.

Bairlein F (1998) The effect of diet composition on migratory fuelling in garden

warblers Sylvia borin. Journal of avian biology 29, 546-551.

203 Baker GB, Dettmann EB, Scotney BT, Hardy LJ and Drynan DAD (1997) Report on

the Australian Bird and Bat Banding Scheme, 1984-95. Australian Bird and Bat

Banding Scheme and Australian Nature Conservation Agency, Canberra.

Balmford A (1996) Extinction filters and current resilience: the significance of past

selection pressures for conservation biology. Trends in Ecology and Evolution

11, 193-196.

Barker RD and Vestjens WJM (1988) The Food of Australian Birds: Non-passerines.

CSIRO, Melbourne

Barker RD and Vestjens WJM (1989) The Food of Australian Birds: Passerines.

CSIRO, Melbourne

Belbin L, Griffith University, CSIRO and University of Queensland (2003) PATN for

Windows. Version 2.28.

Benítez-Malvido J and Martínez-Ramos M (2003) Impact of forest fragmentation on

understory plant species richness in Amazonia. Conservation Biology 17, 389-

400.

Bernays EA and Chapman RF (1994) Host-plant Selection in Phytophagous Insects.

Chapman and Hall, New York.

Bierregaard RO Jr., Lovejoy TE, Kapos V, dos Santos AA and Hutchins RW (1992)

The biological dynamics of tropical rainforest remnants. Bioscience 42, 859-866

Bierregaard RO Jr. and Stouffer PC (1997) Understory birds and dynamic habitat

mosaics in Amazonian rainforests. pp. 138-155 in Laurance WF and

Bierregaard RO Jr. (eds.) Tropical forest remnants: the ecology, conservation,

and management of fragmented communities. University of Chicago Press.

Blakers M, Davies SJJF and Reilly PN (1984) The Atlas of Australian Birds. Melbourne

University Press, Melbourne.

204 Bleher B and Böhning-Gaese K (2001) Consequences of frugivore diversity for seed

dispersal, seedling establishment and the spatial pattern of seedlings and trees.

Oecologia 129, 385-394.

Bleher B and Böhning-Gaese K (2006) The role of birds in seed dispersal and its

consequences for forest ecosystems. Acta Zoologica Sinica 52(Suppl.), 116-119.

Bollen A, Van Elsacker L and Ganzhorn JU (2004) Relations between fruits and

disperser assemblages in a Malagasy littoral forest: a community-level approach.

Journal of Tropical Ecology 20, 599-612.

Bond WJ (1994) Do mutualisms matter? Assessing the impact of pollinator and

disperser disruption on plant extinction. Philosophical Transactions of the Royal

Society of London-Biology Sciences, 344, 83–90.

Bosque C and Calchi R (2003) Food choice by blue-gray tanagers in relation to protein

content. Comparative Biochemistry and Physiology Part A. 135, 321-327.

Bowman DMJS (2000) Australian rainforests: islands of green in a land of fire.

Cambridge University Press, Cambridge.

Brash AR (1987) The history of avian extinction and forest conversion on Puerto Rico.

Biological Conservation 39, 97-111.

Bronstein JL and Hoffman K (1987) Spatial and temporal variation in frugivory at a

Neotropical fig, Ficus pertusa. Oikos 49, 261-268.

Brooks M, Pimm SL, Oyugi JO (1999) Time lag between deforestation and bird

extinction in tropical forest fragments Conservation Biology 13, 1140–1150.

Brown ED and Hopkins MJG (2002) Tests of disperser specificity between frugivorous

birds and rainforest fruits in New Guinea. Emu 102, 137-146.

Buckley YM, Anderson S, Catterall CP, Corlett RT, Engel T, Gosper CR, Nathan R,

Richardson DM, Setter M, Spiegel O, Vivian-Smith G, Voigt FA, Weir JS and

205 Westcott DA (2006) Management of plant invasions mediated by frugivore

interactions Journal of Applied Ecology 43, 848–857.

Burkey TV (1993) Edge effects in seed and egg predation at two Neotropical rainforest

sites. Biological Conservation 66, 139-143.

Burns KC (2006) A simple null model predicts fruit-frugivore interactions in a

temperate rainforest. Oikos 115, 427-432.

Butler DW (2003) Seed dispersal syndromes and the distribution of woody plants in

South-East Queensland's vine-forests. PhD Thesis. University of Queensland,

Brisbane.

Castelletta M, Sodhi NS and Subaraj R (2000) Heavy extinctions of forest avifauna in

Singapore: Lessons for biodiversity conservation in southeast Asia.

Conservation Biology 14, 1870-1880.

Catterall CP (1985) Winter energy deficits and the importance of fruit versus insects in

a tropical island bird population. Australian Journal of Ecology 10, 265-279.

Catterall CP and Kingston M (1993) Remnant bushland of south east Queensland in the

1990’s. Griffith University, Nathan Queensland, Australia.

Catterall CP, Kingston MB, Park K and Sewell S (1998) Deforestation, urbanisation and

seasonality: interacting effects on a regional bird assemblage. Biological

Conservation 84, 65-81.

Catterall CP, Kanowski J, Wardell-Johnson GW, Proctor H, Reis T, Harrison D and

Tucker NIJ (2004) Quantifying the biodiversity values of reforestation:

perspectives, design issues and outcomes in Australian rainforest landscapes. pp.

359-393 in Lunney D (ed.) Conservation of Australia’s Forest Fauna, 2nd

edition. Royal Zoological Society of NSW, Mossman.

Catterall CP, Storey RJ and Kingston MB (1997) Reality versus rhetoric: A case study

monitoring regional deforestation. pp. 367-377 in Hale P and Lamb D (eds)

206 Conservation Outside Nature Reserves. Centre for Conservation Biology,

University of Queensland, Brisbane.

Chapman CA and Onderdonk DA (1998) Forests without primates: Primate/plant

codependency. American Journal of Primatology 45, 127-141.

Chapman CA and Chapman LJ (1995) Survival without dispersers: seedling recruitment

under parent plants. Conservation Biology 9, 675-678.

Charles-Dominique P (1986) Inter-relationships between frugivorous vertebrates and

pioneer plants: Cecropia, birds and bats in French Guyana. pp. 119-135 in

Estrada A and Fleming TH (eds) Frugivores and Seed Dispersal. Dr.W. Junk,

Dordrecht.

Chesson PL and Warner RR (1981) Environmental Variability Promotes Coexistence in

Lottery Competitive Systems The American Naturalist 117, 923-943

Christiansen MB and Pitter E (1997) Species loss in a forest bird community near

Lagoa Santa in southern Brazil. Biological Conservation 80, 23-32.

Christidis L and Boles WE (1994) The Taxonomy and Species of Birds of Australia and

its Territories. Royal Australasian Ornithologists Union, Melbourne.

Church RJ (1997) Avian frugivory in a subtropical rainforest: Eleven years of

observations in Lamington National Park. The Sunbird 27, 85-91.

Churchill S (1998) Australian bats. Reed New Holland: Sydney.

Cipollini ML and Levey DJ (1997) Secondary metabolites of fleshy vertebrate-

dispersed fruits: Adaptive hypotheses and implications for seed dispersal.

American Naturalist 150, 346-372.

Clarke KR (1993) Non-parametric multivariate analyses of changes in community

structure. Australian Journal of Ecology 18, 117-143.

Clarke KR and Green RH (1988) Statistical design and analysis for a ‘biological effects’

study. Marine Ecology Progress Series 46, 213-226.

207 Clarke KR and Warwick RM (2001) Changes in Marine Communities: An Approach to

Statistical Analysis and Interpretation. Second Edition. PRIMER-E Ltd.,

Plymouth.

Cochrane MA, Alencar A, Schulze MD, Souza CM Jr., Nepstad DC, Lefebvre P and

Davidson EA (1999) Positive feedbacks in the fire dynamic of closed canopy

tropical forests. Science 284, 1832-1835.

Connell JH (1971) On the role of natural enemies in preventing competitive exclusion

in some marine animals and in rainforest trees. pp 298-313 in den Boer PJ and

Gradwell GR (eds.) Dynamics of Populations. Centre for Agricultural

Publishing, Wageningen, The Netherlands.

Cooper W and Cooper WT (1994) Fruits of the Rain Forest: A Guide to Fruits in

Australian Tropical Rain Forests. Geo Production, Chatswood.

Cordeiro NJ and Howe HF (2001) Low recruitment of trees dispersed by animals in

African forest fragments. Conservation Biology 15, 1733-1741.

Cordeiro NJ and Howe HF (2003) Forest fragmentation severs mutualism between seed

dispersers and an endemic African tree. Proceedings of the National Academy of

Science 100, 14052-14056.

Corlett RT (1996) Characteristics of vertebrate-dispersed fruits in Hong Kong. Journal

of Tropical Ecology 12, 819-833.

Corlett RT (1998) Frugivory and seed dispersal by vertebrates in the Oriental

(Indomalayan) Region. Biological Reviews 73, 413-448.

Corlett RT (2002) Frugivory and seed dispersal in degraded tropical east Asian

landscapes. pp. 451-465 in Levey DJ, Silva WR and Galetti M (eds) Seed

Dispersal and Frugivory: Ecology, Evolution and Conservation. CAB

International, Oxford.

208 Corlett RT and Ko WP (1995) Frugivory by koels in Hong Kong. Memoirs of the Hong

Kong Natural History Society 20, 221-222.

Corlett RT and Turner IT (1997) Long-term survival in tropical forest remnants in

Singapore and Hong Kong. pp. 333-345 in Laurance WF and Bierregaard RO

(eds) (1997) Tropical Forest Remnants: Ecology, Management and

Conservation of Fragmented Communities. University of Chicago Press,

Chicago.

Corlett RT and Primack RB (2006) Tropical rainforests and the need for cross-

continental comparisons. Trends in Ecology and Evolution 21, 104-110.

Cosson J, Pons J and Masson D (1999) Effects of forest fragmentation on frugivorous

and nectarivorous bats in French Guiana. Journal of Tropical Ecology 15, 515-

534.

Cox PA, Elmqvist T, Pierson ED and Rainey WE (1991) Flying foxes as strong

interactors in south Pacific island ecosystems: A conservation hypothesis.

Conservation Biology 5, 448-454.

Crome FHJ (1975) The ecology of fruit pigeons in tropical northern Queensland.

Australian Wildlife Research 2, 155-185.

Crome FHJ (1978) Foraging ecology of an assemblage of birds in lowland rainforest in

northern Queensland. Australian Journal of Ecology 3, 195-212.

Crome FHJ (1990) Vertebrates and succession. pp. 53-64 in Webb LJ and Kikkawa J

(eds) Australian Tropical Rainforests: Science – Value – Meaning. CSIRO,

Melbourne.

Crome FHJ, Isaacs J and Moore L (1994) The utility to birds and mammals of remnant

riparian vegetation and associated windbreaks in the tropical Queensland

uplands. Pacific Conservation Biology 1, 328-343.

209 Crome FHJ and Shields J (1992) Parrots and Pigeons of Australia. Collins Angus and

Robinson Publishers, Sydney.

Daily GC and Ehrlich PR (1994) Influence of social status on individual foraging and

community structure in a bird guild. Oecologia 100, 53-65.

Daily GC, Ehrlich PR and Sánchez-Azofeifa GA (2001) Countryside biogeography: use

of human-dominated habitats by the avifauna of southern Costa Rica. Ecological

Applications 11, 1-13.

Date EM, Ford HA and Recher HE (1991) Frugivorous pigeons, stepping stones and

weeds in northern New South Wales. pp. 241-244 in Saunders DA and Hobbs RJ

(eds) Nature Conservation 2: The Role of Corridors. Surrey Beatty and Sons,

Chipping Norton, New South Wales.

Date EM, Recher HF, Ford HA and Stewart DA (1996) The conservation and ecology

of rainforest pigeons in northeastern New South Wales. Pacific Conservation

Biology 2, 299-308.

Dennis AJ (1997) Musky rat-kangaroos, Hypsiprymnodon moschatus: cursorial

frugivores in Australia’s wet tropical rain-forests. PhD thesis, James Cook

University of North Queensland, Townsville, Australia

Dennis AJ and Westcott DA (2006) Reducing complexity when studying seed dispersal

at community scales: a functional classification of vertebrate seed dispersers in

tropical forests. Oecologia 149, 620-634.

Develey PF and Metzger JP (2006) Emerging threats to birds in Brazilian Atlantic

forests: the role of forest loss and configuration in a severely fragmented

ecosystem. pp. 269-90 in Laurance WF and Peres CA (eds) Emerging Threats to

Tropical Forests. University of Chicago Press, Chicago.

Dickman C and Fleming M (2002) Pest or passenger pigeon? The New South Wales

Scientific Committee’s assessment of the status of the P. poliocephalus. pp. 20-

210 28 in Eby P and Lunney D (eds) Managing the Grey-headed Flying-fox as a

Threatened Species in New South Wales. Royal Zoological Society of New

South Wales, Mossman.

Didham RK, Ghazoul J, Stork NE and Davis AJ (1996) Insects in fragmented forests: a

functional approach. Trends in Ecology and Evolution 11, 255-260.

Doak DF, Marino PC and Kareiva PM (1992) Spatial scale mediates the influence of

habitat fragmentation on dispersal success: Implications for conservation.

Theoretical Population Biology 41, 315-336.

Duncan RS and Chapman CA (1999) Seed dispersal and potential forest succession in

abandoned agriculture in tropical Africa. Ecological Applications 9, 998-1008.

Duncan RS and Chapman CA (2002) Limitations of animal seed dispersal for

enhancing forest succession on degraded lands. pp. 437-450 in Levey DJ, Silva

WR and Galetti M (eds) Seed Dispersal and Frugivory: Ecology, Evolution and

Conservation. CAB International, Oxford.

Dunning JB (1993) CRC Handbook of Avian Body Masses. CRC Press: Boca Raton,

Florida.

Eby P (1991a) Seasonal movements of Pteropus poliocephalus (Chiroptera:

Pteropidae), from two maternity camps in northern New South Wales. Wildlife

Research 18, 547-549.

Eby P (1991b) “Finger-winged night workers”: managing forests to conserve the role of

Grey-headed Flying-foxes as pollinators and seed dispersers. pp. 91-100 in

Lunney D (ed.) Conservation of Australia’s Forest Fauna. Royal Zoological

Society of New South Wales, Mossman.

Eby P (1995) The Biology and Management of Flying Foxes in New South Wales.

Species Management Report No. 18. New South Wales National Parks and

Wildlife Service, Hurstville.

211 Eby P (1998) An analysis of diet specialisation in frugivorous Pteropus poliocephalus

(Megachiroptera) in Australian subtropical rainforest. Australian Journal of

Ecology 23, 443-456.

Eby P, Collins L, Richards G and Parry-Jones K (1999) The distribution, abundance and

vulnerability to population reduction of a nomadic nectarivore, Pteropus

poliocephalus during a period of resource concentration. Australian Zoologist

31, 240-253.

Eby P and Lunney D (2002) Managing the Grey-headed Flying-fox as a threatened

species: a context for the debate. pp. 1-12 in Eby P and Lunney D (eds)

Managing the Grey-headed Flying-fox as a Threatened Species in New South

Wales. Royal Zoological Society of New South Wales, Mossman.

Edwards W (2005) Within- and between-species patterns of allocation to pulp and seed

in vertebrate dispersed plants. Oikos 110, 109-114.

Eshiamwata GW, Berens DG, Bleher B, Dean WRJ and Böhning-Gaese K (2006) Bird

assemblages in isolated Ficus trees in Kenyan farmland. Journal of Tropical

Ecology 22, 723-726.

Estrada A, Coates-Estrada R, Meritt D Jr, Monteil S and Curiel D (1993) Patterns of

frugivore species richness and abundance in forest islands and in agricultural

habitats at Los Tuxtlas, Mexico. Vegetatio 107/108, 245-257.

Fahrig L and Merriam G (1994) Conservation of fragmented populations. Conservation

Biology 8, 50-59.

Faith DP, Minchin PR and Belbin L (1987) Compositional dissimilarity as a robust

measure of ecological distance. Vegetatio 69, 57-68.

Farwig N, Böhning-Gaese K and Bleher B (2006) Enhanced seed dispersal of Prunus

africana in fragmented and disturbed forests? Oecologia 147, 238-252.

212 Fleming TH (1986) Opportunism versus specialization: the evolution of feeding

strategies in frugivorous bats. pp. 105-135 in Estrada A and Fleming TH (eds)

Frugivores and Seed Dispersal. Dr W. Junk, Dordrecht.

Fleming TH, Breitwisch R and Whitesides GH (1987) Patterns of tropical vertebrate

frugivore diversity. Annual Review of Ecology and Systematics 18, 91-109.

Floyd AG (1989) Rainforest Trees of Mainland South Eastern Australia. Inkata Press,

Sydney.

Floyd AG (1990) Australian rainforests in New South Wales. Volume 2. Surrey Beatty

and Sons, Sydney.

Forget P-M (1993) Post-dispersal predation and scatterhoarding of Dipteryx panamensis

(Papilionaceae) seeds by rodents in Panama. Oecologia 94, 255-261.

Franklin J and Rey SJ (2007) Spatial patterns of tropical forest trees in Western

Polynesia suggest recruitment limitations during secondary succession. Journal

of Tropical Ecology 23, 1-12.

Frawley K (1991) Past rainforest management in Queensland. pp. 85-106 in Werren G

and Kershaw P (eds) The Rainforest Legacy: Australian National Rainforest

Study Volume 3. Rainforest History, Dynamics and Management. Australian

Government Publishing Service, Canberra.

French K (1990) Evidence for frugivory by birds in montane and lowland forests in

south-east Australia. Emu 90, 185-189.

Frith C, Frith D, Barnes E and McGuire M (2004) The Bowerbirds. Ptilonorhynchidae.

Oxford University Press.

Frith HJ (1952) Notes on the pigeons of the Richmond River, NSW. Emu 52, 89-99.

Frith HJ (1957) Food habits of the topknot pigeon. Emu 57, 341-345.

Frith HJ (1982) Pigeons and Doves of Australia. Rigby, Adelaide

213 Frith HJ, Crome FHJ and Wolfe TO (1976) Food of fruit-pigeons in New Guinea. Emu

76, 49-58.

Fuentes M (1994) Diets of fruit-eating birds: what are the causes of interspecific

differences? Oecologia 97, 134-142

Fujita MS and Tuttle MD (1991) Flying foxes (Chiroptera: Pteropididae): threatened

animals of key ecological and economic importance. Conservation Biology 5,

455-463.

Galetti M (2001) The future of the Atlantic forest. Conservation Biology 15, 4.

Galindo-González J, Guevara S and Sosa VJ (2000) Bat- and bird-generated seed rains

at isolated trees in pastures in a tropical rainforest. Conservation Biology 14,

1693-1703.

Galindo-González J and Sosa VJ (2003) Frugivorous bats in isolated trees and riparian

vegetation associated with human-made pastures in a fragmented tropical

landscape. Southwestern Naturalist 48, 579-589.

Gascon C, Lovejoy TE, Bierregaard RO Jr, Malcolm JR, Stouffer PC, Vasconcelos HL,

Laurance WF, Zimmerman B, Tocher M and Borges S (1999) Matrix habitat and

species richness in tropical forest remnants. Biological Conservation 91, 223-

229.

Gautier-Hion A, Duplantier JM, Quiris R, Feer F, Decoux JP, Dubost GF, Emmons L,

Erard C, Hecketsweiler P and Moungazi A (1985) Fruit characters as a basis of

fruit choice and seed dispersal in a tropical forest vertebrate community.

Oecologia 55, 324-333.

Gilmore S (1999) Fauna and rainforest fragmentation – developing improved

conservation planning. pp. 29-66 in Horton S (ed.) Rainforest Remnants: A

Decade of Growth. New South Wales National Parks and Wildlife Service,

Hurstville.

214 Goosem S (2000) Landscape ecology and values. pp. 14-30 in McDonald G and Lane

M (eds.) Securing the Wet Tropics? A retrospective on managing Australia’s

tropical rainforests. Federation Press, Sydney.

Gorchov DL, Cornejo F, Ascorra C. and Jaramillo M (1995) Dietary overlap between

frugivorous birds and bats in the Peruvian Amazon. Oikos 74, 235-250.

Gosper DG and Holmes G (2002) Status of birds in the Richmond River District, New

South Wales 1973-2000. Corella 26, 89-105.

Gould J (1865) Handbook to the Birds of Australia. J. Gould, London.

Graham CH (2001) Factors influencing movement patterns of keel-billed toucans in a

fragmented tropical landscape in southern Mexico. Conservation Biology 15,

1789-1798.

Graham CH and Blake JG (2001) Influence of patch- and landscape-level factors on

bird assemblages in a fragmented tropical landscape. Ecological Applications

11, 1709-1721.

Graham C, Martínez-Leyva JE and Cruz-Paredes L (2002) Use of fruiting trees by birds

in continuous forest and riparian forest remnants in Los Tuxtlas, Veracruz,

Mexico. Biotropica 34, 589-597.

Green RJ (1993) Avian seed dispersal in and near subtropical rainforests. Wildlife

Research 20, 535-557.

Green RJ (1995) Using frugivores for regeneration: a survey of knowledge and

problems in Australia. pp.1-11 in Bissonette JA and Krausman PR (eds)

Integrating People and Wildlife for a Sustainable Future. The Wildlife Society,

Maryland.

Greenberg R, Foster M and Marquez L (1995) The role of white-eyed vireos in the

dispersal of Bursera simaruba fruit. Journal of Tropical Ecology 11, 619-639.

215 Grubb PJ (1977) The maintenance of species richness in plant communities: The

importance of the regeneration niche. Biological Reviews 52, 107-145.

Guevara S, Purata S and van der Maarel E (1986) The role of remnant trees in tropical

secondary succession. Vegetatio 66, 74-84.

Guevara S and Laborde J (1993) Monitoring seed dispersal at isolated standing trees in

tropical pastures: consequences for local species availability. Vegetatio 107/108,

319-338.

Hall LS and Richards G (1979) Bats of Eastern Australia. Queensland Museum Booklet

No.12. Queensland Museum, Brisbane.

Hall LS and Richards G (2000) Flying Foxes: Fruit and Blossom Bats of Australia.

University of New South Wales Press, Sydney.

Hall LS, Richards GC and Spencer HJ (1995) Eastern Tube-nose Bat Nyctimene

robinsoni. pp. 426-8 in Strahan R (ed.) The Mammals of Australia. Australian

Museum and Reed, Sydney.

Hamann A and Curio E (1999) Interactions among frugivores and fleshy fruit trees in a

Philippine submontane rainforest. Conservation Biology 13, 766-773.

Hamilton MB (1999) Tropical tree gene flow and seed dispersal. Nature 401, 129-130.

Harms KE, Wright SJ, Calderón O, Hernández A and Herre EA (2000) Pervasive

density-dependent recruitment enhances seedling diversity in a tropical forest.

Nature 404, 493-495.

Harrington GN, Irvine AK, Crome FHJ and Moore LA (1997) Regeneration of large-

seeded trees in Australian rainforest fragments: a study of higher-order

interactions. pp. 292-303 in Laurance WF and Bierregaard RO (eds) Tropical

Forest Remnants: Ecology, Management and Conservation of Fragmented

Communities. University of Chicago Press, Chicago.

216 Harvey CA (2000) Windbreaks enhance seed dispersal into agricultural landscapes in

Monteverde, Costa Rica. Ecological Applications 10, 155-173.

Hauser J. and Blok J (1998) Fragments of Green: An Identification Field Guide for

Rainforest Plants of the Greater Brisbane Region to the Border Ranges. 2nd

Edition. Australian Rainforest Conservation Society, Bardon.

Henle K, Davies KF, Kleyer M, Margules C and Settele J (2004) Predictors of species

sensitivity to fragmentation. Biodiversity and Conservation 13, 207-251.

Herrera CM (1981) Are tropical fruits more rewarding than temperate ones? American

Naturalist 118, 896-907.

Herrera CM (1982) Seasonal variation in the quality of fruits and diffuse coevolution

between plants and avian dispersers. Ecology 63, 773-785.

Herrera CM (1984) A study of avian frugivores, bird-dispersed plants, and their

interaction in Mediterranean scrublands. Ecological Monographs 54, 1-23.

Herrera CM (1985) Determinants of plant-animal coevolution: the case of mutualistic

dispersal of seeds by vertebrates. Oikos 44, 132-141.

Herrera CM (1987) Vertebrate-dispersed plants of the Iberian Peninsula: A study of

fruit characteristics. Ecological Monographs 57, 305-331.

Herrera CM (1998) Long-term dynamics of Mediterranean frugivorous birds and fleshy

fruits: A 12-year study. Ecological Monographs 68, 511-538.

Herrera CM (2002) Seed dispersal by vertebrates. pp. 185–208 in Herrera CM and

Pellmyr O (eds.) Plant–animal interactions: an evolutionary approach),

Blackwell Publishing, Oxford.

Herrera CM, Jordano P, López-Soria L and Amat JA (1994) Recruitment of a mast-

fruiting, bird-dispersed tree: bridging frugivore activity and seedling

establishment. Ecological Monographs 64, 315-344.

217 Higgins PJ (1999) Handbook of Australian, New Zealand and Antarctic Birds: Volume

4. Parrots to Dollarbirds. Oxford University Press, Melbourne.

Higgins PJ and Davies SJJF (1996) Handbook of Australian, New Zealand and

Antarctic Birds: Volume 3. Snipe to Pigeons. Oxford University Press,

Melbourne.

Higgins PJ, Peter JM and Steele WK (2001) Handbook of Australian, New Zealand and

Antarctic Birds: Volume 5. Tyrant-flycatchers to Chats. Oxford University

Press, Melbourne.

Holbrook KM, Smith TB and Hardesty BD (2002) Implications of long-distance

movements of frugivorous rain forest hornbills. Ecography 25, 745-749.

Holl KD, Loik ME, Lin EHV and Samuels IA (2000) Tropical montane forest

restoration in Costa Rica: overcoming barriers to dispersal and establishment.

Restoration Ecology 8, 339-349.

Hood GM (2003) Pop Tools Version 2.5.9. URL http://www.cse.csiro.au/poptools

Hooper Du, Chapin FS III, Ewel JJ, Hector A, Inchausti P, Lavorel S, Lawton JH,

Lodge DM, Loreau M, Naeem S, Schmid B, Setala H, Symstad AJ, Vandermeer

J and Wardle DA (2005) Effects of biodiversity on ecosystem functioning: a

consensus of current knowledge. Ecological Monographs 75, 3-35.

Hooper ER, Legendre P and Condit R (2004) Factors affecting community composition

of forest regeneration in deforested, abandoned land in Panama. Ecology 85,

3313-3326.

Hopkins MS and Graham AW (1984) Viable soil seed banks in disturbed lowland

tropical rainforest sites in North Queensland. Australian Journal of Ecology 9,

71-79.

Howe HF (1977) Bird activity and seed dispersal of a tropical wet forest tree. Ecology

58, 539-550.

218 Howe HF (1979) Fear and frugivory. American Naturalist 114, 925-931.

Howe HF (1981) Dispersal of a neotropical nutmeg (Virola sebifera) by birds. Auk 98,

88-98.

Howe HF (1984) Implications of seed dispersal by animals for tropical reserve

management. Biological Conservation 30, 261-281.

Howe HF (1986) Seed dispersal by fruit eating birds and mammals. pp. 123-190 in

Murray DR (ed.) Seed dispersal. Academic Press, Sydney.

Howe HF (1993) Specialized and generalized dispersal systems: where does ‘the

paradigm’ stand? Vegetatio 107/108, 3-13.

Howe HF and Primack RB (1975) Differential Seed Dispersal by Birds of the Tree

Casearia nitida (Flacourtiaceae) Biotropica, 7, 278-283.

Howe HF and Estabrook GF (1977) On intraspecific competition for avian dispersers in

tropical trees. American Naturalist 111, 817-832.

Howe HF and van der Kerckhove GA (1980) Nutmeg dispersal by birds. Science 210,

925-927.

Howe HF and Smallwood J (1982) Ecology of seed dispersal. Annual Review of

Ecology and Systematics 13, 210-228.

Howe HF, Schupp EW and Westley LC (1985) Early consequences of seed dispersal for

a neotropical tree (Virola surinamensis). Ecology 66, 781-791.

Howe HF and Miriti MN (2000) No question: seed dispersal matters. Trends in Ecology

and Evolution 15, 434-436.

Howe RW, Howe TD and Ford HA (1981) Bird distributions on small rainforest

remnants in New South Wales. Australian Wildlife Research 8, 637-651.

Hubbell SP (1979) Tree dispersion, abundance and diversity in a tropical dry forest.

Science 203, 1299-1309.

219 Hubbell SP and Foster RB (1990) The fate of juvenile trees in a Neotropical forest:

implications for the natural maintenance of tropical tree diversity. pp. 317-341 in

Bawa KS and Hadley M (eds.) Reproductive Ecology of Tropical Forest Plants.

Parthenon Publishing, Paris.

Hunter J (1999) Fragmentation: the consequences. pp. 19-28 in Horton S (ed.)

Rainforest Remnants: A Decade of Growth. New South Wales National Parks

and Wildlife Service, Hurstville.

Hurtt GC and Pacala GC (1995) The consequences of recruitment limitation:

Reconciling chance, history and competitive differences between plants. Journal

of Theoretical Biology 176, 1-12.

Ingle NI (2003) Seed dispersal by wind, birds, and bats between Philippine montane

rainforest and successional vegetation. Oecologia 134, 251-261.

Innis GJ (1989) Feeding ecology of fruit pigeons in subtropical rainforests of south-

eastern Queensland. Australian Wildlife Research 16, 365-394.

Innis GJ and McEvoy J (1992) Feeding ecology of green catbirds (Ailuroedus

crassirostris) in subtropical rainforests of south-eastern Queensland. Wildlife

Research 19, 317-329.

Izhaki I and Safriel UN (1989) Why are there so few exclusively frugivorous birds?

Experiments on fruit digestibility. Oikos 54, 23-32.

Izhaki I, Tsahar E, Paluy I and Friedman J (2002) Within population variation and

interrelationships between morphology, nutritional content, and secondary

compounds of Rhamnus alaternus fruits. New Phytologist 156, 217-223.

Jansen A (2005) Avian use of restoration plantings along a creek linking rainforest

patches on the Atherton Tablelands, North Queensland. Restoration Ecology 13,

275-283.

220 Janzen DH (1970) Herbivores and the number of tree species in tropical forests.

American Naturalist 104, 501-528.

Janzen DH (1979) How to be a fig. Annual Review of Ecology and Systematics 10, 13-

51.

Janzen DH (1983) No park is an island: increase in interference from outside as park

size decreases. Oikos , 402-410.

Janzen DH (1986) The Eternal External Threat. pp. 286-303 in Soulé ME (ed.)

Conservation Biology: Science of Scarcity and Diversity. Sinauer Associates

Inc., Massachusetts.

Janzen DH and Martin PS (1982) Neotropical anachronisms: the fruits the

Gomphotheres ate. Science 215, 19-27.

Janzen DH and Vásquez-Yanez C (1991) Aspects of tropical seed ecology of relevance

to management of tropical forest wildlands. pp. 137-158 in Gomez-Pompa A,

Whitmore TC and Hadley M (eds.) Rainforest Regeneration and Management.

Man and the Biosphere Series, Volume 6. UNESCO & The Parthenon

Publishing Group, Paris.

Johnson RA, Willson MF, Thompson JN and Bertin RI (1985) Nutritional values of

wild fruits and consumption by migrant frugivorous birds. Ecology 66, 819-827.

Jones RE and Crome FHJ (1990) The biological web – plant/animal interactions in the

rainforest. pp. 74-87 in Webb LJ and Kikkawa J (eds.) Australian Tropical

Rainforests: Science – Value – Meaning. CSIRO, Melbourne.

Jordano P (1987) Patterns of mutualistic interactions in pollination and seed dispersal:

connectance, dependence asymmetries, and coevolution. American Naturalist

129, 657-677.

221 Jordano P (1995) Angiosperm fleshy fruits and seed dispersers: a comparative analysis

of adaptation and constraints in plant-animal interactions. American Naturalist

145, 163-191.

Jordano P and Schupp EW (2000) Determinants of seed dispersal effectiveness: the

quantity component of the Prunus mahaleb –frugivorous bird interaction.

Ecological Monographs 70, 591-615.

Jordano P and Godoy JA (2002) Frugivore-generated seed shadows: a landscape view

of demographic and genetic effects. pp. 305-321 in Levey DJ, Silva WR and

Galetti M (eds) Seed Dispersal and Frugivory: Ecology, Evolution and

Conservation. CAB International, Oxford.

Kanowski J, Catterall CP, Dennis AJ and Westcott DA (2004a) Animal-Plant

Interactions In Rainforest Conservation And Restoration. Rainforest CRC,

Cairns.

Kanowski J, Catterall CP, Reis T and Wardell-Johnson GW (2004b) Animal-plant

interactions in rainforest restoration in tropical and subtropical Australia. pp. 20

–24 in Kanowski J, Catterall CP, Dennis AJ and Westcott DA (eds) Animal-

Plant Interactions In Rainforest Conservation And Restoration. Rainforest CRC,

Cairns.

Karr JR (1976) Seasonality, resource availability, and community diversity in tropical

bird communities. American Naturalist 110, 973-994.

Karr JR (1982) Avian extinction on Barro Colorado Island, Panama: A reassessment.

American Naturalist 119, 220-239.

Kattan GH, Alvarez-López H and Giraldo M (1994) Forest fragmentation and bird

extinctions: San Antonio eighty years later. Conservation Biology 8, 138-146.

Kershaw AP, Sluiter IR, Mason JM, Wagstaff BE and Whitelaw M (1991) The history

of rainforest in Australia – evidence from pollen. pp. 1-15 in Werren G and

222 Kershaw P (Eds.) The rainforest legacy. Australian National Rainforest Study.

Volume 3. Australian Government Publishing Service, Canberra.

Kitamura S, Yumoto T, Poonswad P, Chuailua P, Plongmai K, Maruhashi T and Noma

N (2002) Interactions between fleshy fruits and frugivores in a tropical seasonal

forest in Thailand. Oecologia 133, 559-572

Kitamura S, Suzuki S, Yumoto T, Chuailua P, Plongmai K, Poonswad P, Noma N,

Maruhashi T and Suckasam C (2005) A botanical inventory of a tropical

seasonal forest in Khao Yai National Park, Thailand: implications for fruit-

frugivore interactions. Biodiversity and Conservation 14, 1241-1262.

Kooyman RM (1996) Growing Rainforest: Rainforest Restoration and Regeneration.

Recommendations for the Humid Sub-tropical Region of Northern New South

Wales and South East Queensland. State Forests of New South Wales, Casino

and Greening Australia, Brisbane.

Larson DL (1996) Seed dispersal by specialist versus generalist foragers: the plant’s

perspective. Oikos 76, 113-120.

Laurance WF (1990) Comparative responses of five arboreal marsupials to tropical

forest fragmentation. Journal of Mammalogy 71, 641-653.

Laurance WF (1991) Edge effects in tropical forest fragments: application of a model

for the design of nature reserves. Biological Conservation 57, 205-219.

Laurance WF (2006) Have we overstated the tropical biodiversity crisis? Trends in

Ecology and Evolution 22, 65-70.

Laurance WF and Bierregaard RO (eds) (1997) Tropical Forest Remnants: Ecology,

Management and Conservation of Fragmented Communities. University of

Chicago Press, Chicago.

Laurance WF, Gordon CF and Perry E (1996) Structure of breeding bird communities

in rainforest and regrowth forest in tropical Queensland. Sunbird 26, 1-15.

223 Laurance WF and Peres CA (eds) (2006) Emerging Threats to Tropical Forests.

University of Chicago Press, Chicago.

Laurance WF and Yensen E (1991) Predicting the impacts of edge effects in fragmented

habitats. Biological Conservation 55, 77-92.

Law BS and Spencer HJ (1995) Common blossom bat Syconicteris australis. pp. 423-

425 in Strahan, R. (ed.) The Mammals of Australia. Australian Museum and

Reed, Sydney.

Lawrence D (2004) Erosion of tree diversity during 200 years of shifting cultivation in

Bornean rain forest. Ecological Applications 14, 1855-1869.

Leck CF (1979) Avian extinctions in an isolated tropical wet-forest preserve, Ecuador.

The Auk 96, 343-352.

Leighton M and Leighton DR (1983) Vertebrate responses to fruiting seasonality within

a Bornean rain forest. pp. 181-196 in Sutton SL, Whitmore TC and Chadwick

AC (eds.) Tropical Rainforest Ecology and Management. Blackwell Scientific

Publications, Oxford.

Levey DJ (1987) Seed size and fruit-handling techniques of avian frugivores. American

Naturalist 129, 471-485.

Levey DJ and Grajal A (1991) Evolutionary implications of fruit-processing limitations

in cedar waxwings. American Naturalist 138, 171-189.

Levey DJ and Karasov WH (1989) Digestive responses of temperate birds switched to

fruit and insect diets. Auk 106, 675-686.

Levey DJ and Sargent S (2000) A simple method for tracking vertebrate-dispersed

seeds. Ecology 81, 267-274.

Levey DJ and Martínez del Rio C (2001) It takes guts (and more) to eat fruit: lessons

from avian nutritional ecology. Auk 118, 819-831.

224 Levin SA, Muller-Landau HC, Nathan R and Chave J (2003) The ecology and evolution

of seed dispersal: a theoretical perspective. Annual Review of Ecology and

Systematics 34, 575-604.

Lindsey TR (1995) Encyclopaedia of Australian Birds. Angus & Robertson, Pymble.

Loiselle BA and Blake JG (2002) Potential consequences of extinction of frugivorous

birds for shrubs of a tropical wet forest. pp. 397-406 in Levey DJ, Silva WR and

Galetti M (eds.) Seed Dispersal and Frugivory: Ecology, Evolution and

Conservation. CAB International, Oxford.

Lord JM, Markey AS and Marshall J (2002) Have frugivores influenced the evolution

of fruit traits in New Zealand? pp. 55-68 in Levey DJ, Silva WR and Galetti M

(eds.) Seed Dispersal and Frugivory: Ecology, Evolution and Conservation.

CAB International, Oxford.

Lovejoy TE, Bierregaard RO Jr, Rylands AB, Malcolm JR, Quintela CE, Harper LH,

Brown KS Jr, Powell AH, Powell GVN, Schubart HOR and Hays MB (1986)

Edge and other effects of isolation on Amazon forest fragments. pp. 257-85 in

Soulé ME (ed.) Conservation Biology: Science of Scarcity and Diversity.

Sinauer Associates Inc., Massachusetts.

Luck GW and Daily GC (2003) Tropical countryside bird assemblages: Richness,

composition, and foraging differ by landscape context. Ecological Applications

13, 235-247.

Lunney D and Moon C (1997) Flying foxes and their camps in the remnant rainforests

of north-east New South Wales. pp. 247-77 in Dargavel J (ed.) Australia’s Ever-

changing Forests III. Centre for Resource and Environmental Studies,

Australian National University, Canberra.

Mack AL (1993) The sizes of vertebrate-dispersed fruits: a Neotropical-Paleotropical

comparison. American Naturalist 142, 840-856.

225 Manly BFJ (1994) Multivariate Statistical Methods: A Primer. Chapman and Hall,

London.

Marchant S and Higgins PJ (1993) Handbook of Australian, New Zealand and Antarctic

Birds: Volume 2. Raptors to Lapwings. Oxford University Press, Melbourne.

Marinho-Filho J and Sazima I (1998) Brazilian bats and conservation biology: a first

survey. pp. 282-294 in Kunz TH and Racey PA (eds.) Bat Biology and

Conservation. Sminthsonian Institute Press, Washington D.C.

Markus N and Hall LS (2004) Foraging behaviour of the Black Flying-fox (Pteropus

alecto) in the urban landscape of Brisbane, Queensland. Wildlife Research 31,

345-355.

Martin TE (1985) Resource selection by tropical frugivorous birds: integrating multiple

interactions. Oecologia 66, 563-573.

Martin, H.A. (1990). Tertiary climate and phytogeography in southeastern Australia.

Review of Palaeobotany and Palynology 65, 47-55.

Martínez del Rio C, Stevens BR, Daneke D and Andreadis PT (1988) Physiological

correlates of preference and aversion for sugars in three species of birds.

Physiological Zoology 61, 222-229.

Martínez del Rio C and Restrepo C (1993) Ecological and behavioural consequences of

digestion in frugivorous animals. Vegetatio 197/108:205-216.

Martínez del Rio C and Karasov WH (1990) Digestion strategies in nectar- and fruit-

eating birds and the sugar composition of plant rewards. American Naturalist

136, 618-637.

Martínez-Garza C and Howe HF (2003) Restoring tropical diversity: beating the time

tax on species loss. Journal of Applied Ecology 40, 423-429.

Matlack GR (2005) Slow plants in a fast forest: local dispersal as a predictor of species

frequencies in a dynamic landscape. Journal of Ecology 93, 50-59.

226 McConkey KR and Drake DR (2002) Extinct pigeons and declining bat populations: are

large seeds still being dispersed in the tropical Pacific? pp. 381-395 in Levey DJ,

Silva WR and Galetti M (eds) Seed Dispersal and Frugivory: Ecology,

Evolution and Conservation. CAB International, Oxford.

McDonald-Madden E, Schreiber ESG, Forsyth DM, Choquenot D and Clancy TF

(2005) Factors affecting Grey-headed Flying-fox (Pteropus poliocephalus:

Pteropidae) foraging in the Melbourne metropolitan area, Australia. Austral

Ecology 30: 600-608.

McDonnell MJ and Stiles EW (1983) The structural complexity of old field vegetation

and the recruitment of bird-dispersed plant species. Oecologia 56, 109-116.

McEuan AB and Curran LM (2004) Seed dispersal and recruitment limitation across

spatial scales in temperate forest fragments. Ecology 85, 507-518.

McKey D (1975) The ecology of coevolved seed dispersal systems. pp. 159-191 in

Gilbert E and Raven PH (eds.) Coevolution of Animals and Plants. University of

Texas, Austin.

Medellín RA and Gaona O (1999) Seed dispersal by bats and birds in forest and

disturbed habitats of Chiapas, México. Biotropica 31, 478-485.

Meehan HJ, McConkey KR and Drake DR (2005) Early fate of Myristica hypargyraea

seeds dispersed by Ducula pacifica in Tonga, Western Polynesia. Austral

Ecology 30, 374-382.

Meier L and Figgis P (1985) Rainforests of Australia. Ure Smith Press, Sydney.

Mildenstein TL, Steir SC, Nuevo-Diego CE and Scott Mills L (2005) Habitat selection

of endangered and endemic large flying-foxes in Subic Bay, Philipines.

Biological Conservation 126, 93-102.

Milledge DR (1987) Notes on the occurrence of the Queensland Tube-nosed bat

Nyctimene robinsoni in north-eastern New South Wales. Macroderma 3, 28-29.

227 Moermond TC, Denslow JS (1983) Fruit choice in Neotropical birds: effect of fruit type

and accessibility on selectivity. Journal of Animal Ecology 52, 407-420

Moermond TC, Denslow JS (1985) Neotropical frugivores: patterns of behaviour,

morphology and nutrition with consequences for fruit selection. Ornithological

Monographs 36, 865-897

Moermond TC, Denslow JS, Levey DJ and Santana E (1986) The influence of

morphology on fruit choice in neotropical birds. pp. 137-146 in Estrada A and

Fleming TH (eds.) Frugivores and Seed Dispersal. Dr.W. Junk Publishers,

Dordrecht.

Morton ES (1973) On the evolutionary advantages and disadvantages of fruit eating in

tropical birds. American Naturalist 107, 8-22.

Muller-Landau HC, Wright SJ, Calderón O, Hubbell SP and Foster RB (2002)

Assessing recruitment limitation: Concepts, methods and case-studies from a

tropical forest. pp. 35-53 in Levey DJ, Silva WR and Galetti M (eds.) Seed

Dispersal and Frugivory: Ecology, Evolution and Conservation. CAB

International, Oxford.

Murray KG, Winnett-Murray K, Cromie EA, Minor M. and Meyers E (1993) The

influence of seed packaging and fruit colour on feeding preferences of American

robins Vegetatio 107/108, 217-226.

Murray KG and Garcia-C JM (2002) Contributions of seed dispersal and demography to

recruitment limitation in a Costa Rican cloud forest. pp. 323-338 in Levey DJ,

Silva WR and Galetti M (eds.) Seed Dispersal and Frugivory: Ecology,

Evolution and Conservation. CAB International, Oxford.

Myers N (1984) The Primary Source: Tropical Forests and our Future. W W Norton &

Co Ltd, New York.

228 Nathan R and Muller-Landau C (2000) Spatial patterns of seed dispersal, their

determinants and consequences for recruitment. Trends in Ecology and

Evolution 15, 278-285.

National Land and Water Resources Audit (2001) Australian Native Vegetation

Assessment. Commonwealth of Australia, Canberra.

http://audit.ea.gov.au/ANRA/vegetation/docs/native_vegetation/nat_veg_fact01.cfm

Neilan W, Catterall CP, Kanowski J and McKenna S (2006) Do frugivorous birds assist

rainforest succession in weed dominated oldfield regrowth of subtropical

Australia? Biological Conservation 129, 393-407.

Nelson JE (1965) Movements of Australian flying foxes. Australian Journal of Zoology

13: 53-73.

Nepstad DC, Uhl C, Pereira CA and Silva JMC (1996) A comparative study of tree

establishment in abandoned pasture and mature forest of eastern Amazonia.

Oikos 76:25-39.

Nix HA and Switzer MA (1991) (Eds.) Rainforest Animals: An atlas of vertebrates

endemic to the Australian Wet Tropics. Australian National Parks and Wildlife

Service, Canberra.

Orrock JL, Levey DJ, Danielson BJ and Damschen EI (2006) Seed predation, not seed

dispersal, explains the landscape-level abundance of an early-successional plant.

Journal of Ecology 94, 838-845.

Palmer C and Woinarski JCZ (1999) Seasonal roosts and foraging movements of the

Black Flying-fox (Pteropus alecto) in the Northern Territory: resource tracking

in a landscape mosaic. Wildlife Research 26, 823-838.

Palmer C, Price O and Bach C (2000) Foraging ecology of the black flying fox

(Pteropus alecto) in the seasonal tropics of the Northern Territory, Australia.

Wildlife Research 27, 169-178.

229 Parry-Jones KA and Augee ML (1991) Food selection by Grey-headed Flying-foxes

(Pteropus poliocephalus) occupying a summer colony site near Gosford, New

South Wales. Wildlife Research 18, 111-124.

Parry-Jones KA and Augee ML (1992) Movements of Grey-headed flying foxes

(Pteropus poliocephalus) to and from a colony site on the central coast of New

South Wales. Wildlife Research 19, 331-340.

Patterson BD (1987) The principle of nested subsets and its implications for biological

conservation. Conservation Biology 1, 323-334

Pimm SL, Jones HL and Diamond J (1988) On the risk of extinction. American

Naturalist 132, 757-785.

Piper SD and Catterall CP (2003). A particular case and a general pattern:

hyperaggressive behaviour by one species may mediate avifaunal decreases in

fragmented Australian forests. Oikos 101, 602-614.

Pizo MA (1997) Seed dispersal and predation in two populations of Cabralea canjerana

(Meliaceae) in the Atlantic Forest of southeastern Brazil. Journal of Tropical

Ecology 13, 559-577.

Poschlod P, Bakker J, Bonn S, Fischer S (1996) Dispersal of plants in fragmented

landscapes. pp. 123-127 in Settele J, Margules CR, Poschlod P and Henle K

(eds.) Species Survival in Fragmented Landscapes. Kluwer Academic

Publishers, Netherlands.

Poulsen JR, Clarke CJ, Connor EF and Smith TB (2002) Differential resource use by

primates and hornbills: implications for seed dispersal. Ecology 83, 228-240.

Pratt TK and Stiles EW (1985) The influence of fruit size and structure on composition

of frugivore assemblages in New Guinea. Biotropica 17, 314-21.

Price OF (1999) Conservation of frugivorous birds and monsoon rainforest patches in

the Northern Territory. PhD Thesis, Australian National University, Canberra

230 Price OF, Woinarski JCZ and Robinson D (1999) Very large area requirements for

frugivorous birds in monsoon rainforests of the Northern Territory, Australia.

Biological Conservation 91, 169-180.

Primack RB and Miao SL (1992) Dispersal can limit local plant distribution.

Conservation Biology 6, 513-519.

Pulliam HR (1973) On the advantages of flocking. Journal of Theoretical Biology 38,

419-422.

Pulliam HR (1975) Diet optimization with nutrient constraints. American Naturalist

109, 765-768.

Queensland Department of Natural Resources (1997) GeoScape: Scanned Aerial

Photography. Regional South East Queensland. QDNR, Coopooroo, Brisbane.

Ratcliffe F (1932) Notes on the fruit bats (Pteropus spp.) of Australia. Journal of

Animal Ecology 1, 32-57.

Recher HF, Date EM and Ford HA (1995) The Biology and Management of Rainforest

Pigeons in New South Wales. Species Management Report No. 16, NSW

National Parks and Wildlife, Hurstville.

Reid N (1989) Dispersal of mistletoes by honeyeaters and flowerpeckers: components

of seed dispersal quality. Ecology 70, 137-145.

Renjifo LM (1999) Composition changes in a subandean avifauna after long-term forest

fragmentation. Conservation Biology 13, 1124-1139.

Restrepo C, Renjifo LM and Marples M (1997) Frugivorous birds in fragmented

neotropical forests: landscape pattern and body mass distribution. pp. 171-189 in

Laurence WF and Bierregaard RO (eds.) Tropical Forest Remnants: Ecology,

Management and Conservation of Fragmented Communities. University of

Chicago Press, Chicago.

231 Ribon R, Simon JE and de Mattos GT (2003) Bird extinctions in Atlantic Forest

fragments of the Viçosa region, southeastern Brazil. Conservation Biology 17,

1827-1839.

Richards G (1990) The spectacled flying-fox, Pteropus conspicillatus, in north

Queensland: Part 2, diet and feeding ecology and seed dispersal. Australian

Mammalogist 13, 25-31.

Richards G and Hall LS (1998) Conservation biology of Australian bats: Are recent

advances solving our problems? pp. 271-281 in Kunz TH and Racey PA (eds.)

Bat Biology and Conservation. Sminthsonian Institute Press, Washington D. C.

Richardson DM, Allsopp N, D’Antonio CM, Milton SJ and Rejmanek M (2000) Plant

invasions – the role of mutualisms. Biological Review 75, 65-93.

Ridley HN (1930) The Dispersal of Plants Throughout the World. L.Reeve and Co.

Kent, England.

Roberts B (2005) Habitat characteristics of flying fox camps in south-east Queensland.

Unpublished BSc (Hons). Thesis, Griffith University, Brisbane.

Sallabanks R and Courtney SP (1993) On fruit-frugivore relationships: variety is the

spice of life. Oikos 68: 567-570.

SAS Institute (1999) SAS Procedures Guide. Release 6.11. SAS Inc., Cary.

Saunders DA, Hobbs RJ and Margules CR (1991) Biological consequences of

ecosystem fragmentation. Conservation Biology 5:18-32.

Scanlon T and the Camphor Laurel Taskforce (2000) Camphor Laurel Kit. North Coast

Weed Advisory Committee.

http://www.northcoastweeds.org.au/camphorkit.htm.

Schaefer HM, Schmidt V and Bairlein F (2003) Discrimination abilities for nutrients:

which difference matters for choosy birds and why? Animal Behaviour 65, 531-

541.

232 Schupp EW (1988) Seed and early seedling predation in the forest understorey and in

treefall gaps. Oikos 51, 71-78.

Schupp EW (1993) Quantity, quality and the effectiveness of seed dispersal by animals.

Vegetatio 107/108, 15-29.

Schupp EW, Milleron T and Russo SE (2002) Dissemination limitation and the origin

of species-rich tropical forests. pp. 397-406 in Levey DJ, Silva WR and Galetti

M (eds.) Seed Dispersal and Frugivory: Ecology, Evolution and Conservation.

CAB International, Oxford.

Şekercioğlu CH (2006) Increasing awareness of avian ecological function. Trends in

Ecology and Evolution 21: 464-471.

Şekercioğlu CH, Ehrlich PR, Daily GC, Ayhen D, Goehring D and Sandi RF (2002)

Disappearance of insectivorous birds from tropical forest fragments.

Proceedings of the National Academy of Sciences 99, 263-267.

Şekercioğlu CH, Daily GC and Ehrlich PR (2004) Ecosystem consequences of bird

declines. Proceedings of the National Academy of Sciences 101, 18042-18047.

Shanahan M, So S, Compton SG and Corlett R (2001) Fig-eating by vertebrate

frugivores: a global review. Biological Reviews 76, 529-572.

Shilton LA, Altringham JD, Compton SG and Whittaker RJ (1999) Old World fruit bats

can be long-distance seed dispersers through extended retention of viable seeds

in the gut. Proceedings of the Royal Society B: Biological Sciences 266, 219-

223.

Sigel BJ, Sherry TW and Young BE (2006) Avian community response to lowland

tropical rainforest isolation: 40 years if change at La Selva biological station,

Costa Rica. Conservation Biology 20, 111-121.

Sieving KE and Karr JR (1997) Avian extinction and persistence mechanisms in

lowland Panama. pp. 156-170 in Laurance WF and Bierregaard RO (eds.)

233 Tropical Forest Remnants: Ecology, Management and Conservation of

Fragmented Communities. University of Chicago Press, Chicago.

Silva WR, Marco PD Jr, Hasui E and Gomes VSM (2002) Patterns of fruit-frugivore

interactions in two Atlantic forest bird communities of south-eastern Brazil:

implications for conservation. pp. 423-35 in Levey DJ, Silva WR and Galetti M

(eds.) Seed Dispersal and Frugivory: Ecology, Evolution and Conservation.

CAB International, Oxford.

Silva JMC and Tabarelli M (2000) Tree species impoverishment and the future flora of

the Atlantic forest of northeast Brazil. Nature 404, 72–73.

Silva JMC, Uhl C, Murray G (1996) Plant Succession, Landscape Management and the

Ecology of Frugivorous Birds in Abandoned Amazonian Pastures. Conservation

Biology 10, 491-513.

Snow BK (1970) A field study of the bearded bellbird in Trinidad. Ibis 112, 299-329.

Snow DW (1962) The natural history of the oilbird, Steatornis caripensis, in Trinidad,

W.I. Part 2. Population, breeding ecology and food. Zoologica 47, 199-221.

Snow DW (1971) Evolutionary aspects of fruit-eating by birds. Ibis 113, 194-202.

Snow DW (1981) Tropical frugivorous birds and their food plants: a world survey.

Biotropica 3, 1-14.

Sodhi NS, Liow LH and Bazzaz FA (2004) Avian extinctions from tropical and

subtropical forests. Annual Review of Ecology and Systematics 35, 323-345.

Southerton SG, Birt P, Porter J and Ford HA (2004) Review of gene movement by bats

and birds and its potential significance for eucalypt plantation forestry.

Australian Forestry 67: 44-53.

Spencer HJ and Fleming TH (1989) Roosting and foraging behaviour of the Queensland

Tube-nosed bat, Nyctimene robinsoni (Pteropidae): Preliminary radio-tracking

observations. Australian Wildlife Research 16, 413-420.

234 Spencer HJ, Palmer C and Parry-Jones K (1991) Movements of fruit-bats in eastern

Australia, determined by using radio-tracking. Wildlife Research 18, 463-468.

SPSS (2001) SPSS for Windows. SPSS: Chicago.

Steadman DW and Freifeld HB (1998) Distribution, relative abundance, and habitat

relationships of landbirds in the Vava’u group, kingdom of Tonga. Condor 100,

609-628.

Stiles EW (1993) The influence of pulp lipids on fruit preference by birds. Vegetatio

107/108, 227-235.

Stocker GC and Irvine AK (1983) Seed dispersal by cassowaries (Casuarius casuarius)

in north Queensland rainforest. Biotropica 15, 170–176

Stouffer PC and Bierregaard RO Jr. (1995) Use of Amazonian forest fragments by

understorey insectivorous birds. Ecology 76, 2429-2445.

Sun C, Moermond TC and Givnish TJ (1997) Nutritional determinants of diet in three

turacos in a tropical montane forest. Auk 114: 200-211.

Tabarelli M, Mantovani W and Peres C (1999) Effects of habitat fragmentation on plant

guild structure in the montane Atlantic forest of southeastern Brazil. Biological

Conservation 91, 119-127.

Tabarelli M and Peres C (2002) Abiotic and vertebrate seed dispersal in the Brazilian

Atlantic forest: implications for forest regeneration. Biological Conservation

106, 165-176.

Temple SA (1977) Plant-animal mutualism: Coevolution with dodo leads to near

extinction of plant. Science 197, 885-886.

Terborgh J (1986) Community aspects of frugivory in tropical forests. pp 371-384 in

Estrada A and Fleming TH (eds.) Frugivores and seed dispersal. Dr. W. Junk,

Dordrecht.

235 Terbourgh J and Diamond JM (1970) Niche overlap in feeding assemblages of New

Guinea birds. Wilson Bulletin 82, 29-52.

Terborgh J and Winter B (1980) Some causes of extinction. pp. 119-133 in Soule ME

and Wilcox BA (eds.) Conservation biology: An evolutionary-ecological

perspective. Sinauer Press, Sunderland, Massachusetts,

Terborgh J, Pitman N, Silman M, Schichter H, Núñez PV (2002) Maintenance of Tree

Diversity in Tropical Forests. pp 1-17 in Levey DJ, Silva WR, Galetti M (eds.)

Seed Dispersal and Frugivory. CABI Publishing, Oxon New York.

Terborgh J and Nuñez-Iturri G (2006) Disperser-free tropical forests await an unhappy

fate. pp. 241-52 in Laurance WF and Peres CA (eds) Emerging threats to

tropical forests. University of Chicago Press, Chicago.

Tewksbury JJ, Levey DJ, Haddad NM, Sargent S, Orrock JL, Weldon A, Danielson BJ,

Brinkerhoff J, Damschen EI, and Townsend P (2002) Corridors affect plants,

animals, and their interactions in fragmented landscapes Proceedings of the

National Academy of Sciences 99, 12923–12926.

Tucker NIJ and Murphy TM (1997) The effects of ecological rehabilitation on

vegetation recruitment: some observations from the wet tropics of North

Queensland. Forest Ecology and Management 99, 133-152.

Turner IM (1996) Species loss in fragments of tropical rain forest: a review of the

evidence. Journal of applied Ecology 33, 200-209.

Turner IM and Corlett RT (1996) The conservation value of small, isolate fragments of

lowland tropical rain forest. Trends in Ecology and Evolution 11, 330-333.

Turner IM, Chua KS, Ong JSY, Soong BC, Tan HTW (1996) A century of plant species

loss from an isolated fragment of lowland tropical rain forest. Conservation

Biology. 10, 1229-1244.

236 Uhl C, Clark H, Clark K and Maquirino P (1982) Successional patterns associated with

slash-and-burn agriculture in the upper Rio Negro region of the Amazon Basin.

Biotropica 14, 249-254. van Bael SA, Brown JD, Robinson SK (2003) Birds defend trees from herbivores in a

Neotropical forest canopy. Proceedings of the National Academy of Sciences

100, 8304-8307. van der Pijl L (1982) Principles of Seed Dispersal in Higher Plants. Springer-Verlag,

Berlin.

Walker JS (2006) Resource use and rarity among frugivorous birds in a tropical

rainforest on Sulawesi. Biological Conservation 130: 60-69.

Wang BC and Smith TB (2002) Closing the seed dispersal loop. Trends in Ecology and

Evolution 117, 379-385.

Warburton NH (1997) Structure and conservation of forest avifauna in isolated

rainforest remnants in tropical Australia. pp. 190-206 in Laurance WF and

Bierregaard RO (eds.) Tropical Forest Remnants: Ecology, Management and

Conservation of Fragmented Communities. University of Chicago Press,

Chicago.

Watson D (1989) Clearing the scrubs of south-east Queensland. pp. 365-392 in Frawley

KJ and Semple NM (eds.) Australia’s Ever Changing Forests. Proceedings o f

the First National Conference in Australian Forest History, Canberra.

Watson M, Aebischer NJ and Cresswell W (2007) Vigilance and fitness in grey

partridges Perdix perdix: the effects of group size and foraging-vigilance trade-

offs on predation mortality. Journal of Animal Ecology 76, 211-221.

Webb LJ and Tracey JG (1981) Australian rainforests: patterns and change. pp. 607-694

in Keast A (ed.) Ecological Biogeography of Australia. W. Junk, The Hague.

237 Weir JES and Corlett RT (2007) How far do birds disperse seeds in the degraded

tropical landscape of Hong Kong, China? Landscape Ecology 22, 131-140.

Wenny DG and Levey DJ (1998) Directed seed dispersal by bellbirds in a tropical cloud

forest. Proceedings of the National Academy of Science, U.S.A. 95, 6204-6207.

Westoby M and Burgman M (2006) Climate change as a threatening process. Austral

Ecology 31, 549-550.

Wheelwright NT (1983) Fruits and the ecology of resplendent quetzals. Auk 100, 286-

301.

Wheelwright NT (1985) Fruit size, gape width, and the diets of fruit-eating birds.

Ecology 66, 808-818.

Wheelwright NT (1986) A seven-year study of individual variation in fruit production

in tropical bird-dispersed tree species in the family Lauraceae. pp. 19-35 in

Estrada A and Fleming T (eds.) Frugivores and Seed Dispersal. Dr. W. Junk,

Dordrecht.

Wheelwright NT and Orians GH (1982) Seed dispersal by animals: contrasts with

pollen dispersal, problems of terminology, and constraints on coevolution.

American Naturalist 119, 402-413.

Wheelwright NT, Haber WA, Murray KG and Guindon C (1984) Tropical fruit-eating

birds and their food plants: A survey of a Costa Rican lower montane forest.

Biotropica 16, 173-192.

Whelan CJ and Willson MF (1994) Fruit choice in migrating North American birds:

Field and aviary experiments. Oikos 71, 137-151.

White TCR (1993) The Inadequate Environment: Nitrogen and the Abundance of

Animals. Springer, Berlin.

Whitmore TC (1997) Tropical forest disturbance, disappearance, and species loss. pp. 2-

28 in Laurence WF and Bierregaard RO Jr. (eds.) Tropical forest remnants:

238 Ecology, management and conservation of fragmented communities. The

University of Chicago Press, Chicago.

Wiens JA (1994) Habitat fragmentation: island v landscape perspectives on bird

conservation. Ibis 137, S97-S104.

Williams JB, Harden GJ and McDonald WJF (1984) Trees and Shrubs in Rainforests of

New South Wales and Southern Queensland. University of New England,

Armidale.

Willis EO (1974) Populations and local extinctions of birds on Barro Colorado Island,

Panama. Ecological Monographs 44, 153-169.

Willson MF and Crome FHJ (1989) Patterns of seed rain at the edge of a tropical

Queensland rain forest. Journal of Tropical Ecology 5, 301-108.

Willson MF, Irvine AK, Walsh NG (1989) Vertebrate dispersal syndromes in some

Australian and New Zealand plant communities, with geographic comparisons.

Biotropica 21, 133-147.

With KA, Gardner RH and Turner MG (1997) Landscape connectivity and population

distributions in heterogeneous environments. Oikos 78, 151-169.

Witmer MC and van Soest PJ (1998) Contrasting digestive strategies of fruit-eating

birds. Functional Ecology 12, 728-741.

Woinarski JCZ (1993) A cut-and-paste community: birds of monsoon rainforests in

Kakadu National Park, Northern Territory. Emu 93, 100-120

Wright SJ (2002) Plant diversity in tropical forests: a review of mechanisms of species

coexistence. Oecologia 130, 1-14.

Wright SJ and HC Duber (2001) Poachers and forest fragmentation alter seed dispersal,

seed survival, and seedling recruitment in the palm Attalea butyraceae, with

implications for tropical tree diversity. Biotropica 33, 583-595.

239 Wright SJ, Zeballos H, Dominguez I, Gallardo MM, Moreno MC and Ibanez R (2002)

Poachers alter mammal abundance, seed dispersal, and seed predation in a

neotropical forest. Conservation Biology 14, 227-239.

Young PAR and McDonald WJF (1987) The distribution, composition and status of the

rainforests of southern Queensland. pp. 119-142 in Werren GL and Kershaw AP

(eds.) The Rainforest Legacy: Australian National Rainforest Study Volume 1 -

The Distribution and Status of Rainforest Types. Special Australian Heritage

Publication Series Number 7 (1), Canberra.

Young A, Boyle T and Brown T (1996) The population genetic consequences of habitat

fragmentation for plants. Trends in Ecology and Evolution 11, 413-418.

Zimmerman JK, Pascarella JB and Aide TM (2000) Barriers to forest regeneration in an

abandoned pasture in Puerto Rico. Restoration Ecology 8, 350-360.

240