Resource-Partitioning Between Three Syntopic Thornbills (Acanthizidae: Acanthiza Vigors and Horsfield)

RESOURCE-PARTITIONING BETWEEN THREE SYNTOPIC THORNBILLS (ACANTHIZIDAE: ACANTHIZA VIGORS AND HORSFIELD)

HENRY LAWRIE BELL

(B.App.Sc., Canberra College of Advanced Education; M.Sc., University of Papua New Guinea)

A Thesis submitted for the degree of Doctor of Philosophy of The University of New England

January 1983 iii

TABLE OF CONTENTS

LIST OF TABLES LIST OF FIGURES LIST OF PLATES LIST OF APPENDICES ACKNOWLEDGEMENTS ABSTRACT INTRODUCTION Chapter 1 1-7 PART I Environment, Availability of Food, and the Bird Community Chapter 2 The Environment 8-33 Chapter 3 Availability of Arthropods 33-47 Chapter 4 The Bird Community 48-52 PART II Social Organization of Thornbills Chapter 5 Social and Territorial Organization 53-88 Chapter 6 Mixed-species Flocks of Insectivorous Birds 89-106 PART III Foraging Behaviour Chapter 7 Substrate Foraged Upon by Thornbills 107-125 Chapter 8 Plant Species Foraged Upon by Thornbills 126-153 Chapter 9 Vertical Distribution 154-174 Chapter 10 Foraging Method 175-189 Chapter 11 Diet 190-204 Chapter 12 Morphology 205-220 PART V Chapter 13 Food Limitations in Thornbills 221-241 Chapter 14 General Summary 242-248 REFERENCES 249-272 APPENDICES iv

LIST OF TABLES Number Page

Table 1. Results of Transect Counts of Vegetation at Wollomombi 18 Showing Percentage of Numbers and Canopy Cover by Plant Species, and the Extent of Defoliation During Drought.

Table 2. Mean Density of Plant Species at Study Area, Wollomombi 19 (individuals/ha) (data from point-centred quarter sampling method).

Table 3. Proportions, by Height Categories, of Volume of Foliage 20 of Plant Species at Study Area (died or defoliated shown in brackets).

Table 4. Proportions, by Eucalypt Species, of Trunk Bark Available 22 to Birds; and of Trunk Sizes by dbh at Wollomombi.

Table 5. Seasonal Abundance of Arthropods Inside and on the 46 Surface of Bark of Eucalypts at Wollomombi, 1978/79. (Data supplied by R.A. Noske).

Table 6. Distribution, by Months, of Agonistic Displays and 61 Breeding Records of Brown Thornbills at Wollomombi, and Division of Labour at the Nest.

Table 7. Distribution, by Months, of Agonistic Displays and 69 Breeding Records of Buff-rumped Thornbills at Wollomombi and Division of Labour at the Nest.

Table 8. Distribution, by Months, of Agonistic Displays and 77 Breeding Records of Striated Thornbills at Wollomombi, and Division of Labour at the Nest.

Table 9. Interactions (see paragraph 111) in Mixed-species 98 Feeding Flocks, Between Thornbills and Other Species.

Table 10. Composition, by Time, of a MSF Following Buff-rumped 101 Thornbills 2 July 1979 (see Figure 32b). v.

Number Page

Table 11. Differences Between Mean Annual and Seasonal 117 Frequencies of Use of Substrate by Thornbills. Significance Values and Direction of the Seasonal Value from the Mean (+: higher than mean, -: lower than mean) are shown.

2 Table 12. Significance by X Test (d.f.l) for Increases 133 Between Seasonal Use of Plant Species for Foraging by Thornbills Against Mean Values for Each Plant Species Over the Whole Study Period.

Table 13. Preferences of Brown Thornbills for Plant Species 134 Compared with Frequency of Plant Species in Study Area.

Table 14. Preferences of Buff-rumped Thornbills for Plant 138 Species Compared with Frequency of Plant Species in the Study Area at Wollomombi.

Table 15. Preferences of Striated Thornbills for Plant 139 Species Compared with Frequency of Plant Species in the Study Area at Wollomombi.

Table 16. Preferences of Brown Thornbills for Different 143 Groups of Eucalypts.

Table 17. Preferences of Buff-rumped Thornbills for Different 144 Groups of Eucalypts.

Table 18. Preferences of Striated Thornbills for Different 145 Groups of Eucalypts.

Table 19. Foraging by Thornbills on Main Groups of Eucalypts, 146 When in MSF and Overall (including MSF).

2 Table 20. Significance, by X Test (d.f. 1) of Foraging by 160 Thornbills, by Seasons, at Heights More Frequently Than Expected, 1978/81. vi

Number Page

Table 21. Proportions of Foliage at Wollomombi, by Height 165 Categories, Compared to the Frequency of Observat- ion of Thornbills in those Categories.

Table 22. Frequency of Foraging by Thornbills on the Foliage 168 of Plant Species Below 2m in Height and the Availability of Foliage at that Level.

Table 23. Mean Numbers and Biomass (per 10 ha) by Height 172 Distribution of Insectivorous Birds at Wollomombi, 1978-81.

2 Table 24. Significance, by X Test (d.f. 1) of Foraging Methods 179 Used by Thornbills, by Seasons, at Wollomombi, 1978-81.

Table 25. Distribution (%) of Observations of Thornbills at 181 Wollomombi 1978-1981, by Substrate and Foraging Method Used.

Table 26. Distribution (%) of Observations of Thornbills at 182 Wollomombi 1978-81, by Plant Species and Foraging Method Used.

Table 27. Significance Levels (by Wilcoxsons Sum of Ranks Test) 185 of Differences by Time of Day in Speeds and Distance Covered by Thornbills in Foraging.

Table 28. Prey Items of Thornbills at Wollomombi Identified 198 by Direct Observation.

Table 29. Behaviour of Thornbills at Wollomombi When Offered 199 Sugared Water.

Table 30. Differences in Stomach Contents of Thornbills 201 Between Warmer Months (October-March) and Cooler Months (April-September). vii

Number Page

Table 31. Results of Comparisons of Morphological Data Between 213 Species Pairs of Thornbills (each pair compared by T-test except where otherwise shown; *: Mann-Whitney U-test; : mere comparison of means for each species).

Table 32. Results of Grasping-strength Tests on Thornbills at 214 Wollomombi (value for each individual in g).

Table 33. Comparison of the Seasons of 1978/79 in Which Each 227 Thornbill Species Carried out its Specialised Foraging Behaviour Significantly More Than Expected, and the Availability of Arthropods by Seasons.

Table 34. Comparisons, by T-test, of Mean Overlaps and S.D. 228 Between Thornbill Species in Months of Scarce and Abundant Resources.

Table 35. Percentage Overlaps in Foraging Between Species Pairs 232 of Thornbills Compared When in and When Not in MSF. viii

LIST OF FIGURES Page

Figure 1. Study Areas at Wollomombi Falls Reserve of R.A. Noske 10 and L. Huddy (Edgaris Spur and Quail Thrush Spur) and H.L. Bell (EdgaA Spur).

Figure 2. Mean monthly maximum and minimum temperatures, Monthly 12 and mean monthly rainfall, monthly Plant Growth Index and monthly Leaf Growth Index at Wollomombi Falls Reserve (September 1978 to August 1981).

Figure 3. Distribution, by height categories, of foliage present 17 at Wollomombi showing proportions killed or defoliated by drought. Figure 4. Cross-section of bark on trunks, at breast height, 23 of three Eucalyptus spp. (life-size). Figure 5. Leaves of (a) Stringybark Eucalyptus caliginosa, 26 (b) Box, E. sp. aff. cypellocarpa, (c) Gum, E.viminalis, (d) bipinnate Acacia., A. irrorata.

Figure 6. Leaves of (a) Cassinia quinquefaria, (b) Olearia viscidula, 27 (c) Jacksonia scoparia, and (d) Exocarpos cupressiformis.

Figure 7. Monthly biomass (mg wet weight/1000g foliage) of arthropods38 taken from grass sweeps and foliage sampling at Wollomombi, September 1978-August 1981. Figure 8. Monthly biomass (mg wet weight/1000g of foliage) of 39 arthropod samples taken from eucalypts, bipinnate Acacia, Olearia, Exocarpos, Jacksonia and Cassinia at Wollomombi, September 1978-August 1981 . Figure 9. Monthly biomass (mg wet weight/1000g of foliage) of 41 Coleoptera, Hemiptera, Formicoidea and Araneidea taken in foliage samples at Wollomombi, September 1978-August 19 81 . Figure 10. Monthly biomass (mg wet weight/100 X 2 m grass sweeps) 42 of Araneidea, Hymenoptera, Diptera and Lepidoptera taken in grass sweeps at Wollomombi, September 1978-August 1981 . ix

Page

Figure 11. Proportions of arthropods, by lengths, taken in foliage 44 samples and grass sweeps at Wollomombi.

Figure 12. Monthly mean lengths of arthropods taken on grass sweeps 45 and on foliage samples of eucalypts, bipinnate Acacia, OZearia and Jacksonia at Wollomombi, September 1978- August 1981). Figure 13. Mean monthly biomass of all insectivores at Wollomombi 50 (13a) mean daily energy demand of insectivores compared with arthropod samples (13W-; Resource demand (Kjoules/ 10 ha per mg arthropods/1000 g of foliage) and distribution of breeding records of insectivores at Wollomombi(13c). Figure 14. Percentage occurrences, by months (September 1978-August 56 1981), of Brown Thornbills by size of party (total n: 4656 individuals, monthly range 63-112).

Figure 15. Territories of Brown Thornbills at Wollomombi, 1978/79, 58 1979/80, 1980/81 (September-August) and 1981/82 (September-February). Letters denote colour-banded birds. Top bird of each pair is female. (Note: BuR daughter of YY, BuW son of LB). Figure 16. Inverse correlation (r: 0.855, P: < 0.01) between 59 frequency of observed agonistic encounters by Brown Thornbills against colour-banded neighbours and length of territorial boundary. Letters identify females, number in brackets: number of breeding seasons.

Figure 17. Sizes of Brown Thornbills territories at Wollomombi, 60 and the proportion of each covered by understorey (Spearman Rank Correlation, smallness of territory with proportion covered in understorey, r: 1.000, P: 0.01). Letters refer to female of each pair. Figure 18. Percentage occurrences, by months (September 1978- 64 August 1981), of Buff-rumped Thornbills by size of party (total n: 4674 individuals, Monthly Range 30-232). Page

Figure 19. Territories of clans of Buff-rumped Thornbills at 66 Wollomombi 1978-181. Area to North cleared grazing land, to South gorge country, unoccupied by species. Figure 20. Clan and group territories of Buff-rumped Thornbill 67 at Wollomombi, 1978/79, 1979/80, 1980/81 (September- August) and 1981/82 (September-February). Letters refer to banded birds (paragraph 75) top bird is female.

Figure 21. History of individual Buff-rumped Thornbills of Clan A, 68 at Wollomombi 1978/82. Broken edge at left of bar indicates age unknown at banding, at right indicates known to be alive after leaving. Figure 22. Percentage occurrences, by months (September 1978- 73 August 1981) of Striated Thornbills by size of parties (total n: 5933, monthly range (120-404). Figure 23. Dendrogram showing degree of association between 74 Striated Thornbills observed at Wollomombi during 1978/79 and probable membership of clans.

Figure 24. Observed sightings of clans and presumed clan territories 75 of Striated Thornbills at Wollomombi 1978/79, 1979/80, 1980/81 (September-August) and 1981/82 (September- February). (Closed circles: Clan A, open circles: Clan B, upright crosses: Clan C, diagonal crosses Clan D).

Figure 25. Sightings during breeding season (1 August-30 November) 76 of 1979 of individuals of five groups of Striated Thornbills at Wollomombi, and location of nests (N). (Note that vacant areas were occupied, by unidentified groups). Figure 26. Abundance (mg/1000 g of foliage) of arthropods in 80 foliage samples, by seasons compared to (a) deaths of thornbills, (b) numbers of thornbills present, and (c) death-rate of thornbills (deaths as % of numbers), at Wollomombi. xi

Page

Figure 27. Numbers of MSF/10 h of observation; by months, 91 (September 1978-August 1981), and percentage of all birds seen in transect counts, by months, that were in MSF (May 1979-August 1981). Figure 28. Number of MSF/10 h in which Brown, Buff-rumped and 93 Striated Thornbills participated (September 1978- August 1981).

Figure 29. Percentage of all thornbills seen that were in MSF, 94 September 1978-August 1981. (n of individuals: Brown 4656, Buff-rumped 4690, Striated 5491). Figure 30. Mean monthly number of individual thornbills per MSF 95 (a), and mean monthly number of bird species per MSF of which thornbills were members (b). (n of MSF: Brown 626, Buff-rumped 493, Striated 470).

Figure 31. Combined total occurrences of following by thornbill 97 species of each other.

Figure 32. Routes and timings of movements by four MSF at Wollomombi 100 in which Buff-romped Thornbills were the nuclear species.

Figure 33. Distribution of observations of foraging moves, by 113 substrate foraged upon, by thornbills at Wollomombi.

Figure 34. Distribution of observations of foraging moves, by 115 substrate, of Brown Thornbills at Wollomombi. (Total n: 24369, monthly mean 676, monthly range 390-955).

Figure 35. Distribution of observations of foraging moves, by sub- 116 strate, of Buff-rumped Thornbills at Wollomombi. (Total n: 23, 536; monthly mean 654; monthly range 300-825) (lower box of bark substrates are trunks and limbs).

Figure 36. Substrates foraged upon by Buff-rumped Thornbills, by 119 2 size of party. (X Test observations of substrate in party-size category against overall mean). x i i

Page

Figure 37. Distribution of observations of foraging moves, by 120 substrates, of Striated Thornbills at Wollomombi (Total n: 22 232; monthly mean 617; monthly range 280-835).

Figure 38. Distribution of observations of foraging, by substrate, 121 2 y thornbills when with, or not with, MSF. (X probabilities, : P < 0.005; : P < 0.01). (n shown: number of moves).

Figure 39. Distribution of observations of foraging moves, by plant 129 species foraged upon, of thornbill species at Wollomombi. (n: Brown, 24 369; Buff-rumped, 23 536; Striated, 22 232). (Note: all others for Striated includes all plant species other than eucalypts and Acacia.).

Figure 40. Distribution of observations of foraging moves, by plant 131 species foraged upon, of Brown Thornbills at Wollomombi (Total n: 24 369; monthly mean 676; monthly range 390-955) (months where eucalypts not grouped; open; stringybark; shaded: boxes and gums).

Figure 41. Correlation between extent of cover of- plant species 132 within Brown Thornbill territories, and the extent to which the resident pair foraged on each species (letters indicate '7 of pair; n: YY 2375, LB 1010, RY 2100, YL 920, LL 280, WY 855). Figure 42. Distribution of observations of foraging moves (less 136 ground) by plant species foraged upon, of Buff-rumped Thornbills at Wollomombi (Total n: 23 536; monthly mean 654; monthly range 300-825) (Months where eucalypts not grouped, top open: stringybark, shaded: boxes, bottom open: gums). Figure 43. Distribution of observations of foraging moves, by 137 plant species foraged upon, of Striated Thornbills at Wollomombi (Total n: 22 232; monthly mean 617; monthly range 280-835) where top open, Gums; bottom open: Stringybarks. Page

Figure 44. Distribution of foraging observed of (a) Brown, (b) 141 Buff-rumped and (c) Striated Thornbills by major substrates of plant species foraged upon.

Figure 45. Distribution of observations of foraging by thornbills, 142 by plant species foraged upon, with and not with MSF 2 (X probabilities; : P < 0.005; : P < 0.01; : P < 0.05) (n shown: number of moves).

Figure 46. Vertical distribution afobservations of foraging moves 157 of thornbills at Wollomombi. (Note height scale is not linear). Figure 47. Distribution of observations of foraging moves, by 158 height categories, of Brown Thornbills at Wollomombi (n: 24369, monthly mean 676, monthly range 390-955).

Figure 48. Distribution of observations of foraging moves, by height 159 categories, of Buff-rumped Thornbills at Wollomombi. (n: 23536, monthly mean 654, monthly range 300-825)).

Figure 49. Correlation between size of party of Buff-rumped Thorn- 162 bills not in MSF and (a) ground-foraging (Spearman Rank Correlation, r: 0.965, P 4.0.01); and (b) foraging over 2 m from ground (r: -1.000, P es, 0.01). (n: individual moves: single birds 850; pairs 4426; trios 1176; quartets 2592, quintets 1406, sextets 1777, > six birds 9810)..

Figure 50. Distribution of observations of foraging moves, by height 163 categories, of Striated Thornbills at Wollomombi. (n: 22 232; monthly mean 617; monthly range 280-825).

Figure 51. Correlation between size of party of Striated Thornbills 164 not in MSF and (a) foraging below 5 m (Spearman Rank Correlation, r: 0.943, P < 0.05); and (b) foraging above 10 m (Spearman Rank Correlation, r: 0.943, P <0.05). (n: individual moves: single birds 1913; pairs 4004; trios and quartets 1960, quintets 682; sextets 687, parties > six birds 2392). xiv

Page

Figure 52. Vertical distribution of thornbills not in and in NSF 167 2 (X probabilities; : P40.005; : P 4C 0.01) (n shown: number of moves).

Figure 53. Distribution, by foraging method, of observations of 178 thornbills at Wollomombi (n: Brown 24369; Buff-rumped 2 23536; Striated 22232) (X probabilities, more than expected; : P < 0.005, : P < 0.01). Figure 54. Foraging speed (secs/move) (a), and distance covered 184 (m/move) (b), compared with abundance of arthropods (histogram), for thornbills at Wollomombi 1978/81.

Figure 55. Proportions, by dietary items, of stomach and faecal samples taken from thornbills (number of items given for each species). 193 Figure 56. Percentage frequency of dietary items in thornbill 195 stomachs taken near Armidale and Wollomombi (number of stomachs in brackets).

Figure 57. Proportions, by size classes, of arthropods in stomach 196 samples of thornbills (mean and S.D. given for each species).

Figure 58. Percentage frequency, by size classes of beetles, 197 Coleoptera (n: stomach contents 253, foliage samples 3965) and ants Formicoidea (n: stomach contents 60, foliage samples 5660) taken in foliage sampling (histogram) and stomach samples of thornbills (graph).

Figure 59. Morphological data taken from right feet measured as 207 follows; h: foot span; i: front toe length; j , rear toe length; k: front claw length; 1: curvature of front claw; m: rear claw length; n: curvature of rear claw.

Figure 60. Tongues of Brown, Buff-rumped and Striated Thornbills 211 taken at Wollomombi (top view, tips at right). XV Page

Figure 61. Comparison of foraging overlaps between species pairs 225 of thornbills with PGI (Overlaps: x: all observations; y: all those not in MSF5 shaded part of Brown/Buff- rumped overlap is ground-foraging not in MSF). Figure 62. Comparison of foraging overlaps between species pairs 226 of thornbills with abundance of arthropods on foliage overlaps: x: all observations; y: all those not in MSF; shaded part of Brown/Buff-rumped overlap is ground- foraging not in MSF).

Figure 63. Comparison, by seasons, of overlaps between species pairs 230 of thornbills with abundance of arthropods on foliage (a) (overlaps: x: all observations; y: all those not in MSF; shaded part of Brown/Buff-rumped overlap is ground-foraging not in MSF).

Figure 64. Percentage overlap in foraging between thornbill species, 243 by the four variables (histogram) and degrees of overlap when variables are combined (graph). xvi

LIST OF PLATES

Number Title Page

1 Typical Trunk Bark of (Top) Stringybark, (Centre) 24 Box and (Bottom) Gum.

2 (Top left) Acacia irrorata, (Top Right) Cassinia sp. 28 (Bottom Left) Exocarpos cupressiformis, (Bottom Right) Olearia viscidula.

3 Typical Views of Wollomombi Reserve. Top: Territory 32 of Brown Thornbill YELLOW/YELLOW, Bottom: Territory of Brown Thornbill LIGHT GREEN/LIGHT GREEN.

4 Typical Beaks of (Top) Brown Thornbill, (Centre) 212 Buff-rumped Thornbill and (Bottom) Striated Thornbill. Scale line: 1 cm. xvii

LIST OF APPENDICES

Number Title

A Distribution of Foraging Moves, by Substrate, of Brown Thornbills at Wollomombi 1978-81.

B Distribution of Foraging Moves, by Substrate, of Buff-rumped Thornbills at Wollomombi 1978-81.

C Distribution of Foraging Moves, by Substrate, of Striated Thornbills at Wollomombi 1978-81.

D Distribution of Foraging Moves, by Plant Species, of Brown Thornbills at Wollomombi 1978-81.

E Distribution of Foraging Moves, by Plant Species, of Buff-rumped Thornbills at Wollomombi 1978-81.

F Distribution of Foraging Moves, by Plant Species, of Striated Thornbills at Wollomombi 1978-81.

G Distribution of Foraging Moves, by Height, of Brown Thornbills at Wollomombi, 1978-81.

H Distribution of Foraging Moves, by Height, of Buff- rumped Thornbills at Wollomombi 1978-81.

I Distribution of Foraging Moves, by Height, of Striated Thornbills at Wollomombi 1978-81.

J Distribution of Foraging Moves, by Method Used, of Brown Thornbills at Wollomombi 1978-81.

K Distribution of Foraging Moves, by Method Used, of Buff-rumped Thornbills at Wollomombi 1978-81.

L Distribution of Foraging Moves, by Method Used, of Striated Thornbills at Wollomombi 1978-81. xviii

M Composition of Stomach Contents and Faeces of Brown Thornbbills.

N Composition of Stomach Contents and Faeces of Buff-rumped Thornbills.

0 Composition of Stomach Contents and Faeces of Striated Thornbills.

P Total Foraging Observations of Thornbills at Wollomombi 1978-81, by Plant Species, Height, Foraging Method and Substrate.

Means, Standard Deviations and Numbers of Each Sex From Which Means Taken, of Morphological Data of Thornbills at Wollomombi.

R Scientific Names of Australian Bird Species Referred to in Text. xix

ACKNOWLEDGEMENTS

I thank the Australian Department of Education and the University of New England for granting me a Commonwealth Post-graduate Research Award, which made the study possible. I also thank my supervisors, Dr H.A. Ford and Dr P.J. Jarman, for their guidance and encouragement throughout the project.

The study would have been much less productive without the stimulat- ion of discussions within the University, with staff, in particular Associate Professors H. Heatwole and K. Rhode, Drs S.C. Cairns, M. Notestine, D.J. Woodland and P. Watters; and colleagues, particularly V. Aslin, R. Howe, D. McFarland, R. Nias and J. Taylor. To R.A. Noske and L. Huddy I owe a great debt of thanks for the mutual benefits that arose from our working at the same study area at Wollomombi.

To my correspondents I thank them for patiently answering my queries; Dr James H. Brown, Dr Jerram I. Brown, Dr Jared M. Diamond. Dr Douglas D. Dow, Dr Peter Grant, Dr Richard T. Holmes, Dr Hilary Fry, Dr Eric Pianka, Dr David L. Pearson. Mrs Mary Le Croy, Mr Nick Reid, Mr Ian Rowley, Dr Thomas W. Schoener, Dr Jamie Smith and Mr John Woinarski.

I am grateful for the support given by the staff of the Department of Zoology at the University of New England, in particular that of Mrs V. Watt, Mr R.P. Hobbs, Mr E. Brewster and Mr G. Hodges.

I thank Dr I. Davies and Dr V. Boffinger for advice on statistical methods, and Dr J. Williams for botanical advice.

Mr H.A. Nix kindly made available to me the Plant Growth Index Program of the CSIRO Division of Land and Water Resources. Mr D. Purchase, Secretary of the CSIRO Australian Bird-banding Scheme assisted with banding equipment and information. The specimens of the National Collection in Canberra were made available by Dr R. Schodde, and those in the Australian Museum by Mr W. Boles. The National Parks and Wildlife Service, and the State Forestry Commission, both of New South Wales, made available the permits necessary to work on their lands and to collect. XX

Mrs S. Lucas typed the manuscript, Ms S. Harrington drew the figures, Mr R.P. Hobbs assisted with photography and Mrs M. Hitchcock drew the pictures of plant specimens. Their skills, and more importantly, their interest in the study itself were greatly appreciated.

Lastly, but with gratitude, I thank Mr and Mrs R. Edgar, of Rosewood, Wollomombi, whose tolerance towards both wildlife, and the students of wildlife, made my (and other colleagues) studies possible. Nor will I forget their kindnesses during numerous minor emergencies. xxi

ABSTRACT

i. Resource-partitioning among three syntopic species of thornbill (Acanthiza pusina, A.reguloides and A.lineata) were studied at Wollomombi on the Northern Tablelands of New South Wales. All three species are mainly insectivorous and weigh about 7 grams. Abundance of arthropods, on foliage and grass, was measured monthly. There were great differences betweensummerpeaks and winter troughs (> 10:1) and also between the normal year of 1978/79 and the two drought years that followed (ca. 4:1).

ii. Pairs of A.pusilla held territories year-round, and evicted their young in autumn. A.reguloides and A.Zineata by contrast, were highly social, living for most of the year in a clan of 10-20 birds which held a territory. In spring the clan broke up into breeding groups of one female and one to three males. Unlike A.pusilla they retained their young at least until the following breeding season. All three species associated in mixed-species feeding flocks of which A.reguloides and A.lineata were often the nuclear species (ie. that around which other species of birds gathered). A.pusilla and A.lineata bred mainly from October to December. Nest success was very low, about 16% for all species, but new attempts were made after failure. For 25 pair/years of A.pusilla, 11 clutches were fledged. For 17 pair or group/years of A.reguloides 7 clutches were fledged. However most young of A.pusilla were predated within a few weeks of fledging but all young of A.reguloides survived Eor at least six months.

ABSTRACT i. Resource-partitioning among three syntopic species of thornbill (Acanthiza pusilla, A.reguloides and A.Zineata) were studied at Wollomombi on the Northern Tablelands of New South Wales. All three species are mainly insectivorous and weigh about 7 grams. Abundance of arthropods, on foliage and grass, was measured monthly. There were great differences betweenzummerpeaks and winter troughs (> 10:1) and also between the normal year of 1978/79 and the two drought years that followed (ca. 4:1).

ii. Pairs of A.pusilla held territories year-round, and evicted their young in autumn. A.reguloides and A.Zineata by contrast, were highly social, living for most of the year in a clan of 10-20 birds which held a territory. In spring the clan broke up into breeding groups of one female and one to three males. Unlike A.pusilla they retained their young at least until the following breeding season. All three species associated in mixed-species feeding flocks of which A.reguloides and A.Zineata were often the nuclear species (ie. that around which other species of birds gathered). A.rusilla and A.Zineata bred mainl y from Amendment: Paragraph ii, line 9 "A. pusilla and A. lineata bred mainly from August to October but A. reguloides bred from October to December".

were predated within a few weeks of fledging but all young of A.reguloides survived Eor at least six months.

iii. The three species were ecologically separated as they fed on different substrates and on different species of plants and at different heights. A.pusilla foraged mainly on dense foliage of understorey shrubs, A.reguZoides foraged mainly on the ground or on the bark of trees and shrubs, and A.Zineata foraged mainly on the foliage of eucalypts. Foraging behaviour changed seasonally and with the size and composition of the group. For example A.reguZoides fed high in trees when in small groups but on the ground when in large groups. All three species foraged differently when in mixed-species flocks, even in areas where they were the nuclear species. iv. The three species take much the same food except that A.lineata takes fewer ants and no seeds, but possibly more scale insects. The species were selective in that they mainly took arthropods over 2 mm in length, whereas only 25% of arthropods taken in foliage samples were longer than 2 mm. They took smaller insects when food was scarce.

v. Morphological differences between each species are slight but are related to specialisations in foraging. A.pusilla is adapted to manoeuvre in dense vegetation and to probe into leaf clusters. A.reguloides is adapted to foraging off smooth surfaces and to hanging upside down on leaves of eucalypts. However, all three did use most or all of the foraging methods used by each other, but specialised most in times of scarcity: thus adaptations appear to be to conditions when competition is likely to be strongest. vi. The monthly overlap in foraging, between species, ranged from 3 to 43%. Overlaps were lowest when food was scarce but during prolonged periods of scarcity rose to comparatively high levels well before food increased. I believe that at such times either that a species is forced by intra-specific competition, in its specialised foraging role, to broaden its niche; or that overall scarcity is so severe that all species have to abandon specialisation and search for whatever prey is available, with a consequent increase in overlap. Also, mortality of adults was significantly highest in those periods overlaps rose while food was still scarce. I suspect that direct competition between species probably occurs only during such periods, specialised foraging behaviour reducing competition in other times of scarcity, and abundance of food obviating competition in times of plenty. To David Leithhead (1900- ), wildlife writer and photographer; my primary school teacher in 1939, without whose encouragement at that time this thesis would probably not have been written.