The Invertebrate World of Australia's Subtropical Rainforests

By Geoff Williams

The Invertebrate World of Australia’s Subtropical Rainforests is a comprehensive review of Australia’s Gondwanan rainforest invertebrate fauna, covering its taxonomy, distribution, biogeography, fossil history, plant community and insect–plant relationships. This is the first work to document the invertebrate diversity of this biologically important region, as well as explain the uniqueness and importance of the organisms.

This book examines invertebrates within the context of the plant world that they are dependent on and offers an understanding of Australia’s outstanding (but still largely unknown) subtropical rainforests. All major, and many minor, invertebrate taxa are described and the book includes a section of colour photos of distinctive species. There is also a strong emphasis on plant and habitat associations and fragmentation impacts, as well as a focus on the regionally inclusive Gondwana Rainforests (Central Eastern Rainforest Reserves of Australia) World Heritage Area.

The Invertebrate World of Australia’s Subtropical Rainforests will be of value to professional biologists and ecologists, as well as amateur entomologists and naturalists in Australia and abroad.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Sep 1, 2020
• ISBN: 9781486312931

Geoff Williams

Geoff Williams OAM, AM is a pollination ecologist, conservation biologist and entomologist with a PhD from the University of New South Wales, and a Research Associate of the Australian Museum. He was awarded the Medal of the Order of Australia and appointed a Member of the Order of Australia in recognition of his contributions to science and biodiversity conservation. He is the author of The Invertebrate World of Australia’s Subtropical Rainforests (CSIRO Publishing, 2020) and The Flowering of Australia’s Rainforests: Pollination Ecology and Plant Evolution (CSIRO Publishing, 2021).

Book preview

To Alfred Russel Wallace (1823–1913)

for giving us the theory of the origin and divergence of species through natural selection, and the theory of biogeography.

But first and foremost for being an inspiring natural historian.

© Geoff Williams 2020

All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO Publishing for all permission requests.

The author asserts their moral rights, including the right to be identified as the author.

A catalogue record for this book is available from the National Library of Australia.

ISBN: 9781486312917 (hbk)

ISBN: 9781486312924 (epdf)

ISBN: 9781486312931 (epub)

How to cite:

Williams G (2020) The Invertebrate World of Australia’s Subtropical Rainforests. CSIRO Publishing, Melbourne.

Published by:

CSIRO Publishing

Locked Bag 10

Clayton South VIC 3169

Australia

Telephone: [phone]

Email: [e-mail]

Website: www.publish.csiro.au

Front cover: (main) Richmond birdwing butterfly, Ornithoptera richmondia (Papilionidae) (photo: Carl Bento, Australian Museum). (top, left to right) Hadronyche sp. (Hexathelidae), Scaptia barbara (Tabanidae), Paraschizognathus ?ocularis (Scarabaeidae), Cabasa ?pulchella (Asilidae), Calodema regale (Buprestidae) (photos: Geoff Williams)

Back cover: (top) Eborophyllus middletoni (Cerambycidae), (middle) Regent Skipper, Euschemon rafflesia rafflesia (Hesperiidae), (bottom) Australodon nearnsi (Cerambycidae) (photos: Kevin Mitchell)

All photographs by the author except Ornithoptera richmondia (Carl Bento/Australian Museum), Euschemon rafflesia rafflesia, Australodon nearnsi, Eborophyllus middletoni (Kevin Mitchell) and Thersites mitchellae (Mark V. Robinson).

Edited by Joy Window (Living Language)

Cover design by Andrew Weatherill

Typeset by Envisage Information Technology

Printed in China by Leo Paper Products Ltd.

CSIRO Publishing publishes and distributes scientific, technical and health science books, magazines and journals from Australia to a worldwide audience and conducts these activities autonomously from the research activities of the Commonwealth Scientific and Industrial Research Organisation (CSIRO). The views expressed in this publication are those of the author(s) and do not necessarily represent those of, and should not be attributed to, the publisher or CSIRO. The copyright owner shall not be liable for technical or other errors or omissions contained herein. The reader/user accepts all risks and responsibility for losses, damages, costs and other consequences resulting directly or indirectly from using this information.

The paper this book is printed on is in accordance with the standards of the Forest Stewardship Council® and other controlled material. The FSC® promotes environmentally responsible, socially beneficial and economically viable management of the world’s forests.

Apr 20_01

Contents

Acknowledgments

1Introduction

2Australia’s subtropical rainforests – the plant context

Colour plates

3The invertebrates – the ‘other 99 per cent’ of subtropical rainforest biodiversity

The role of invertebrates in subtropical rainforest ecosystems

Pollination in Australia’s subtropical rainforests and the pollinator function of insects

4Invertebrate taxa of Australian subtropical rainforests

PLATYHELMINTHES (flatworms, land planarians)

Temnocephalida (flatworms)

Tricladida (land planarians)

NEMERTEA (ribbon worms)

NEMATODA (round worms)

ANNELIDA (earthworms, leeches)

Oligochaeta (earthworms)

Hirudinea (leeches)

CHELICERATA (spiders, scorpions, harvestmen, mites, ticks, pseudoscorpions, whip spiders)

ARACHNIDA (spiders)

Araneomorphae (‘true’ spiders)

Mygalomorphae (trap-door and funnel-web spiders)

SCHIZOMIDA (whip spiders)

SCORPIONIDA (scorpions)

PSEUDOSCORPIONIDA (pseudoscorpions)

OPILIONES (harvestmen)

ACARINA (mites, ticks)

CRUSTACEA (amphipods, isopods, syncarids, crayfish, shrimp and related groups)

PERACARIDA (amphipods, isopods)

SYNCARIDA

DECAPODA (crayfish, shrimp, freshwater prawns)

UNIRAMIA (peripatus, centipedes, millipedes, springtails, insects and related groups)

ONYCHOPHORA (peripatus, velvet worms)

CHILOPODA (centipedes)

DIPLOPODA (millipedes)

COLLEMBOLA (springtails)

PROTURA

DIPLURA (two-pronged bristletails)

INSECTA (‘true’ insects)

ARCHAEOGNATHA (bristletails)

EPHEMEROPTERA (mayflies)

ODONATA (dragonflies, damselflies)

Zygoptera (damselflies)

Anisoptera (‘true’ dragonflies)

PLECOPTERA (stoneflies)

BLATTODEA (cockroaches, termites)

Blattaria (cockroaches)

Isoptera (termites)

MANTODEA (praying mantids, mantises)

DERMAPTERA (earwigs)

ORTHOPTERA (grasshoppers, katydids, crickets)

PHASMATODEA (stick insects)

EMBIOPTERA (web-spinners, embiids)

PSOCOPTERA (psocids, book lice)

PHTHIRAPTERA (lice)

HEMIPTERA (true bugs, leafhoppers, cicadas, mealybugs, lerps)

THYSANOPTERA (thrips)

MEGALOPTERA (alder flies)

NEUROPTERA (lacewings, ant-lions)

COLEOPTERA (beetles)

MECOPTERA (scorpion flies)

SIPHONAPTERA (fleas)

DIPTERA (flies, midges, mosquitoes)

TRICHOPTERA (caddis flies)

LEPIDOPTERA (moths and butterflies)

Glossata (Hesperoidea and Papilionoidea) (butterflies)

Glossata (part – remaining superfamilies), Aglossata and Zeugloptera (moths)

HYMENOPTERA (wasps, bees, ants, sawflies)

Symphyta (sawflies)

Apocrita (Aculeata – Apiformes) (bees)

Apocrita (Aculeata and Parasitica) (wasps)

Apocrita (ants)

MOLLUSCA (land and freshwater snails, slugs)

Terrestrial fauna

Freshwater fauna

Appendix 1. Additional ecosystem values – the stygofauna

Appendix 2. The Gondwana Rainforests of Australia (CERRA) World Heritage Area

Appendix 3. Major occurrences of rainforest subforms in subtropical eastern Queensland and New South Wales

Appendix 4. Higher divisions of Coleoptera, Diptera, Hemiptera, Hymenoptera (Insecta) and Mollusca

Appendix 5. Divisions of geological time

References

Index

Acknowledgments

Foremost I have to thank my partner Thusnelda for countless hours of great company and untiring help in the field, my children Beren, Sarid and Arwen for lightening the days in so many ways, and my parents Arthur and Patricia for their encouragement (and cheerful acceptance when in earlier years many a strange creature was brought into their lives). To Eloise Moir-Ford, Lauren Webb, Tracey Kudis (CSIRO Publishing) and Joy Window (copy editor, Living Language) I also express special thanks. Without their diligence and encouragement, the publication of this book would not have been possible.

Over the years of field investigation and research I have benefited from discussions and interactions with numerous friends and colleagues: notably Paul Adam at the University of New South Wales and Dan Bickel at the Australian Museum. Many others have contributed to my knowledge and understanding of rainforests and Australia’s plant communities and their biota in general, or have given support in diverse ways: Rebecca Johnson, Alexandra Durham, Mike Gray, Glenn Hunt, David McAlpine, Shane McEvey, Jacquieline Rescei, Russell Cox, Chris Reid, Courtenay Smithers, Derek Smith, Helen Smith, Max Moulds, Dave Britton, John Macdonald, Graham Milledge, Natalie Kees, Timothy Lee, (Australian Museum, Sydney), John Hunter, Terry Evans (NSW NPWS), Steve Nikitin, Clarry Chadwick (NSW Department of Agriculture), Ross Storey (Queensland Department of Primary Industries), Günther Theischinger (Grays Point), Ev. Britton, Phil Carne, John Lawrence, Barry Moore, Laurence Mound, Tom Weir (CSIRO, Canberra), Eric Matthews (South Australian Museum), Shelley Barker (University of Adelaide), Ken Walker, (Museum of Victoria), Allen Sundholm, Roger de Keyser (Sydney), Hugh and Nan Nicholson (Terania Creek), Paul Brock (Natural History Museum, London), Terry Houston (Western Australian Museum, Perth), and Geoff Monteith, Christine Lambkin and John Stanisic (Queensland Museum). Kevin Mitchell (Wherrol Flat), Mark Robinson (Sawtell) and the Australian Museum kindly allowed me to use their photos of Euschemon rafflesia, Australodon nearnsi, Eborophyllus middletoni, Thersites mitchellae and Ornithophtera richmondia respectively. Eric Matthews kindly allowed the use of Fig. 3 and Jane Hosking (Taree) is thanked for drawing Fig. 2. Not least of all, I acknowledge my colleague and friend, the late Charles ‘Chuck’ Bellamy, previously with the University of Pretoria (Republic of South Africa) and the California Department of Food and Agriculture (Sacramento, USA).

Too many are now gone, but all are thanked for their diverse free giving of time and thoughts. In conclusion this book is indebted to the often unsung research undertaken by Australian invertebrate taxonomists, systematists and field naturalists.

1Introduction

The Invertebrate World of Australia’s Subtropical Rainforests is about the ‘lesser animals’: the insects, crustaceans, spiders, harvestmen, mites, scorpions, centipedes, millipedes, true worms, flatworms, nematodes, snails and slugs, and the ancient Peripatus or velvet worms. These organisms without backbones, inconspicuous as most are, dominate Earth’s terrestrial animal diversity.

The scope of this book takes a broad latitudinal sweep spanning ~23.4–37°S (this to about the southern-most occurrence of littoral rainforest), stretching inland to the dry vine thickets of Queensland and New South Wales, and addressing rainforests as ‘subtropical’ loosely in a geographical sense (see Fig. 1). Figure 2 shows the locations of the places referred to in this book. Within this geographical range there are several overlapping plant thermal response groups:

•a humid mesotherm dominated (mean annual air temperature >14–<24°C) province centred on coastal southern Queensland and northern New South Wales, but extending to the south coast of New South Wales

•a subtropical – lower montane province with a microtherm–mesotherm humid climate (mean annual air temperature >12–<14°C) whose floristic elements encompass upland Nothofagus moorei –dominated rainforest stands

•and a geographically extensive subhumid–humid (mesotherm–megatherm/>20–<24°C) warm subtropical zone that extends from the Northern Tablelands of New South Wales, largely reaching the Queensland coast north of Brisbane and extending to the subhumid tropics inland from Cairns ( Adam 1992 ; Nix 1982 ).

These capture warmer lowland and lower latitude forests that are subtropical by measure of climate (see Corlett 2013), and cooler upland and higher latitude forests that are ‘temperate’ (see Adam 1992; Nix 1982; Webb and Tracey 1981). Ultimately, however, altitude and geological history will confound characterisations of rainforest based solely on latitude.

Subtropical rainforest in the context of this definition includes ‘closed forest’ formations of great floristic and structural diversity, existing under marked differences of altitude, mean air temperature range, rainfall and soil type. A floristically circumscribed characterisation of subtropical rainforest is probably elusive, for they contain combinations of the original Gondwana flora (alluding to plants that evolved in areas of the Gondwana supercontinent, now mainly represented by Antarctica, South Ameria, Africa, Madagascar, Australia, India, New Zealand and New Caledonia) (see ‘The formation of Gondwanan landmasses’ on p. 5 for a cautionary note on the use of the often conflated term ‘Gondwanan’) that stretch back to at least the Early Miocene (Adam 1987, 1992; Floyd 1990, 2010; Kitching et al. 2010; Appendix 5), autochthonous elements, as well as taxa that represent later ‘tropical’ intrusions. These rainforests have acted as refugia over lengthy geological cycles of expansion and contraction (e.g. Dorrigo and the Nightcap–Border Ranges, in New South Wales) and include those (e.g. Washpool) that are the result of colonisation events (Rossetto et al. 2015). Such patterns of retreat to refugia followed by recolonisation and expansion during environmentally conducive periods characterise the history of Australia’s diverse rainforests (Hilbert et al. 2007; Moritz et al. 1997; Price and Sobbe 2005; Shapcott et al. 2015; VanDerWal et al. 2009), for these rainforests have been subject over long periods of time to the oscillating influence of changing climatic events (i.e. periods of warm–wet and cool–dry), in particular the progressive aridification trend from the Late Miocene (McLaren and Wallace 2010) that has led to the fragmentation of an earlier, more widespread and diverse mesic-adapted biota (that existed during warm–wet conditions 20–10 million years ago (mya)), and the emergence of an extensive xeric-adapted one (Adam 1992; Byrne et al. 2008; Martin 2006; Price and Sobbe 2005; Truswell 1993). The distribution and areal extent of subtropical rainforest first encountered by European colonists strongly mirrored that influence; that which survived is set, as floristically and structurally distinct plant communities that reflect varying geological episodes (Benson and Redpath 1997; Bowman 2000; Rossetto et al. 2015), in a landscape context in which the intervening vegetation matrix is largely one of eucalypt-dominated ‘wet sclerophyll’ and ‘dry sclerophyll’ forests (Adam 1987; McDonald and Hunter 2010) (see Plate 312). However, there is no simple conceptual demarcation of Australian rainforest, for wet sclerophyll forest (a floristically unstable entity) (see Plates 302, 303) has a greater ecological and functional similarity to rainforest than to dry (open) sclerophyll forest (Adam 1987, 1992; Tng et al. 2013; Warman and Moles 2009; Warman et al. 2013), the nature of the understorey plant community, and edaphic and microclimate conditions, allowing the diffusion of otherwise rainforest-restricted animals and ecological interactions. And, indeed, if wet sclerophyll forest is considered a transient seral stage in the establishment of rainforest (with species such as Eucalyptus grandis and Lophostemon confertus being relegated to ‘early mesic forest colonizer’ status), then wetter mixed forest communities with sclerophyll overstoreys may be considered within the category of rainforest (Dale et al. 1980; P. Richards 1952; Schimper 1903), albeit, as Adam (1992) notes, one ‘with special features’. Ecological processes are not constrained by floristic boundaries, and the invertebrate biota of eucalypt-dominated mesic forests often reflects close ancestral relationships; thus the discussion that follows often considers the interrelated natures of rainforest communities and mesic sclerophyll forest, frequently acknowledging associated ecological and biotic attributes entwined between the two.

Figure 1: Distribution of forests and rainforests in Australia, 2013. This book focuses on the subtropical rainforests of New South Wales and Queensland, shown in black shading within the boxed area. Source: ABARES, Commonwealth of Australia, 2016, CC BY 3.0. http://www.agriculture.gov.au/abares/forestsaustralia/profiles/rainforest

Figure 2: Locations mentioned in text. Queensland: 1. Blackdown Tableland, 2. Kroombit Tops, 3. Eurimbula NP, 4. Goodnight Scrub, 5. Fraser Island, 6. Cooloola NP, 7. Conondale NP, Conondale Range, Imbil State Forest, Kenilworth SF, 8. D’Aguilar NP, D’Aguilar Range, Mt Glennie, Mt Glorious, Mt Mee SF, Mt Nebo, 9. Cunninghams Gap, Lamington NP, McPherson Range, Main Range, Mount Barney NP, Mt Tamborine, Springbrook NP. New South Wales: 10. Acacia Plateau, Boatharbour NR, Border Ranges NP, Brunswick Heads, Cambridge Plateau, Koreelah NP, Lismore, Mallanganee, Moore Park NR, Mount Nothofagus, Mt Warning NP, Murwillumbah, Nightcap NP, Nightcap Range, Richmond Range NP, Stotts Island NR, Terania Creek, Tooloom Scrub, Unumgar SF, Victoria Park NR, Whian Whian SF, Wiangarie SF, Yabbra SF, 11. Iluka NR, 12. Forest Lands SF, Gibraltar Range NP, Washpool NP, 13. Chaelundi NP, Mt Hyland NR, 14. Bruxner Park, Carrai SF, Dorrigo NP, Dorrigo Plateau, Glenugie Peak, Hastings Valley, Mt Banda Banda, New England NP, League Scrub, Nulla Five-day SF, Styx River State Forest, Werrikimbe NP, Willi Willi NP, 15. Boonanghi SF, Crescent Head, Kerewong SF, Laurieton, Macleay Valley, Sea Acres NR, Way Way SF, 16. Black Head, Boorganna NR, Cape Hawke, Comboyne Plateau, Coocumbac Island NR, Coorabakh NP, Dingo Tops, Dooragan NP, Harrington, Kiwarrak SF, Lansdowne SF, Lansdowne NR, Lansdowne-Comboyne Escarpment, Lorien Wildlife Refuge, Tapin Tops NP, Wingham Brush NR, 17. Allyn River, Barrington Tops, Bulahdelah, Chichester SF, Mt Royal Range, Tuglo Wildlife Refuge, Woko National Park, 18. Cedar Brush NR, Liverpool Range, 19. Watagans SF, 20. Blue Mountains, Kanangra Boyd NP, Mt Irvine, Mt Wilson, Wollemi NP, 21. Barrengarry Mountain, Cambewarra Mountain, Illawarra Escarpment, Jamberoo Pass, Minnamurra Falls, Moreton NP, Mount Keira, Robertson Plateau, 22. Brown Mountain, Clyde Mountain, Deua NP, Durras Mountain, Mount Dromedary.

What is ‘subtropical’ rainforest?

This book considers rainforests and their included biota in a region extending in eastern Australia from about Rockhampton in the north to the south coast of New South Wales. In this sense ‘subtropical’ is a convenient geographical construct within the latitudinal scope of which there is considerable heterogeneity of rainforest vegetation, stand disjunction and size, climate, soil types, and landscape relief. The included biota is a rich one, reflecting diverse evolutionary histories. Using a structural–physiognomic classification the term ‘subtropical’ spatially reduces the concept to a suite of notophyll vine forest subformations categorised by complexity of leaf and growth forms, and diversity of strata (see discussions in Adam 1987, 1992; Winter et al. 1987). Although the floristic differences in individual stands, owing to changes in soil type, can be striking, rarely can a clear floristic or structural boundary be drawn between what are frequently intergrading rainforest subformations and associations. High numbers of tree species and diverse canopy architecture mosaics define more complex stands, with localised patterns of species recruitment and dominance reflecting episodes of chance dispersal, colonisation and extinction. Altitude alone may allow coarse categorisation of stands, but by itself is not a predictor of species presence/absence or structural complexity.

Although there is a large suite of trees and shrubs that are characteristic of subtropical rainforest, as classically considered in New South Wales (e.g. Adam 1987, 1992; Floyd 1990; Harden et al. 2006; McDonald and Hunter 2010; and others), a strict floristic definition is not possible. Variation is the dominating theme and so assigning individual stands to the various floristic alliances that have been proposed (e.g. Floyd 1990) is problematic, and may overlook the dynamic nature of plant composition. Numerous tree species are common to a range of subformations, and are found at greatly varying altitudes and latitudes, such that the occurrence of ‘within stand’ populations of many cannot always be predicted. Thus the use of floristic criteria to recognise and classify the various alliances that have been proposed, as Adam (1992) notes, may be difficult to apply. However, if ‘dry rainforest’, ‘warm temperate rainforest’ and ‘littoral rainforest’ are considered as floristic derivatives of ‘subtropical rainforest’, then the distinction can be more narrowly stated as formations that are subtropical by virtue of obvious floristic and structural diversity and upland stands (in which microphyll–nanophyll leaf forms dominate) that are intuitively ‘cool temperate’. But regardless of how individual formations are classified or perceived Australia’s subtropical rainforests, as a geographical entity, are a rich storehouse of biological diversity.

Defined solely by latitude, the greatest concentration of Australia’s subtropical rainforests occurs mainly from southern Queensland to northern New South Wales (Adam 1987, 1992; Floyd 1990); this is also a region where southern temperate faunas and floras overlap in rainforests occurring along the Great Dividing Range. However, in addition to this core area there are geographically isolated rainforest stands that extend archipelago-like from about Gladstone and Rockhampton in central eastern Queensland to the western slopes and south coast of New South Wales (Adam 1992; Curran et al. 2008; Harden et al. 2006; Helman 1987; Mills 1987; Kooyman et al. 2012), with floristically depauperate rainforests occurring in eastern Victoria, at altitudes <500 m (which are considered temperate examples [Cameron 1987]). There is a distinct north–south loss of floristic richness, but within this trend there are southern stands of high floristic diversity (e.g. Sea Acres Nature Reserve in Port Macquarie, northern New South Wales).

The formation of Gondwanan landmasses

The term ‘Gondwanan’ can imply patterns or processes, such as vicariance events due to the fragmentation of the supercontinent, or distribution in the Southern Hemisphere by more recent processes such as long-distance dispersal of younger age.

The diversification of numerous plants and animals relates to the breakup of the megacontinent Pangaea (Jokat et al. 2003; McLoughlin 2001; Taylor et al. 2009), which itself formed from the coalescence of earlier landmasses, with the evolution of many living families of insects, and other terrestrial invertebrates and vertebrates, originating during the Mesozoic (collectively the Triassic, Jurassic, Cretaceous periods) (Grimaldi and Engel 2005), an era of geological time extending from ~250 to 65 mya. The breakup of Pangaea, beginning in the Early–Middle Jurassic (~175 mya) and giving rise to the supercontinents Laurasia (North America, Europe, Asia) and Gondwana, shaped the development and subsequent widespread separation (vicariance) of closely related invertebrate lineages (e.g. Borda et al. 2008; Martínez-Aquino et al. 2017).

Following the rifting of Pangaea, Gondwana further split, initially into East Gondwana (Africa, Madagascar, India) and West Gondwana (Australia, Antarctica, South America, New Zealand, New Caledonia). During the Early Cretaceous (121 mya) Madagascar and India broke away from Africa, rafting to the south-east, with India separating from Madagascar in the Late Cretaceous (88–84 mya). India then drifted northward to eventually collide with Asia in the Early Tertiary (~50 mya). South America began to separate from Africa in the Early Cretaceous (135 mya), moving south-west to join Antarctica, together with Australia and New Zealand, as a single landmass. Subsequently, these separated from each other in the Oligocene (30–28 mya). Australia had finally separated from Antarctica by ~35–33 mya, moving northwards to its present position (Neall and Trewick 2008). There are variations on this time scale, and competing arguments as to the nature of plate/landmass associations and divergences (e.g. Ali and Aitchison 2008; McCarthy et al. 2007; Sanmartín and Ronquist 2004; Waters and Craw 2006). In particular, molecular phylogenetic studies have challenged the traditional timetable of the sequential continental breakup, indicating that the connection among major landmasses might have been prolonged until the Late Cretaceous based on the timing of divergence (e.g. Van Bocxlaer et al. 2006; Krause et al. 1997); this implies an extended opportunity for biotic interchange.

Nevertheless, as Dennis McCarthy and his colleagues reflected (McCarthy et al. 2007), Gondwana is a biological and paleontological phenomenon, ‘and today remains an elegant symbol of the intricate relationship between geological and biological history’.

These closed forest communities possess an invertebrate biota that reflects diverse evolutionary histories (see Williams 1993, 2002, 2003a; Williams and Bickel 2010), upon which, together with my own field records over the last 40 or so years, much of this book is founded. In addition to the impact of climate oscillations over long spans of geological time (see Appendix 5) the evolution of the Australian subtropical biota has been influenced by colonisation from the north as the Australian continental plate rafted towards Malesia; this is the biogeographical region that includes the Malay Peninsula, Indonesia, New Guinea and the Bismark Archipelago. Nevertheless, a high proportion of the fauna appears to be autochthonous (considered to have originated in this region).

The rainforests represent a significant zoogeographical refugium in terms of the evolution of Australian invertebrate taxa (see Bryant and Krosch 2016 for a review of eastern Australian biogeographical barriers to rainforest taxa). This is especially true of the Gondwana Rainforest World Heritage Area (formerly known as the Central Eastern Rainforest Reserves of Australia, or ‘CERRA’) located from the Barrington Tops in north-eastern New South Wales to south-eastern Queensland (Adam 1987, 2017; Kitching et al. 2010; see Appendix 2). In this book I use ‘Gondwana rainforests’ (as in Kitching et al. 2010) to indicate taxa recorded from the World Heritage Area. It replaces the acronym ‘CERRA’ used in Williams (2002) and Hunter (2003). ‘Gondwana rainforests’ is a useful identifier, specifically for World Heritage Area sites, that highlights the biodiversity and biogeographical significance of the World Heritage reserves (Adam 1987, 2017), rather than the alternative of the repetitive and unwieldy ‘northern New South Wales–southern Queensland’, and allows a cross-reference to the individual invertebrate records previously and more fully cited in Williams (2002, 2003a).

Examples of evolutionarily significant invertebrate taxa include the terrestrial snail family Charopidae, the freshwater snail family Hydrobiidae, the beetle taxa Adeliini, Cyphaelini, Dichotomini, Denticollinae, Rutelinae, Melolonthinae, Lucanidae, flightless beetles in the family Carabidae, the fly family Drosophilidae and dolichopodid subfamily Sciapodinae, parastacid crayfish and their ectosymbiotic flatworms, aradid bugs, the Onychophoran family Peripatopsidae and mygalomorph spiders (Williams 2002). Many taxa are thought to be ‘Gondwanan’ in origin (see Heatwole 1987; Main 1987), this term taking its name from the indigenous Gond tribe of India. Their extant distributions suggest a Gondwanan origin, but the occurrence of taxonomically related organisms in these southern rainforests does not necessarily establish a Gondwanan ancestry. Some groups are cosmopolitan or have widely distributed rainforest-associated relatives. Others have invaded through periods of geological time from the north or are related to southern Australian temperate rainforest invertebrate faunas, and are restricted to rainforests at higher elevations; these are captured within the ‘subtropical’ net of this book simply by virtue of the presence of distinctive faunistically overlapping zones in this range of latitudes.

Apparent extant biogeographical patterns are seductive, but may not truthfully reflect past events and lineages. In some instances, as with the conspicuous rainforest ant genus Leptomyrmex, a series of selective extinctions have resulted in an apparent Gondwanan affinity simply because organisms have disappeared from elsewhere (Lucky and Ward 2010; Williams and Bickel 2010). This applies equally to the bombyliid fly genus Comptosia, for although the extant fauna is restricted to Australia, with close relatives in southern South America, Eocene (56–33.9 mya) fossils from Germany and North America argue against a previously held view that the genus was of Gondwanan origin (Wedmann and Yeates 2008). And sometimes plants and animals display an asymmetrical ancestry, as in the instance of several leafhopper bugs (Membracidae), which occur on the distinctly Gondwanan plant families Myrtaceae and Proteaceae yet the bugs themselves are more geologically recent invaders (Wallace and Deitz 2006). In addition, many northern invaders reach their limits of distribution in the subtropical rainforests of northern New South Wales and southern Queensland. There is often a latitudinal ‘sifting’ or ‘filtering’ as particular taxa extend south or north along the east coast (as in the beetle family Carabidae and numerous land snails).

Alfred Russel Wallace, the father of evolutionary biogeography

This book starts with a dedication to a scientist that few may have heard of: Alfred Russel Wallace (1823–1913). He was a self-educated optimistist and Utopian socialist of impoverished means, and ultimately obliging and generous to his detriment. By unanticipated twists of circumstance Wallace declined credit for first establishing the theory of evolution by natural selection, yet founded the science and theory of biogeography (or zoogeography for those who study only animals), the study of the geographical distribution of plants and animals (Wallace 1876). Wallace independently discovered natural selection (Costa 2014), and is notoriously known for writing the letter that prompted Charles Darwin to finally construct a hurried outline of his theory that would be formulated in his book, On the Origin of Species. Darwin had digested his thoughts on the subject for many years, but had written nothing concrete or finished. Driven by economic necessity, the unmarried and relatively obscure Wallace had spent years in the Moluccas (what is now largely Indonesia), studying and collecting animals on the many islands there. In February of 1858, alone on the small island of Ternate and suffering from the debilitations of malaria, Wallace wrote On the Tendency of Varieties to Depart Indefinitely from the Original Type; the so-called ‘Ternate paper’. It was a complete and lucid picture of evolution by natural selection. His paper was 3764 words long. At the time Darwin’s thoughts were unformulated, for he was still grappling with the riddle of divergence, and any final resolution seemed beyond his grasp. In early March of that same year, Wallace, for seemingly good personal reasons, sent his paper to Darwin (whom he revered) for comment, and in the hope that Wallace’s idea would assist Darwin by supplying the missing factor that Darwin needed to explain the origin of species. The rest is contentious history (Brackman 1980). Wallace returned to Britain to write many things about many subjects (Berry 2002), but until the end he stood in loyal defence of Darwin – not least of all in the pages of Darwinism, Wallace’s authoritative summary of evolutionary biology published in 1889, several years after Darwin’s death. Wallace’s Line (also called the Wallace Line), that demarcation (though now perhaps a scientific curiosity) between the particular absence and presence of species, and located between Borneo and Sulawesi and Bali and Lombok, still bears fitting tribute to him, as does the transitional zone known as ‘Wallacea’ (see Holt et al. 2013a, 2013b; Kreft and Jetz 2010; 2013; and Proche and Ramdhani 2012 for a discussion of zones). But the islands in the time of Wallace, then so rich in forests and wildlife, live in their former biological glory now only on the pages of his monumental The Malay Archipelago (van Oosterzee 1997).

Nevertheless, there is a significant number of taxa with Gondwanan or ‘Old Southern endemic’ affiliations (Table 1). This group includes taxa with relatives on other Gondwanan landmasses such as South America, Africa and the Indian subcontinent, and those restricted to Australasia. Some are restricted to Australia but with close relatives in either New Caledonia or New Zealand, but not both. Within known ranges individual taxa can be widespread or relictual. Examples of ‘southern’ fauna exist across all taxonomic levels – from species to that of subphylum. Notable higher taxonomic rank examples are the megascolecid earthworms, mygalomorph and amphectid spiders, harvestmen in the family Acropsopilionidae, the mite family Pheroliodidae, and the terrestrial snail families Athoracophoridae, Charopidae, Cystopeltidae and Rhytididae. The insects are well represented, and include the beetle tribes Pamborini, Adeliini, Epistomentini and Stigmoderini, the subfamily Spilopyrinae, the family Migadopinae, flies in the family Pelecorhynchidae, the subfamily Arachnocampinae and the tribe Pangoniini, the moth families Hepialidae and Micropterigidae, the birdwing butterfly genus Ornithoptera, the hemipteran bug families Idiostolidae and Peloridiidae, the neuropteran lacewing subfamilies Kempyninae and Stenosmylinae, ambrositrine and hyptiogastrine wasps, and the plecopteran families Austroperlidae, Eustheniidae and Griptopterygidae.

The region is also one in which the fauna exhibits a high degree of short-range endemism (Table 1) (notably in stream-, cave- and karst-associated species); this is defined by species with a range <10 000 km² and further characterised by poor powers of dispersal and confinement to discontinuous habitats (Harvey 2002), though not always rainforest. Groups in this category include Oligochaeta, Onychophora, Mygalomorphae, Schizomida, Diplopoda and Decapoda. With respect to subtropical rainforest only, much of the terrestrial and freshwater invertebrate fauna is endemic at various hierarchical levels, particularly those of species and genus but also extending to that of family. In addition, many species are currently known only from single localities. Instances of generic and species endemism are particularly high in the snail families Hydrobiidae and Charopidae; the earthworm family Megascolecidae; the crayfish family Parastacidae; the flatworm family Temnocephalidae; the subphylum Onychophora; the mite family Platyameridae; and the spider suborder Mygalomorphae. Examples within the insects are the fly families Dolichopodidae, Platystomatidae, Exeretonevridae, Pelecorhynchidae and superfamily Tipuloidea; the beetle tribes Onthophagini and Canthonini; the beetle families Buprestidae, Lucanidae, Carabidae, Rhinorphidae and Lamingtoniidae, and scarabaeid subfamilies Melolonthinae and Rutelinae; australembiid webspinners; cicadelloid and mezirine Hemiptera; oechophorine moths; and the lacewing family Hemerobiidae and the subfamily Kempyninae (Williams 2002).

However, significant invertebrate biodiversity values in south-eastern mainland Australia are not restricted to taxa inhabiting subtropical rainforest. Although subtropical rainforest, before European occupation, did include notable examples of expansive rainforest stands (e.g. lowland Illawarra, the Comboyne and Dorrigo plateaux, and the Big Scrub rainforests in New South Wales), subtropical rainforest more normally is encountered as patch-like stands of varying size and floristic and structural composition in an intervening matrix of sclerophyllous forests, and more rarely woodlands, shrub, heath and swamp-mire complexes, and associated freshwater ecosystems. These habitats sustain distinctive non-rainforest invertebrate faunas that nevertheless are often ecologically interactive (e.g. the recruitment of certain pollinators and decomposers) or share a close evolutionary ancestry. Important higher taxa inhabiting these ecosystems are ruteline Christmas beetles in the tribe Anoplognathini, the family Elateridae, the speciose buprestid genus Castiarina, the tenebrionid tribe Heleini, the fly genera Pelecorhynchus and Trichophthalma, petalurid dragonflies, lestoideid damselflies, and athoracophorid, helicarionid and glacidorbid snails. Limestone outcrops, often set within a rainforest context, are important focal points of terrestrial snail endemism and diversity (Stanisic 1994, 1997; Stanisic et al. 2010). Cave systems are important for the conservation of associated, often endemic or localised, arachnid and insect faunas. Boggy seepages and swamp-mires are important habitat for the conservation of several endangered endemic dragonflies (Theischinger and Hawking 2006). Dissecting freshwater streams (Plate 310) are critical habitats for the larval and adult stages of many rainforest invertebrates, and for the functioning of broader ecosystem food webs. However, seasonal changes in flow rates, climate stability, substrate dynamics and individual life histories strongly influence the biomass, trophic structure, functionality and species richness of in-stream invertebrate assemblages (Cummins and Lauff 1969; Griswold et al. 2008; Lancaster et al. 2009; Pearson 2014).

Table 1. Examples of terrestrial invertebrate taxa exhibiting high levels of endemism, Gondwana affiliations or zoogeographical significance within the Gondwana Rainforests region (after Williams and Bickel 2010 ).

The natural and evolutionary history of Australia’s subtropical rainforests has been previously discussed by Adam (1987, 1992), Barlow (1981), Baur (1957), Connell et al. (1984), Floyd (1990, 2010), Hunter (2003), Rossetto et al. (2015), Webb (1978), Webb et al. (1984), Young and McDonald (1987), and others. These provide an overview and analysis of Australian subtropical rainforest communities. A introductory review of the significance and conservation of invertebrates in New South Wales rainforests was given by Nadolny (1984), and invertebrates of the Gondwana rainforests region have been discussed in Adam (1987), DASET (1992), Williams (1993, 2002, 2003a) and Williams and Bickel (2010).

The invertebrates discussed here constitute an indication of the known fauna. It is not an exhaustive treatment, nor is the structure of presentation a strictly systematic one; and indeed it is somewhat idiosyncratic (though not too dissimilar to the artificial categorising of ‘moths’ and ‘butterflies’). Some groups are absent. Families, and in several instances higher ranks or informal groups (such as Sphaerotheriida, Psylloidea, Tipuloidea, Hydradephaga), are listed alphabetically. This is intended to facilitate access by non-specialist readers and reduce unnecessary complexity in the Index. However, Appendix 4 indicates the position of families or superfamilies that are discussed in the text within higher divisions of Coleoptera, Diptera, Hemiptera, Hymenoptera and Mollusca.

Individual distribution records can be found in Williams (1993, 2002, 2003a, and papers cited therein); however, these represent only a glimpse of the published records. A considerable body of later work has been published, and this I have drawn upon. I have also cited references that may be considered dated, but these have more than historial significance (nor are they necessarily obselete) for they demonstrate the foundations upon which current trends and understandings of the fauna stand. Nevertheless references such as the Zoological Catalogue of Australia series and Insects of Australia are increasingly out of date (and thus need to be considered cautiously), yet remain gateways to an understanding of the Australian fauna. The discussion also draws heavily on the many years of my own field work in the region, subsequent to the publication of earlier reviews. This personal field work has encomposed a much broader group of forests and included several significant New South Wales north coast rainforests that are not in the World Heritage Area (the focus of Williams 2002, 2003a; Williams and Bickel 2010). These rainforests include endangered floodplain rainforest remnants, isolated pockets in the Liverpool Range, mountain sites in close proximity to the coast (e.g. Glenugie Peak south-east of Grafton, Way Way State Forest near Nambucca Heads, North and Middle Brother mountains at Laurieton), and extensive areas in the New South Wales Lower Hastings, Camden Haven and Manning River Catchments (notably the montane forests of the Dingo–Tapin Tops, Comboyne Plateau, and Lansdowne–Comboyne Escarpment). Other areas include the rainforests in the Myall River at Bulahdelah, and Newcastle regions, small stands such as Mountain Lagoon and Sassafras Gully in the Blue Mountains west of Sydney, forests on the New South Wales central and south coast, and Queensland rainforests extending to the outlying Bunya Mountains and north to Kroombit Tops, Eurimbula and the Blackdown Tableland; as well as Araucarian vine thickets located in central eastern Queensland (such as Goodnight Scrub). Littoral rainforests, a series of distinctive subtropical–dry rainforest subformations now occurring as remnants on headlands and Holocene sands (see Adam et al. 1989; Floyd 1990; Williams 1993), have been extensively visited.

The mixed warm temperate and subtropical dominated rainforests of the Lansdowne–Comboyne Escarpment have been investigated over many years (owing to the presence there of my own property, the Lorien Wildlife Refuge and Conservation Area); the long-term study through various seasons, as well as opportunistic encounters, has resulted in the recognition of the escarpment’s role as a refugium for numerous species originally thought only to occur on the Comboyne massif itself; recognition of many previously unrecorded plant–animal relationships; new and rare plant discoveries; and the discovery of invertebrate species new or poorly known to science, and species previously believed restricted to rainforest communities at much higher altitudes or at lower latitudes. In this Lorien has served the function of a long-term field studies station by default, a purpose similar to that of privately owned forests at Mt Glorious and Mt Tamborine (Queensland), and Tuglo Wildlife Refuge (north-west of Singleton) (Smithers 1981) in the Mount Royal Range (New South Wales).

Much rudimentary inventory work remains to be done to assess the invertebrate faunas of individual Australian subtropical rainforest stands, but, even with the knowledge at hand, it is evident that even quite small, physically degraded and geographically isolated remnants can possess many ecologically specialised invertebrate species and numerous Australian endemics. These possess high heritage and scientific reference attributes not necessarily suggested by the often disturbed state of the rainforest structure, and the impoverished abundance and diversity of the surviving plant and vertebrate communities. The endangered floodplain rainforests of Wingham Brush and Lansdowne nature reserves, surviving in an urban and agricultural context on the lower north coast of New South Wales, are cases in point. In particular, littoral rainforests, listed as endangered ecological communities under state and federal legislation, retain high invertebrate biodiversity values although now just remnants of a once greater area (following sand-mining and clearing for agriculture or residential development).

Despite being the subject of more than 200 years of scientific investigation there remain significant inventory data gaps in our knowledge of higher taxa (families and above in the taxonomic hierarchy) within Australia’s floristically diverse subtropical rainforests. Many taxa have limited dispersal capabilities and, as a consequence, serve as useful tools for better identifying and understanding geological refugia and speciation events (e.g. carabid, tenebrionid and scarabaeine ground beetles, Phasmatodea). Other taxa have specialised and obligate host dependencies, or can function as indicators of environmental impacts (e.g. ants, in-stream fauna), and as such can serve to highlight inadequacies and the need for changes in habitat management strategies. Examples of rainforest-inhabiting taxa with inadequate knowledge bases are hydrobiid freshwater snails, terrestrial leeches, land planarians, flatworms, millipedes, and the cricket families Anostostomatidae and Rhaphidophoridae. Even the composition and distribution of such conspicuous invertebrate groups as bees (Apoidea/Apiformes) in montane rainforests are poorly known. Historically, considerable collecting effort has been targeted towards butterflies, jewel beetles (Buprestidae) and ground beetles generally, but little published data, specific to individual rainforest reserves and stands, are available. This restricts the foundation upon which management plans can be built.

Various authors have used a variety of terms to describe and define biogeographical zones and provinces (e.g. Allsopp 1995; Burbidge 1960; CSIRO 1996; Matthews 1972; Matthews and Bouchard 2008; McAlpine 2001; McMichael and Hiscock 1958; Smith and Kershaw 1979). Frequently these overlapped in range or were synonymous, and often had doubtful validity when applied across major taxonomic boundaries (i.e. utility is restricted to individual taxa). The terms ‘Eyrean’, ‘Bassian’ and ‘Torresian’ can now be viewed as obsolete and clouding perceptions of the development and origins of the Australian fauna, and may be of little use in delineating faunal zones to the level of definition demanded by most modern taxonomists and biogeographers. More recently C. Reid (2016) has used the term ‘Australopapua’ to define a faunal province that includes Papua, the Australian mainland and its immediate off-shore islands (such as Kangaroo Island and those of Bass Strait), and Tasmania, but excludes Lord Howe Island and other more distant islands that are defined as ‘Australian’ by virtue of political inclusion. However, rather than attempt to bring the various definitions into a standardised system, with risk of distorting or misinterpreting the original concepts, these have been retained.

The very concept of zoogeoraphic divisions, however, is open to question. This is because the degree of faunal overlap may defy recognition of distinct regions, we lack the benefit of hindsight to accurately relate extinction, radiation and speciation histories, and the high number of undescribed and poorly studied invertebrate taxa may place us in a position premature for understanding taxonomic affiliations and zoogeographical boundaries. Concepts such as Matthew’s and Stanisic’s ‘refugium’ areas (see discussion under Coleoptera – Tenebrionidae, and terrestrial Mollusca) may have greater validity for characterising more discrete biogeographical zones important in the evolution of the Australian invertebrate fauna.

As a caveat, it should be highlighted that biogeographical patterns, and the inferences and analyses that can be drawn from them, are only as good as (1) the level of the underlying taxonomy, (2) the adequacy of field sampling at individual sites and landmasses, and (3) known fossil history. Anomalous disjunct distributions, for example, may prove with time to be artefacts of insufficient sampling and deficient taxonomic rigour, or the result of accidental introduction.

2Australia’s subtropical rainforests – the plant context

Rainforests within the region of subtropical Australia’s east coast include formations of considerable structural and floristic diversity (Adam 1987, 1992; Baur 1965; Floyd 1990; Harden et al. 2006; Helman 1987; Shapcott et al. 2015; Williams et al. 1984; Young and McDonald 1987). Floristic alliances and suballiances for New South Wales rainforests are detailed in Floyd (1990). Tidally inundated mangrove communities (Plate 315) may also be considered a distinctive rainforest formation (Adam 1992) (see ‘Mangroves – a distinctive maritime rainforest formation’ for discussion).

Mangroves – a distinctive maritime rainforest formation

Mangroves, and mangrove communities (Plate 31) more inclusively, are generally not thought of as a rainforest ecosystem. Traditionally, mangroves have been considered a distinct vegetation type, but on structural grounds (that of a ‘closed canopy’ in more developed stands), their placement within the broader conceptual notion of ‘rainforest’ may be appropriate. This is not a new idea, for their consideration as but one of numerous rainforest formations has previously been discussed and advocated by Paul Adam (1992), and regarded as such by Whitmore (1984). Citing the structural schemes of Specht (1970, 1981), in which ‘closed forest’ and ‘rainforest’ are treated as synonyms, Adam opines that mangroves should be regarded as a distinctive salt-tolerant maritime rainforest formation.

Mangroves are usually restricted to the intertidal zone. In Australia they are widely found along the eastern and northern coasts but are essentially tropical and subtropical in distribution (Adam 1992; Specht et al. 1995). Only five species reach the subtropical region of the east coast and only two, Avicennia marina (Avicenniaceae) and Aegiceras corniculatum (Myrsinaceae), are commonly encountered at higher latitudes. They vary structurally from closed canopy forest to open woodland, being floristically diverse at lower latitudes and most extensively developed in protected estuary zones. Regional variation in estuary length and size, temperature, salinity, groundwater, tidal amplitude and patterns of inundation, and possibly rainfall, influence species richness (Reef and Lovelock 2015; Saenger and Moverley 1985; Smith and Duke 1987). Some mangrove stands may represent a transitional community, but exposed shores and hypersalinity can act as a barrier to colonisation (Adam 1992; Mitchell and Adam 1989).

Their tidal-prone location, however, greatly restricts the kinds of terrestrial invertebrates that can occupy the mangrove environment. Various crustaceans and semi-aquatic and semi-arboreal molluscs characterise resident populations, but truly terrestrial forms (such as ants), particularly those species that at some stage of their life cycle are dependent upon the terrestrial substrate, are subject to tidal inundation, and so are limited to incorporating mangrove communities only within ephemeral foraging patterns (e.g. visiting flowers), but not occupation. Those that may be encountered comprise species capable of active flight (e.g. butterflies, cicadas, wasps, bees, flies, certain beetles) and largely sessile taxa, which, once able to reach foliage and branches (e.g. spiderlings that may disperse in air currents), are then able to survive because their immature life stages are not dependent on temporal residency under or on the intertidal-zone ground surface, prey items are dependable, and their habitat remains fully or partly above cyclical tide levels.

Climate, soil type and fertility, underlying geography, altitude and latitude, aspect and gradient, disturbance history (such as fire and wind impacts [Bowman 2000; Lange et al. 1981]), plant spatial and temporal heterogeneity, and random variation in conditions affecting reproductive or mortality rates influence both the extent and nature of individual stands (Adam 1992; Connell 1978; Floyd 1990; Williams and Adam 1999b). Rainforests trend to less floristic and structural complexity with increases in altitude and at higher latitudes, with decreased rainfall, and with lower soil fertility. Rainforest at high (temperate) altitudes experience greater seasonal variation in temperature, those at Barrington Tops being subject to snowfall. Individual plant species (and invertebrates [Lambkin et al. 2011; Majer et al. 2001; McKie et al. 2005; Ollerton and Cranmer 2002; Ødegaard and Diserud 2011; Yeates 1985]) respond to gradients in these (e.g. Kooyman et al. 2012; Laidlaw et al. 2011), such that rainforest plant communities at high latitudes, and at sites of low or irratic rainfall and low soil fertility, are floristically depauperate. Epiphytes respond to micro-environmental variation (Sanger and Kirkpatrick 2017), but in rainforests growing in such environmental extremes epiphytes may be noticeably few, or absent. Rainfall alone is a major factor in determining the complexity of rainforest, with annual rainfall in the region exceeding 1400–2000 mm (though not uniform throughout the region). But patterns of precipitation, not just total annual rainfall, also compound rainforest establishment and survival (and that of individual plant species). Chance events of seed dispersal (or conversely seed theft and predation) can also influence the floristic composition of stands as much as a predisposition to successful foundation, possibly more so. To all of these, individual invertebrate taxa, in addition to plants, respond accordingly.

Rainforest canopies occupy the interface between the biosphere and the atmosphere. They are exposed to impacts of wind, rain, humidity, and seasonal and daily temperature regimes to an extent far greater than the strata below. The frontal canopy of wind-sheared rainforest on exposed headlands (see Plate 305) may not exceed 1 m in height; deformed trees found there (e.g. Syzygium smithii – Myrtaceae) are able to grow to greater height in protected areas of the stand (Adam et al. 1989). Canopies also experience greater fluctuations in air chemistry, and for littoral rainforests there is the added burden of salt impacts (Parsons and Gill 1968; P. Adam pers. comm.). Plants and animals living at the edges of the canopy must be adapted physiologically and behaviourally to tolerate or avoid these extremes (Ozanne 2013).

Differences, often marked, in tree species dominance and occurrence in separate stands is normal and the intergrading of plant communities is commonplace. There can be considerable differences in structure ranging from complex canopy and subcanopy formations, to simple low vine thickets that, nevertheless, meet Andreas Schimper’s (1903) original definition of ‘rainforest’(regenwald in German). The rainforests of Australia’s subtropical region are characterised by floristic alliances (Floyd 1990; Webb et al. 1984; and the forestry-based classification of Baur 1965) and physiognomic/structural features (see e.g. Adam 1992; Beadle and Costin 1952; Webb 1959, 1978). McDonald and Hunter (2010) note the close correlation between Baur’s classification and the physiognomic/structural classification of Webb (1959, 1978). Historically, the rainforest subforms in New South Wales have been frequently termed ‘subtropical’, ‘dry’, ‘warm temperate’ and ‘cool temperate’ (Baur 1965; Harden et al. 2006; Williams et al.1984, and others). Significant occurrences in Queensland and New South Wales are given in Appendix 3. These four simple subform terms are used throughout this work, but it is important to highlight that rarely are they found as distinct structural and floristic units (dry rainforest–vine thicket patches within woodland and savannah dominated landscapes being examples of exceptions), instead occurring as intergrading communities with fluid boundaries.

Subtropical rainforest is the most structurally and floristically complex of the four types (Adam 1992; Floyd 1990), and altitudinally extends from sea level to >900 m (see Plates 301, 306, 310, 311). It includes rainforests alternatively termed ‘complex notophyll vine forest’, ‘notophyll vine forest’ and ‘araucarian vine forest’ (Webb 1978; Young and McDonald 1987). The canopy is uneven with up to three layers, trees are diverse with 10–>60 species in the canopy, entire (untoothed) leaf margins dominate, there are many species with compound leaves, leaf size is generally large with notophylls and mesophylls dominating (>0.5–25 cm long), trunk buttressing and the presence of palms and figs are common, and vines, ferns and epiphytes are conspicuous and diverse. Subtropical rainforest is favoured by high rainfall (>1300 mm), sheltered aspect, and richer volvanic and alluvial soils (after Harden et al. 2006; McDonald and Hunter 2010; Williams et al. 1984). Littoral rainforest is a subform predominantly of subtropical rainforest, found on headlands, behind dunes, and in coastal estuaries, but may include trees that characterise dry rainforest, such as the presence of Backhousia sciadophora (Myrtaceae) and associates at Black Head on the lower north coast of New South Wales (Williams 1993) (see Plates 304, 305, 313).

Trees that are typical of subtropical rainforest in northern New South Wales and southern Queensland include numerous species of Cunoniaceae, Elaeocarpaceae, Escalloniaceae/Rousseaceae, Lauraceae, Euphorbiaceae/Phyllanthaceae, Icacinaceae/Cardiopteridaceae, Meliaceae, Atherospermataceae/Monimiaceae, Moraceae, Myrtaceae, Proteceae, and Sapindaceae (Floyd 1990; Young and McDonald 1987). In northern New South Wales Dendrocnide excelsa (Urticaceae), Brachychiton acerifolius, Argyrodendron actinophyllum (Sterculiaceae/Malvaceae), and Gmelina leichhardtii (Verbenaceae) are frequently encountered as members of the upper canopy. Araucaria cunninghamii (Araucariaceae) sometimes occurs as an emergent, and Archontophoenix cunninghamiana palms (Arecaceae) are common, frequently dominating the canopy of sheltered gullies. At higher latitudes (e.g. Durras Mountain, Mount Dromedary) mixed subtropical rainforest–warm temperate rainforest occurs at lower elevations in sites of higher soil fertility. Trees are less diverse with Syzygium smithii (Myrtaceae), Dendrocnide excelsa (Urticaceae), Citronella moorei (Icacinaceae/Cardiopteridaceae), Ficus obliqua (Moraceae), and the palm Archontophoenix cunninghamiana (Arecaceae) being characteristic species (Helman 1987).

Endangered subtropical rainforest in New South Wales and Queensland

The total rainforest in Australia is ~2 000 000 ha, this surviving mainly as scattered patches.This is a very small area, less than the total Amazon rainforest cleared annually. Australian rainforest is spread over a latitudinally extensive area, its origin predating that of the continent’s now iconic eucalypt forests and woodlands. Of the moister coastal rainforest it is estimated that two-thirds of the original Queensland forest and three-quarters of that present in New South Wales has been cleared; the area of original rainforest at European occupation therefore being ~80 000 km² (Webb and Tracey 1981; Winter et al. 1987). What remains are fragmented and often highly disturbed and isolated ‘habitat islands’(e.g. Bradshaw 2012; P. Rose 2014; Scanlan et al. 2018).

About 250 000 ha of rainforest survives in New South Wales (Pople and Cowley 1982). Of the remaining subtropical rainforest subform in New South Wales, ~73 400 ha is structurally intact. This is only equal to the estimated size of ‘The Big Scrub’ rainforest in 1842, of which only ~300 ha in small remnants now survives (Floyd 1990). Situated near Lismore in the far north-east corner of New South Wales, this was once the largest stand of tall subtropical rainforest in Australia. It took <40 years for European farmers to almost totally clear, its loss testimony to Australia’s thoughtless and ongoing destruction of native vegetation. A similar fate befell the luxuriant New South Wales north coast rainforests of the Dorrigo Plateau and Comboyne Plateau, and the lowland rainforests of ‘The Illawarra’ south of Sydney. Of the estimated original littoral rainforest only 1200 ha remains, this in numerous weed infested fragments, surviving largely from coastal sand mining. Dry rainforest is estimated to have ~76 100 ha remaining, but this includes floristically depauperate vine thickets and remnant scrubs on the Western Slopes, in addition to more complex stands. Warm temperate rainforest totals 46 600 ha; however, about half of this has been heavily logged with consequent structural and floristic changes occurring. The conservation status of cool temperate rainforest has been relatively unaffected by logging, though cool temperate stands on the Comboyne Plateau, in particular, have suffered from clearing. Now only relict stands of Nothofagus moorei survive there (Bale and Williams 1994).

Despite their acknowledged rich biodiversity (Adam 1987, 1992; Department of the Environment, Sport and Territories 1994; Kitching et al. 2010), however, Australia’s subtropical rainforests (and their included biota) remain under threat; directly or indirectly they are ensnared by the impacts of what has been termed the ‘Anthropocene human-induced mass extinction event’ (see Braje and Erlandson 2013; Ceballos et al. 2015). Throughout their range in Queensland and New South Wales, stands continue to be vulnerable to weed invasion, pathogen attack (e.g. Austropuccinia psidii, Myrtle Rust) (Williams 2018; Williams and Adam 2019), ongoing loss of connectivity (see Beier and Noss 1998), the deleterious impact of ‘edge effects’ (e.g. reduction in core area and vegetation buffers, changes to wind, temperature and light regimes, altered drainage patterns and nutrient inflow) (Fox et al. 1997), fire damage initiated from adjacent open forest and woodland, climate change and clearing (including underscrubbing) (Saunders et al. 1991). Logging operations in New South Wales state forests threaten isolated small stands due to boundary encroachment, and vine thicket communities in Queensland remain at risk from broadscale land clearing (see Plate 314). Feral animals also make significant impacts on the biota, potentially catastrophic ones, and though the threats to native vertebrates are generally understood, or can be predicted with a high degree of accuracy, the impact on the invertebrate fauna (e.g. by exotic rodents and cane toads [Shine 2010]) remains to be rigorously examined.

In response to this ongoing vulnerability state and federal legislation gives a measure of recognition and protection to surviving stands. In New South Wales various rainforest communities (e.g. ‘Lowland Rainforest in the New South Wales North Coast and Sydney Basin Bioregions’, ‘Lowland Rainforest on Floodplain in the New South Wales North Coast Bioregion, ‘Illawarra Subtropical Rainforest in the Sydney Basin Bioregion’) are considered endangered ecological communities under the Biodiversity Conservation Act 2016. In addition to being listed as an endangered ecological community under this Act, threats to lttoral rainforest were further recognised under the NSW State Environmental Planning Policy (S.E.P.P.) No. 26 – Littoral Rainforest. The federal Environment Protection and Biodiversity Conservation Act 1999 also lists several floristically diverse subtropical rainforest communities in New South Wales and Queensland as ‘critically endangered’ ecological communities (e.g. ‘Lowland Rainforest of Subtropical Australia’, ‘Littoral Rainforest and Coastal Vine Thickets of Eastern Australia’, ‘Semi-evergreen Vine Thickets of the Brigalow Belt [North and South] and Nandewar Bioregions’). How these communities will be protected in the face of increasing socio-economic pressure remains to be tested.

Biological consequences of rainforest fragmentation

The historical, anthropogenic, fragmentation of subtropical rainforest ecosystems has resulted in the creation of isolated, island-like remnant patches of various sizes existing within agricultural, and sometimes urban, landscapes (Fox et al. 1997; Horton 1999; P. Rose 2017; Scanlan et al. 2018; Williams and Gerrand 1998b, 1998c). These altered landscapes form matrices, usually with abrupt boundaries, which are hostile to forest organisms (Harris and Silva-Lopez 1992; Zipperer 1993) (Plate 314).

Fragmentation causes significant, and often dramatic, changes to the physical environment of forest remnants (e.g. alteration to fluxes in radiation, wind and water [Nobel 1981; Saunders et al. 1991]) and this in combination with the reduction in the area of remaining habitat has important consequences for the survival of individual relict plant and animal populations (Didham 1997; Cagnolo et al. 2009; Krosch 2011; Laurance and Bierregaard 1997). These impacts are often subtle and extend over long periods; examples are the breakdown and loss of species mutualisms (Harris and Johnson 2004), invasion by exotic plants and animals (Christian 2001; Taylor et al. 2011; Williams and Gerrand 1998a, 1998b, 1998c; see also Crichton et al. 2019 for a discussion of pollinator ‘super-generalism’ potentially facilitating weed invasion), loss of gene flow and genetic drift (Honnay and Jacquenym 2007), and the eventual failure of trees to reproduce (Hobbs and Yates 2003). Although population size might place rare species at a heightened extinction risk in fragments (Thomas 1993), common species may be as, or more, susceptible to the genetic consequences of habitat fragmentation as rare species (Honnay and Jacuenym 2007). Stork et al. (2017) highlight the fact that trends in species richness at local scales may not reflect the resilience of ecosystems to disturbance, and functional feeding groups (e.g. herbivores, coprophages, predators, pollinator guilds) may respond differently to individual disturbance events and fragmentation more broadly. And so consideration of the effect of fragmentation on functional composition, rather than just on numbers of species and their abundance within remnants, is also important when assessing disturbance threats.

The longevity of many plants can mask the loss of ecological function due to the extinction of insect pollinators and the subsequent absence of seed set (Williams and Adam 1999b). Surviving patches of subtropical rainforest are frequently characterised by large numbers of vascular plant species often at low population levels with individuals commonly scattered widely about in the stand. Pollen flow between distantly spaced conspecifics is predominantly non-directional (pollinators usually exhibit low fidelity) and so levels of pollen transfer between conspecifics are likely to be low and seasonally unreliable. This threat is magnified when plants are obligate out-crossers (Williams and Adam 1994, 1999b, 2010), or do not flower annually (Williams 1995). Even when a degree of pollen transfer and fertilisation is successful, but with lowered fecundity, levels of seed production can be insufficient to sustain recruitment back into the stand (e.g. in the face of seed predation pressures). The surrounding cleared land experiences alterations to the radiation balance, resulting in increased surface temperatures during the day and the potential for frost at night. This had adverse impacts on remnants, especially along edges, and higher latitudes can impose significantly more solar radiation (and thus higher temperatures) than that in unfragmented landscapes. A consequence of this is that rainforest edges may have a predominance of colonising plant species, and shade-tolerant species are restricted to the interior ‘core’ or at different distances along a tolerance gradient. Indeed, edge effects may be so intrusive that no area of the remnant remains unaffected. Nutrient cycling; soil microorganism function, abundance and diversity; soil moisture retention; the decomposition of leaf litter; plant growth and plant phenological patterns; and the survival of individual animal and plant species can also be affected directly and indirectly by soil heating and increased atmospheric temperature regimes (e.g. Klein 1989; Norton 1994; Saunders et al. 1991). Reduction in the area of rainforest, and its retention in otherwise cleared landscapes, alters the nature of air flow with resultant flow-on impacts involving changes to wind turbulence patterns, exposure and severity. These changes in turn lead to modification of the canopy structure (e.g. crown shearing, lower overall canopy height) and atmospheric gas fluxes above the canopy (Nobel 1981; see also Pincebourde and Casas 2016), as well as canopy destruction and crown damage (e.g. salt burn of littoral rainforest [see Adam et al. 1989; Parsons and Gill 1968]), increased vulnerability to isolated tree windthrow and consequent increased light regimes at ground level (with concomitant impacts on leaf chemistry [Frankel and Berenbaum 1999]), increased litter fall and thus alteration of animal microhabitat. Substantial increases in spring and summer temperatures, seasonal episodes of prolonged and repeated drought, and vulnerability to hot dry winds can lead to greater tree mortality and the suspension of invertebrate maturation and emergence cycles. Thermal tolerances in individual animal species may be exceeded (Hemmings and Andrew 2017; Kaspari et al. 2015), with localised microclimates acting as amplifiers of macroclimatic conditions, exacerbating even seemingly small fluctuations in temperature (Pincebourde et al. 2016). Canopy destruction can also occur where large populations of flying foxes concentrate in small remnants (as at Wingham Brush Nature Reserve and Maclean, northern New South Wales), resulting in loss of herbivore and pollinator resources. Fragmentation also alters the hydrological cycle, changing the rates of rainfall interception and evapotransportation, modifying the penetration of water into the soil profile, and changing decomposition rates and altering seed-bed characteristics. Attendent impacts on the suitability of habitat for the biota follow. The reduction in the area of habitat that fragmentation intrinsically involves results in a loss of carrying capacity in both species abundance and diversity (Howden and Nealis 1975; Recher et al. 1987), fewer microhabitat and niche opportunities, and increased vulnerability to