For Tony Harman, my dear friend and passionate entomologist, who nurtured in me a love of all things creepy crawly

Acknowledgements

I am grateful to many different people for their part in the completion of this work. Firstly, I owe a debt of gratitude to my principal supervisor Scott Johnson for his guidance throughout my time here. In particular, I am thankful for Scott’s unrivalled speed in turning around drafts, his writing acumen and the obvious investment he places in the interests of his students. I am grateful too for the support and encouragement provided by my supervisory panel – both near and far – by David Ellsworth and Jo Staley. I am thankful for the work of Sabine Nooten; her infectious passion for ants, her valuable experience of community ecology studies and our impromptu cups of tea. I am grateful to Philip Smith for his meticulous proof-reading skills.

Throughout my time here, I have had the honour of working with several enthusiastic volunteers and undergraduate students – including David Fidler, Marcel Torode, Jenni Kremer, Rhiannon Rowe and Aqeel Abbas – whose tireless, exemplary efforts in the field and lab made this work a possibility. I have had the pleasure of working alongside some wonderful students and post-docs, many of whom I will maintain life-long connections with. I am particularly thankful to my office-mate Jessica Rigg for her humour and friendship, as well as our (probably medically inadvisably-frequent) motivational cups of tea with Kylie Brice and Caroline Fromont. I am grateful that I got to share this PhD experience with my good friend and PhD-counterpart Kirk Barnett – his scientific rigour, statistical genius and philosophical discussions have been invaluable and helped to form the basis of a deep friendship.

Words cannot express the gratitude I have for the unwavering support of my partner William Balmont – for taking a risk with me and uprooting to the other side of the world; for listening to me talk about insects for years with sustained, sincere interest; for his tireless work in the field and laboratory in his free time; and for his unquestionable love and support throughout the emotional rollercoaster that completing a PhD entails. I am thankful too for the love and support of our dear friends Lanilà and Ivan Hiltpold, without whom life both in and out of work would have been a whole lot less spirited

and fulfilling. My thanks go to the Yoga Shed Richmond – especially Catherine Sherlock and our inspiring teacher and friend Anneriek Favelle. It has been a privilege to be a part of such a welcoming and enriching community, at a time when we needed it most.

I am thankful for the long-range, Skype enabled support provided by family and friends back in the UK, particularly my mother and nan who have a seemingly never-ending supply of encouragement and love. I am especially grateful for the advice and infectious excitement of my inspiring friend, the late Tony Harman – to whom this work is dedicated – who encouraged me to take a chance, move to his beloved Australia and study the bugs he cherished so much.

Last, but by no means least, I am grateful to the staff and students at the Institute for having me these past years. I have very much enjoyed being an active part of such a vibrant, international research community. In particular, I am thankful for the tireless, often silent work of the administrative, laboratory and site staff who were always there to advise and help when I needed it – especially David Harland, Patricia Hellier, Jenny Harvey, Gavin McKenzie and Vinod Kumar.

This work was supported by a Higher Degree Research Scholarship from the Hawkesbury Institute for the Environment, Western Sydney University.

Declaration of Authenticity

The work presented in this thesis is, to the best of my knowledge and belief, original except as acknowledged in the text. I hereby declare that I have not submitted this material, either in full or in part, for a degree at this or any other institution.

Contents

List of Tables ...... i

List of Figures ...... iii

Abstract ...... v

Preface...... viii

Chapter One: General Introduction ...... 1

1.1 The importance of invertebrates in terrestrial ecosystems ...... 1

1.2 Climatic and atmospheric change...... 1

1.3 Invertebrates and environmental change ...... 2

1.4 Altered precipitation impacts on invertebrate communities ...... 2

1.5 Elevated CO2 impacts on invertebrate communities ...... 6

1.6 Rationale ...... 8

1.7 Thesis overview ...... 9

1.8 Thesis outline ...... 10

Chapter Two: Grasslands, invertebrates and precipitation: a review of the effects of climate change ...... 13

2.1 Summary ...... 13

2.2 Introduction ...... 14

2.3 Invertebrate responses to precipitation change ...... 15

2.4 Invertebrate-mediated feedbacks on plant communities ...... 21

2.5 Conclusions and future directions ...... 22

Chapter Three: Predicted changes in rainfall patterns cause temporary perturbations in a grassland invertebrate assemblage which quickly recover ...... 25

3.1 Summary ...... 25

3.2 Introduction ...... 26

3.3 Materials and Methods ...... 29

3.4 Results ...... 32

3.5 Discussion ...... 54

Chapter Four: The effects of atmospheric change on forest invertebrates ...... 59

4.1 Summary ...... 59

4.2 Why are forest invertebrate communities important? ...... 60

4.3 Atmospheric change and invertebrates ...... 61

4.4 Responses of forest invertebrates to elevated CO2 concentrations ...... 62

4.5 Responses of forest invertebrates to elevated ozone concentrations ...... 75

4.6 Interactions between carbon dioxide and ozone ...... 78

4.7 Conclusions and future directions ...... 80

Chapter Five: Atmospheric change causes declines in woodland arthropods and impacts specific trophic groups ...... 82

5.1 Summary ...... 82

5.2 Introduction ...... 83

5.3 Materials and methods ...... 85

5.4 Results ...... 89

5.5 Discussion ...... 98

Chapter Six: Atmospheric change causes declines and compositional changes in populations of ants in Eucalypt woodland ...... 104

6.1 Summary ...... 104

6.2 Introduction ...... 105

6.3 Materials and Methods ...... 107

6.4 Results ...... 109

6.5 Discussion ...... 116

Chapter Seven: General Discussion ...... 119

7.1 Main findings and synthesis ...... 119

7.2 Effects of climatic and atmospheric change on primary consumers ...... 120

7.3 Effects of climatic and atmospheric change on higher trophic levels ...... 124

7.4 Climatic and atmospheric change impacts on invertebrate communities ...... 126

7.5 Interactions with other climate factors ...... 128

7.6 Constraints, caveats and further study ...... 129

7.7 Conclusions and future outlook ...... 132

Acknowledgements ...... 135

Bibliography ...... 136

Appendices ...... 171

Appendix I – Chapter Three Supplementary Material ...... 171

Appendix II – Chapter Five Supplementary Material ...... 178

List of Publications ...... 181

List of Tables

Table 2.1: A summary of the major precipitation manipulation experimental platforms assessing both plant and invertebrate responses to altered rainfall regimes...... 16 Table 3.1: Results from repeated measure LME models carried out on the square root transformed abundance of the different Orders ...... 34 Table 3.2: Results from LMs carried out on the square root transformed abundance of the different Orders from each watering treatment ...... 35 Table 3.3: Results from repeated measure LME models carried out on the square root transformed abundance of the different feeding guilds ...... 39 Table 3.4: Results from LMs carried out on the square root transformed abundance of the different feeding guilds ...... 43 Table 3.5: Results from PERMANOVA analyses on invertebrate community abundance-based composition...... 46 Table 3.6: Significant correlation results from Spearman’s rank correlations of invertebrate abundance with measures of plant composition and quantity ...... 50 Table 3.7: summary of the findings for invertebrate primary consumers in relation to the hypotheses discussed in this study...... 51 Table 4.1: Summary of the literature observing individual insect responses to elevated concentrations of CO2 and O3 ...... 65

Table 4.2: A summary of the literature considering the effects of elevated CO2 and/or

O3 concentrations on multiple species of invertebrates...... 69 Table 5.1: Results from likelihood ratio tests performed on GLMMs (abundance) and LMMs (rank-transformed biomass)...... 90 Table 5.2: Results from likelihood ratio tests performed on GLMMs (abundance) and LMMs (rank-transformed biomass) ...... 91

Table 6.1: The genera found in this study and their total abundances across both CO2 treatments, as well as their functional guild assignments...... 110 Table 6.2: Results from likelihood ratio tests performed on LME models with and without the fixed effect of CO2 treatment...... 111 Table 6.3: Results from multivariate permutational analysis (PERMANOVA) of the effect of CO2 treatment on ant assemblage...... 113

i

Table S3.1: a list of the total abundances of the groups identified in this study and their functional guild classifications ...... 171 Table S5.1: a list of the total abundances of the groups identified in this study and their functional guild classifications ...... 178 Table S5.2: Results from multivariate permutational analysis (PERMANOVA) of the effect of CO2 treatment on community data...... 179

ii

List of Figures

Fig 1.1: A schematic showing the direct and indirect effects of climate change on plants, herbivores and their natural enemies ...... 3 Fig 1.2: A conceptual diagram linking together the different chapters of this thesis, highlighting the common invertebrate communities and global change theme...... 12 Fig 2.1: A summary diagram of the general trends expected or found in the literature (theoretical, experimental and observational studies) ...... 23 Fig 3.1: the mean total abundance of invertebrates captured by sticky trapping, per plot (a) and vacuum sampling (scaled to 1 m2) (b) ...... 38 Fig 3.2: the mean abundance of invertebrates from different Orders (a, c-f) and functional groupings (b) captured by vacuum sampling (scaled to 1 m2)...... 40 Fig 3.3: the mean abundance of invertebrates from different Orders (a, c) and functional groupings (b, d) captured by sticky sampling ...... 41 Fig 3.4 (overleaf): NMDS plots of the invertebrate communities occurring under the different watering treatments in April 2014...... 47 Fig 3.5: Total plant biomass recovered from plots subjected to the five watering treatments over the course of the experiment, scaled to 1 m2...... 52

Fig 3.6: Mean C3:C4 ratios of plant biomass recovered from plots subjected to the five watering treatments over the course of the experiment...... 53 Fig 4.1: Conceptual diagram summarising the main directions of the responses of invertebrates to elevated CO2 and O3...... 79 Fig 5.1 (overleaf): Mean abundance of different functional guilds (a, b) and taxonomic groups (c, d) of ground-dwelling arthropods, split by CO2 treatment ...... 93 Fig 5.2 (overleaf): Mean abundance of different functional guilds (a, b) and taxonomic groups (c, d) of understorey arthropods, split by CO2 treatment ...... 93 Fig 5.3 (overleaf): Mean abundance of different, the functional guilds (a, b) and taxonomic groups (c, d) of aerial arthropods, split by CO2 treatment...... 93 Fig 5.4 (overleaf): a schematic diagram summarising the main findings in this study, and showing a scaled drawing of one of the EucFACE arrays...... 93 Fig 6.1: The abundances – absolute (bars), and relative (%, points) – of the different ant genera sampled in the different CO2 treatments...... 112

iii

Fig 6.2: NMDS plot of the genus-level ant communities under the two CO2 treatments

(ambient CO2 shown in white ellipse/pale grey points, elevated CO2 in grey)...... 114

Fig 6.3: NMDS plot of the sampled ant communities split by functional guild and CO2 treatment...... 115 Fig 7.1: A conceptual diagram linking together the different chapters of this thesis, highlighting the main findings from the empirical research chapters...... 134

Fig S3.1: soil water content and rain applied under the four watering regimes...... 172 Fig S3.2: the mean abundance of invertebrates captured by sticky trapping, divided into Order groups, over the course of the experiment...... 174 Fig S3.3: the mean abundance of invertebrates captured by sticky trapping, divided into feeding groups, over the course of the experiment...... 175 Fig S3.4: the mean abundance of invertebrates captured by vacuum sampling, divided into Order groups, over the course of the experiment, scaled to 1 m2 ...... 176 Fig S3.5: the mean abundance of invertebrates captured by vacuum sampling, divided into feeding groups, over the course of the experiment, scaled to 1 m2 ...... 177 Fig S5.1 (overleaf): NMDS plots of arthropod community data in each of the three niche types, partitioned by functional guild classification and Order identity ...... 179

iv

Abstract

Invertebrates form the foundation of terrestrial ecosystems, far outnumbering their vertebrate counterparts in terms of abundance, biomass and diversity. As such, arthropod communities play vitally important roles in ecosystem processes ranging from pollination to soil fertility. Given the importance of invertebrates in ecosystems, predicting their responses – and those of the communities they form – to global change is one of the great challenges facing contemporary ecology. Our climate is changing as a result of the anthropogenic release of greenhouse gases, including carbon dioxide (CO2), produced from burning fossil fuels and land use change. The concentration of CO2 in the atmosphere now exceeds the range the Earth has seen in the last 800,000 years. Through the effect of such gases on radiative forcing, sustained greenhouse gas emissions will continue to drive increases in global average temperatures. Additionally, precipitation patterns are likely to change across the world, with increases in the occurrence of extreme weather events, such as droughts, as well as alterations in the magnitude and frequency of rainfall events.

Climate change is already causing measurable changes in the Earth’s biotic environment. Past work has been heavily focused on the responses of plants to various climate change parameters, with most studies including invertebrates limited to highly controlled studies of pair-wise interactions between one arthropod species and its host plant. Relatively little work to date, however, has looked at the potential impacts of climatic and atmospheric change for invertebrate communities as a whole. The overarching goal of this project was to help remedy this research gap, specifically by investigating the effects of precipitation and atmospheric change on invertebrate communities in grassland and woodland habitat, respectively.

Chapters 2 and 4 synthesised recent work on climate change-driven alterations in precipitation and atmospheric change impacts on invertebrates in grassland and woodland systems, respectively. These chapters both highlighted the need for more community-level studies looking at the effects of global change on invertebrates, coupled with greater geographical representation across ecosystems. Particularly for

v

atmospheric change studies, there has been a strong bias toward Northern Hemisphere plantation systems in previous work.

Empirical research chapters 3, 5 and 6 used two state-of-the-art field-scale experimental platforms to address the question of how climatic and atmospheric changes will impact invertebrate communities in two Southern Hemisphere systems. Specifically, chapter 3 investigated how a subtropical grassland invertebrate community will respond to five climate change precipitation scenarios, including alterations in the seasonality, frequency and magnitude of rainfall events. Chapters 5 and 6 determined the effects of elevated concentrations of CO2 gas on the overall invertebrate (chapter 5) and ant community (chapter 6) of a critically endangered Eucalyptus woodland.

Altered precipitation regimes caused highly variable responses in the abundance of invertebrates across the community, which were strongly seasonal and only weakly related to changes in the underlying plant community. The short-term, transient nature of the observed responses suggests that the invertebrate community – which has evolved against a background of strong precipitation variability in Australia – will be resilient to changes in rainfall. Elevated CO2 on the other hand, caused widespread declines in the populations of various arthropods across the community, including herbivore (-48.3%) and parasitoid (-14.7%) functional groups, with overall declines in total arthropod abundance of up to 14.7%. Despite these reductions, elevated CO2 did not measurably affect overall invertebrate community composition; the widespread declines across the community resulted in compositionally-similar communities comprised of fewer total individuals, compared with those under ambient conditions. However, for the ant community, shifts in the dominant genus-level ant populations occurring under elevated CO2 drove changes in ant assemblage structure. This, coupled with the general declines witnessed in the ant and broader invertebrate community, supports the notion that elevated CO2 could lead to changes in the ecosystem processes these organisms support.

Taken together, these results present contrasting evidence for invertebrate community- level responses to climatic and atmospheric change. On the one hand, communities may be able to cope with future increases in precipitation variability, suggesting that the ecosystem processes underpinned by invertebrates may remain stable in this system. On the other hand, exposure to levels of CO2 not recently experienced within evolutionary

vi

timescales, could result in declines in the abundance of organisms that could play important roles in ecological processes.

Avenues for future research are discussed, as well as the limitations and challenges inherent in field-scale, community-level climate change research.

vii

Preface

This thesis comprises original work conducted by myself, with guidance from my supervisory panel, Scott Johnson, David Ellsworth and Jo Staley, with whom I conceptualised the research project and established the experimental design. I have collected, analysed and interpreted all the data herein with guidance from my supervisory panel.

Five of the chapters in this thesis (Two to Six) are presented in a format appropriate for peer-review publications and are therefore structured to be standalone. Some repetition was unavoidable where recurrent methodologies were used. Contributors and details of are listed at the beginning of each chapter. Referencing follows the style of the Journal of Animal Ecology throughout this thesis.

viii

Chapter One: General Introduction

1.1 The importance of invertebrates in terrestrial ecosystems Invertebrates form a central component of terrestrial ecosystems, far outnumbering their vertebrate counterparts in terms of abundance, biomass and diversity (Stork et al. 2015). Invertebrates have been described as ‘little things that run the world’, making reference to their immense, often overlooked, functional importance (Wilson 1987). As competitors for limited food resources, and with a burgeoning human population, much of the attention directed toward invertebrates has historically cast a spotlight on inconvenient herbivore pest species. However, many invertebrates have supporting, stabilising roles in the maintenance of diverse ecosystems on which human life depends, playing central roles in ecosystem processes through activities such as pollination, pest control, nutrient cycling and soil formation (Schowalter 2006; Whiles & Charlton 2006). Further, invertebrates underpin the existence of other animals in terrestrial ecosystems, forming the diet of insectivorous vertebrates such as birds and reptiles, as well as invertebrate predators and parasitoids.

1.2 Climatic and atmospheric change Our climate is changing as a result of the anthropogenic release of greenhouse gases, including carbon dioxide (CO2), ozone (O3) and nitrous oxide (N2O), produced from burning fossil fuels and land use change (Karl & Trenberth 2003; IPCC 2013). The concentration of CO2 in the atmosphere now exceeds the range the Earth has seen in the last 800,000 years (IPCC 2013), reaching a record high of 400ppm in May 2013 (NOAA 2013), with the rate of emission predicted to increase by 25-90% between 2000 and 2030 (IPCC 2007). Indeed, the largest ever annual increase in CO2 emission occurred in 2015 (LePage 2016). Concentrations of tropospheric O3 have doubled over the last century (Vingarzan 2004), and are predicted to continue to increase by up to 25% before 2050 (IPCC 2007). The sustained emission of greenhouse gases, including

CO2 and O3, will continue to raise global average temperatures, with projected increases ranging from 0.3-4.8°C by the end of this century. Along with changes in global average temperatures, climate models predict changes in precipitation patterns, in terms of the total amount and the frequency and intensity of rainfall events (IPCC 2013). In addition, 1

CHAPTER ONE GENERAL INTRODUCTION there is likely to be an increase in the occurrence and severity of extreme events like droughts, particularly within south eastern Australia (Chiew et al. 2011; Dai 2011). Indeed, the region has already seen increases in aridity over recent years, even against a backdrop of high climate variability, revealed by paleo-climatic records (Chiew et al. 2011).

1.3 Invertebrates and environmental change Given the importance of invertebrates in underpinning terrestrial ecosystems, it is in our interest to understand how climatic and atmospheric changes will impact them. Predicting the responses of species, and the communities they comprise, to global change is one of the great challenges facing ecology (Gilman et al. 2010; Andrew et al. 2013; Facey et al. 2014). Climatic and atmospheric change can influence plants, invertebrate herbivores and higher trophic levels either directly – such as by causing alterations in temperature-sensitive insect metabolism and behaviour – or indirectly, where changes occurring in one group impact another trophic level (Figure 1.1, overleaf). This introduction briefly outlines some of the major direct and indirect ways in which changes in precipitation and atmospheric CO2 levels will impact invertebrate communities – the focus of this thesis, reviewed in more detail in Chapters 2 and 4 respectively.

1.4 Altered precipitation impacts on invertebrate communities

1.4.1 Direct effects Water availability is recognised as being important for invertebrate survival. On one hand, invertebrates are at risk from excessive water loss due to their small size – especially during dormant life stages when they are unable to actively seek out water (Danks 2000; Addo-Bediako, Chown & Gaston 2001). Soft-bodied invertebrates may be particularly at risk of desiccation, as they lack a waxy cuticle that would otherwise prevent evaporative water loss (Berridge 2012). On the other hand, excessive rainfall (and subsequent flooding) can also potentially increase invertebrate mortality, especially for species occurring belowground for some or all of their life cycles, such as overwintering moth pupae (Slosser et al. 1975).

As a result of their sensitivity to moisture change, invertebrates both above and belowground have evolved various behavioural and morphological adaptations to alterations in water availability. These strategies include migration, hiding in the soil,

2

CHAPTER ONE GENERAL INTRODUCTION aestivation, shelter-building and water-repelling hairs (Willmer 1982; Zalucki, Clarke & Malcolm 2002). The extent to which different functional groups exploit and benefit from such strategies could influence invertebrate community composition (Barnett & Facey 2015). For instance, flooding is known to structure hygrophilic beetle and spider assemblages in areas prone to flooding, with water inundation favouring specialist species over generalist ones (Ballinger, Mac Nally & Lake 2005).

Fig 1.1: A schematic showing the direct and indirect effects of climate change on plants, herbivores and their natural enemies. Direct effects result from the action of climatic variables on species traits and physiology. Indirect effects are the result of altered species interactions caused by climate-sensitive responses of one or both reciprocal partners, which can cascade across trophic levels. Adapted from Jamieson et al. (2012).

1.4.2 Indirect (plant-mediated) effects on invertebrate herbivores Plant community composition plays a bottom-up role in structuring arthropod communities (Koricheva et al. 2000; Perner et al. 2005; Hertzog et al. 2016). Differences

3

CHAPTER ONE GENERAL INTRODUCTION in the ability of plant species to cope with altered rainfall regimes will lead to changes in plant community composition (Knapp et al. 2002), which could have implications for the invertebrates which inhabit them, particularly specialist monophagous species which may be reliant on the presence of just one or two host plant species (Koricheva et al. 2000). Alterations in water availability could impact the relative abundances of plants utilising C3 and C4 photosynthetic pathways. C4 plants are known to have hydraulic benefits over their C3 counterparts when drought-stressed, due to their ability to maintain higher levels of stomatal conductance and greater water potentials, and thus greater rates of photosynthesis (Taylor et al. 2014). Such changes in plant functional types could have implications for invertebrates in the system. Under reduced water availability scenarios, there may be an increase in the number of drought-tolerant C4 plants, which could force invertebrates to consume relatively more C4-derived material, as found in a North American desert system (Warne, Pershall & Wolf 2010). However, plants utilising the C4 photosynthetic pathway are generally thought to be less palatable for insect herbivores, having tougher leaves with lower relative amounts of nitrogen compared with C3 foliage which is therefore favoured by herbivores under normal conditions (Caswell & Reed 1976; Boutton, Cameron & Smith 1978; Barbehenn et al.

2004; Nokelainen et al. 2016). Thus, changes in C3:C4 plant composition under altered precipitation regimes could have knock on effects for invertebrate performance and abundance.

As well as plant community composition, plant quantity and quality are also important determinants of invertebrate community structure. Generally, reduced rainfall and drought conditions are associated with decreased plant growth rate and size, whereas increases in rainfall, to a certain point, are associated with stimulated growth and productivity (Fay et al. 2003; Byrne, Lauenroth & Adler 2013). Given that secondary production by invertebrate herbivores is reliant on the primary productivity occurring in the ecosystem (Begon, Townsend & Harper 2006), we might therefore expect to see an increase in herbivore abundance under increased rainfall. Increased food resources will be able to support greater populations of secondary consumers which may flow up through the food chain to increase the abundance of organisms at higher trophic levels. Further, fast growing plants, as may be encountered during rapid increases in water availability following longer dry periods, may represent more attractive resources for herbivores than slower growing individuals (plant vigour hypothesis (PVH) Price 1991; Endara & Coley 2011).

4

CHAPTER ONE GENERAL INTRODUCTION

Conversely, under reduced rainfall we may expect to see reductions in invertebrate abundance, as herbivore food resources become limited. However, drought stress can cause changes in the concentrations of phytohormones and secondary metabolites (Hsiao 1973; Jamieson et al. 2012), along with increases in amino acids in the phloem due to proteolysis of proteins into free amino acids (Brodbeck & Strong 1987; Girousse, Bournoville & Bonnemain 1996). This increase in the availability of nitrogenous compounds in stressed plants was the basis for the plant-stress hypothesis (PSH, White 1969), which predicts that stressed plants will be more susceptible to herbivory and population outbreaks will result. Contrary to this, a meta-analysis by Huberty and Denno (2004) revealed that generally speaking, leaf chewing insects showed no trend towards performing better – and in some cases perform worse – on water-stressed plants. This could be due to negative effects of increases in leaf toughness and reductions in water content (Huberty & Denno 2004) which, along with nitrogen concentration, are important determinants of palatability. Sap feeding insects had lower performance on stressed plants, likely as a result of reduced turgor pressure and water content impairing feeding (Huberty & Denno 2004). They therefore proposed the pulsed-stress hypothesis (PuSH), whereby the positive effects of water stress on nutrient content can be realised by phloem feeders during intermittent periods of relatively less stress, when turgor is recovered. Under such pulsed conditions, which readily occur in nature and may well increase under climate change, we could therefore expect phloem feeders such as aphids to experience enhanced levels of success compared with other feeding guilds.

1.4.3 Indirect (plant-mediated) effects on higher trophic levels Given the predominantly negative effects of reduced rainfall availability on herbivores, the indirect effects on higher trophic levels are also likely to be negative, and empirical studies support this. The rate of parasitism by Hymenopterans has been shown to be negatively related to precipitation variability, with host species becoming decoupled from their enemies (Stireman et al. 2005). A recent study by Aslam et al. (2013) reported a significant decrease in aphid (Rhopalosiphum padi) parasitism rates by Aphidius ervi on drought stressed plants, most likely as a result of altered aphid demography on these plants. Similarly, Tariq et al. (2013) found reduced aphid parasitism rates by two different species of parasitoid wasp under drought conditions. Romo and Tylianakis (2013) reported a significant decrease in parasitoid longevity when exposed to drought conditions, although the number of successfully eclosed individuals remained constant.

5

CHAPTER ONE GENERAL INTRODUCTION

A study by Johnson et al. (2011b) showed that drought had a non-linear negative effect on parasitism rates by A. ervi by reducing the numbers and quality of R. padi hosts.

1.5 Elevated CO2 impacts on invertebrate communities

1.5.1 Indirect (plant-mediated) effects on invertebrate herbivores

Whilst instances of direct effects of CO2 on insect herbivores are scarce in the literature (but see Stange 1997), the indirect consequences, mediated by physiological and morphological changes in plants, can be substantial and multi-faceted. Plants grown under elevated CO2 conditions generally experience stimulated rates of photosynthesis (Couture & Lindroth 2013), due to so-called ‘carbon fertilisation’ (Lamarche et al. 1984), as well as increased water use efficiency as a result of lower stomatal conductance and transpiration rates (e.g. Bazzaz 1990; Drake, Gonzalez-Meler & Long 1997). Plants often have increased biomass (up to nearly 40% compared with plants grown under ambient conditions, Stiling & Cornelissen 2007) with greater leaf areas and tougher leaves when grown in enriched CO2 (decreased specific leaf area) (Pritchard et al. 1999; Robinson, Ryan & Newman 2012). Larger plants with greater leaf area represent greater resources to invertebrate herbivores. Thicker, tougher leaves, however, present a disadvantage to foraging herbivores; such leaves are harder for herbivores to consume (Bernays 1986) and represent lower quality resources (Knepp et al. 2005; DeLucia et al. 2012).

Elevated CO2 has the capacity to influence plant primary and secondary metabolism. Carbohydrates such as starch and sugars often accumulate in plant tissues under elevated CO2 as a result of increased rates of photosynthesis. This is usually coupled with a decrease in foliar nitrogen concentration as a result of dilution by excess carbohydrates and reallocation within tissues (Lawler et al. 1996; Curtis & Wang 1998; Stiling & Cornelissen 2007; Robinson et al. 2012). This can lead to increased carbon to nitrogen (C:N) ratios of plant tissues, with a recent meta-analysis by Robinson et al.

(2012) finding that, on average, plants grown under elevated CO2 experience a 19% increase in C:N ratio. This decline in the relative amount of available nitrogen has critical implications for insect herbivores, which rely on food intake to obtain nitrogen from plant tissues. Nitrogen plays a vital role in metabolic processes – indeed, nitrogen availability is often a major factor limiting insect growth (Mattson 1980) and has been shown to be strongly positively linked with fecundity (Awmack & Leather 2002), survival and adult size (e.g. Myers & Post 1981).

6

CHAPTER ONE GENERAL INTRODUCTION

In response to the lower quality of plant tissues, herbivores may feed in a compensatory way (i.e. consume more plant tissue) to acquire the critical levels of nitrogen required for their development and survival, and this phenomena has been documented many times in the literature (e.g. Lincoln, Sionit & Strain 1984; Lincoln, Couvet & Sionit 1986; Fajer, Bowers & Bazzaz 1989; Docherty & Hurst 1996; Johnson & McNicol 2010). However, increased consumption of plant materials by herbivores to meet nitrogen requirements under elevated CO2 entails ingesting higher levels of other compounds present in the leaves, such as the products of plant secondary metabolism (Couture et al. 2016). Secondary metabolites (also known as allelochemicals or defensive compounds) play a crucial role in plant defence against herbivores, interfering with their growth, development and survival (Slansky & Wheeler 1992; Ryan, Rasmussen & Newman 2010).

Therefore, under elevated CO2 conditions, with lower relative nitrogen availability and changes in exposure to secondary metabolites, we can expect a decrease in insect herbivore performance, in terms of pupal weight, survival, fecundity etc. and this has generally been shown to be the case, although it is worth noting that results can differ between study systems and feeding guilds (Bezemer & Jones 1998). For instance, aphids and other phloem feeders may stand to benefit from enhanced CO2-grown foliage (Bezemer, Jones & Knight 1998). Nevertheless, in general, herbivore survival, abundance, pupal weights and development rates have all been shown to be negatively affected by CO2-related changes in foliar chemistry (Bezemer & Jones 1998; Stiling, Rossi & Hungate 1999; Stiling & Cornelissen 2007).

1.5.2 Indirect (plant-mediated) effects on higher trophic levels

As for insect herbivores, the impacts of future CO2 concentrations on higher trophic levels are more likely to manifest indirectly, through climate change induced responses in their host species, than through direct CO2-related physiological changes (Ryan et al. 2010). Moreover, the consequences of climate change are likely to be even greater for higher trophic levels as they are dependent on the responses of the levels below them (Hance et al. 2007). Increased herbivore development times, as a result of declining foliar quality under elevated CO2 conditions, could present an opportunity to parasitoids and predators as hosts become more susceptible to attack (Roth & Lindroth 1995), remaining in vulnerable larval life stages for longer periods. The immune response of herbivore hosts to parasitoids may also be weakened when reared on a lower quality diet

7

CHAPTER ONE GENERAL INTRODUCTION

(Barbosa, Saunders & Waldvogel 1982). This would suggest that natural enemy fitness could improve under elevated CO2 conditions in the future – this has shown to be the case experimentally, with leaf miners experiencing higher rates of parasitism when reared under elevated CO2 (Stiling et al. 1999).

However, despite potential increased host-susceptibility, plant-mediated changes in nutritional quality may have negative knock-on effects for higher trophic levels as hosts, in turn, provide lower quality resources for their natural enemies. In a recent study by

Klaiber et al. (2013), cabbage aphids (Brevicoryne brassicae) reared under elevated CO2 conditions were smaller and presented lower quality resources to Hymenopteran parasitoids (Diaeretiella rapae) which suffered a corresponding decrease in longevity and parasitism rate.

1.6 Rationale Biotic changes as a result of environmental change will be intricate; the responses of different species will be inextricably tied to those of other species occurring within and across trophic levels. A recent review by Jamieson et al. (2012) highlighted the current need for studies across multiple species within systems; they found that studies on plants alone vastly outnumbered plant-insect studies nearly ten to one, and that studies incorporating higher trophic levels and complexity (i.e. natural enemies and other herbivore competitors) represented roughly one tenth of those on plant–insect systems, despite the ecological and economic importance such interactions hold in ecosystem processes. Therefore, studies with more than one functional group or species – such as those at the community-level – are needed to enhance our understanding and build a realistic picture of the way species might respond in the future (Harrington, Woiwod & Sparks 1999).

Further, much of the previous work looking at invertebrate responses to climatic and atmospheric change has been carried out under highly controlled laboratory conditions. Such reductionist studies, whilst useful in revealing the mechanistic bases of detected changes (Lindroth & Raffa 2016), may have a tendency to overestimate the magnitude of treatment effects on invertebrates (Zvereva & Kozlov 2012) and fail to incorporate the complexity inherent in natural systems. Field-level climate change studies, whilst logistically challenging, and not without their own limitations, allow the incorporation of indirect effects of the imposed changes via the inclusion of multispecies interactions

8

CHAPTER ONE GENERAL INTRODUCTION

(Lindroth & Raffa 2016), as well as the inclusion of variability in background environmental conditions.

Due to limited resources and funding, one approach increasingly used by ecologists is to utilise carefully selected indicator taxa or ‘bioindicators’ that function as representatives of the broader ecosystem in which they are found (Gardner et al. 2008). Given the breadth of ecological roles played by invertebrates, as well as their sensitivity to stress and disturbances, they are frequently used as indicator taxa (e.g. Kremen et al., 1993). This makes the study of invertebrate community responses to climatic and atmospheric change particularly worthwhile, as the results from such studies can be used to inform the management of the wider ecosystem in which the study is performed.

1.7 Thesis overview Using a series of field experiments, this work addresses how predicted precipitation and atmospheric changes will impact invertebrate communities in the future.

The aims of this project were to:

- Review the current literature concerning invertebrate population- and community-level responses to altered rainfall in grassland ecosystems.

- Determine the effects of a range of predicted altered precipitation regimes on the invertebrate community of a south east Australian grassland.

- Review the literature investigating the effects of climate change-altered atmospheres on forest invertebrate communities.

- Determine the effects of concentrations of CO2 predicted for the middle of this century on the ant and wider invertebrate community of a native Eucalyptus woodland system.

An additional objective was to determine the effects of the studied climate change factors (altered precipitation and elevated CO2) on different functional groups (i.e. feeding guilds) occurring in the community, in order to relate any detected changes to potential alterations in ecosystem functioning under future climates. Furthermore, given the strong relationship between invertebrates and the underlying plant community, changes occurring in the invertebrate communities of both study systems were

9

CHAPTER ONE GENERAL INTRODUCTION examined in the context of broad changes in plant quantity and quality. These included biomass, C:N ratio and other plant community metrics.

1.8 Thesis outline This work is structured into seven chapters (the overarching structure of chapters 2-6 is depicted in Fig 1.2). This introductory chapter has reviewed general and contemporary theory concerning the importance of invertebrates within ecosystems and the ways in which invertebrate communities may be affected by predicted changes in atmospheric composition and precipitation variability.

Chapter 2 reviews findings from manipulative and observational studies which have examined grassland invertebrate responses to altered rainfall, with a particular focus on large-scale field experiments employing precipitation manipulations. This chapter identifies a need for the inclusion of invertebrate responses in a greater number of precipitation manipulation experiments, with improved geographical representation across ecosystems. This review entitled ‘Grasslands, invertebrates and precipitation: a review of the effects of climate change’ (Kirk L. Barnett and Sarah L. Facey – joint first authors) was accepted for publication in Frontiers in Plant Science, on July 26th 2016.

Chapter 3 uses a novel precipitation manipulation experiment – ‘DRI-Grass’ – to investigate how predicted changes in rainfall will affect invertebrate communities in an eastern Australian grassland. This study assesses the effects of alterations in the absolute amount of rainfall inputs (increased and reduced), as well as changes in the seasonality (summer drought) and frequency of rainfall (reduced frequency) on the abundance and composition of the invertebrate community. This work entitled ‘Climate-change altered rainfall patterns cause short-term perturbations in a grassland invertebrate assemblage’ (Sarah L. Facey, Kirk L. Barnett, Aqeel Abbas, William M. Balmont, David S. Ellsworth, Joanna T. Staley & Scott N. Johnson) has been submitted to a peer-reviewed journal and is currently under review.

Chapter 4 reviews the findings from the body of work looking at the responses of forest invertebrates to atmospheric changes, focusing on the two most studied gases –

CO2 and O3 – with a special focus on the results from Free-Air Enrichment studies. This chapter highlights the need for more work looking at the responses of invertebrate communities to atmospheric change in field settings within natural systems, particularly outside of the Northern Hemisphere where the bulk of previous work has been

10

CHAPTER ONE GENERAL INTRODUCTION completed. This review was accepted for publication as a peer-reviewed book chapter (Sarah L. Facey & Andrew N. Gherlenda) in Global Climate Change and Terrestrial Invertebrates (2016) (eds S.N. Johnson & T.H. Jones). John Wiley & Sons, Chichester, UK.

Chapter 5 uses the first Free-Air Enrichment study based in a native forest system and in the Southern Hemisphere (‘EucFACE’), to address the question of how the invertebrate community of a Eucalyptus woodland will respond to the concentration of atmospheric CO2 expected by the middle of this century. Specifically, this chapter looks at alterations in abundance and community structure occurring across the invertebrate community. This research entitled ‘Atmospheric change causes declines in woodland arthropods and impacts specific trophic groups’ (Sarah L. Facey, David B. Fidler, Rhiannon C. Rowe, Lisa M. Bromfield, Sabine S. Nooten, Joanna T. Staley, David S. Ellsworth and Scott N. Johnson) has been accepted in Agricultural and Forest Entomology.

Chapter 6 addresses changes in the ant assemblages occurring under current and future levels of CO2. Specifically, this work looks at changes in genera-level abundance and ant assemblage composition occurring at EucFACE. Ants have been frequently used as an indicator taxon, given their broad ecological roles and responsiveness to environmental change. This research entitled ‘Atmospheric change causes declines and compositional changes in populations of ants in a Eucalypt woodland’ (Sarah L. Facey, Jenni M.M. Kremer, Sabine S. Nooten, David S. Ellsworth, James M. Cook, Joanna T. Staley and Scott N. Johnson) has been submitted to a peer-reviewed journal and is currently under review.

Chapter 7 synthesises the key findings from this project against the backdrop of previous invertebrate community-level climate change research. Avenues for future research are discussed, as well as the limitations and challenges inherent in field-scale, community-level climate change research.

11

CHAPTER ONE GENERAL INTRODUCTION

Fig 1.2: A conceptual diagram linking together the different chapters of this thesis, highlighting the common invertebrate communities and global change theme.

12

Chapter Two: Grasslands, invertebrates and precipitation: a review of the effects of climate change

Published as Barnett, K.L. & Facey, S.L. (2016) Frontiers in Plant Science, 7, 01196. Overall, SLF and KLB undertook 70% and 30%, respectively, of the work associated with this chapter; however in the published version, the author names appear in alphabetical order.

2.1 Summary Invertebrates are the main components of faunal diversity in grasslands, playing substantial roles in ecosystem processes including nutrient cycling and pollination. Grassland invertebrate communities are heavily dependent on the plant diversity and production within a given system. Climate change models predict alterations in precipitation patterns, both in terms of the amount of total inputs and the frequency, seasonality and intensity with which these inputs occur, which will impact grassland productivity. Given the ecological, economic and biodiversity value of grasslands, and their importance globally as areas of carbon storage and agricultural development, it is in our interest to understand how predicted alterations in precipitation patterns will affect grasslands and the invertebrate communities they contain. Here, the findings from manipulative and observational studies which have examined invertebrate responses to altered rainfall are reviewed, with a particular focus on large-scale field experiments employing precipitation manipulations. Given the tight associations between invertebrate communities and their underlying plant communities, invertebrate responses to altered precipitation generally mirror those of the plants in the system. However, there is evidence that species responses to future precipitation changes will be idiosyncratic and context dependent across trophic levels, challenging our ability to make reliable predictions about how grassland communities will respond to future climatic changes, without further investigation. Thus, moving forward, the increased consideration of invertebrate communities in current and future rainfall manipulation platforms, as well as the adoption of new technologies to aid such studies, is recommended.

13

CHAPTER 2 GRASSLAND INVERTEBRATES AND CLIMATE CHANGE

2.2 Introduction Grasses cover more of the earth’s surface than any other vegetation type (Tscharntke and Greiler 1995; Wang and Fang 2009) and are often of high economic, ecological and biodiversity value, providing forage for livestock and high levels of carbon storage (Lee et al. 2014; Lenhart et al. 2015). Many grasslands exist in seasonal states of water limitation, and are highly responsive to changes in water availability in terms of biomass and composition (Knapp et al. 2002; Fry et al. 2014; Lenhart et al. 2015). Climate models predict changes in precipitation patterns, in terms of the total amount and the frequency and intensity of rainfall events (IPCC 2013), therefore leading to alterations in grassland plant composition and primary production.

Invertebrates are the most diverse and abundant constituent of terrestrial ecosystem fauna (Stork et al. 2015). Often overlooked, these organisms contribute to structuring grassland communities, through activities such as pollination and nutrient cycling (Whiles and Charlton 2006) and contribute to shaping grasslands through top-down processes. For instance, herbivores can modify plant species richness by altering competitive dynamics between plant species (Olff and Ritchie 1998). Likewise, plant community composition plays a bottom-up role in structuring arthropod communities (Perner et al. 2005; Hertzog et al. 2016), as do abiotic factors like temperature and water availability (e.g. Bale et al. 2002). Thus, both grassland plant and invertebrate communities can be directly impacted by alterations in climate. In addition, precipitation changes can have indirect impacts on both plants and invertebrates as the interactions occurring between the two communities are also climate-sensitive; the effect of herbivory on plant diversity varies across precipitation gradients, for instance (Olff and Ritchie 1998).

It is in our interest to understand how climate change-driven alterations in precipitation will affect valuable grassland systems and the invertebrate communities they both support and rely on. Over the past twenty years, multiple experiments and observational studies have addressed the responses of grasslands to changes in precipitation, with a subset of these also examining invertebrate responses across a range of precipitation scenarios and spatial scales (summarised in Table 2.1). To our knowledge, there has been no synthesis of the relevant literature examining insect responses to precipitation changes, making a review of these studies timely. This mini-review looks at the effects of altered precipitation patterns – including reductions and increases in average rainfall,

14

CHAPTER 2 GRASSLAND INVERTEBRATES AND CLIMATE CHANGE and changes in rainfall frequency – on grassland invertebrates and the plant communities they inhabit. This review focuses on findings from field- based/observational studies and precipitation manipulation experiments conducted in grasslands, including steppe and savannah habitats.

2.3 Invertebrate responses to precipitation change

2.3.1 Direct impacts In general, terrestrial arthropods are sensitive to changes in moisture, given their high surface-to-volume-ratio (Kimura et al. 1985). Under reduced rainfall, most aboveground arthropods avoid desiccation behaviourally by migrating, aestivating (dry season diapause), hiding in the soil, or, in a few cases, building a shelter (Willmer 1982; Zalucki et al. 2002; Benoit 2010; Berridge 2012). Structurally speaking, soft-bodied arthropods (isopods and myriapods) lack the waxy cuticle found in arachnids and insects that prevents or reduces evaporation (Berridge 2012). This, in combination with differences in excretion-related water losses (Horne 1968), suggests that soft-bodied arthropods will be more vulnerable to reductions in water availability, and, in some cases, to increases (Sylvain et al. 2014). Thus, changes in rainfall could be expected to affect hard and soft- bodied groups differently, resulting in shifts in the arthropod community.

On the other end of the spectrum, average increases in rainfall may negatively impact arthropods by disrupting flight, reducing foraging efficiency and increasing migration times (Peng et al. 1992; Drake 1994; Kasper et al. 2008). Some arthropods can vary their behaviour to combat the effects of extreme rainfall events and flooding by shelter- seeking and utilising submersion tolerance strategies (Lambeets et al. 2008). The effects of increased rainfall on arthropods are also dependent on invertebrate morphology and are group-specific, with larger winged insects like Lepidopterans having a much higher degree of ‘unwettability’ (i.e. requiring greater volumes of water to become wet) than smaller winged insects (Wagner et al. 1996). Altered rainfall frequency can be positive or negative for invertebrates depending on the size of the event (Nielsen and Ball 2015), but on the whole is expected to impact more rain-sensitive orders like Lepidoptera (Palmer 2010).

Arthropods that spend all or some of their life stages belowground have evolved behaviours to manage water stress in times of drought and flood (Verhoef and Witteveen 1980). Under reduced water availability, most soil invertebrates combat

15

Table 2.1: A summary of the major precipitation manipulation experimental platforms assessing both plant and invertebrate responses to altered rainfall regimes. Climate classifications follow the Global Agro-Ecological Zones set out by the Food and Agriculture Organization (gaez.fao.org). Invert. Groups; Name; Ecosystem; Method; Manipulation Collection Outcome Reference Location Plant groups Shelter design method Auchenorrhyncha, +/- water, Under rainfall and nitrogen addition, plants did not Grassland; Araneae, Coleopt., Irrigation with Silwood summer respond. In the third year plant biomass declined in (Lee et al. forbs and Collembola, Dipt., rain water; (UK) drought and drought plots. Auchenorrhyncha and Araneae 2014) grasses Heteropt., Isopoda; Removeable roof winter increase declined with plant biomass. Vacuum sampling Enhanced summer rainfall increased leaf miner abundance, but not when root herbivores were also +/- water, +/- Calcareous Lepidopt., present. Root herbivores reduced the parasitism rates (Staley et al. root grassland; Coleopt.; Manual of moths above ground (smaller pupal size). Plants 2007) Wytham - herbivores forb Manual with DI under drought were overall less susceptible to leaf- TIGER IV water; Mobile miners regardless of root damage. 2c. Calcareous shelters (UK) Added water increased plant cover and grassland; +/- water, Auchenorrhyncha; Auchenorrhyncha abundance; though drought (Masters et al. forbs, + winter heat Vacuum sampling reduced vegetation cover, the abundance of 1998) legumes and Auchenorrhyncha remained at ambient levels. grasses Mixed-grass Water stress reduced plant biomass but not nutrient (Lenhart, + water, prairie and Not mentioned; content and species diversity. Drought reduced forb BCNWR1 Eubanks & natural oak savannah; Orthopt.; Manual No shelter, protein content and grasshopper abundance and (USA) Behmer drought forbs and natural drought diversity. There was increased abundance and species 2015) grasses richness of certain grasshoppers in irrigated plots OCCAM2 +/- heat, +/- Old field – 163 Irrigation with No strong trends in terms of water effects; there was (Villalpando, (USA) water, +/- fescue; forbs, morphospecies; rain water; Fixed a greater peak plant biomass in wet compared to dry. Williams &

CO2 legumes and Sticky traps, roof Weak effects of soil moisture on invertebrate Norby 2009) grasses vacuum sampling community composition; more parasitoids in the dry treatment – temperature more important. Measurements were taken 1 year after drought Calcareous Not mentioned; - water, application. Drought significantly increased the Agroscope pasture; forbs, Annelida; Mustard Temporary (Mariotte et diversity biomass of earthworms in plots where subordinate (Switzerland) legumes and extraction shelter: summer al. 2016) increments plant species were present. Drought caused shifts in grasses only earthworm community in terms of individual species. + water, Grassland; Coleopt., Hemipt., Irrigation with Spring water addition caused diminishing increases in (Suttle, Berkeley winter forbs, Hymenopt., spring water; No winter forbs/legumes resulting in lower plant and Thomsen & (USA) increase, legumes and Orthopt., Araneae; shelter invertebrate species richness at the end of 5 years. Power 2007) spring increase grasses Manual, pitfall +/- water, Coleopt., Hemipt., Automatic Pasture; Summer drought caused strong declines in plant (Power et al., altered Hymenopt., Irrigation with DRI-Grass3 forbs, biomass, which rebounded when watering was 2016 frequency, Orthopt., Araneae; tap water; fixed (Australia) legumes and resumed. The subsequent regrowth supported an Torode et al., summer Vacuum sampling, shelter grasses increased population of Hemipterans and parasitoids. 2016) drought sticky trap 1 Balcones Canyonlands National Wildlife Refuge (Marble Falls, TX) 2Oldfield Community Climate and Atmospheric Manipulation (Oak Ridge, TN) 3Drought and Root herbivore Impacts on Grassland (Richmond, NSW)

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE fluctuating moisture by relocating to places that are more favourable within the soil- matrix. Such movement, however, is dependent on suitable soil moisture and texture (Lees 1943; Brust and House 1990). Some invertebrates build earthen chambers, controlling the microclimate, similar to shelter-builders aboveground (Haile 2001; Barnett and Johnson 2013). Prolonged drought conditions may favour those species capable of such behaviours. Similarly, larvae with morphological adaptations to flooding may fare better in areas predicted to experience increases in rainfall. Species that have evolved in flood-prone environments in particular, like the cranberry root grub, with water-repellent hairs along its body that can trap air (King et al. 1990; Villani et al. 1999), may stand to have competitive advantages over flood-intolerant species. Thus, invertebrates both above and belowground have evolved a range of behavioural and morphological adaptations to alterations in water availability. Differences in the use of such strategies between species and functional groups will likely lead to alterations in invertebrate community composition.

2.3.2 Indirect (plant-mediated) impacts

Invertebrate responses to plant quantity and quality: There is strong evidence in the literature for resource quantity-driven changes in invertebrate herbivore populations under altered precipitation regimes. Reduced rainfall results in reduced plant biomass, aboveground net primary productivity (ANPP), forage quality and cover, with increases in canopy light penetration and root:shoot ratios (Fay et al. 2003; Wu et al. 2011), leaving less plant biomass to support herbivores; however, there is strong evidence that this response is ecosystem dependent (Byrne et al. 2013). Accordingly, declines have been reported – across various ecotypes – in the abundances of Orthoptera (Kemp & Cigliano 1994); earthworms and scarabs (Davis et al. 2006; Mariotte et al. 2016); belowground herbivores (Staley et al. 2007); and across herbivore communities generally (Lee et al. 2014). Plant quality changes could also play a role in these declines. In a feeding experiment, army worm larvae reared on droughted Yorkshire fog grass (Holcus lanatus) took longer to develop and had higher mortality rates than those feeding on non-droughted grass, due to lower soluble protein content (Walter et al. 2012). While some trends can be identified in the responses of invertebrates to reductions in rainfall, there is a high degree of variation between species. For instance, gastropod species in a UK study had highly individual responses

18

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE to changes in water availability, with some benefitting from drought and others instead occurring in greater abundance under supplemented rainfall (Sternberg 2000).

Increases in average precipitation (to a degree – the negative effects of flooding in grasslands have been reviewed elsewhere – see Plum (2005)) may result in benefits to invertebrate herbivores, except in cases where increased moisture facilitates pathogens and disease (Grant and Villani 2003). On the whole, increases in precipitation lead to increases in ANPP (Zaller and Arnone 1999; Byrne et al. 2013). Consequently, studies have reported improved grasshopper nymph survival (Guo et al. 2009) and increased abundance and richness of grasshoppers (Lenhart et al. 2015). However, in contrast to the increases in grasshopper abundance reported in Lenhart et al., (2015) two other studies reported reductions in grasshopper survival under similar increased rainfall treatments (Barton et al. 2009; Guo et al. 2009).

Hence, a recurrent theme in the literature is that the responses of herbivorous invertebrates to altered precipitation will likely be idiosyncratic in nature, making it difficult to make generalised predictions about the directions of their responses under different scenarios (González-Megías and Menéndez 2012; Nielsen and Ball 2015). The responses of herbivores to both reduced and increased water availability are likely to be linked to the responses of their individual food-plant(s), as well as the invertebrate species own physiological precipitation optimum (Schowalter et al. 1999).

Invertebrate responses to plant community composition: The responses of a grassland plant community to alterations in rainfall depend on the type of grassland (i.e. average water state – mesic, xeric etc.) (Heisler-White et al. 2009), as well as the plant functional types (PFTs) that dominate the system (Collins et al. 2012; Andrey et al. 2014). For example, under altered rainfall frequency, with longer dry periods between more intense rainfall events, mesic grasslands generally experience a decrease in ANPP, whereas xeric grasslands show an increase (Fay et al. 2002; Fay et al.

2003; Heisler-White et al. 2009). In terms of PFTs, grasslands dominated by C4 grasses tend not to show stimulations in ANPP under increased rainfall, perhaps because they are likely to be less water limited than their C3 counterparts (Niu et al. 2008; Wu et al. 2011; Wilcox et al. 2015). Indeed, there is evidence that herbivorous insects consume relatively more C4 plants in years with reduced rainfall frequencies (Warne et al. 2010), possibly due to improved quality or increased quantities of these plants under such scenarios. Thus, we could expect that reorganisations occurring at the plant community

19

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE level in response to altered rainfall regimes will have consequences for herbivores, particularly for specialist feeders which may be reliant on the presence of just one or two plant species.

So far, experimental evidence directly linking precipitation-mediated changes in plant diversity to changes in the herbivore community is lacking. However, a five-year field experiment by Suttle et al. (2007) showed that whilst increased summer rainfall enhanced plant biomass, increased dominance and reduced grassland plant species richness had eventual negative consequences for the invertebrate community. Specifically, herbivore and consumer abundance declined and the invertebrate food web became simplified, potentially pointing to the loss of more specialised herbivores. This study demonstrates the importance of longer-term studies in detecting plant community shifts as opposed to more immediate biomass responses. Furthermore, Wilcox et al. (2016) recently showed that short-term plant community shifts in response to increased water availability may be misleading when considering shifts over a decadal time scale.

Secondary consumer responses to altered rainfall: Alterations occurring in the abundance and diversity of primary consumers can flow up through the food chain to affect populations of predators and parasitoids (Suttle et al. 2007; Lee et al. 2014), which may themselves be more sensitive to climatic change (Voigt et al. 2003). Buchholz et al. (2013) found reductions in semi-dry grassland spider and carabid diversity and abundance under water-limited conditions. However, at a similar site three years earlier, the same authors found no change in spider species richness, composition or abundance under precipitation manipulation (Buchholz et al. 2010). Similarly, in a Chinese steppe, reduced rainfall caused declines in herbivore abundance with no corresponding decline in secondary consumers (Zhu et al. 2014). Clearly, as with herbivores, there will be differences in the individual responses of higher trophic levels to changes in precipitation patterns.

Precipitation-sensitive species interactions: The idiosyncratic nature of invertebrate responses can be at least partially explained by complex precipitation-driven alterations in the interactions occurring between species within the system. The handful of studies which have tackled pairwise species interactions under precipitation manipulations have found complex, unpredictable responses with the potential to affect multiple trophic levels. For instance, spatially separated above- and belowground herbivores may influence each other through their

20

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE effects on the shared host plant, such as by altering secondary chemistry (Johnson et al. 2012). Staley et al. (2007) found that enhanced summer rainfall increased the abundance of leaf mining moths on wild basil, but not when root herbivores were present. The negative effects of root herbivores on leaf miner pupal size reduced the parasitism rates of moths above ground, indicating the potential for precipitation-altered species interactions to have knock-on consequences for higher trophic levels.

In a separate study, detritivorous tenebrionid beetles belowground had negative effects on the abundances of generalist sap sucking and chewing herbivores when summer precipitation was supplemented, similar to the findings of Staley et al. (2007) (González- Megías and Menéndez 2012). In contrast, the presence of belowground herbivores had positive impacts on aboveground leaf-mining flies, restoring their pupal weight and development time to ambient levels, when reared under drought conditions on milk- thistle (Staley et al. 2008). Taken together, these studies suggest that belowground organisms could serve to moderate the effects of reduced or increased water availability on aboveground herbivores, which may otherwise be expected to decrease or increase in abundance, respectively, in response to such rainfall regimes. As with the responses of individual species, the directions of the responses of the interactions between multiple species may also prove to be species- and system-specific. Further work is needed in the area to determine whether or not generalisations can be made, and to determine whether other interactions such as competition may also be affected by alterations in precipitation.

2.4 Invertebrate-mediated feedbacks on plant communities At the ecosystem scale, invertebrate herbivores generally exert weak control over grassland plant communities (Whiles and Charlton 2006; Coupe et al. 2009), though their impacts on plant species richness, for instance, may be stronger during herbivore outbreaks (Olff and Ritchie 1998). Altered precipitation has the capacity to change the relative strength of the interactions occurring between grassland plant and invertebrate communities, by altering the abundance and composition of species within the system. In a Canadian grassland, invertebrates caused short-term reductions in plant cover, increases in root mortality and altered plant composition, compared with pesticide treated plots (Coupe et al. 2009). The effects of the invertebrate community only became apparent under naturally-occurring drought conditions. This suggests that invertebrate

21

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE herbivores may exacerbate the negative effects of drought for grassland plants, and that grasslands may become more vulnerable to herbivores under drought.

In an American temperate old-field study, experimentally increased rainfall caused declines in grasshopper abundance, which translated into a 15% reduction in grasshopper-inflicted plant damage for every 1 cm of increase in mean monthly precipitation (Barton et al. 2009). Conversely, in a limestone grassland, the plant community sustained an increase in biomass under supplemented rainfall scenarios, despite a significant increase in the abundance of Auchenorrhyncha herbivores (leaf, plant and frog hoppers) (Masters et al. 1998). Assuming that this greater abundance of insects inflicted comparable levels of damage on a per capita basis as those in ambient plots, this would suggest that grassland plants may be able to maintain increased growth despite higher levels of herbivory under increased rainfall scenarios. These two studies demonstrate that the strength of indirect, invertebrate community-mediated effects of altered precipitation on grasslands will depend on the identities of both the plant community and invertebrate species involved.

2.5 Conclusions and future directions Depending on the underlying water-status of the ecosystem, alterations in rainfall may have generally negative direct and indirect consequences for invertebrates (summarised in Figure 2.1). Changes in precipitation will also have the potential to cause impacts spanning multiple trophic levels, moderating the outcomes of species interactions. Reductions in rainfall may exacerbate the negative effects of herbivores for the plant communities they inhabit, though other players in the system might alter this response (e.g. Staley et al. 2008). In order to better understand how grassland invertebrates – and the important ecological processes they underpin – will respond to altered precipitation, the following four areas should be targets for future research:

1. The incorporation of invertebrate responses in the design of current and future precipitation manipulation experiments: Invertebrate responses remain under-studied in rainfall manipulation experiments, with the majority of studies considering only the responses of plants to short-term rainfall alterations of limited scope – altered frequency scenarios, for instance, remain critically under-represented (Johnson et al. 2016). This under-representation, coupled with the idiosyncratic nature of the responses detailed to date, makes it difficult to identify solid trends and predict how grasslands will respond to a wide range of precipitation scenarios. Achieving such a goal will require an

22

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE increased number of studies from which to draw patterns, across a broader range of precipitation scenarios.

Fig 2.1: A summary diagram of the general trends expected or found in the literature (theoretical, experimental and observational studies) for plant, primary consumers (herbivores, detritivores etc.) and secondary consumers (predators and parasitoids), in response to altered precipitation regimes. Arrows: ▲ indicates increases in the given metric, ▼represents declines and ▲▼ denotes more varied results. References are given by the numbers on the diagram:1 Barnett & Johnson 2013, 2 Buchholz et al. 2013, 3 Coupe et al. 2009, 4 Davis et al. 2006, 5 Fay et al. 2002, 6 Fay et al. 2003, 7 Guo et al. 2009, 8 Heisler-White et al. 2009, 9 Kasper et al. 2008, 10 Lambeets et al. 2008, 11 Lee et al. 2014, 12 Lenhart et al. 2015, 13 Masters et al. 1998, 14 Palmer 2010, 15 Staley et al. 2007, 16 Staley et al. 2008, 17 Suttle et al. 2007, 18 Walter et al. 2012, 19 Warne et al. 2010, 20 Zhu et al. 2014.

2. A focus on long-term studies: Aside from the identities of the different components of the system, the timescale over which precipitation alterations are studied may also be important. The relatively short term nature of many field experiments to date obfuscates

23

CHAPTER TWO GRASSLAND INVERTEBRATES AND CLIMATE CHANGE our ability to make more realistic predictions about how grassland communities will respond to future changes (Beier et al. 2012). Such studies are needed in order to capture changes in the direction of responses over time and lags in the manifestation of responses – particularly given the potential for short-term studies to have misleading results compared with those over longer timescales (Suttle et al. 2007; Wilcox et al. 2016).

3. Greater geographical representation: This should be prioritised to determine the extent to which findings can be extrapolated across different biomes (the studies reviewed here are mostly from the UK and USA) (Beier et al. 2012; White et al. 2012). Given how many plant and insect responses are likely to depend on the underlying water-status of the system, research across ecotypes will be an essential target for enabling progress in the field.

4. Examination of multiple climate factors at once: There is a need for experiments reflecting the reality of global change which will involve the simultaneous alterations of many factors (Villalpando et al. 2009; Beier et al. 2012). This is especially important given the potential for synergisms between factors, as may be expected between increased temperatures and reduced water availability. On the other hand, the effects of one factor may serve to moderate those of another (e.g. Lee et al. 2014).

Such studies will not be without logistical difficulty, though future developments in technology will help to ease this, including improvements in long-term, sensor-based data gathering. Continued development of DNA-based methods like metabarcoding will assist community-level studies by reducing time-consuming work and taxonomic expertise (Cristescu 2014). Results from studies like those suggested above will provide critical information about grassland community responses for use in theoretical modelling approaches such as structural equation modelling, enabling the testing of theories at scales not yet possible experimentally. Such experimental and modelling approaches, carried out with broader geographical and precipitation-scenario representation, will be necessary in order for us to form more accurate predictions about the fate of these ecologically important grassland ecosystems under climate change.

24

Chapter Three: Predicted changes in rainfall patterns cause temporary perturbations in a grassland invertebrate assemblage which quickly recover

3.1 Summary Globally, grasslands form an important basis of agricultural production, biodiversity, carbon storage and sequestration. Grassland invertebrates, by virtue of their abundance and diversity, contribute to structuring these economically valuable ecosystems.

Increasing levels of carbon dioxide gas in the atmosphere are causing increases in global average temperatures and changes in precipitation patterns. Alterations in the average amount of rainfall, as well as the seasonality and frequency of rainfall, are likely to have implications for grassland ecosystems – many of which are water-limited. Currently, we know little about how changes in precipitation will affect grassland invertebrate communities, despite the potential for rainfall-induced disruptions in the valuable ecosystem processes these organisms underpin.

This study used a novel precipitation manipulation platform to investigate how changes in rainfall will affect invertebrate communities in an eastern Australian grassland. Two sampling methods were used to collect invertebrates over a sixteen month period from plots subjected to five watering treatments; i) ambient rainfall, ii) increased amount, iii) reduced amount, iv) reduced frequency and v) summer drought. Of interest was how predicted alterations in rainfall would affect invertebrates within the framework of the plant stress, pulsed stress and plant vigour hypotheses.

Watering treatments exerted weak and temporary influences on invertebrate abundance across sampling dates, which varied in the direction of change between taxonomic and feeding groups in an unpredictable manner. Treatment effects were strongest immediately after the cessation of the summer drought treatment when associated reductions in plant biomass were most apparent. Some limited evidence for the plant stress and plant vigour hypotheses was found within different taxonomic groups. Herbivores showed mixed responses; Hemiptera and Lepidoptera were increased in

25

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES abundance under reduced rainfall, whereas Diptera and Orthoptera were more abundant in increased amount plots.

Overall, the season-dependent, idiosyncratic effects of the altered precipitation scenarios tested suggest that the invertebrate community – and the ecosystem processes underpinned by it – may remain relatively stable as our climate continues to change.

3.2 Introduction Grasses are the most abundant plants on the planet, covering over 40% of the ice-free land surface (Tscharntke & Greiler 1995; Wang & Fang 2009). Grasses, and other associated plants, form vast areas of grasslands which are of great importance both economically – as centres of agricultural production and carbon sequestration, and ecologically – as areas harbouring high levels of biodiversity (Lee et al. 2014; Lenhart et al. 2015). Invertebrates are the dominant faunal components of many grassland ecosystems, in terms of abundance and diversity (Buchholz, Hannig & Schirmel 2013). These organisms play important parts in various ecosystem processes such as nutrient recycling and pollination, as well as pest control in agricultural landscapes through the provision of so-called ‘beneficial insects’ (Whiles & Charlton 2006; Werling et al. 2014). The roles performed by invertebrates in such ecosystem processes, including the pressure exerted by invertebrate herbivores, contribute to shaping the structure of the underlying plant community (Crawley 1989; Olff & Ritchie 1998).

Changes in water availability have the capacity to modify the interactions occurring between invertebrate herbivores and the plant communities that support them, as well as those occurring between members of the invertebrate community (Olff & Ritchie 1998; Huberty & Denno 2004; Staley et al. 2007; González-Megías et al. 2012). Global climate change, associated with the anthropogenic release of greenhouse gases and the linked increases in average temperatures, will lead to changes in global circulation and precipitation patterns (IPCC 2013). Current projections are highly variable; climate models predict changes in the average amount of rainfall – both increases and decreases – coupled with alterations in the seasonality of rainfall and the frequency with which rain events occur. Specifically, in many locations around the world there is likely to be an increase in the occurrence of extreme events like droughts, coupled with a reduction in the frequency of rainfall events (i.e. longer dry periods in between heavier downpour events) (Fischer, Beyerle & Knutti 2013; IPCC 2013).

26

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Alterations in water availability will affect grassland ecosystems – many of which exist primarily in water-limited states – by altering primary production and plant species composition (Knapp et al. 2002; Fry, Power & Manning 2014). Previous studies have reported declines in plant biomass and net primary productivity under reduced rainfall (Fay et al. 2003), with corresponding increases in the same metrics under increased rainfall (Byrne et al. 2013). Aside from the total amount of precipitation inputs, the seasonality and frequency of rainfall is also an important factor controlling processes like germination, species turnover and interspecific competition (Gao & Reynolds 2003; Huxman et al. 2003).

Alterations in the amount and seasonality of rainfall will also affect plant quality from the perspective of invertebrate herbivores. For instance, drought stress can cause changes in the concentrations of phytohormones and secondary metabolites (Hsiao 1973; Jamieson et al. 2012), along with increases in amino acids in the phloem due to the proteolysis of proteins into free amino acids (Brodbeck & Strong 1987; Girousse et al. 1996). This could have benefits for herbivores (plant stress hypothesis (PSH, White, 1969)), though alterations in other plant characteristics associated with plant quality under drought may negate such benefits, including increases in leaf toughness and reductions in turgor pressure (Huberty & Denno 2004). Reductions in rainfall frequency may provide benefits to herbivores, whereby the positive effects of water stress on nutrient content can be realised by herbivores – particularly sap-sucking taxa – during intermittent periods of relatively less stress, when turgor is recovered. This principle forms the basis of the ‘pulsed stress hypothesis’ (PuSH, Huberty & Denno, 2004). At the other end of the spectrum, fast growing, healthy plants, as may be encountered during increases in water availability, may represent more attractive resources for herbivores than slower growing individuals (plant vigour hypothesis (PVH) Price 1991; Endara & Coley 2011).

Invertebrate communities are underpinned by the plant production and species composition occurring in the system (Perner et al. 2005; Borer, Seabloom & Tilman 2012; Hertzog et al. 2016). Thus, any changes in the plant community, as well as changes in plant quality, resulting from alterations in precipitation regimes will impact the invertebrate communities they support. Given the range of ecosystem processes that are underpinned by invertebrates, there is a need to determine how precipitation changes – and the resulting alterations in the underlying plant community – will affect grassland

27

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES invertebrate communities, in order to facilitate the effective management of these valuable ecosystems (Zhu et al. 2014; Barnett & Facey 2015). Despite the functional and ecological importance of invertebrates in grassland systems, there is currently a lack of community-level studies considering invertebrate responses to altered rainfall, with many of the field-based precipitation manipulation experiments to date focusing on plant responses to rainfall (Barnett & Facey 2015). Of those studies which do look at invertebrate responses, many have simulated rainfall regimes associated with increases and/or reductions in rainfall compared with present-day (ambient) conditions, investigating one or two scenarios at most (e.g. Villalpando et al., 2009; Lee et al., 2014; Lenhart et al., 2015). However, the reality of the predicted changes in precipitation will be far more complex, with changes in the frequency of rainfall events and increases in the occurrence of extreme events like droughts which may be just as important in determining grassland responses as changes in the absolute amount of rainfall (Heisler- White et al. 2009; Peng et al. 2013; Johnson, Ryalls & Staley 2016b).

In part of the effort to remedy this gap in knowledge, this study used a novel precipitation manipulation experiment set in south east Australia to determine the effects of four different rainfall regimes on grassland invertebrate communities, in line with those predicted by climate change models. These regimes included both changes in the absolute amount of rainfall inputs (increased and reduced), as well as changes in the seasonality (summer drought) and frequency of rainfall (reduced frequency). Given the negative effects of reductions in rainfall on plant biomass reported across numerous different systems (e.g. Tilman & Haddi, 1992; Fay et al., 2003; Hoover et al., 2014), it was hypothesized that rainfall treatments resulting in reductions in water availability (reduced amount and summer drought) would cause reductions in the abundance of invertebrates. It was predicted that these declines would be particularly noticeable in herbivorous groups of arthropods (Lee et al. 2014), as their food plant resources become limited, with subsequent reductions in the consumer (predator and parasitoid) community (Buchholz et al. 2013). In contrast, it was predicted that reduced frequency rainfall would stimulate herbivore performance, leading to an increase in invertebrate abundance (PuSH), particularly for sap-sucking groups such as Hemipterans (which benefit from drought-induced alterations in plant chemistry during periods of increased turgor pressure). Given the naturally water-limited state of the study system, situated in a subtropical warm/moderate cool grassland, it was expected that increased rainfall

28

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES inputs would stimulate plant growth and lead to a greater abundance of invertebrates (PVH).

3.3 Materials and Methods

3.3.1 Experimental site The study was carried out at the DRI-Grass (Drought and Root herbivore Impacts on Grasslands) experimental site in Richmond, New South Wales, Australia (33°36′35″S, 150°44′18″E). The site has been more thoroughly described in Power et al. (2016). In brief, the experiment consists of an historically grazed (until 2001), unimproved grassland, containing 48 steel-framed rainout shelters with slanted, clear Perspex roofs (max height 140 cm) which each intercept rainfall over a 1.8 × 2.0 m area (3.6 m2), with 2.0 m between plots. Root barriers to a depth of 30 cm around the outside edge of each plot prevent lateral water and root movement. Since June 2013, rainwater has been reapplied to the 48 plots using automated mist irrigation systems fitted to the four corners of each plot, according to each of five watering treatments: AMB ambient rainfall; Increased Amount IA – ambient plus 50%; Reduced Amount RA – ambient less 50%; Reduced Frequency RF – accumulated ambient rainfall applied once every three weeks; and Summer Drought SD – interception of all rainfall for three months from December to March, with ambient rainfall levels applied outside of this time. The soil water content and rainfall amount applied under each of the water regimes is shown in Fig S3.1. Three of the treatments (AMB, RA and RF) are crossed with a belowground herbivore addition treatment as part of a co-occurring experiment (for details see Power et al. 2016). Each treatment combination is replicated six times – i.e. 12 × AMB, RA and RF plots (6 with supplemented root herbivores, 6 without) and 6 × IA and SD plots. In April and October of each year, beginning in the October of 2013, destructive biomass harvests have been carried out for each plot. For this, the central 1 m2 of vegetation is cut to ground level and dried at 80°C for 48 hrs before weighing. The ratio of material from C3 and C4 plant species is calculated for each plot.

The performance and abundance of C3 and C4 plants may be sensitive to altered rainfall as the distribution of plants has been shown to vary across rainfall gradients according to plant photosynthetic pathway (Teeri & Stowe 1976). Generally, C4 plants tend to perform better under water stress than C3 plants (Ward et al. 1999; Taylor et al. 2014). Such differences in plant functional types could confound the effects of altered precipitation on invertebrate abundance; the C3–C4 hypothesis posited by Caswell et al.

29

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

(1973) states that herbivores will tend to select more palatable and nutritionally-rich C3 plants over their C4 counterparts. Thus, the ratio of the biomass of plants in each of the two groups was correlated with the abundance of invertebrates in different taxonomic and functional groups. This allowed us to determine whether alterations in plant functional composition drove any of the changes in the invertebrate community in response to altered water availability.

3.3.2 Invertebrate sampling Aboveground invertebrates were sampled using two methods – suction sampling and yellow sticky card traps. Suction sampling was carried out three times in total (October (spring) 2013, April (autumn) and October 2014) using a petrol powered ‘G-Vac’ vacuum device (SH 86C, Stihl AG & Co. KG, Germany) passed over the plots in a zigzag pattern for 20 seconds within the vegetation, fitted with an organza bag to capture dislodged material and invertebrates. Suction samples were euthanized using

CO2 gas upon return to the laboratory and sorted under a dissection microscope (SZ51, Olympus, Japan). On a quarterly basis from October 2013, yellow sticky card traps (Bugs for Bugs, Mundubbera, Australia) were suspended from the centre of each rain- out shelter roof and left in place for one week, for a total of six sampling periods. This design ensured that invertebrates were sampled over four seasons – summer (January), autumn (April), winter (July) and spring (October). Traps were frozen in the laboratory prior to identification under a dissection microscope. All invertebrate sampling took place immediately before the corresponding biomass harvests, to ensure that removal of the vegetation did not affect invertebrate capture. All invertebrates were identified to at least Order level (except for two groups taken to Subclass only – Acari and Collembola). In order to more reliably determine feeding guilds, invertebrates were identified to Family level where possible (see Table S3.1 for the groups identified in the study and their guild classifications).

3.3.3 Statistical analyses All analyses were performed in R, version 3.2.3 (R Core Team 2015). Separate analyses were carried out on data from the two sampling techniques. Differences in the abundance of individual taxonomic groups and feeding groups occurring between the watering regimes were assessed using repeated measures linear mixed effects models (LME) in the package lme4 (function lmer, Bates et al., 2014). All data were square root- transformed prior to analysis in order to normalise residuals. Models contained the fixed

30

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES effects of watering treatment and time (date sampled) with the interaction between them, with plot as the random factor. Previous work has shown that abundance of invertebrates sampled using the G-Vac sampler is sensitive to the structure of the underlying vegetation (Facey & Torode 2016), thus all models for vacuum-caught invertebrates contained scaled plot-level plant biomass as a covariate. To account for the belowground herbivore treatment applied as part of a co-occurring experiment, herbivore treatment was also included in all statistical models as a co-factor. Model fit was assessed by inspecting residual plots. The effect of the interaction between watering treatment and date sampled was tested using likelihood ratio tests (LRTs) with the anova function comparing two models (one including the interaction and another with only additive effects of the two factors). Where significant, the interaction and main effect of watering treatment alone were tested using the Anova function from the car package (Fox & Weisberg 2011) to attain χ2 values. Where the interaction term was not significant, a LRT was instead used to determine the effect of watering treatment on invertebrate abundance, between the additive effects model and a reduced model without watering treatment as a fixed effect.

In addition, separate linear models were carried out for each invertebrate group (i.e. taxonomic or feeding guild) for each sampling date, on square root-transformed abundance data (to normalise residuals) (Zhu et al. 2014). To ensure adequate sample size, poorly represented groups (i.e. those with fewer than ten individuals sampled for a given time point) were not analysed. As with the repeated measures analyses, scaled plot biomass (for vacuum-caught samples) and root herbivore treatment were included in the models, and model fit was assessed by inspecting residual plots. The significance of the effect of watering regime on invertebrate abundance was determined using a LRT between the full model and a reduced model without watering treatment as a main effect. For both repeated measures and individual sample date models, where the effect of watering treatment was significant, Tukey post-hoc tests were used to determine differences between individual watering treatments using the glht function from the multcomp package, with p values adjusted for multiple comparisons (Hothorn, Bretz & Westfall 2008).

Permutational multivariate analysis of variance (PERMANOVA, adonis function from the vegan package; Oksanen et al., 2015) was used to assess the effect of watering treatment on invertebrate community composition, both in terms of Order and

31

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES functional guild identity. As with the previously described analyses, plot-level biomass (for vacuum-caught samples) and root herbivore treatment were included in the models alongside watering treatment as fixed effects. Analyses were carried out on Bray-Curtis dissimilarity matrices, permuted 999 times. Individual analyses were carried out for each sampling date. Additionally, to determine the overall effects of watering treatments on invertebrate community composition through time, invertebrate abundances in each of the taxonomic and feeding guild groupings were summed across sampling points and analysed in the same way. The community composition occurring under the different watering regimes was visualised using non-metric multi-dimensional scaling (NMDS, metaMDS function in the vegan package). Stress values remained below 0.2 across multiple runs with three dimensions (k). Plots were constructed for each sampling date separately; data points represent the abundance-based invertebrate community captured in each plot.

In order to relate changes in the abundance of the different invertebrate groups (both taxonomic and feeding guild-based) to changes in biomass related to altered rainfall regimes, the total abundance of invertebrates within each group, summed across sampling periods (October 2013, April and October 2014), was correlated with the total biomass collected during the corresponding harvests, for both sticky and vacuum samples. Similarly, total invertebrate abundances were correlated with the mean C3:C4 ratios over the same time period. All correlations were carried out using the cor.test function to generate Spearman’s rank correlation coefficients.

3.4 Results

3.4.1 Repeated measures analyses A total of 73,290 invertebrates were sampled and identified over the course of the study using the two sampling methods (Table S3.1). At the Order level, watering treatment had a season-dependent effect on the total number of invertebrates sampled using sticky traps (Table 3.1). Individual analyses by sampling date showed that the abundance of invertebrates varied significantly in response to altered watering regimes during April 2014 only (Table 3.2, Fig 3.1a). Specifically, invertebrates experienced a reduction in abundance under RA and SD conditions compared with IA plots (Tukey, p<0.05). For suction samples, the total abundance of invertebrates was also responsive to watering treatment, in both the April and October 2014 sampling campaigns (Table 3.2, Fig 3.1b). These differences were driven by increases in total invertebrate abundance under

32

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

RA conditions compared with IA plots in April (Tukey, p<0.05). Conversely, in October, invertebrate abundance was highest under the IA treatment compared with all other treatments, though there was no detectable difference between individual treatments despite an overall effect of watering regime on invertebrate abundance (Tukey, all p>0.05).

33

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Table 3.1: Results from repeated measure LME models carried out on the square root transformed abundance of the different Orders from each watering treatment over the duration of the experiment, for the two sampling methods. Values are derived from LRTs on the full model and a reduced model (i.e. one without the interaction term or fixed effect of watering treatment). Significant p values are highlighted in bold (p<0.05). † denotes that the sample size for this group was too small to permit analysis. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

Group Variable Sticky traps Vacuum samples Tukey post- hoc 2 2 χ (d.f.) p χ (d.f.) p Total Treatment*run 35.376 (20) 0.018 12.465 (8) 0.132 Treatment 3.140 (4) 0.535 11.510 (4) 0.021 non sig. post- hoc Diptera Treatment*run 62.398 (20) <0.001 48.460 (8) <0.001 Treatment 4.501 (4) 0.342 15.039 (4) <0.001 IA>SD Coleoptera Treatment*run 19.651 (20) 0.480 7.301 (8) 0.505 Treatment 7.134 (4) 0.129 12.478 (4) 0.014 RF>IA+AMB Araneae Treatment*run 15.839 (20) 0.727 27.322 (8) <0.001 Treatment 5.811 (4) 0.214 7.375 (4) 0.117 Acari Treatment*run † † 23.583 (8) 0.003 Treatment † † 2.433 (4) 0.657 Neuroptera Treatment*run 18.887 (20) 0.529 † † Treatment 4.544 (4) 0.337 † † Hemiptera Treatment*run 16.674 (20) 0.674 18.335 (8) 0.019 Treatment 6.228 (4) 0.183 1.668 (4) 0.797 Hymenoptera Treatment*run 28.287 (20) 0.090 9.147 (8) 0.330 Treatment 5.174 (4) 0.270 7.104 (4) 0.131 Thysanoptera Treatment*run 12.221 (20) 0.908 15.002 (8) 0.059 Treatment 2.454 (4) 0.653 0.655 (4) 0.957 Lepidoptera Treatment*run 20.810 (20) 0.408 3.059 (8) 0.931 Treatment 3.040 (4) 0.551 2.634 (4) 0.621 Psocoptera Treatment*run 13.416 (20) 0.859 13.394 (8) 0.099 Treatment 6.685 (4) 0.154 4.185 (4) 0.381 Orthoptera Treatment*run † † 15.452 (8) 0.051 Treatment † † 10.902 (4) 0.028 IA>RA Collembola Treatment*run † † 11.655 (8) 0.167 Treatment † † 4.803 (4) 0.308

34

Table 3.2: Results from LMs carried out on the square root transformed abundance of the different Orders from each watering treatment, for each individual sampling period, for the two sampling methods. Values are derived from LRTs on the full model and a reduced model (i.e. one without the fixed effect of watering treatment). Differences between individual treatments were calculated using Tukey post-hoc tests. The sign ‘>’ denotes that abundance from the preceding group(s) was significantly greater than that of the group(s) following the symbol. Significant p values are highlighted in bold (p<0.05). † denotes that the sample size for this group was too small to permit analysis. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

Sticky traps Tukey post-hoc Vacuum samples Tukey post-hoc Date Group Treatment χ2 p Treatment χ2 p (4 d.f.) (4 d.f.) Oct-13 Acari † † 7.625 0.106 Araneae 3.559 0.469 15.165 0.004 SD>RF Coleoptera 5.854 0.210 4.667 0.323 Collembola † † 3.084 0.544 Diptera 4.087 0.394 9.56 0.049 non sig. post-hoc Hemiptera 0.722 0.949 7.476 0.113 Hymenoptera 0.956 0.916 6.329 0.176 Lepidoptera 1.857 0.762 0.932 0.920 Neuroptera 6.312 0.177 8.524 0.074 Orthoptera † † 10.941 0.027 IA>RA Psocoptera 4.877 0.300 14.273 0.006 SD>Amb+RF Thysanoptera 2.581 0.630 1.627 0.804 Total abundance 1.696 0.792 4.551 0.337 Jan-14 Araneae 3.308 0.508 - - Coleoptera 0.792 0.939 - -

Diptera 1.143 0.887 - - Hemiptera 4.406 0.354 - - Hymenoptera 4.199 0.380 - - Thysanoptera 3.079 0.545 - - Total abundance 1.404 0.843 - - Apr-14 Acari † † 11.113 0.025 non sig. post-hoc Araneae 2.648 0.618 21.749 <0.001 SD+RF+RA>IA, SD>Amb Coleoptera 1.251 0.870 11.509 0.021 RA>IA Collembola † † 17.711 0.001 RA+RF>IA Diptera 14.471 0.006 IA>RA+SD 20.241 <0.001 Amb>RF+SD, IA>SD Hemiptera 2.085 0.720 11.654 0.020 RA>IA Hymenoptera 10.246 0.036 non sig. post-hoc 6.806 0.146 Lepidoptera 13.738 0.008 SD>RF 2.746 0.601 Orthoptera † † 4.193 0.380 Psocoptera 0.315 0.989 3.496 0.478 Thysanoptera 7.819 0.098 4.937 0.294 Total abundance 13.611 <0.001 IA>RA+SD 10.618 0.031 RA>IA Jul-14 Araneae 1.867 0.760 - - Coleoptera 0.960 0.916 - - Diptera 7.561 0.109 - - Hemiptera 6.226 0.183 - - Hymenoptera 9.135 0.058 - - Thysanoptera 0.792 0.940 - - Total abundance 5.393 0.249 - - Oct-14 Acari † † 2.243 0.691 Araneae 2.819 0.589 3.775 0.437 Coleoptera 7.549 0.110 5.105 0.277

Collembola † † 13.721 0.008 IA>RA Diptera 4.302 0.367 7.444 0.114 Hemiptera 1.962 0.743 5.998 0.199 Hymenoptera 4.886 0.299 13.493 0.009 SD>RA Lepidoptera 1.296 0.862 3.087 0.543 Neuroptera 1.178 0.882 † † Orthoptera † † 12.099 0.017 IA>RA Psocoptera 3.052 0.549 † † Thysanoptera 2.396 0.663 14.811 0.005 IA>RA Total abundance 3.270 0.514 10.749 0.030 non sig. post-hoc Jan-15 Araneae 5.349 0.253 - - Coleoptera 4.007 0.405 - - Diptera 4.552 0.336 - - Hemiptera 7.029 0.134 - - Hymenoptera 0.486 0.975 - - Psocoptera 0.579 0.965 - - Thysanoptera 3.917 0.417 - - Total abundance 4.239 0.375 - -

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Fig 3.1: the mean total abundance of invertebrates captured by sticky trapping, per plot (a) and vacuum sampling (scaled to 1 m2) (b) over the course of the experiment. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

For both sampling methods, various invertebrate Orders showed responses to the applied watering treatments. Seven out of the 13 groups tested across both sampling techniques showed significant effects, three of which were dependent on the date of

38

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES sampling (i.e. interactive effects between watering treatment and date sampled), with only vacuum-caught Diptera, Orthoptera, Coleoptera and overall invertebrate abundance showing general, non-seasonal effects of watering treatment (Table 3.1). Orthoptera were significantly more abundant under IA compared with RA plots, with Diptera from IA plots outnumbering those from SD plots (Tukey, all p<0.05, Fig 3.2c & f). Coleoptera, on the other hand, were more abundant in RF plots compared with AMB and IA plots (Tukey, all p<0.05, Fig 3.2d).

At the feeding guild level, fewer significant effects of watering treatment on invertebrate abundance were apparent, with three groups – detritivores, scavengers and predators – showing season-dependent responses to altered rainfall across the two sampling methods (Table 3.3, Figs 3.2 & 3.3). No feeding groups showed evidence of non- seasonal changes in abundance in response to the applied watering treatments.

Table 3.3: Results from repeated measure LME models carried out on the square root transformed abundance of the different feeding guilds from each watering treatment over the duration of the experiment, for the two sampling methods. Values are derived from LRTs on the full model and a reduced model (i.e. one without the interaction term or fixed effect of watering treatment). Significant p values are highlighted in bold (p<0.05). AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

Group Variable Sticky traps Vacuum samples 2 2 χ (d.f.) p χ (d.f.) p

Scavengers Treatment*run 11.360 (20) 0.936 22.770 (8) 0.004

Treatment 8.060 (4) 0.089 1.630 (4) 0.803

Detritivores Treatment*run 46.921 (20) <0.001 11.146 (8) 0.194

Treatment 4.382 (4) 0.357 8.819 (4) 0.066

Predators Treatment*run 13.913 (20) 0.835 23.145 (8) 0.003

Treatment 6.667 (4) 0.155 6.190 (4) 0.185

Omnivores Treatment*run 18.434 (20) 0.559 13.764 (8) 0.088

Treatment 0.666 (4) 0.954 2.661 (4) 0.616

Chewers Treatment*run 17.546 (20) 0.617 14.988 (8) 0.059

Treatment 0.613 (4) 0.962 1.756 (4) 0.781

Suckers Treatment*run 16.674 (20) 0.674 18.724 (8) 0.016

Treatment 6.228 (4) 0.183 1.602 (4) 0.808

Parasitoids Treatment*run 28.764 (20) 0.093 5.248 (8) 0.731

Treatment 5.054 (4) 0.282 6.069 (4) 0.194

39

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Fig 3.2: the mean abundance of invertebrates from different Orders (a, c-f) and functional groupings (b) captured by vacuum sampling (scaled to 1 m2) over the course of the experiment. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

40

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Fig 3.3: the mean abundance of invertebrates from different Orders (a, c) and functional groupings (b, d) captured by sticky sampling over the course of the experiment. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD. O = October, Ja = January, A = April, Ju = July.

3.4.2 Individual time-point analyses For sticky traps, three Orders showed changes in abundance in response to the different watering regimes. All responses occurred in April, with Diptera, Hymenoptera and Lepidoptera showing evidence of altered abundance (Figs 3.3 & S3.2). Two groups showed evidence of significant differences in abundance between treatment combinations: Diptera and Lepidoptera. For Diptera, this was driven by an increase in abundance under IA compared with RA and SD plots, which contributed to the same pattern seen in overall invertebrate abundance during this sampling campaign, and in the abundance of detritivores and omnivores (Tables 3.2 and 3.4, Figs 3.3 & S3.3). Lepidoptera were significantly more abundant under SD conditions compared with the

41

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

RF treatment; the abundance of chewing herbivores at this time point followed the same pattern (Table 3.4, Fig 3.3).

For vacuum samples, a greater number of significant changes occurred in Order-level abundance across the different sampling dates (Table 3.2, Fig 3.2), reflecting the greater number of significant interactions and direct effects of watering treatment found in the repeated measures analyses (Table 3.1). As with sticky traps, most changes occurred in the April 2014 sampling campaign. Araneae were reduced in abundance in treatments with greater water availability in both October 2013 and April 2014; SD plots contained significantly higher numbers of spiders compared with IA, AMD and RF plots (Fig 3.2a). The abundance of the predator feeding group reflected the responses seen for spiders, the dominant invertebrates in this group (Table 3.4, Fig 3.2b). Similarly, Psocoptera were significantly more abundant under SD conditions compared with RF and AMB plots during October 2013 (Fig S3.4). In April 2014, Coleoptera, Collembola, Hemiptera and the sucking invertebrate feeding group were all significantly more abundant in RA plots compared with IA plots (Tables 3.2 and 3.4, Figs 3.2, S3.4 & S3.5). This, in combination with the similar pattern observed for Araneae, drove the overall increase in total arthropod abundance seen in RA plots compared with IA ones during this sampling period (Fig 3.1).

Conversely, in the following October, Collembola and Thysanoptera were significantly more abundant in IA plots compared with RA plots, with changes in the former driving the same pattern in detritivore abundance found during the same sampling campaign (Figs S3.4 & S3.5). Also during the October 2014 sampling event, Hymenoptera were significantly more abundant under SD conditions compared with RA plots, though the same effect was not seen for the corresponding feeding group – the parasitoids (Figs S3.4 & S3.5).

42

Table 3.4: Results from LMs carried out on the square root transformed abundance of the different feeding guilds from each watering treatment, for each individual sampling period, for the two sampling methods. Values are derived from LRTs on the full model and a reduced model (i.e. one without the fixed effect of watering treatment). Differences between individual treatments were calculated using Tukey post-hoc tests. The sign ‘>’ denotes that abundance from the preceding group(s) was significantly greater than that of the group(s) following the symbol. Significant p values are highlighted in bold (p<0.05). † denotes that the sample size for this group was too small to permit analysis. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

Sticky traps Vacuum samples Date Group Treatment χ2 (4 d.f.) p Tukey post-hoc Treatment χ2 (4 d.f.) p Tukey post-hoc Oct-13 chewers 1.857 0.762 6.297 0.178 omnivores 1.991 0.737 4.611 0.330 parasitoids 1.205 0.877 6.604 0.158 predators 4.852 0.303 16.883 0.002 SD>RA+RF detritivores 2.814 0.589 3.372 0.498 scavengers † † 6.814 0.146 suckers 0.722 0.949 6.742 0.150 Jan-14 omnivores 2.718 0.606 - - parasitoids 4.349 0.361 - - predators 4.051 0.399 - - detritivores 3.588 0.465 - - scavengers 7.775 0.100 - - suckers 4.406 0.354 - - Apr-14 chewers 10.748 0.030 SD>RF 2.274 0.685 omnivores 11.758 0.019 IA>RA+SD 5.506 0.239 parasitoids 10.396 0.034 non sig. post-hoc 4.762 0.313

predators 1.425 0.840 19.471 0.001 SD>Amb+IA, RA+RF>IA detritivores 10.297 0.036 non sig. post-hoc 5.826 0.213 scavengers 0.630 0.960 8.708 0.069 suckers 2.085 0.720 12.194 0.016 RA>IA Jul-14 omnivores 1.935 0.748 - - parasitoids 9.048 0.06 - - predators 1.378 0.848 - - detritivores 8.592 0.072 - - suckers 6.226 0.183 - - Oct-14 chewers 2.622 0.623 10.455 0.033 non sig. post-hoc omnivores 2.048 0.727 8.955 0.062 parasitoids 3.639 0.457 7.607 0.107 predators 1.244 0.871 3.352 0.501 detritivores 2.320 0.677 14.620 0.006 IA>RA>RF scavengers 4.396 0.355 3.836 0.429 suckers 1.962 0.743 6.169 0.187 Jan-15 omnivores 2.485 0.647 - - parasitoids 0.483 0.975 - - predators 4.279 0.37 - - detritivores 7.001 0.136 - - scavengers 0.686 0.953 - - suckers 7.029 0.134 - -

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

3.4.3 Community level results

At the community level, altered rainfall had significant effects on abundance-based invertebrate community structure, at certain time points (Table 3.5). Specifically, both Order and feeding guild community composition were significantly altered during the April 2014 sampling campaign for both sticky trap and vacuum-caught samples. In all cases, this was linked to divergence occurring between RF and SD plots (Fig 3.4 a-d). SD plot communities were particularly different for vacuum-caught samples, differing from other treatments including IA and AMB plots (Fig 3.4 c and d). However, when the analyses were carried out with all sampling dates combined over time, all significant effects of watering treatment on community composition were lost for both sampling types and classification methods (Order or feeding guild, all p>0.05).

45

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Table 3.5: Results from PERMANOVA analyses on invertebrate community abundance-based composition, assigned either by Order-identity or feeding guild. Analyses were carried out with invertebrate abundance summed across all sampling dates, as well as for each sampling date separately. Significant p values are highlighted in bold (p<0.05).

d.f. Variable (treatment, SS MS Pseudo-F R2 p (perm) residual) Sticky traps: Orders: summed 4, 41 0.040 0.010 1.008 0.086 0.432 Oct-13 4, 41 0.039 0.010 0.372 0.033 0.963 Jan-14 4, 41 0.058 0.014 0.513 0.046 0.920 Apr-14 4, 41 0.200 0.050 2.616 0.199 0.021 Jul-14 4, 41 0.175 0.044 1.363 0.113 0.222 Oct-14 4, 41 0.056 0.014 0.853 0.075 0.593 Jan-15 4, 41 0.142 0.035 1.064 0.090 0.374

Guilds: summed 4, 41 0.025 0.006 0.722 0.063 0.725 Oct-13 4, 41 0.035 0.009 0.334 0.030 0.965 Jan-14 4, 41 0.041 0.010 0.429 0.039 0.943 Apr-14 4, 41 0.182 0.046 2.132 0.168 0.041 Jul-14 4, 41 0.203 0.051 1.370 0.114 0.201 Oct-14 4, 41 0.040 0.010 0.631 0.057 0.806 Jan-15 4, 41 0.124 0.031 1.126 0.096 0.326

Vacuum samples:

Orders: summed 4, 41 0.508 0.127 1.586 0.129 0.064 Oct-13 4, 41 0.988 0.247 1.267 0.105 0.190 Apr-14 4, 41 0.921 0.230 3.273 0.220 0.001 Oct-14 4, 41 0.527 0.132 1.404 0.107 0.113

Guilds: summed 4, 41 0.262 0.065 0.929 0.080 0.515 Oct-13 4, 41 0.897 0.224 1.221 0.101 0.232 Apr-14 4, 41 0.529 0.132 2.356 0.164 0.003 Oct-14 4, 41 0.456 0.114 1.386 0.106 0.139

46

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Fig 3.4(overleaf): NMDS plots of the invertebrate communities occurring under the different watering treatments in April 2014, divided either by Order identity (a and c) or feeding guild assignment (b and d). Stress values remained below 0.2 for three dimensions in all analyses. Sticky trap samples are shown in panels a and b, with vacuum samples in c and d. Points represent the invertebrate community occurring in each plot, with the colours representing the different watering regimes; grey = AMB, black = IA, yellow = RA, red = RF, blue = SD. Ellipses show the standard deviation of the centroid points for each treatment level. Scavengers = Sc, Saprotrophs = Sa, Predators = Pr, Omnivores = Om, Chewing herbivores = Ch, Sucking herbivores = Su, Parasitoids = Pa; Diptera = Di, Coleoptera = Col, Araneae = Ar, Acari = Ac, Neuroptera = Ne, Hemiptera = He, Hymenoptera = Hy, Thysanoptera = Th, Orthoptera = Or, Lepidoptera = Le, Psocoptera = Ps, Blattodea = Bl, Collembola = Coll.

47

a) b)

c) d)

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

3.4.4 Correlations of invertebrate abundance with plant data Invertebrate abundance was generally weakly correlated with measures of plant community composition (C3:C4 ratio) and resource availability (biomass). Both biomass and C3:C4 ratio varied significantly in response to the imposed watering treatments (Figs 3.5 & 3.6 respectively). All significant correlations of invertebrate abundance with plant biomass occurred for vacuum caught samples, with Diptera, Orthoptera and Psocoptera, as well as chewing invertebrates, showing evidence of positive relationships with plant biomass (Table 3.6). A greater number of groups showed correlations with the mean C3:C4 ratio of the plant community, in varying directions. Diptera, scavengers and detritivores showed negative correlations with C3:C4 ratios (i.e. were more abundant when C4 plants made up a greater proportion of the community). On the other hand, Thysanoptera, Coleoptera, omnivores and sucking herbivores showed evidence of positive correlations with the same metric (Table 3.6).

A summary of the groups showing support for the various hypotheses discussed in this study is shown in Table 3.7.

49

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Table 3.6: Significant correlation results from Spearman’s rank correlations of invertebrate abundance with measures of plant composition and quantity. Total invertebrate abundance was summed over the three sampling periods for which data on C3:C4 and biomass were available (Oct 13, Apr and Oct 14) into Orders and functional groups. These data were then correlated with total biomass or mean C3:C4 ratios over the same period. Positive ρ values indicate positive correlations, with negative values indicating negative correlations between the two variables.

Group Trapping method Spearman’s ρ S statistic p value

C3:C4 RATIO: Thysanoptera Sticky 0.453 10081 0.001 Vacuum 0.355 11891 0.013 Diptera Sticky -0.334 24572 0.020 Coleoptera Sticky 0.405 10965 0.004 Scavengers Sticky -0.308 24090 0.033 Detritivores Sticky -0.373 25296 0.009 Omnivores Sticky 0.367 11663 0.010 Suckers Vacuum 0.347 12038 0.016

BIOMASS: Diptera Vacuum 0.351 11956 0.014 Orthoptera Vacuum 0.554 8218.9 <0.001 Psocoptera Vacuum 0.392 11195 0.006 Chewers Vacuum 0.533 8609.9 <0.001

50

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Table 3.7: summary of the findings for invertebrate primary consumers in relation to the hypotheses discussed in this study.

Hypothesis Description Supportive groups in our study

Orders Guilds Plant Stress Reductions in water promote Psocoptera, Chewing Hypothesis (PSH) beneficial alterations in plant Lepidoptera herbivores (White 1969) quality for invertebrate Coleoptera Suckers Hemiptera Detritivores herbivores Collembola Pulsed Stress The benefits of drought-altered Collembola Hypothesis foliage can be realised by (PuSH) (Huberty herbivores during periods of & Denno 2004) plant recovery following rain events

Plant Vigour Herbivores perform best on Orthoptera Omnivores Hypothesis vigorously growing, non-stressed Diptera Detritivores (PVH) (Price (i.e. well-watered) plants Collembola 1991)

C3–C4 Hypothesis Herbivores perform best on Thysanoptera Omnivores

(Caswell et al. more palatable C3 species, Coleoptera Suckers 1973) tending to avoid feeding on less

nutritious C4 species.

51

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

Fig 3.5: Total plant biomass recovered from plots subjected to the five watering treatments over the course of the experiment, scaled to 1 m2. Watering treatment had a significant effect on plant biomass 2 across all treatment dates (repeated measures LMM, χ 4 = 10.988, p = 0.027). This was driven by 2 the significant effect of watering treatment on plant biomass during April 2014 (LM, χ 4 = 18.000, p = 0.003). Specifically, summer drought plots contained significantly less plant biomass than those plots under the ambient, increased and reduced frequency regimes (Tukey HSD, all p<0.05). AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

52

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

3.0 AMB RA 2.5 IA * RF SD 2.0

1.5

ratio 4

:C 3 1.0

C

0.5

0.0

Oct 13 Apr 14 Oct 14

Fig 3.6: Mean C3:C4 ratios of plant biomass recovered from plots subjected to the five watering treatments over the course of the experiment. Watering treatment showed no evidence of having an effect 2 on the C3:C4 ratio of plant biomass across all treatment dates (repeated measures LMM, χ 4 = 6.674, p = 0.154). In October 2014, watering treatment had a significant effect on the C3:C4 ratio of plant 2 biomass (LM, χ 4 = 11.854, p = 0.018). Specifically, summer drought plots contained significantly higher amounts of plant biomass from C3 plants relative to C4 plants compared with those plots under the reduced amount and reduced frequency regimes (Tukey HSD, all p<0.05). AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

53

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

3.5 Discussion This study has demonstrated the capacity for grassland invertebrate communities to show complex responses to the varied precipitation scenarios we can expect as a result of climate change. In general, watering treatments exerted weak and temporary influences on invertebrate abundance across sampling dates, which varied between taxonomic and feeding groups. Indeed, overall, most groups were unresponsive to changes in moisture conditions. Of those groups that did show responses to the imposed watering treatments, the majority showed season-specific responses, with most effects occurring during April when the impacts of the summer drought treatment – applied during December to March – will have been most apparent.

It was anticipated that the abundance of invertebrates would be reduced under scenarios leading to reductions in water availability, driven by corresponding declines in plant productivity. A significant decline in plant biomass under summer drought conditions was found in April, as expected, with corresponding declines in the abundance of Diptera during the same period. Thysanoptera, Collembola, Orthoptera, omnivores and detritivores also showed evidence of reduced abundance under more water-limited scenarios compared with ambient levels or increased water availability over the course of the experiment, in line with findings from other researchers (Guo et al. 2009; Zhu et al. 2014; Lenhart et al. 2015). This, in combination with the positive correlations found between the abundance of Diptera, Orthoptera, Psocoptera, chewing herbivores and plant biomass lend some support to the hypothesis and the PVH, given the stimulated plant biomass in increased amount plots. Moreover, the positive correlations between invertebrate abundance and plant biomass were found for vacuum caught samples, despite the documented negative effect of increased structural sward complexity on invertebrate capture using this sampling method (Brook et al. 2008; Facey & Torode 2016). Thus, this study may have underestimated the true positive effect of plant biomass – and increased water availability – on these invertebrate groups. Nonetheless, the effect of watering regime on plant biomass was no longer apparent by the following harvest, due to a reduction in biomass across all treatments coupled with renewed growth in the summer drought plots following the cessation of the drought treatment

(as evidenced by the sharp increase in C3:C4 plant ratios in this treatment during October 2014). This suggests that alterations in water availability are unlikely to have strong, long-lasting impacts on plant productivity in the ecosystem, and the invertebrates which rely on it.

54

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

No evidence was found for the increases in herbivore/primary consumer abundance (Diptera, Orthoptera, Thysanoptera and Collembola) under increased water availability leading to increases in the abundance of predators – indeed, the predatory functional group (spiders) instead increased under progressively more water-limited conditions. Habitat complexity is known to be an important factor structuring spider assemblages (Rypstra et al. 1999). In this study, such a pattern could be caused by increases in the abundances of spiders less reliant on complex vegetation structure for web-building, such as cursorial wandering spiders (e.g. Thomisidae, Lycosidae), given the reduction in plant biomass found in these plots. Greater taxonomic resolution would be needed in this study to determine whether or not there is a shift occurring in the Family or species identity of spiders under the different rainfall regimes. These spiders are known to feed on Collembola and Hemipterans, among other small, soft-bodied arthropods (Nyffeler & Breene 1990) which also increased in abundance under reduced water availability during the same sampling period. This suggests that these predators may be tracking the availability of their prey, and that any increases in the abundance of predators did not mask changes in the abundance of organisms likely to constitute their prey.

Aside from predators, other groups showed evidence of increases in abundance under decreasing water availability, contrary to expectations, including groups with plant-based diets like Lepidoptera, Hemiptera, Psocoptera and sticky trap-caught chewing herbivores. Reduced water availability and increased drought stress could result in increases in plant quality for invertebrate herbivores (PSH) (White 1974, 1984). This could have led to increases in the abundance of these groups relative to increased rainfall conditions. The findings for these taxa provides support for the PSH (summarised in Table 3.7), suggesting that these organisms may be able to overcome any associated reductions in plant quality (e.g. increased leaf toughness, reductions in turgor pressure) and quantity, instead benefitting from altered plant chemistry under reduced water availability. A similar number of groups showed negative responses to increased rainfall/positive responses to reduced rainfall as those indicating the opposite (i.e. increased abundance under less water-limited conditions). Previous work by other researchers has found declines in invertebrate abundance under increased watering regimes despite increases in plant biomass, driven primarily by declines in plant species richness and potentially by declines in plant nutritional quality and increases in disease prevalence (Grant & Villani 2003; Suttle et al. 2007; Gao et al. 2008). The same

55

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES mechanisms could be operating at DRI-Grass – further research looking at plant chemical and species composition would be needed to confirm this.

This study predicted that sap-feeding herbivores like Hemiptera would be less negatively affected by reductions in water availability (PSH) and may stand to benefit from reduced rainfall frequency (PuSH). This study found support for the PSH in this group; Hemiptera were positively affected by reductions in the total amount of applied water, to a point. Under more extreme reductions in water availability – the summer drought treatment – Hemiptera abundance did not differ compared with ambient conditions, suggesting that past a certain level of drought-stress, any positive effects of plant water- stress on Hemiptera may be lost. This may be as a result of reductions in turgor pressure inhibiting feeding by sap-sucking insects (Huberty & Denno 2004). These findings are in line with those found in a glasshouse study by Tariq et al. (2012), who reported improved aphid performance on plants receiving moderate drought stress only; when plants were highly drought stressed, benefits to aphid performance were no longer apparent. As for alterations in rainfall frequency, the mean abundance of Hemiptera in reduced frequency plots was comparable to that in reduced amount plots, but did not differ significantly from ambient levels. Thus, this study found no support for the PuSH in this group, though with continued data collection over longer time scales, the differences in abundance between ambient and reduced frequency plots may reach significance.

In terms of what changes in the plant community may be underpinning the observed alterations in invertebrate abundance, the abundance of various groups varied in accordance with plant biomass and C3:C4 ratio. Based on the C3–C4 hypothesis (Caswell et al. 1973), it was expected that herbivorous groups would be positively correlated with the biomass of C3 plants in the system, as these plants are often nutritionally favourable.

Of eight significant correlations of abundance against C3:C4 ratio, five were positive, including for groups with plant-based diets such as omnivores, sucking herbivores and Thysanoptera (Table 3.7), providing some support for this hypothesis. The negative correlations of the remaining three groups with the same metric, including Diptera which are also likely to have plant-based elements in their diets, suggest that the preferences of invertebrate groups for different plant functional types will be group- specific. This is supported by findings from other studies which have found contrasting patterns between invertebrate consumption and plant functional types in different

56

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES

systems (Pinder & Kroh 1987). Alterations in C3:C4 ratios between watering treatments were only apparent during the October 2014 sampling campaign, during which period very few groups showed evidence of responses to the applied treatments. Biomass, however, varied during April, when most responses in the invertebrate community were apparent, suggesting that alterations in biomass may be more important for the invertebrate community than those in plant functional type-composition.

For all the groups in this study, differences in abundance between watering treatments may have been confounded by the movement of highly-mobile invertebrates between plots. This is an inherent issue in open, plot-level field experiments (Moise & Henry 2010) and one that is a necessary compromise given limited funding and the urgent need for climate change studies. Previous work has found alterations in invertebrate communities at fine (metre) scales (e.g. Muff et al., 2009) – this in combination with the findings for highly mobile Diptera for instance, consistent between sampling methods, suggest that this study will have been able to detect any effects of the applied watering regimes on invertebrate abundance occurring in the community. Indeed, changes in invertebrate community composition were detected in April, even with relatively coarse- scale taxonomic and feeding guild assignments. This indicates that changes in water availability, particularly prolonged dry periods like summer droughts, could lead to alterations in the composition of the invertebrates within the ecosystem, potentially leading disruptions in the processes underpinned by these organisms.

However, the effects of watering regimes on invertebrate community structure, like those seen for abundance, were short lived and not apparent when data were summed over time. This indicates that the applied drought was not strong enough to alter the plant community to the point of ‘community closure’, whereby the community can no longer support certain species after they have been lost, even if they are reintroduced (Lundberg, Ranta & Kaitala 2000). Such resilience may be a product of the selection pressures exerted by the strong climatic variability experienced over geological time in the area (Chiew et al. 2011), and suggests that the community will be resilient to change. The possibility that geographical areas with more stable climates will be more greatly affected by alterations in water availability cannot be ruled out, especially those experiencing more extreme drought conditions, sustained over longer periods and across greater areas. Such changes could have the potential to alter grassland

57

CHAPTER THREE ALTERED RAINFALL AND INVERTEBRATE COMMUNITIES invertebrate communities, particularly if there is less potential for migration of animals to rebalance the community following such extreme events.

In conclusion, this study has provided support for three widely researched hypotheses in the field of water availability impacts on invertebrates. Overall, the generally weak, season-dependent effects of the precipitation scenarios adopted in this study on invertebrate abundance and community structure suggest that the invertebrate community in the study system – and the ecosystem processes underpinned by it – may remain relatively stable as our climate continues to change. The resilience of the invertebrate community may be a result of the strong base-line seasonal variation in rainfall availability over geological time scales in the study system – areas with more stable climates may sustain greater alterations to their invertebrate communities than those documented here.

58

Chapter Four: The effects of atmospheric change on forest invertebrates

Published as Facey, S.L. & Gherlenda, A.N. (2016) In: Global Climate Change and Terrestrial Invertebrates (in press) (eds Johnson, S.N. & Jones, T.H.). John Wiley & Sons, Chichester, UK. Overall, SLF and ANG undertook 80% and 20%, respectively, of the work associated with this chapter.

4.1 Summary Predicting the responses of invertebrate species, and the communities they form, to global change is one of the great challenges facing modern ecology. Invertebrates play vitally important roles in forests, underpinning fundamental ecosystem processes like nutrient cycling and pollination. Changes in the composition of our atmosphere, associated with increased levels of carbon dioxide (CO2) and ozone (O3), have the potential to affect the abundance, diversity and structure of invertebrate communities and the ecosystems they support. This chapter reviews the findings from the body of work looking at the responses of invertebrates to changes in CO2 and O3 concentrations with a special focus on the results from Free-Air Enrichment studies. The most consistent finding across the studies reviewed here is the idiosyncratic nature of the responses of invertebrate species to the elevation of CO2 and/or O3. This finding can be explained to some extent by bottom-up and top-down processes. These include the species- and genotype-specific responses of host plant chemistry and differences in the abilities of individual insect species to physiologically and behaviourally overcome changes in resource quality. Although evidence is clearly mixed, certain general conclusions can be made regarding the influence of CO2 and/or O3 on invertebrates. Forest invertebrate herbivores tend to respond negatively to elevated concentrations of

CO2. This response is likely due to diminished food-plant quality. Conversely, predators and parasitoids may benefit under enriched-CO2 conditions as prey susceptibility increases. Elevated O3 concentrations generally have opposing effects: herbivores show a tendency to consume more and develop faster while higher trophic levels experience declines in performance. Therefore, simultaneous elevation of both gases, such as is found in reality, may moderate the effects of either gas in isolation. There also appears 59

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE to be some capacity for invertebrate communities to rebound over time, as evidenced by long-term studies. From the few community-level studies available, the current conclusion is that the structure of invertebrate communities will not be strongly disrupted by increases in CO2 and O3. This suggests that the ecosystem processes underpinned by these communities may be maintained under future atmospheres in these systems, though more work is needed.

Looking forward, the critical need for long term studies of invertebrate responses at the population and community-level within natural systems is emphasized. Such studies will be particularly important in tropical regions where no such information currently exists. Studies incorporating multiple climatic and atmospheric factors will also be of great value, such as those looking at the combined effects of atmospheric change and alterations in water availability. These studies will allow us to better predict the effects of future climates on these fundamental ecological systems.

4.2 Why are forest invertebrate communities important? Invertebrates form the foundation of terrestrial ecosystems, far outnumbering their vertebrate counterparts in terms of abundance, biomass and diversity (Wilson 1987). In addition to comprising unrivalled levels of biodiversity – such as that found in the tropics – forest arthropod communities have multi-faceted roles in ecosystem functioning. Herbivorous arthropods play a central part in nutrient cycling through consuming plant material and returning nutrients to the forest floor as frass (Hunter 2001; Meehan et al. 2014; Gherlenda et al. 2015a), with others decomposing leaf litter and non-living materials (Speight, Hunter & Watt 1999). The feeding and burrowing of these detritivorous invertebrates contributes to soil formation, fertility and health which, in turn, determine the capacity of the system to sustain primary production (Schowalter 2006). Many insects are also involved in the pollination of a wide range of plant species, further supporting plant growth and diversity (Potts et al. 2010). At higher trophic levels, arthropods form the diet of insectivorous vertebrates such as birds and reptiles, as well as invertebrate predators and parasitoids. In addition to underpinning and contributing to the biodiversity and functioning of forest ecosystems, forest invertebrates play roles in other systems. Arthropod natural enemies, such as parasitoid wasps for instance, can have beneficial functions as pest-suppressors in agricultural systems and forests may act as reservoirs of these insects at the landscape-scale (Bianchi, Booij & Tscharntke 2006; Bianchi, Goedhart & Baveco 2008).

60

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

Whilst the majority of invertebrates arguably have supporting, stabilising roles in the maintenance of diverse forest ecosystems (Schowalter 2006), there is the potential for invertebrates to become pests, particularly in species-poor plantations (reviewed in Hartley 2002), when they consume abnormally high amounts of plant material. Further, natural and managed systems alike are at increasing risk from non-native, invasive arthropods (Tobin, Parry & Aukema 2014); a recent study estimated that wood-boring invasive insects cost 1.7 billion USD annually in the USA, in government expenditure alone (Aukema et al. 2011). Thus, forest invertebrates have strong ecological and economic importance.

Clearly, forest ecosystems – and the invertebrate communities which they encompass – represent invaluable global resources from several different viewpoints. Forests are vital from an ecological perspective, providing clean air and other ecosystem services. In addition, forests often harbour great biodiversity – indeed, 15 of the 25 global biodiversity hot spots are areas of tropical forest (Myers et al. 2000). On the other hand, forests can be viewed as economic sources of capital from forestry and other related industries. Perhaps most importantly in the context of this chapter, the world’s major forests represent vast carbon stores and the largest terrestrial sink at a time when there is increasing pressure to reduce emissions of carbon dioxide (CO2). Tropical forests for instance, comprise 50% of the carbon stored in terrestrial biomass (Ainsworth & Long 2005). Further, it has been estimated that, globally, forests are responsible for sequestering up to 60% of the carbon emitted from burning fossil fuels (Pan et al. 2011), highlighting the potential for the world’s forests to slow increases in atmospheric CO2 concentrations.

4.3 Atmospheric change and invertebrates Changes in the composition of the Earth’s atmosphere, associated with emissions of

CO2 and other greenhouse gases, have the potential to affect invertebrates and the ecosystems they underpin. Predicting the responses of species, and the communities they comprise, to global change is one of the great challenges facing modern ecology (Gilman et al. 2010; Andrew et al. 2013; Facey et al. 2014). Many studies have looked at the implications of altered atmospheres for natural and farmed systems, with most measuring plant responses (Coviella & Trumble 1999). So far, the majority of studies considering the effects of elevated greenhouse gas concentrations on plants and the

61

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE animals they support have been concerned with the direct and indirect effects of carbon dioxide (CO2) and tropospheric ozone (O3).

As for other less-studied atmospheric gases, mono-nitrogen oxides (NOx) and sulphur dioxide (SO2) are two indirect greenhouse gases known to be detrimental to plant growth (Allen 1992), through wet and dry deposition and the resulting acidification of ecosystems. Continued emissions of these gases are likely to have negative indirect consequences for heterotrophic food webs, the specifics of which have been reviewed elsewhere (Greaver et al. 2012; Park 2014). In addition, atmospheric NOx also play a role in the formation of tropospheric ozone, by reacting with other gases in the presence of sunlight, further exacerbating the negative effects of these gases.

The effects of the two direct greenhouse gases nitrous oxide (N2O) and methane (CH4) on invertebrates will be indirect through their general warming effects on average global temperatures. The consequences of global warming will be profound for ectothermic arthropods (reviewed in Bale et al. 2002), but are beyond the scope of this chapter. This chapter reviews invertebrate community responses to changes in CO2 and O3 concentrations with a special focus on the results from Free-Air Enrichment studies. Where possible, this review also compares findings with experiments from more controlled settings, as well as those from non-forest ecosystems.

4.4 Responses of forest invertebrates to elevated carbon dioxide concentrations

4.4.1 Herbivores

Whilst instances of direct effects of elevated CO2 on insect herbivores are scarce in the literature (but see Awmack, Woodcock & Harrington 1997; Stange 1997; Mondor et al. 2004), the indirect consequences, mediated by physiological and morphological changes in plants, can be extensive. Elevated atmospheric concentrations of CO2 have been shown to stimulate plant growth across a range of systems, through so-called ‘carbon fertilisation’ effects (Lamarche et al. 1984). These larger plants often have changed primary and secondary chemistry as a result of physiological changes occurring in the plant in response to increased CO2 availability; indeed, a recent meta-analysis reported an average increase in plant tissue C:N ratios of 19% when grown under elevated CO2

62

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE conditions (Robinson et al. 2012). Given the reliance of arthropods on their food sources to acquire nitrogen for growth (Mattson 1980), changes occurring in plants in response to elevated CO2 conditions often lead to negative effects for herbivorous invertebrates. Numerous studies in multiple systems have shown decreases in insect herbivore development rates, conversion efficiency, pupal mass and survival under elevated CO2, often coupled with increases in consumption rates – so called ‘compensatory feeding’ (Bezemer & Jones 1998; Zvereva & Kozlov 2006; Stiling & Cornelissen 2007; Robinson et al. 2012). In addition to compensatory feeding, insect herbivores may display other behavioural or physiological responses to altered foliar nutrition in an attempt to minimise impacts on their growth and development resulting from nutritionally sub-optimal food. Generalist herbivores may choose to preferentially feed on tree species or genotypes which are minimally impacted by changes in CO2 concentration (Osier, Hwang & Lindroth 2000; Agrell et al. 2005). Further, some insect herbivores can increase their nitrogen use efficiencies when feeding on leaves grown at elevated CO2 in order to overcome nutrient reductions (Williams, Lincoln & Norby 1998; Hattenschwiler & Schafellner 2004).

Insects belonging to the order Lepidoptera are the most commonly studied within forested FACE sites (Table 4.1). In these systems, elevated CO2 has been observed to increase lepidopteran larval developmental time and reduce larval survival, relative growth rate and pupal mass (Roth et al. 1998; Lindroth, Wood & Kopper 2002; Stiling et al. 2002; Knepp et al. 2007). Evidence has also been found for compensatory feeding by leaf mining Lepidoptera in the field (Stiling et al. 1999, 2002, 2003), further supporting the findings from glasshouse work. Thus, as seen in more controlled experimental systems, the overall effect of elevated CO2 on forest invertebrate larval performance has proved to be negative. However, other field-based studies have found no effect (Kopper et al. 2001; Kopper & Lindroth 2003) or positive effects (Hattenschwiler & Schafellner 2004; Couture, Meehan & Lindroth 2012; Couture & Lindroth 2012) on larval performance parameters under elevated CO2. Such variation in the direction of change and/or the responsiveness of insect performance to elevated CO2 may be due to differences in the responses of their food sources; i.e. variation between individual tree species. For instance, Hattenschwiler & Schafellner (2004) observed a 30% decrease in the relative growth rate of gypsy moth larvae under elevated CO2 when feeding on oak, while relative growth rate increased by 29% when feeding on hornbeam. In a similar

63

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE study, Lindroth & Kinney (1998) reported a 30% decrease in the growth rate of gypsy moth larvae under elevated CO2 conditions when feeding on aspen, while growth rate was not altered when larvae fed on maple. Different genotypes of the same tree species can also display this disparity in insect herbivore responses (Roth et al. 1998; Lindroth et al. 2002; Agrell et al. 2005). Experiments conducted with plants grown in greenhouse conditions have observed similar, generally negative, trends in insect performance (Stiling & Cornelissen 2007; Robinson et al. 2012), as well as host-plant dependent differences in insect performance (Gherlenda et al. 2015b), mirroring those results obtained from field experiments under elevated CO2 conditions. Thus, studies in various settings have revealed generally negative effects of elevated CO2 on invertebrate herbivores, though there is considerable variation between study species. The differing responses of insects to various food-plant species and genotypes may be explained by the individualistic responses of plant chemistry to elevated CO2. Secondary compounds present within leaves may increase in response to elevated CO2, and the degree of this increase may be species dependent (Ryan et al. 2010). Some secondary compounds are known to bind nitrogen, reducing the availability of this critical nutrient for assimilation by insects (Schweitzer et al. 2008), thus impacting their growth and development. Secondary compounds may also deter or prohibit insects from compensatory feeding (Roth et al. 1998), limiting their ability to mitigate CO2-induced reductions in foliar nutritional quality. Thus, species and/or genotype-specific differences in foliar chemistry under elevated CO2, coupled with the plasticity of herbivorous insect species in dealing with these changes in foliar chemistry, may account for the varying response in insect performance under elevated CO2 conditions.

4.4.2 Natural Enemies Work from glasshouse experiments has shown that for invertebrate predators and parasitoids, elevated CO2 could lead to improved fitness through increases in the susceptibility of their prey/hosts to attack, as mediated by changes in host-plant quality (e.g. Chen, Ge & Parajulee 2005; Coll & Hughes 2008). Parasitoids and predators have generally received little attention in forested FACE sites owing to the difficulties in monitoring their performance. One study by Stiling et al. (2002) conducted in open top chambers found that parasitoid attack rates on leaf mining Lepidoptera increased by

80% under elevated CO2 in scrub forest. This was potentially due to prolonged larval development times which lead to increased susceptibility to parasitoid attack; this finding is in line with expectations from previous work in controlled settings. Increased

64

Table 4.1: Summary of the literature observing individual insect responses to elevated concentrations of CO2 and O3 conducted within forested FACE sites. Cells highlighted in green indicate positive impacts, orange cells indicate no overall change and red cells indicate negative response of insects to elevated CO2 and O3.

Factor(s) Insect species Tree species Insect parameters Summary of insect outcomes Reference Birch Increased consumption at eCO and eO . (Betula papyrifera) 2 3 Larvae preferred aspen over birch at eCO Forest tent 2 which was opposite at eO . No CO × O (Agrell et al. CO × O caterpillar Aspen Consumption 3 2 3 2 3 interaction on consumption. Aspen 2005) Malacosoma disstria (Populus tremuloides, genotype affected consumption at elevated genotype 216 and CO and O 259) 2 3 Oak (Quercus petraea) RGR decreased on oak (Hattenschwil Gypsy moth Beech RGR increased on hornbeam er & CO RGR 2 Lymantria dispar (Carpinus betulus) Schafellner 2004) Hornbeam RGR did not change on beech (Fagus sylvatica) Herbivore eCO alone had little effect on herbivore forest tent 2 Developmental time performance and was reduced when CO caterpillar Aspen 2 (Holton, Larval mass and O were combined on genotype 216. Malacosoma disstria (Populus tremuloides, 3 Lindroth & CO × O Pupal mass Parasitoid 2 3 genotype 216 and Nordheim Survivorship Survivorship overall decreased at O and was dipteran parasitoid 259) 3 2003) (parasitoid) offset in combination with CO . An O × Compsilura concinnata 2 3 CO2 × genotype interaction was detected for survival

O3 alone improved herbivore performance Developmental time White oak Survival Reduced performance at elevated CO2. (Quercus alba) RGR (Knepp et al. CO Antheraea polyphemus 2 ECD 2007) Black oak Larvae increased ECI at elevated CO2 for AD (Quercus velutina) white oak only ECI Whitemarked Birch Developmental time eCO and eO had no effect on insect (Kopper et al. CO × O Tussock Moth 2 3 2 3 (Betula papyrifera) Pupal mass performance 2001) Orgyia leucostigma

eCO2 had minimal effect on insect Aspen performance. CO2 × O3 interaction Forest tent Developmental time (Kopper & (Populus tremuloides, observed in which CO2 negated the effect of CO × O caterpillar Consumption Lindroth 2 3 genotype 216 and eO3. Significant CO2 × O3 × genotype Malacosoma disstria Pupal mass 2003) 259) observed

Larval performance improved at eO3. Aspen Whitemarked (Populus tremuloides, Developmental time eCO reduced insect performance. CO × (Lindroth et al. CO Tussock Moth 2 2 2 genotype 216 and Pupal mass genotype interaction observed 2002) Orgyia leucostigma 259) Forest tent Forest tent caterpillar pupal mass reduced at caterpillar eCO2 which was magnified in combination (Percy, Malacosoma disstria Aspen Pupal mass with eO3. Awmack & CO2 × O3 (Populus tremuloides) Abundance eO alone increased pupal mass Lindroth Aphid 3 2002) Chaitophorus stevensis Aphid abundance unaffected by CO2 or O3

Aspen Developmental time eCO reduced insect growth and food Forest tent (Populus tremuloides) RGR 2 processing efficiencies. Negative effect of (Roth et al. CO caterpillar RCR 2 eCO on insects were larger on maple than 1998) Malacosoma disstria Maple AD 2 aspen (Acer saccharum) ECD (Awmack, Aphid Developmental time Birch No effect of CO or O on aphid Harrington & CO × O Cepegillettea Adult weight 2 3 2 3 (Betula papyrifera) performance Lindroth betulaefoliae RGR 2004)

AD - approximate digestibility; ECD - efficiency of conversion of digested food; ECI - efficiency of conversion of ingested food; eCO2 - elevated CO2 concentration; eO3 - elevated O3 concentration; RGR - relative growth rate.

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE parasitism rates, in combination with decreased herbivore performance under elevated

CO2, may suggest that insect herbivores will experience population reductions in forest systems as CO2 concentrations continue to rise. On the other hand, predatory and parasitic organisms will also be subjected to lower quality food sources in terms of the abundance, size and chemistry of their prey/hosts (reviewed in Facey et al. 2014). A study by Percy, Awmack & Lindroth (2002) found that although the density of aphid natural enemies was increased under future levels of CO2, the synchrony between the two trophic levels was disrupted, leading to aphid population release and thus reductions in pest regulation. Conversely, in a separate study by Holton, Lindroth & Nordheim (2003), parasitoid fitness (survival, developmental time and adult mass) was not affected by elevated CO2. Clearly, as with herbivorous insects, the responses of higher trophic levels to altered CO2 concentrations appear to be idiosyncratic.

4.4.3 Community-level responses Given the idiosyncratic nature of the responses of all levels of the system (i.e. plants, herbivores and higher trophic levels) to elevated concentrations of CO2, making informed predictions about the fate of forest ecosystems under future atmospheres becomes increasingly daunting (e.g. Hillstrom, Couture & Lindroth 2014). It is problematic to scale up and extrapolate the responses of one individual species seen under glasshouse conditions to the world at large around us, given the likelihood that the response of the species in question does not adequately represent those of the multiple components in natural systems. Instead, these types of studies have greater value from a mechanistic perspective and allow the formation of ideas and hypotheses for testing in the field (discussed further in Lindroth & Raffa 2016). In reality, organisms exist in nature as parts of intricately interwoven communities, linked to one another by trophic and other biotic interactions, with each subject to change in the abiotic environment. Studies from FACE sites integrate the complex biotic and abiotic factors occurring in the system (Hillstrom & Lindroth 2008), though these too, are not without limitations inherent to plot-level studies of this type (Moise & Henry 2010; Lindroth & Raffa 2016). Community-level assessments of forest invertebrate responses to elevated

CO2 have been conducted in various systems and will be the focus of this section. As with experiments under more controlled and simplified settings, the results of community level field studies are mixed, though some patterns can be gleaned from the literature. Community and population-level studies have, on the whole, confirmed the

68

Table 4.2: A summary of the literature considering the effects of elevated CO2 and/or O3 concentrations on multiple species of invertebrates in forest/woodland ecosystems. Results show the effects of elevated concentrations of the given gas i.e. a reported decrease represents a decline under elevated compared with ambient conditions. Cells highlighted in green indicate positive trends, orange cells indicate no overall change and red cells indicate mostly negative effects on different groups of invertebrates. FACE = Free Air Carbon dioxide Enrichment; OTC = Open Top Chamber.

Factor(s) Ecosystem Organism Method Summary of outcomes Reference Loblolly pine Decrease in herbivore abundance, Increase CO plantation Arthropods Sticky traps (Hamilton et al. 2012) 2 in carnivorous orders (Duke FACE) Loblolly pine

CO2 plantation Micro-arthropods Soil samples Reduced micro-arthropod abundance (Hansen et al. 2001) (Duke FACE) Sweetgum Pitfall traps and No effect on herbivory, total abundance or (Sanders, Belote & CO Liquidambar styraciflua Arthropods 2 sweep-netting richness Weltzin 2004) (Oak Ridge FACE) Scrub-oak community Reduced abundance of leafminers. Increased CO (Kennedy Space Centre Leafmining insects Census (Stiling et al. 1999) 2 mortality and leaf consumption OTC) Scrub-oak community Leaf damage, Reduced herbivory and survival of

CO2 (Kennedy Space Centre Insect herbivores census of leafminers. Increased mortality by (Stiling et al. 2002) OTC) leafminers parasitoids Scrub-oak community Leaf damage, Reduced leaf damage by leafminers and CO (Kennedy Space Centre Insect herbivores census of (Stiling et al. 2003) 2 chewers. OTC) leafminers Scrub-oak community (Stiling & Cornelissen CO Leafmining insects Census Reduced abundance 2 (Kennedy Space Centre 2007)

OTC)

Scrub-oak community Pitfall traps and Increased herbivore abundance (Stiling, Forkner & CO (Kennedy Space Centre Arthropods 2 Sticky traps Drake 2010) OTC) No effect on abundance of other arthropods Mature mixed forest Reduced abundance CO2 Soil Collembola Soil samples (Xu et al. 2013) (WebFACE) No effect on community richness Mature mixed forest CO Arthropods Canopy beating Reduced arthropod diversity (Altermatt 2003) 2 (WebFACE) Reduced abundance under single exposure Mixed Aspen to CO2 or O3 (Loranger, Pregitzer CO × O plantation Soil Invertebrates Soil samples 2 3 & King 2004) (Aspen FACE) No effect on abundance when CO2 and O3 are combined

Reduced pupal mass under elevated CO2 (forest tent caterpillar). Increased aphid predator abundance Trembling aspen Forest tent Increased pupal mass under elevated O3 (forest tent caterpillar). O increased aphid CO2 × O3 Populus tremuloides caterpillar, aphids Census 3 (Percy et al. 2002) (Aspen FACE) and their predators abundance and reduced aphid predator abundances

No effect on aphid abundance when CO2 and O3 are combined Mixed Aspen Weak, idiosyncratic effects – unlikely to CO × O plantation(Aspen Insect herbivores Census (Hillstrom et al. 2014) 2 3 influence composition or abundance FACE)

Elevated O3 reduced total abundance. Mixed Aspen Altered community composition (Hillstrom & Lindroth CO × O plantation Arthropods Pan traps 2 3 2008) (Aspen FACE) Elevated CO2 had on effect on abundance. No effect on species richness Loblolly pine Leaf chewing CO plantation Leaf damage Reduced herbivory at elevated CO (Knepp et al. 2005) 2 insects 2 (Duke FACE) Loblolly pine Herbivorous CO by plant species interaction, either CO plantation Leaf damage 2 (Hamilton et al. 2004) 2 insects reduced or no effect on herbivory (Duke FACE) Mixed Aspen Herbivorous Increased herbivory at elevated CO2 (Meehan et al. 2014; CO × O plantation Leaf damage 2 3 insects Couture et al. 2015) (Aspen FACE) Reduced herbivory at elevated O3 Eucalyptus woodland Leaf chewing Frass (Gherlenda et al. CO No effect on frass deposition 2 (EucFACE) insects collections 2015a) Reduced total arthropod abundance, reduced abundance of certain trophic groups Pitfall traps, Eucalyptus woodland including herbivores, omnivores, scavengers CO Arthropods Sticky traps and (Facey et al. 2016) 2 (EucFACE) and parasitoid wasps. No change in Vortis sampling community structure.

Eucalyptus woodland Leaf chewing (Gherlenda et al. CO Leaf Damage No effect of CO on leaf damage 2 (EucFACE) insects 2 2016b) Eucalyptus woodland Reduced abundance of psyllid community at (Gherlenda et al. CO2 Psyllid community Census (EucFACE) elevated CO2. 2016a)

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

negative effects of elevated CO2 concentrations on forest invertebrate herbivores (Table 4.2). Other studies have shown alterations in the species richness of certain groups under future CO2 levels, leading to increases in species dominance and changes in feeding guild composition (Altermatt 2003; Sanders et al. 2004). In a Eucalyptus FACE system, Facey et al. (2016) found declines in the abundance of several Orders and feeding guilds of invertebrates under elevated conditions, including chewing herbivores and omnivores. This effect was also seen for Eucalyptus-feeding psyllids at the same site (Gherlenda et al. 2016a). As for the mechanisms behind the patterns we see, the effects of elevated CO2 concentrations on populations of herbivorous invertebrates may be dependent on underlying chemical changes occurring in food plants (e.g. Stiling et al. 2003), and has been shown to vary between host species (Altermatt 2003); both of these findings are in line with expectations from previous work in the field and glasshouses. However, evidence for this mechanism in the most commonly used plant quality measure, C:N ratio, is still somewhat elusive, with some studies showing reductions in herbivore performance with no concurrent expected increase in this metric (Hamilton et al. 2004, 2012; Facey et al. 2016).

The assessment of CO2-induced changes in the abundances of leaf chewing insects in forested FACE sites has often be determined by the measurement of leaf damage. Based on evidence from glasshouse experiments, two contrasting hypotheses could be drawn; i) leaf damage will decrease as a result of reduced survival and performance of leaf chewers under elevated CO2, or ii) leaf damage and frass inputs will increase as leaf chewing insects display compensatory feeding at elevated CO2. Declines in leaf damage have been observed in forested FACE sites, indicating reduced abundances of leaf chewing insects (Hamilton et al. 2004; Knepp et al. 2005). Conversely, increased leaf damage has also been observed, indicating either an increase in leaf-chewer or increased compensatory feeding behaviour by comparable numbers of leaf-chewers (Hillstrom et al. 2010; Meehan et al. 2014; Couture et al. 2015). Further complicating matters, no difference in leaf damage or frass production from leaf chewing insects has been observed in an evergreen Eucalyptus forest under elevated CO2 (Gherlenda et al. 2015a, 2016b). Eucalyptus trees generally grow in nutrient-poor soils and are often water limited (Crous, Ósvaldsson & Ellsworth 2015; Duursma et al. 2015) which may pose challenges for insect leaf chewers. Under these conditions, water availability has been shown to be

72

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

a stronger driver than elevated CO2 in altering the abundance of leaf chewing insects (Gherlenda et al. 2015a, 2016b).

For higher trophic levels, evidence from community-level studies supporting the findings from population and species-level experiments is somewhat less apparent. In support of the predicted increases in natural enemy performance, increases in the abundance of the carnivorous Orders Araneae and Hymenoptera (Hamilton et al. 2012) and parasitism rates (Stiling et al. 1999, 2002, 2003) have been reported in myrtle oak and loblolly pine forests, coupled with declines in herbivorous groups. However, predator abundance and richness declined in a sweetgum plantation under future concentrations of CO2 (Sanders et al. 2004). In a Eucalyptus woodland, the abundance of parasitic wasps declined significantly under elevated CO2, but the abundance of more generalist predators like spiders was unresponsive (Facey et al. 2016). Another study in a scrub oak forest found no effects of elevated CO2 on predatory group abundance (Stiling et al. 2010). Hence, the evidence concerning the responses of natural enemies to elevated CO2 remains mixed.

Studies looking at soil fauna under elevated CO2 conditions have consistently found reductions in the abundance of micro arthropods like springtails and mites (Hansen et al. 2001; Loranger et al. 2004; Xu et al. 2013), perhaps as a result of microbial changes occurring in the soil (Hansen et al. 2001). These organisms play important roles in litter decomposition and nutrient cycling in forest ecosystems (Hansen et al. 2001); reductions in their population densities could thus feasibly have ramifications for soil health and primary production.

Given that most studies to date report changes in the abundances of different functional groups in one direction or another, one might feasibly expect to see alterations in overall invertebrate community composition occurring as a result of elevated CO2. Very few studies have considered the effects of future CO2 levels on composition, with only one study reporting significant effects of CO2 fumigation on community composition which varied over time (Hillstrom & Lindroth 2008). Of the other three studies looking at community composition, none reported significant changes occurring in response to

CO2 elevation (Sanders et al. 2004; Hillstrom et al. 2014; Facey et al. 2016). The same lack of CO2 effect was true for a similar field-based experiment in a grassland ecosystem

73

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

(Villalpando et al. 2009). Thus, there is limited evidence for the idea that elevated CO2 will have strong effects on invertebrate community structure. The most emergent trend in the literature is the idiosyncratic nature of responses of forest invertebrates to elevated CO2, at all levels of organisation from individual species to communities. Indeed, it is this finding which could explain the general lack of community shifts under elevated CO2, as the many components within natural systems respond in opposing ways, effectively ‘cancelling each other out’ (Hamilton et al. 2012). Documented responses are individualistic both in terms of being specific to the organisms involved as well as the time point sampled; community shifts have been reported to vary from month to month and year to year in several studies (e.g. Hillstrom & Lindroth 2008; Gherlenda et al. 2015a, 2016b).

Looking forward, there is the potential for any emerging responses to adjust to increasing levels of CO2. Indeed, it could be expected that the higher levels of parasitism and predator abundance reported in some studies (e.g. Hamilton et al. 2012) may not be supported in the long term if populations of herbivores continue to decline, as food resources become limiting for them. On the other hand, observed negative trends could recover – for instance, the reduced abundances of soil micro arthropods may be expected to rebound over time as litter inputs increase in the system (Hansen et al. 2001). Additionally, concurrent studies in a scrub oak ecosystem by Stiling et al., have found that although herbivore (leaf miners and tiers) densities initially decreased under elevated CO2 at the beginning of their experiment (Stiling et al. 2002), the total number of herbivores eventually showed a trend towards increasing under these conditions (Stiling et al. 2009, 2010). This was due to boosted plant growth, supporting higher numbers of insects at reduced densities per leaf (Stiling et al. 2009). Indeed, this supports findings from more recent work in aspen plantations documenting increases in herbivore-mediated nutrient flux (Meehan et al. 2014; Couture et al. 2015). These examples, both theoretical and empirical, support the need for long-term monitoring in order to produce the most realistic insights into how forest invertebrate communities will respond to atmospheric change (Couture & Lindroth 2013).

74

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

4.5 Responses of forest invertebrates to elevated ozone concentrations

4.5.1 Herbivores

In contrast to the generally positive effects of CO2 on plant growth, tropospheric O3 causes a cascade of negative effects on plant physiology, leading to growth inhibition and accelerated senescence (reviewed in Lindroth 2010). Plants grown in the presence of

O3 have reduced biomass, lower photosynthetic capacity and reduced leaf area; a recent meta-analysis indicated that, at current levels of O3, plants have decreased biomass of

7% compared with those grown at pre-industrial levels of O3 (Wittig et al. 2009). As with elevated concentrations of CO2, O3 can cause altered plant secondary chemistry, but in generally different ways  O3 exposure often leads to increased levels of compounds like phenolics and reduced carbohydrate concentrations (Valkama, Koricheva & Oksanen

2007; Lindroth 2010). On one hand, this changed plant chemistry under O3 fumigation can stimulate compensatory feeding behaviour by herbivores - potentially due to the reduced quality of nitrogenous compounds in plants - and consequently lead to higher levels of herbivory and generally reduced insect performance (Coleman & Jones 1988a; Jones & Coleman 1988). On the other hand, some herbivorous insects achieve higher pupal masses and develop faster under O3 fumigation (Percy et al. 2002; Kopper &

Lindroth 2003; Holton et al. 2003; reviewed in Valkama et al. 2007). Hence, as with CO2, there is variation in the strength and/or direction of the responses of insect herbivores to elevated O3. In addition to positive and negative responses in herbivore performance, herbivorous insects have also been observed to show no response to elevated O3 in a forested (Awmack et al. 2004) and a soybean (Hamilton et al. 2005; Dermody et al. 2008)

FACE site. In those systems where herbivore performance is sensitive to O3 fumigation, changes in the abundance and success of these groups may lead to alterations in competitive dynamics with other species, highlighting the need for studies integrating the biotic interactions occurring within systems (Coleman & Jones 1988b).

In common with elevated CO2, the effects of O3 fumigation can vary between plant species, and this has been shown to lead to changes in the feeding preferences of herbivorous insects (Agrell et al. 2005). Likewise, there is a need to consider the effects of O3 fumigation over longer timescales as systems may adapt to changes in O3 levels and fumigation may have variable effects at different stages of invertebrate life cycles. For instance, willow leaf beetles showed a preference for ozone treated cotton-wood foliage and consumed higher levels of leaf material, suggesting that they may benefit

75

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

from O3 exposure; however over the longer term, these beetles were less fecund on O3 treated plants and showed a preference for ovipositing on non-fumigated controls (Coleman & Jones 1988a; Jones & Coleman 1988). Hence, while studies have generally revealed positive effects of elevated O3 on invertebrate herbivores, there is considerable variation between study species, as seen with studies looking at elevated CO2.

4.5.2 Natural Enemies

For higher trophic levels, increased levels of O3 may have negative consequences as a result of the O3-catalysed breakdown of biogenic volatile organic compounds (BVOCs) which play a role in host location (Butler, Beckage & Trumble 2009; Himanen et al. 2009; Pinto et al. 2010). Whilst studies have indeed shown that ozone can reduce parasitoid larval survivorship and searching efficiency (Gate, McNeill & Ashmore 1995; Holton et al. 2003), others in crop systems have shown that ozone fumigation does not significantly affect BVOC-dependent predatory mite or parasitoid search behaviour and success (Vuorinen, Nerg & Holopainen 2004; Pinto et al. 2007, 2008). These studies suggest that the foraging abilities of some invertebrates may be conserved under a higher O3 atmosphere, particularly those sensitive to the presence of BVOCs which are resistant to O3-breakdown (Pinto et al. 2007). Studies looking at parasitoid foraging efficiency and performance under elevated O3 conditions in forest systems are currently lacking (with the exception of Holton et al. 2003), though based on these studies from other systems it could be expected that any responses of individual parasitoid species will be idiosyncratic in nature.

Predatory and parasitic invertebrates responses to elevated O3 environments will depend on the behavioural and physiological responses of their prey (Lindroth 2010) which will also be variable. One might expect that elevated O3 may tend to benefit predators and parasitoids in those instances where prey/hosts achieve higher pupal masses. However,

Holton et al. (2003) showed that while forest tent caterpillars growing under elevated O3 were larger, parasitoid larval survivorship decreased despite access to larger food resources. This was potentially a result of chemical limitations in the host relative to the parasitoids’ requirements (Holton et al. 2003). Additional causes for the reduction in parasitoid survival under elevated O3 could be the result of improved host immunological function (Hättenschwiler & Schafellner 1999) and faster host developmental times (Karowe & Schoonhoven 1992). In terms of host behaviour,

76

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

Mondor et al. (2004) found that aphids on trembling aspen had improved predatory escape behaviour under O3 fumigation as they became more sensitive to alarm pheromones released by conspecifics indicating the presence of danger. This could lead to higher population densities of these insects and potential pest outbreaks as the concentration of O3 continues to increase. Overall, there is consistent evidence for negative effects of O3 fumigation on forest invertebrate natural enemies.

4.5.3 Community-level studies Few studies exist looking at the responses of forest systems at the population and community level to elevated O3 concentrations, with those that do often also looking at the simultaneous elevation of CO2 (the focus of the next section). The results from such studies have provided support for the negative effects of O3 elevation on higher trophic levels in some cases. For instance, Percy et al. (2002) found that aphid populations at

Aspen FACE were more likely to grow under O3 fumigation as natural enemies became less abundant, pointing to a loss of effective top-down aphid regulation under future O3 scenarios. On the other hand, in a separate study on a different aphid species at Aspen

FACE, Awmack et al. (2004) showed that although aphid populations grew under O3 fumigation, they were unaffected when predators and parasitoid populations were also present. This suggests that, for this species, effective pest-suppression will be maintained under future atmospheres.

One of the three existing community-level studies by Hillstrom and Lindroth (2008) found that O3 fumigation resulted in declines in the abundance of parasitoid wasps whilst tending to positively influence the abundance of sucking herbivores. They found time-dependent alterations in arthropod community composition in response to O3 elevation, as a result of these population changes. In contrast, in a more recent herbivore-only study by Hillstrom and Lindroth (2014) at the same site, the opposite pattern emerged with sucking herbivores experiencing population declines relative to chewing herbivores. In an environmental gradient study, Jones and Paine (2006) also showed that chewing herbivores became more dominant compared with sucking herbivores on oak trees, as levels of O3 pollution increased. And so, as with CO2, elevated O3 may cause alterations in invertebrate population abundance to the extent that species (or group) dominance alters within the ecosystem. At the same time, there is

77

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

also a general lack of evidence for community shifts occurring as a result of O3 fumigation, though studies are very limited.

Clearly, like CO2-related changes in invertebrate communities, the effects of O3 will likely be idiosyncratic and difficult to predict, with changes in the direction of community responses documented through time and between systems within experimental sites. In contrast with the effects of elevated CO2 (general reductions in herbivorous groups, some increases in predators), there is evidence for enriched O3 environments benefitting herbivores and disadvantaging parasitoids. For soil micro arthropods however, there is evidence for more consistent declines in the abundance of microbial feeding mites under both CO2 and O3 exposure (Loranger et al. 2004).

4.6 Interactions between carbon dioxide and ozone

Given the generally opposing directions of the effects of elevated CO2 and O3 on insect performance (Fig 4.1), the study of the simultaneous elevation of these gases will be important in order to gain realistic insights into the effects of future atmospheres on invertebrate communities (e.g. Valkama et al. 2007; Lindroth 2010). Despite the logistical issues associated with studies investigating multiple global change drivers, several studies have considered the effect of the elevation of these two gases on invertebrates.

In some cases, elevated O3 has been demonstrated to mitigate the negative effects of elevated CO2 on insect performance. The benefits of O3 fumigation on forest tent caterpillar performance were offset when CO2 concentrations were also elevated

(Kopper & Lindroth 2003; Holton et al. 2003), as were the negative effects of O3 elevation on its parasitoid (Holton et al. 2003). Conversely, other studies have found evidence for interactive effects of CO2 and O3. A study by Percy et al. (2002), also on the forest tent caterpillar, found that the negative effects of elevated CO2 on female pupal mass were worsened under higher levels of O3. They also found no evidence for a moderating effect of CO2 and O3 on aphid infestations, which were just as severe under simultaneous fumigation as when either gas was elevated singly (Percy et al. 2002). In a separate study, whitemarked tussock moths showed a non-significant trend toward reduced pupal mass under combined elevation of CO2 and O3, with no trends seen when either gas were elevated in isolation (Kopper et al. 2001).

78

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

Fig 4.1: Conceptual diagram summarising the main directions of the responses of invertebrates to elevated CO2 and O3. Arrows show the direction of responses based on the literature; the length of the arrows give some idea of the degree of confidence (i.e. the number of studies showing the same result). Adapted from Facey et al. (2014).

At the community level, Loranger et al. (2004) found that soil micro-arthropod abundance was not significantly different compared with ambient conditions when CO2 and O3 were simultaneously elevated. Similarly, Hillstrom and Lindroth (2014) found that simultaneous elevated CO2 and O3 exposure had little effect on invertebrate community composition in aspen and birch woodland. In an earlier study, they found generally opposing responses of invertebrate groups to O3 and CO2. In addition, 79

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE

fumigation with CO2, O3 and both CO2 and O3 caused idiosyncratic effects on invertebrate community composition which varied over time. Thus, simultaneous elevation of CO2 and O3, as will likely be experienced in nature, can cause invertebrates to respond in ways that are inconsistent with studies elevating either gas in isolation.

4.7 Conclusions and future directions The use of forested FACE sites to assess the impacts of atmospheric change on insects is fundamental in understanding how insects will respond to future atmospheres, from the individual to the community level. Whilst the responses of invertebrates to elevated

CO2 and O3 are highly species dependent, some general trends can be gleaned from the literature. Forest invertebrate herbivores tend to respond negatively to elevated concentrations of CO2, whilst predators and parasitoids may benefit under enriched-

CO2 conditions. Under O3 enrichment, on the other hand, herbivores show a tendency to consume more and develop faster, with higher trophic levels showing more consistent declines in performance. Therefore, simultaneous elevation of both gases, such as will occur in reality, could serve to moderate the effects of either gas in isolation (Kopper & Lindroth 2003; Agrell et al. 2005; Couture & Lindroth 2012), though evidence for this is currently mixed. This, coupled with the general lack of observed community shifts occurring under CO2 and/or O3 elevation, suggests that the ecosystem processes provided by invertebrate communities may not be strongly responsive to simultaneous changes in atmospheric concentration of these gases (Hillstrom et al. 2014). However, there are still too few experiments at sufficient scales, or across the range of systems needed, to draw solid conclusions about the fate of invertebrate communities. Tropical ecosystems, for instance, remain unrepresented despite their immense significance as both biological hotspots and prominent carbon sinks (Ainsworth & Long 2005).

As pointed out by other authors (e.g. Lindroth 2010; Couture & Lindroth 2013), the integration of other environmental factors beyond CO2 and O3 will now be required to better understand how invertebrate communities and the ecosystems they form, will respond to global change. For instance, increases in temperature and changes in the timing and amount of rainfall will have profound effects on invertebrates. Effectively assessing these environmental variables will be a logistically difficult but necessary challenge at the scale of forested FACE systems. The incorporation of multiple global

80

CHAPTER FOUR FOREST INVERTEBRATES AND ATMOSPHERIC CHANGE change variables over long time periods, coupled with greater representation across forest ecosystems, will further advance our understanding of how invertebrate communities will respond to climate change. This will allow improvements in model accuracy and more effective conservation strategy development, to ensure the sustainable management of the world’s forests and the invertebrate communities that underpin them.

81

Chapter Five: Atmospheric change causes declines in woodland arthropods and impacts specific trophic groups

Published as Facey, S.L., Fidler, D.B., Rowe, R.C., Bromfield, L.M., Nooten, S.S., Staley, J.T., Ellsworth, D.S. & Johnson, S.N. (2016) Agricultural & Forest Entomology (accepted). The writing of this chapter was carried out entirely by SLF, with advice and editorial input from SSN, JTS, DSE and SNJ.

5.1 Summary Arthropod assemblages form a fundamental part of terrestrial ecosystems, underpinning ecosystem processes and services. Yet, little is known about how invertebrate communities, as a whole, respond to climatic and atmospheric changes, including predicted increases in carbon dioxide concentrations (CO2).

To date, woodland Free Air CO2 Enrichment (FACE) studies have focused entirely on northern hemisphere managed plantations. This study manipulated atmospheric CO2 in a mature, native Eucalyptus woodland (0.15ha, >32,000m3) in Australia, using the EucFACE facility. Using three complementary sampling methods (vacuum sampling, pitfall and sticky trapping), this study recorded invertebrate abundances under ambient and elevated levels of CO2 (400 vs. 550 ppm).

Based on the collection of over 83,000 invertebrates, this study found significant declines in the overall abundance of ground-dwelling (14.7%) and aerial (12.9%) arthropods under elevated CO2, with significant decreases in herbivore, omnivore, scavenger and parasitoid functional groups. Even though several groups showed varying declines in abundance, elevated CO2 did not measurably affect community composition.

Our results indicate that atmospheric CO2 predicted within the next 35 years may cause declines in arthropod abundances in Eucalyptus woodland. Declines found in several functional groups suggest that elevated atmospheric CO2 has the potential to affect ecosystem processes, possibly including nutrient cycling by herbivores and omnivores and biocontrol by parasitoids.

82

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

5.2 Introduction With over one million described species, arthropods comprise the majority of terrestrial multicellular life on Earth (Mora et al. 2011; Scheffers et al. 2012) and are the main players in the bulk of terrestrial plant-based food-webs (Price 2002). Aside from their impressive contribution to biodiversity, arthropod communities are important in a functional context, underpinning a variety of ecosystem processes (Wilson 1987). For instance, invertebrates perform substantial roles in nutrient cycling through the consumption and break down of plant material (Hunter 2001).

Arthropod communities are shaped by complex combinations of abiotic and biotic factors, as well as biotic interactions including trophic associations (Polis 1998), which are themselves sensitive to environmental change (Tylianakis et al. 2008). Consequently, perturbations occurring in either the biotic or abiotic environment have the capacity to alter the structure of communities, and the interactions occurring between the species that form them, by virtue of the fact that not all species in a system will respond to change in the same way (Sanders et al. 2003; Raffaelli 2004; Pocock, Evans & Memmott 2012).

While previous community-level studies have shown that the responses of different taxa to environmental change can be highly individualistic and species-specific (e.g. Altermatt 2003; Sanders, Belote & Weltzin 2004), invertebrate taxa sharing the same feeding strategy are likely to be affected by change in similar ways to each other, allowing some generalisations to be made (Altermatt 2003; Hillstrom & Lindroth 2008). For instance, sap-feeding invertebrates may be positively affected by changes in the quality of their food plants under elevated CO2 (Bezemer & Jones, 1998). Conversely, folivores, by virtue of their different feeding habits, tend to have reduced performance under elevated CO2 (Stiling et al. 2003; Stiling & Cornelissen 2007), leading to reductions in folivory (Hamilton et al. 2004; Knepp et al. 2005). Organisms at higher trophic levels, including predators, may be more sensitive to environmental perturbations, perhaps as a result of higher metabolic costs and their dependency on the responses of organisms at lower trophic levels (Voigt et al. 2003; Hance et al. 2007). Moreover, specialist species such as endoparasitoid wasps may be more at risk from changes to the environment than generalists because they are dependent on a smaller group of hosts and therefore might be disproportionately affected if they cannot utilise alternative hosts (Hance et al. 2007; Vanbergen et al. 2010).

83

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

The concentration of carbon dioxide (CO2) in the atmosphere now exceeds the range the Earth has seen in the last 800,000 years (IPCC 2013), and as such is considered an abiotic perturbation with the potential to alter ecological communities. Numerous studies have reported CO2-induced changes in plant biomass and morphology (Pritchard et al. 1999; Stiling & Cornelissen 2007; Zhu et al. 2016), altered botanical composition (Vasseur & Potvin 1998), coupled with reductions in plant quality

(Robinson et al. 2012). Elevated CO2-related changes in plants could therefore have crucial implications for invertebrate herbivores (Hentley et al. 2014) and their arthropod consumers, as well as the ecosystems these communities support (Tylianakis et al. 2008). In spite of their recognised importance, relatively little is known about how invertebrate communities, as a whole, will respond to climatic and atmospheric changes (Jamieson et al. 2012; Facey et al. 2014). In order for us to adequately predict the consequences of climatic and atmospheric change on ecosystems, large scale experiments considering community-level responses to change will be necessary, complementing work in more controlled settings (Stiling et al. 2003; Facey et al. 2014). Free Air CO2 Enrichment

(FACE) experiments have been invaluable for assessing the impacts of elevated CO2 for plants and invertebrates in temperate forest systems (Hamilton et al. 2004; Knepp et al. 2005; Stiling & Cornelissen 2007; Couture & Lindroth 2012; Couture et al. 2015; Facey & Gherlenda 2016).

Thus far, however, forest FACE invertebrate community studies have been limited to experiments on relatively young, managed plantation trees in the northern hemisphere. Gaining an adequate understanding of the responses of the terrestrial biosphere to elevated CO2 will require greater habitat representation among the next generation of FACE experiments (Ainsworth & Long 2005; Facey & Gherlenda 2016; Norby et al., 2016). The present study redresses this gap two ways. Firstly, it is the first field-based experiment investigating arthropod responses to atmospheric change in a southern hemisphere forest system, allowing comparisons and generalisations to be made across studies in other systems. Secondly, this study is the first experiment established in native mature natural woodland. The site consists of Eucalyptus woodland, the second most dominant habitat type in Australia, after grasslands. This habitat is estimated to cover over 890, 000 km2 of the continent (Department of the Environment and Water Resources 2007). Further, the Eucalyptus genus is the most widely planted hardwood globally (Frew et al. 2013), yet information concerning the responses of Eucalyptus

84

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 communities to climatic and atmospheric change is scant owing to a lack of field studies.

The aim of this study was to characterise the arthropod community occurring in this woodland study system and assess the extent to which this community may be affected by rising atmospheric CO2 concentrations. This study used a variety of sampling methods applied from ground-level to the forest canopy, to obtain representative samples of the invertebrate community. Of interest was how different functional groups

(i.e. feeding guilds) responded to elevated CO2 i.e. if specialists like parasitoids were more sensitive than generalist predator groups.

Given the generally negative effects of elevated CO2 on plant quality, it was predicted that i) folivorous herbivores would decline in abundance under elevated CO2, whereas those in different feeding guilds including sap-suckers would be positively affected by alterations in food quality; ii) arthropods at higher trophic levels would show greater declines than groups from lower trophic levels; iii) specialised taxa (e.g. parasitoids) would be more strongly affected by CO2 manipulation than generalists, and; iii) As a result of changes in the abundances of different taxa, invertebrate community composition would be altered under elevated CO2 conditions.

5.3 Materials and methods

5.3.1 Experimental site

The study was carried out at the Eucalyptus Free-Air CO2 Enrichment (‘EucFACE’) site in western Sydney, Australia (33°36′59″S, 150°44′17″E), described in Duursma et al. (2015). In brief, the site consists of ~15 ha within a 167 ha tract of mature, native Cumberland Plain woodland, dominated by E. tereticornis (over 90% coverage). There are six 25 m diameter ring arrays; since September 2012, the CO2 levels have been manipulated in three randomly selected rings (ambient +150ppm, corresponding to the concentration predicted by the middle of this century under the emission scenario A1F1

(IPCC 2007)), with the other three receiving ambient CO2 levels. Diluted CO2 or air (in ambient plots) is released into the vegetation within the ring from valves in the vertical vent pipes around the outside edge of the ring during the day time.

5.3.2 Invertebrate collections Invertebrates were collected using three different methods across seasons to obtain a broad, representative, sample of the arthropod community occupying different niches.

85

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

Pitfall traps were used to sample ground-dwelling arthropods, with suction sampling to capture invertebrates from understorey vegetation and sticky traps to sample aerial (canopy) invertebrates. Suction sampling is a proven quantitative technique for sampling invertebrate populations (Brook et al. 2008). Sticky and pitfall trapping allow for relative comparisons of invertebrate abundance between CO2 treatments, within sampling method (Buntin 1993; Woodcock 2005). Pitfall traps were first used in November 2013; pitfall and suction sampling was then carried out quarterly from January 2014 to January 2015 (six pitfall campaigns, five suction sampling campaigns). Sticky trapping was carried out six times throughout the experiment, on a monthly basis between the end of September and the start of December during 2013 and 2014, when most flying arthropods would be active. In each of the three niche-types, sampling was carried out in fixed locations across all sampling dates. i) Ground-dwelling arthropods (pitfall sampling)

Within each ring, two locations were selected at random on the woodland floor. In each of these, a 500 ml 9 cm diameter transparent plastic pot was buried flush with the soil level. Traps were left dry and open for one week prior to the initial sampling period in November 2013 in order to account for digging-in effects (Woodcock 2005). Thereafter, traps were active for two weeks at the beginning of each of the six sampling periods; for this they were filled to approximately one third full with water (around 5cm deep), with a droplet of scentless detergent to break surface tension. A piece of chicken- wire mesh was pegged over the mouth of the trap to prevent by-catch of non-target mammals and reptiles (Woodcock 2005), whilst only potentially excluding the very largest of beetle species. A transparent plastic roof was suspended above each trap for protection during rain events. A lid was placed over each trap in between sampling events. ii) Understorey arthropods (Suction sampling)

Two 1×1m plots (selected at random) within each ring were used on the woodland floor. A petrol-powered vacuum ‘G-Vac’ device (SH 86C, Stihl AG & Co. KG, Germany, Bell et al. 2000), fitted with an organza bag to capture dislodged debris and invertebrates, was passed over the understorey herbaceous vegetation in a zig-zag pattern for 20 seconds, within the vegetation, during each sampling event. Sampling was carried out when the vegetation was dry to the touch.

86

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 iii) Aerial arthropods (Sticky trapping)

In each of the six rings, 16 yellow card sticky traps (Bugs for Bugs, Mundubbera, Australia) were secured to the central scaffold at four height intervals (2, 5, 10 and 20m) facing each compass direction. This allowed a full range of arthropods occurring at different strata to be sampled. Traps were left in place for one week prior to collection.

5.3.3 Identification and processing Arthropods were counted and identified under a dissecting microscope (SZ51, Olympus, Japan) to at least Order level (except for three groups taken to Subclass only – Acari, Collembola and Chilognatha), and in some cases, Family level, to more reliably determine functional guild in as many cases as possible (Hamilton et al. 2012; for a full list of identified groups and guild assignments, see Table S5.1). Psyllidae (Hemiptera) were excluded from the study as they are the focus of a concurrent study occurring at the site (Gherlenda et al. 2016a). Better estimations of the energy flow occurring through different trophic levels within communities can be achieved through the assessment of biomass (Saint-Germain et al. 2007). Thus, after abundances were taken, pitfall and suction samples were dried at 60°C to constant weight before weighing using a microbalance with 1μg accuracy (model XP6, Mettler-Toledo GmbH, Germany).

5.3.4 Statistical analyses All statistics were performed in R, version 3.2.0 (R Core Team 2015). To avoid pseudoreplication, the subplots in each ring were pooled for all analyses, giving one sample per ring, per time point (n = 6, 36 plot-time samples in total for pitfall and sticky traps, 30 for suction). Separate analyses were carried out on data from each of the three sampling methods to enable assessment of the effect of elevated CO2 on the arthropod communities in the different niches.

Abundance analyses

Total arthropod abundance, and the abundance of individual taxa and functional groups, was analysed firstly using generalized liner mixed models (GLMM) with Poisson error distributions using glmer. Models contained CO2 treatment as a fixed effect and date sampled as a random factor. Model fit was verified by inspection of residual plots and overdispersion parameters from the overdisp_fun function (specified at http://glmm.wikidot.com/faq). In the majority of cases, data were overdispersed and so models were refitted using the negative binomial extension of GLMM, glmer.nb, in

87

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

lme4 (Bates et al. 2014). The significance of CO2 treatment as a predictor was assessed using likelihood ratio tests between the full model and a reduced model without the fixed effect of CO2 treatment (Faraway 2006).

Orders which were poorly represented – found in fewer than ten percent of samples or had fewer than 50 individuals – were removed from the individual Order analyses. Sanguivores were also not analysed due to small sample size. In one case (aerial Thysanoptera), a negative binomial model did not adequately fit the data and so an observation-level random effect was included in the model to account for overdispersion (Harrison 2014).

Biomass analyses

Similar to the abundance analyses, arthropod biomass data (in terms of total sample biomass, not individual biomass) were analysed for separate functional guilds and Orders, with the same poorly-represented groups removed. Total arthropod biomass across all groups was also analysed for each sampling method. Data were modelled using linear mixed models (LMM) with the lmer function, with CO2 treatment as a fixed effect and date sampled as a random factor. In all cases, biomass was rank transformed prior to analysis in order to meet assumptions of homoscedasticity of residuals. For groups with tied ranks (where zeros were present in the variable), the analysis was iterated 1000 times on retransformed data with randomly broken ties to attain stable mean P and χ2 values.

Community composition

To assess the effects of elevated CO2 on overall community composition, permutational multivariate analysis of variance (PERMANOVA) was used, coupled with non-metric multidimensional scaling (NMDS) to visualise the data (Hillstrom et al. 2014), with the package vegan (adonis and metaMDS functions, Oksanen et al. 2015). For community- level analyses, poorly-represented taxa were included. PERMANOVA was carried out on the three niche-types separately, with the fixed effect of CO2 treatment, on both functional guild and Order-level abundance data. Analyses were carried out on Bray- Curtis dissimilarity matrices, permuted 999 times. The number of dimensions, k, used in each NMDS analysis was determined by visual inspection of stress plots and stress values. Stress values were <0.2 across multiple runs for all analyses.

88

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

Due to the low replication inherent in FACE designs, a critical P value of 0.1 was set to avoid type II errors, as recommended by Lindroth & Raffa (2016) and consistent with previous studies of this type (Sanders et al. 2004; Villalpando et al. 2009; Hamilton et al. 2012).

5.4 Results A total of 83,528 arthropods from 19 different taxa (16 Orders and three Subclasses) were collected and identified during the experiment (14,459 ground-dwelling, 19,153 understorey and 49,916 aerial arthropods; Table S5.1). Total arthropod abundance was lower in elevated CO2 in all three of the sampled niches; this effect was significant for ground-dwelling and aerial invertebrates (Table 5.1), which decreased by 14.7% and 12.9% respectively (ground-dwelling total individuals ± SD: ambient 7,803 ± 280.17, elevated 6,656 ± 280.11; understorey: ambient 11,362 ± 792.56, elevated 7,791 ± 437.55; aerial: ambient 26,672 ± 384.29, elevated 23,244 ± 403.32, Table 5.1). Across all groups, total arthropod biomass did not significantly differ between CO2 treatments (p > 0.1, Table 5.1).

Ground-dwelling arthropods

The abundance of ground-dwelling chewing herbivores was significantly reduced under elevated CO2 conditions (Fig. 5.1b, Table 5.2), though their biomass remained unchanged (p > 0.1). Detritivores and omnivores showed a decrease in biomass under elevated CO2, but did not show measurable declines in abundance (Fig. 5.1a, b, Table 5.2). The abundances of ground-dwelling Hymenoptera, Isopoda and Orthoptera were significantly reduced under elevated CO2 (Fig. 5.1c and d, Table 5.1), with latter two groups also showing decreases in biomass (Fig. 5.1c, d, Table 5.1). Acarina showed an increase in biomass (Fig. 5.1c, Table 5.1), with no evidence of change in abundance (p > 0.1).

Understorey arthropods

Declines were also seen in the abundance of certain groups in the understorey, though different groups were affected. Omnivores showed a significant decrease in mean abundance and this was coupled with a marked decline in population biomass (Fig. 5.2, Table 5.2). None of the other feeding guilds showed a significant response to elevated

CO2 in this niche (p > 0.1).

89

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

Table 5.1: Results from likelihood ratio tests performed on GLMMs (abundance) and LMMs (rank- transformed biomass) with and without the fixed effect of CO2 treatment, with the abundance or biomass of each of the groups collected over the course of the experiment as the dependent variable. † denotes that for this taxa, strong variation in the data made analysis unreliable. Those groups with sample sizes too small for analysis in all three niches are not shown. Significant P values are highlighted in bold (ɑ = 0.1).

Abundance Biomass Group Niche Type 2 2 χ 1 P χ 1 P Overall Ground-dwelling 3.442 0.064 0.190 0.66 Understorey 2.412 0.12 0.087 0.77 Aerial 3.878 0.049 - -

Coleoptera Ground-dwelling 1.122 0.29 0.219 0.64 Understorey 3.129 0.077 2.540 0.11 Aerial 0.493 0.48 - - Diptera Ground-dwelling 0.443 0.51 0.623 0.43 Understorey 2.002 0.16 2.005 0.16 Aerial 0.194 0.66 - - Araneae Ground-dwelling 1.546 0.21 0.000 1.00 Understorey 0.023 0.88 0.338 0.56 Aerial 0.658 0.42 - - 2.783 0.09 Acarina Ground-dwelling 0.085 0.77 5 Understorey 1.854 0.17 2.624 0.11 Aerial 16.486 <0.001 - - Hemiptera Ground-dwelling 0.047 0.83 0.046 0.83 Understorey 1.188 0.28 0.163 0.69 Aerial 1.989 0.16 - - Hymenoptera Ground-dwelling 4.646 0.031 0.785 0.38 Understorey 2.580 0.11 0.941 0.33 Aerial 3.441 0.064 - - Thysanoptera Understorey 0.367 0.55 0.013 0.91 Aerial 2.418 0.12 - - 5.356 0.02 Orthoptera Ground-dwelling 11.347 <0.001 1 14.469 <0.0 Isopoda Ground-dwelling 9.010 0.0027 01 Understorey † † † † Blattodea Understorey 0.205 0.65 0.141 0.71 Collembola Ground-dwelling 0.824 0.36 0.081 0.78 Understorey 0.148 0.70 0.014 0.91 Aerial 2.708 0.10 - - Lepidoptera Aerial 1.190 0.28 - - Psocoptera Aerial 26.389 <0.001 - - Neuroptera Aerial 5.982 0.014 - -

90

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

Table 5.2: Results from likelihood ratio tests performed on GLMMs (abundance) and LMMs (rank- transformed biomass) with and without the fixed effect of CO2 treatment, with the abundance of the arthropods in each of the recognised guilds as the dependent variable. Significant P values are highlighted in bold.

Abundance Biomass Group Niche type 2 2 χ 1 P χ 1 P Scavengers Ground-dwelling 2.661 0.10 0.003 0.96 Understorey 2.619 0.11 1.709 0.19 Aerial 9.961 0.0016 - - Detritivores Ground-dwelling 1.712 0.19 3.379 0.066 Understorey 0.116 0.73 1.005 0.32 Aerial <0.001 0.98 - - Omnivores Ground-dwelling 1.643 0.11 3.481 0.062 Understorey 3.448 0.063 8.471 0.0036 Aerial 1.303 0.25 - - Chewing herbivores Ground-dwelling 2.845 0.092 1.419 0.24 Understorey 0.091 0.76 0.323 0.57 Aerial 0.252 0.62 - - Sucking herbivores Ground-dwelling 0.751 0.39 0.800 0.38 Understorey 1.095 0.30 0.078 0.78 Aerial 1.989 0.16 - - Predators Ground-dwelling 0.597 0.44 0.069 0.79 Understorey 0.890 0.35 0.149 0.70 Aerial 0.535 0.47 - -- Parasitoids Ground-dwelling 0.305 0.58 0.296 0.59 Understorey 0.026 0.87 0.012 0.91 Aerial 3.422 0.064 - -

91

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

At Order-level, Coleoptera were significantly decreased in abundance under elevated

CO2, but their biomass was not significantly different from ambient CO2 (Fig. 5.2d, Table 5.1). While the data for understorey Isopoda could not be accurately modelled, this group appeared to show a trend towards lower abundance under elevated CO2, as found for the same group in ground-dwelling samples (Fig. 5.2d vs. Fig. 5.1c).

Aerial arthropods

Elevated CO2 generally resulted in decreased abundances of aerial arthropods. At feeding guild level, both scavengers and parasitoids experienced a significant decline in abundance (Fig. 5.3a, b, Table 5.2). At Order-level, significant decreases were seen for four of the recorded taxa; Hymenoptera, Neuroptera, Acari and Collembola (Fig. 5.3c, d, Table 5.1). However, in contrast to these declines, aerial Psocoptera showed a significant increase in abundance under elevated CO2 (Fig. 5.3d, Table 5.1).

Community composition

While elevated CO2 resulted in significant changes in the abundances of several different feeding guilds and Orders (summarised in Fig. 5.4), this did not significantly affect the community composition occurring in any of the three niche types, either in terms of functional guild or Order composition (Table S5.2, Fig. S5.1).

92

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

Fig 5.1 (overleaf): Mean abundance of different functional guilds (a, b) and taxonomic groups (c, d) of ground-dwelling arthropods, split by CO2 treatment (across all dates). Ambient samples are shown with white bars; those from elevated conditions are in grey. To the right of each bar total mean biomass ± SE is shown for the corresponding group. Significant differences (from GLMMs (abundance) and LMMs (biomass), Table 1 and Table 2) are denoted by asterisks (* P < 0.1, **P < 0.05, *** P < 0.01). Error bars show ± SE of the mean.

Fig 5.2 (overleaf): Mean abundance of different functional guilds (a, b) and taxonomic groups (c, d) of understorey arthropods, split by CO2 treatment (across all dates). Ambient samples are shown with white bars; those from elevated conditions are in grey. To the right of each bar total mean biomass ± SE is shown for the corresponding group. Significant differences (from GLMMs (abundance) and LMMs (biomass), Table 1 and Table 2) are denoted by asterisks (* P < 0.1, **P < 0.05, *** P < 0.01). Error bars show ± SE of the mean.

Fig 5.3 (overleaf): Mean abundance of different, the functional guilds (a, b) and taxonomic groups (c, d) of aerial arthropods, split by CO2 treatment (across all dates, with no biomass data due to the sampling method). Ambient samples are shown with white bars; those from elevated conditions are in grey. Significant differences (from GLMMs, Table 1 and Table 2) are denoted by asterisks (* P < 0.1, **P < 0.05, *** P < 0.01). Error bars show ± SE of the mean.

Fig 5.4 (overleaf): a schematic diagram summarising the main findings in this study, and showing a scaled drawing of one of the EucFACE arrays. CO2 (or air in the case of ambient rings) is pumped in to the ring from the vertical vent pipes surrounding each array. The crane is used to access the tree canopy. Arrows show the direction of significant changes in the abundances of the taxa shown, in response to elevated CO2. The widths of the arrows indicate their level of significance, with wider arrows representing smaller P values, and thus greater evidence against the null hypothesis.

93

(a) Ground-dwelling (>1000 individuals in total) (c) Ground-dwelling (>500 individuals in total)

4.54mg ± 2.31 Diptera 10.41mg ± 7.40 Predators 209.34mg ± 40.57 203.76mg ± 46.90 47.91mg ± 9.70 Araneae 51.65mg ± 11.81

Detritivores 52.24mg ± 22.18 75.01mg ± 22.05 Isopoda 14.37mg ± 4.75 * *** ***52.22mg ± 12.60

2.56mg ± 0.46 102.22mg ± 22.85 Acarina Scavengers 133.17mg ± 36.27 * 1.81mg ± 0.40 3.01mg ± 0.85 Collembola 2.05mg ± 0.43

0 100 200 300 Hymenoptera 89.09mg ± 20.34 (b) Ground-dwelling (<1000 individuals in total) ** 122.65mg ± 33.19

6.20mg ± 2.15 28.71mg ± 11.74 Chewers * 0 50 100 150 200 250 300

2.99mg ± 1.29 (d) Ground-dwelling (<500 individuals in total) Suckers 8.14mg ± 6.41 Orthoptera 3.51mg ± 2.04 *** ** 16.16mg ± 7.36 1.75mg ± 0.59 Parasitoids 7.51mg ± 5.55 Hemiptera 7.54mg ± 3.10 10.31mg ± 6.42

6.62mg ± 2.67 Omnivores Coleoptera 180.98mg ± 49.08 *16.06mg ± 10.18 177.44mg ± 52.99

0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 Abundance Figure 5.1

(a) Understorey (>1000 individuals in total) (c) Understorey (>500 individuals in total) Araneae 7.31mg ± 2.55 3.38mg ± 0.86 8.83mg ± 1.98 Omnivores 6.65mg ± 1.52 * *** Hemiptera 20.66mg ± 9.37 8.62mg ± 2.04

0.87mg ± 0.44 2.91mg ± 0.76 Thysanoptera Detritivores 4.47mg ± 1.52 0.64mg ± 0.20

2.00mg ± 0.70 Diptera 3.57mg ± 1.08 9.74mg ± 1.47 Scavengers 13.09mg ± 1.93 5.90mg ± 1.19 Hymenoptera 6.25mg ± 0.61

2.24mg ± 0.63 Collembola 2.84mg ± 1.07 0 200 400 600 800 Acarina 2.88mg ± 0.85 5.29mg ± 1.71 (b) Understorey (<1000 individuals in total)

3.79mg ± 0.99 0 200 400 600 800 Chewers 5.79mg ± 3.03 (d) Understorey (<500 individuals in total)

1.09mg ± 0.42 Parasitoids 2.05mg ± 0.72 0.69mg ± 0.10 Blattodea 2.23mg ± 0.94

20.62mg ± 9.37 Suckers 0.09mg ± 0.07 7.86mg ± 1.91 Isopoda 0.87mg ± 0.57

7.51mg ± 2.57 Predators Coleoptera 2.85mg ± 0.72 10.10mg ± 2.19 * 4.68mg ± 1.11

0 10 20 30 0 2 4 6 8 10 12 14 16 Abundance Figure 5.2

(a) Aerial (>5000 individuals in total) (c) Aerial (>1000 individuals in total) Araneae Detritivores Coleoptera

Omnivores Hemiptera

Thysanoptera Parasitoids * Diptera

Hymenoptera 0 200 400 600 800 * (b) Aerial (<5000 individuals in total) 0 200 400 600 800 Chewers (d) Aerial (<1000 individuals in total) Psocoptera *** Scavengers *** Lepidoptera

Neuroptera Predators ** Acari *** Suckers Collembola *

0 20 40 60 80 100 0 5 10 15 20 25 Abundance Figure 5.3

Figure 5.4

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2

5.5 Discussion

Elevated CO2 caused widespread changes in arthropod abundance and biomass

To our knowledge, this is the first study of its kind to find significant declines in the abundance of a wide range of woodland arthropods under elevated CO2; out of the 21 taxonomic and functional groups which satisfied the analysis criteria, over half (11 groups) experienced significant declines in abundance. Previous work on soil micro- arthropod communities has found similar decreases (Hansen et al. 2001; Loranger et al. 2004), yet most previous studies looking at aboveground invertebrate communities have revealed no significant changes in abundance as a result of elevated CO2 (Sanders et al. 2004; Hillstrom & Lindroth 2008; Hillstrom et al. 2014).

The declines in total arthropod abundance did not translate into overall declines in total biomass, though this is not unexpected as the two metrics are known to not necessarily correlate well (Saint-Germain et al. 2007). However, significant changes in arthropod biomass at the individual functional group/Order level were found, five out of six of which were negative. This reinforces the findings from the abundance analyses and indicates the potential for changes in ecosystem functioning. Reductions in biomass point to the loss of larger bodied organisms, especially in groups which did not see a corresponding reduction in abundance, such as ground-dwelling detritivores and omnivores. Larger organisms are likely to be of greater importance for trophic interactions occurring within the ecosystem (Saint-Germain et al. 2007), as energy flow through trophic levels is tied to body mass (Brown et al. 2004). Conversely, in cases where declines in abundance were not reflected by biomass data (e.g. chewing herbivores and Coleoptera), there may be a greater proportion of larger-bodied individuals occurring under elevated CO2 compared with ambient conditions, suggesting that ecological functionality may be more likely to be maintained for these groups, despite population declines. At the same time, larger-bodied organisms tend to have longer life spans (Speakman 2005) and so the increased representation of this group in elevated rings may be explained by slower responses to the elevation in CO2.

Elevated CO2 had variable effects on feeding guilds

It was predicted that chewing herbivores would suffer a decrease in abundance under elevated CO2 compared with other feeding guilds with different feeding methods, such as sap-suckers. Sap feeders may even stand to benefit from elevated CO2 via changes in

98

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 phloem chemistry, such as increases in carbohydrate concentrations (Bezemer & Jones 1998). Significant reductions in the abundances of ground-dwelling chewing herbivores were found, though this effect was not seen in either the understorey or aerial niches. Declines in the abundance and biomass of omnivorous taxa in the understorey and at ground level were found; these animals will also have partially plant-based diets. These findings are consistent with those reported in other studies of this type (Stiling et al. 2002, 2003; Hamilton et al. 2012), and could indicate a reduction in herbivore-pressure and herbivore-mediated nutrient cycling in the system.

Given the decline in parasitoids and stable levels of other predatory taxa, the reduction in the abundance of herbivorous taxa at ground and understorey level is unlikely to be explained by changes in top-down regulation. This study based the prediction that herbivore abundance would be reduced under elevated CO2 conditions on the widely reported decrease in plant resource quality observed elsewhere under the same conditions (Robinson et al. 2012). However, work carried out at the EucFACE site concurrently with this study has revealed no change in various plant quality metrics, including canopy C:N ratios (Gherlenda et al. 2015) and leaf area index (Duursma et al. 2015). This is not entirely unexpected; Hamilton et al. (2012) also observed changes in arthropod populations under elevated conditions with no accompanying alteration in C:N ratios of plant tissues. One as yet undetermined plant-mediated mechanism for these declines could be alterations in plant secondary chemistry occurring under elevated CO2, as found in other studies and known to affect invertebrate herbivores (Robinson et al. 2012). Further work, following a greater fumigation period, is needed to link the observed changes in invertebrate abundance with plant quality changes occurring in the woodland at EucFACE.

In contrast to the declines in chewing herbivores and omnivores, sap-sucking herbivores (Order Hemiptera) did not decline in any of the three niche types. However, work by other researchers at EucFACE has shown decreased abundance in the abundance of three species of psyllids under elevated CO2 (Gherlenda et al. 2016a). While controlled environment studies tend to report enhanced abundance and performance of sap- feeders, linked with CO2-induced changes in phloem and sap chemistry (Bezemer & Jones 1998), these often do not consider natural enemies. Hentley et al. (2014) showed that aphid populations under elevated CO2 were supressed to population levels at ambient CO2 when a predatory ladybird was also included in the experiment. On the

99

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 other hand, Percy et al. (2002) found that the severity of aphid infestations on aspen was increased under long-term CO2 exposure, as a result of asynchrony between aphid and natural enemy populations. In this study, given the significant reduction found in parasitoid abundance in the canopy, there is the potential for reduced top-down regulation of sap-feeding insects in the future, and thus population growth. Such growth could increase herbivory levels under elevated CO2, as found by Couture et al. (2015). Long term monitoring would be needed at the EucFACE experimental site to substantiate this. Presently, however, the decline in chewing herbivores and omnivores and comparable levels of (non-psyllid) Hemiptera suggests that herbivory will decline in

Eucalyptus woodland as emissions of CO2 rise, as found in other northern hemisphere systems (Hamilton et al. 2004; Knepp et al. 2005).

A significant decline in the abundance of scavengers in the canopy was also found, likely driven by the significant decrease in the abundance of mites (Acari) under elevated CO2, and perhaps also linked to undetected alterations in plant quality. The significant reduction in mites is consistent with findings from other studies (Hansen et al. 2001; Loranger et al. 2004). However, the same decline was not seen in the ground-dwelling and understorey samples which contained far greater abundances of this group; indeed, the total biomass of ground dwelling mites actually increased under elevated CO2, potentially as a result of larger individuals of greater body size.

Specialist vs. generalist natural enemies

It was predicted that arthropods at higher trophic levels would show greater declines in abundance than groups from lower trophic levels, given the generally greater sensitivity of higher trophic levels to environmental change (Voigt et al. 2003). It was expected that this would be particularly true for more specialised feeding groups such as parasitoids, because they are more restricted by tightly-coupled relationships with a limited number of host species compared with generalist predators which can exploit a greater range of prey species. Significant reductions in the abundance of aerial parasitoid wasps were found, as expected, and as such this study adds to the body of evidence that specialised taxa may be more susceptible to environmental change (e.g. Hance et al. 2007). Conversely, previous studies from similar sites have found increases in the numbers of parasitoids or parasitism rates under elevated CO2 (Percy et al. 2002; Stiling et al. 2002, 2003; Hillstrom & Lindroth 2008). Stiling et al. (2003) attributed their findings to the host plant quality-mediated increases development time of host species, leaving them

100

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 vulnerable to parasitoid attack for longer periods. In the present study, host species may well be experiencing reduced development rates – this would require further work to determine – but the reductions seen in absolute host abundance may be more important for parasitoids. The declines found across a range of groups from lower trophic levels, both in terms of abundance and biomass, could be responsible for the declines seen in parasitoid abundance, as their larval food sources become limited.

Contrary to expectations, the abundance and biomass of generalist predators such as spiders did not decline in any of the niche types, despite declines in the number of many of the groups likely to constitute their prey. This study found a significant decline in the abundance of aerial Neuroptera, though this was the only predatory Order to show a response. Previous findings concerning the responses of predatory taxa to elevated CO2 are mixed, with some studies finding increases in the abundance of carnivorous groups (Sanders et al. 2004; Hamilton et al. 2012) and one reporting no change (Hillstrom & Lindroth 2008). In this study, it is possible that highly mobile predators, such as ground- walking spiders and carabid beetles could access prey external to the rings from which they were caught, enabling the maintenance of ambient population levels within elevated rings; however this is also true of winged parasitoids for which this study still detected an effect. Alternatively, the effects of elevated CO2 at the plot level may have deterred certain insects from entering the rings, resulting in the population declines seen for many of the groups studied; this is an inherent issue in plot-level experiments of this type (Moise & Henry 2010) that needs consideration when interpreting the results presented here. Either way, reduced densities of these invertebrate groups in elevated

CO2 suggest that conditions were less favourable for them than those under ambient

CO2 levels. In addition, it could be possible that the predator population is yet to respond to declines in prey availability under elevated CO2, given the relatively short fumigation time (since late 2012).

The level of taxonomic identification (Order/Family) using in this study did not allow for estimations of the abundance of arthropods from the fourth trophic level (intra- guild predators). The inclusion of intra-guild predators within the predator group could potentially mask any CO2 effects on the abundance of third-level predatory taxa, though we might expect that fourth-level predatory taxa would be negatively impacted by elevated CO2 over the long term. The design of this study also did not allow for the level of host specificity of herbivorous arthropods to be determined – this could be an

101

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 interesting consideration for further study, particular in the plant species-rich understorey, as monophagous specialist herbivores have been shown to be more strongly negatively affected by increases in CO2 than polyphagous species (Stiling & Cornelissen 2007).

Community composition did not change under elevated CO2

Despite widespread overall declines within individual trophic groups and Orders, there was no evidence for an effect of elevated CO2 on community composition, contrary to the predictions. Similarly, other studies of this type have shown weak to non-existent effects of elevated CO2 on community composition (Sanders et al. 2004; Hillstrom et al.

2014). Given that the majority of the responses of the different groups to elevated CO2 in the present study, both in terms of abundance and biomass, were negative in nature, this could have resulted in a compositionally-similar communities comprised of fewer total individuals compared with those under ambient conditions.

The range of sampling methods used here mean that this study gained a broad, representative sample of the community occurring in Eucalyptus woodland. Many differences in the responses of the individual trophic and taxonomic groups to elevated

CO2 were found between sampling methods, indicating the potential for studies using only one sampling technique to overlook effects of elevated CO2. This study therefore demonstrates the importance of using multiple sampling methods in future work in other such studies, to ensure that the results more accurately reflect the responses occurring in the system.

Conclusions

There is a growing body of evidence from community-level studies that the responses of invertebrates to climatic and atmospheric change will likely be taxon-specific and idiosyncratic (Sanders et al. 2004; Hamilton et al. 2012; Hillstrom et al. 2014). In support of this, this study found differences in the directions and/or strength of change for certain groups between niche types, as well as differences in the responsiveness of the taxa comprising the individual feeding guilds, highlighting the importance of studies across multiple trophic levels (Pocock et al. 2012). However, overall the present study found evidence for a consistent decline across a broad range of groups under elevated

CO2. Particularly for those groups showing corresponding declines in biomass such as detritivorous Isopoda and omnivores, these declines could indicate reductions in the

102

CHAPTER FIVE INVERTEBRATE COMMUNITIES UNDER ELEVATED CO2 energy flow attributed to these organisms in the system. Significant reductions in the abundance and biomass of several groups with roles in nutrient cycling and biocontrol suggest that woodland ecosystem processes could potentially be affected as global concentrations of atmospheric CO2 continue to rise.

103

Chapter Six: Atmospheric change causes declines and compositional changes in populations of ants in Eucalypt woodland

6.1 Summary Predicting the impacts of global changes, such as elevated concentrations of atmospheric carbon dioxide (CO2), on biodiversity and ecosystems is a priority in contemporary ecology. The use of indicator taxa is an increasingly popular approach for assessing the likely impacts of global change. Ants fill numerous ecological roles, and are highly sensitive to stress and disturbance, making them good indicators of ecosystem- level invertebrate responses to global change.

This study used Free Air CO2 Enrichment to look at the effect of elevated CO2 on the ant assemblage of an Australian native woodland. Pitfall traps were deployed to sample ground-foraging ants under current (400ppm) and future CO2 levels (2050, 550ppm). To our knowledge, this is the first study to examine assemblage-level responses of ants to atmospheric change.

Total ground-foraging ant abundance declined under elevated CO2, driven primarily by reductions in the abundance of Ochetellus and Rhytidoponera ants. Genus-level ant assemblage composition was significantly altered under elevated CO2, and diversity was increased. This increase in diversity came as a result of reduced numerical dominance and increased evenness under elevated CO2; genus richness remained comparable between the two treatments.

This study found evidence supporting the use of ants as indicators of broader-scale invertebrate community responses to atmospheric change. The significant declines in the abundance of key ant genera – and the abundance of ants across the community – suggest that elevated CO2 could cause alterations in the important ecosystem functions that these insects provide, such as soil formation and nutrient cycling.

104

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

6.2 Introduction Anthropogenic climatic and atmospheric changes are causing measurable perturbations in the Earth’s ecosystems (Walther 2010). Increasing concentrations of carbon dioxide gas (CO2), have been shown to alter plant growth, morphology and chemical content (Pritchard et al. 1999; Stiling & Cornelissen 2007; Robinson et al. 2012), as well as plant community composition (Owensby et al. 1999; Morgan et al. 2011). Invertebrates are generally not directly affected by increased concentrations of CO2 (but see Stange,

1997). However, CO2-induced changes via indirect mechanisms, particularly in plant communities, could have significant impacts on insects, including herbivores, pollinators, natural enemies and detritivores. Herbivore performance, for example, can be affected by changes in plant quantity and quality (reviewed in Facey et al., 2014). In addition, elevated concentrations of CO2 may impact pollinators by altering nectar chemistry (Hoover et al. 2012), as well as soil micro arthropods which underpin soil formation and decomposition (Hansen et al. 2001; Loranger et al. 2004). Such changes in primary consumers can then affect higher trophic levels, such as predators and parasitoids, which may themselves be more sensitive to environmental perturbations (Voigt et al. 2003; Hance et al. 2007). As the effects of atmospheric change can differ across trophic levels, elevated concentrations of CO2 may cause shifts in the structure and composition of communities and ecosystems (van der Putten et al. 2004; Hillstrom et al. 2014), with associated implications for ecosystem functioning (van der Putten et al. 2004; Del Toro, Ribbons & Pelini 2012).

Predicting the impacts of global changes on biodiversity and ecosystems is a key goal for ecology, but one that is particularly challenging (van der Putten et al. 2004). One approach increasingly used by ecologists is to utilise carefully selected indicator taxa or ‘bioindicators’ that function as representatives of the responses of other organisms in the broader ecosystem in which they are found (Gardner et al. 2008). Ants (Hymenoptera: Formicidae) have been widely used as an indicator taxon across geographical locations and habitat types (Perfecto et al. 1997; Read & Andersen 2000; Bestelmeyer & Wiens 2001; Andersen & Majer 2004; Andersen et al. 2004), with previous work by Majer (1983) showing that the responses of ant communities are generally representative of those occurring in the broader invertebrate community. Ants are ubiquitous, abundant keystone organisms within ecosystems and form the main component of animal biomass in many habitats (Hölldobler & Wilson 1990; Folgarait 1998). Ants occupy a wide range of ecological niches, functioning as herbivores, 105

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

scavengers, pollinators, predators and ecosystem engineers, with many of their roles involving interactions with other species (Del Toro et al. 2012).

The distributions of ant species are determined by a range of biotic factors, such as vegetation and soils (Andersen 1995a; Vasconcelos et al. 2008), inter- and intraspecific ant competition (Savolainen & Vepsäläinen 1988) and the presence of mutualistic animals like Hemipterans (Helms & Vinson 2008). In addition, abiotic factors like temperature stress and micro-climatic conditions play a role in structuring ant communities (Nakamura et al. 2015), as well as human-induced disturbance and land-use change (Majer 1983; Gómez et al. 2003; Hoffmann & Andersen 2003). Previous work has looked at the effects of habitat disturbances like grazing, fire and mining on ant communities (Andersen et al. 2014; Barton et al. 2016), with several recent studies looking at the effects of climatic change on ants using temperature gradients and modelling approaches (Diamond et al. 2012; Andersen, Del Toro & Parr 2015; Del Toro, Silva & Ellison 2015; Nakamura et al. 2015; Gibb et al. 2015). To our knowledge, no previous work has addressed the impact of atmospheric change on ant communities.

Research from Free Air CO2 Enrichment (FACE) experiments has looked at invertebrate community responses to elevated CO2, including ants as part of the broader assemblage, but we are unaware of any published field study that addresses the effect of elevated concentrations of CO2 specifically on ant communities at the genus level. Further, these previous FACE studies have been based in managed plantations, with no work yet looking at the responses of ant communities within native natural systems.

Here, ants were used as a focal group to assess the effects of future levels of CO2 on the invertebrate assemblage of an Australian native Eucalyptus woodland. Forests are known to be sensitive to climatic and atmospheric change (Fowler et al. 1999; Hillstrom & Lindroth 2008) making the study of their community responses pressing. Further, trees from the Eucalyptus genus are the most widely planted hardwoods globally (Paine, Steinbauer & Lawson 2011), making studies of the responses of Eucalypt communities of broad interest. Given the generally negative effects of elevated CO2 on plant quality and invertebrate performance reported in the literature (Stiling & Cornelissen 2007;

Robinson et al. 2012; Facey & Gherlenda 2016), it was predicted that elevated CO2 would cause a reduction in woodland ant abundance, based on the distribution of ground-foraging ants. It was also predicted that such reductions in abundance would lead to alterations in generic richness, diversity, composition and functional group-based structure. 106

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

6.3 Materials and Methods

6.3.1 Experimental site and ant collections

The study was carried out at the Eucalyptus Free Air CO2 Enrichment (‘EucFACE’) site in western Sydney, Australia (33°36′59″S, 150°44′17″E), which consists of six 25 m diameter ring arrays (three × ambient CO2, three × ambient +150ppm) constructed in a mature, Eucalyptus-dominated woodland. The EucFACE site has been more thoroughly described in Duursma et al. (2015). Ants were sampled using pitfall traps in November 2013, and then on a quarterly basis from January 2014 to January 2015 (i.e. six sampling events). Within each of the six rings, two 500 ml (9 cm diameter, 9 cm depth) transparent plastic pitfall traps were placed randomly into the woodland floor, flush with the soil level. The absolute abundance of foraging ants caught during pitfall trapping campaigns was used to describe assemblage structure, which has the potential to be biased by high abundances of ants when placed near nest sites (Andersen 1991a). Such effects were minimised by leaving traps dry and open for one week prior to the first sampling period in November 2013 in order to account for digging-in effects (Woodcock 2005). Thereafter, traps were active for two weeks at the beginning of each sampling period during which time they were filled with around 5cm depth of water with added scentless detergent. A piece of chicken-wire mesh was pegged over the mouth of the trap to prevent bycatch of non-target mammals and reptiles (Woodcock 2005); this did not disrupt the capture of even the largest ants (Myrmecia spp., >3 cm). A clear plastic roof above each trap prevented the trap from overflowing during rain events. Lids covered the traps between sampling periods.

6.3.2 Sample processing and classification Ants were identified under a dissection microscope (SZ51, Olympus, Japan) to genus level using the dichotomous keys from Andersen (1991b) and Shattuck (1999), the nomenclature of which this study follows. The ant genera of Australia are well-described and relatively easily discriminated, unlike species which require highly specialised knowledge to identify, with some species still undescribed (Andersen 1995b; Schnell, Pik & Dangerfield 2003; Piper et al. 2009). While the current study has lower taxonomic resolution than those identifying samples to species/morphospecies level, a strong relationship between genus and species diversity has been found within well-forested sites in Australia, and within biogeographical regions (Andersen 1995b; Pik et al. 2002; Schnell et al. 2003). Biomass measurements can allow for better estimations of energy

107

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

flow occurring within communities (Hawes et al. 2003; Saint-Germain et al. 2007). Therefore, all samples were dried at 60°C for 48hr before weighing on a microbalance with 1μg accuracy (model XP6, Mettler-Toledo GmbH, Germany).

Functional groupings, devised by Andersen (1995a) for Australian ants, categorize ant genera based on a broad range of characteristics including life-history traits, morphology and behaviour (Andersen 1997). Such groupings give an indication of the structure of the ant assemblage and have been shown to vary predictably with climate, soil type and vegetation cover (Andersen 1997). While such groupings may be best suited to comparisons between biogeographical regions (Andersen & Majer 2004), in the present studied data are analysed using both taxonomic and functional guild classifications, since they have proved useful in terms of detecting changes in ecosystem functioning and ant competition.

6.3.3 Statistical analyses All analyses were performed in R, version 3.2.3 (R Core Team 2015). The two traps in each ring were aggregated in order to avoid pseudoreplication, giving one sample per ring, per time point (n= 6 × 6 = 36 samples in total). Abundance data were normalised using a square root transformed prior to analysis. Total ant abundance was analysed using a linear mixed effect model (LME, package lme4 (Bates et al. 2014)), with CO2 treatment as the fixed effect and date sampled as the random effect. The full model was compared with a reduced model containing only date sampled in order to determine the significance of the CO2 treatment, using a likelihood ratio test (LRT, Faraway, 2006).

The responses of individual ant genus abundances to CO2 were also investigated. To ensure adequate sample sizes, only the most abundant genera were analysed – those which occurred in more than 10% of all samples or with an abundance of 50 or more individuals (sensu Hillstrom & Lindroth, 2008; Facey et al., 2016). The rank-transformed biomass of the samples, both in total and split by genus, was analysed using the same process and model structure. Genera with zeros present in the data were analysed 1000

2 times with randomly broken ties to attain stable mean χ and p values (Facey et al., 2016).

The Shannon-Weiner (SW) genus diversity of the samples was analysed using an LME with the same model structure as above. We chose to use the SW index as it is a compound measure of diversity based on the richness and abundance of elements in a community (Morris et al., 2014), and therefore considers community evenness as well as absolute diversity. The SW index is commonly used in community ecology studies (e.g. 108

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

Zhu et al., 2014). For generic richness, sample-based rarefaction curves were created in R using the package vegan (Oksanen et al. 2015). To determine the difference in genera richness between the two CO2 treatments, a t-test was used to analyse the rarefied genera richness based on the maximal abundance shared by all six rings, at which rarefied richness was asymptotic (Hillstrom et al. 2014). Permutational multivariate analysis of variance (PERMANOVA, package vegan) was used to assess the effect of elevated CO2 concentrations on ant assemblages in two ways. Firstly, assemblage composition in terms of genus identity was analysed, and second, assemblage structure based on functional guild assignments following classifications from Andersen (1995a), with poorly represented taxa included in the analyses. Analyses were also carried out with the abundance of ants from the Rhytidoponera genus excluded, to confirm that alterations in the abundance of this group were driving observed alterations in ant assemblage composition. For these analyses, CO2 treatment was entered as a fixed effect. Analyses were carried out on Bray-Curtis dissimilarity matrices, with 1000 permutations. The assemblage composition occurring under the two CO2 treatments was visualised using non-metric multi-dimensional scaling (NMDS). Stress values remained below 0.2 across multiple runs with two or three dimensions. Data points represent the ant assemblage captured in each ring, during each of the six two-week sampling periods.

Given the low replication in FACE studies, an alpha of p = 0.1 was used to minimise the occurrence of type II errors, following the approach of other workers in similar systems (Sanders et al. 2004; Hamilton et al. 2012).

6.4 Results A total of 4,046 ants from 18 genera were collected during the study (Table 6.1). Total ant abundance was reduced by 35.5% under elevated CO2 conditions (total abundance – across study ± SD ambient 2458 ± 142.2 vs. elevated 1588 ± 100.6; mean ± SD: 136.6 ± 142.2 vs. 88.2 ± 100.6, Table 6.2, Fig 6.1). Specifically, Rhytidoponera and Ochetellus ants had reduced abundances under elevated CO2 compared with ambient conditions (Table 6.2, Fig 6.1). Ants from the genus Rhytidoponera also showed a corresponding decline in biomass under elevated CO2 conditions (mean ± SD: ambient 19.62mg ± 12.00 vs. 2 elevated 12.94mg ± 9.19, χ 1 = 5.417, p = 0.020, Table 6.2); this was the only genus to show an alteration in biomass in the study. None of the genera studied showed significant increases in abundance or biomass under future levels of CO2. Ants from the

109

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

genus Pheidole were numerically dominant under both elevated and ambient conditions, both in terms of having high abundance and occurring in a high percentage of traps (Table 6.1). Ants from the genus Rhytidoponera were the most frequently caught,

occurring in 100% of traps over the course of the study in both CO2 treatments, despite

reductions in their abundance under elevated CO2 conditions.

Table 6.1: The genera found in this study and their total abundances across both CO2 treatments, as well as their functional guild assignments, based on (Andersen 1995a; Andersen & Majer 2004) DomDoli = Dominant Dolichodorinae, Opp = Opportunists, Cold = Cold climate specialists, Hot = Hot climate specialists, Tropical = Tropical climate specialists, Subord = Subordinate Camponotini, GenMyrm = Generalised Myrmecinae, Predators = Specialised predators, Cryptic = Cryptic genera.

Total individuals Traps present (%) Subfamily Genus Functional Guild (ambient + elevated) ambient elevated Dolichoderinae Ochetellus Opp 448 (322+126) 77.78 66.67 Iridomyrmex DomDoli 216 (77+139) 50.00 55.56 Tapinoma Opp 13 (11+2) 22.22 11.11 Papyrius DomDoli 1 (1+0) 5.56 0.00 Formicinae Nylanderia Opp 574 (324+250) 83.33 83.33 Prolasius Cold 192 (114+78) 44.44 72.22 Camponotus Subord 165 (126+39) 50.00 55.56 Polyrhachis Subord 124 (108+16) 38.89 38.89 Melophorus Hot 4 (3+1) 11.11 5.56 Anonychomyrma DomDoli 4 (3+1) 16.67 5.56 Myrmicinae Pheidole GenMyrm 1212 (746+466) 83.33 83.33 Crematogaster GenMyrm 360 (173+187) 27.78 61.11 Myrmecia Predators 82 (38+44) 66.67 66.67 Solenopsis Cryptic 42 (3+39) 11.11 55.56 Monomorium GenMyrm 42 (26+16) 44.44 22.22 Mayriella Tropical 10 (1+9) 5.56 16.67 Ponerinae Rhytidoponera Opp 555 (381+174) 100.00 100.00 Hypoponera Cryptic 2 (1+1) 5.56 5.56 Total: 4046

110

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

Table 6.2: Results from likelihood ratio tests performed on LME models with and without the fixed effect of CO2 treatment, on the square root transformed abundance and rank transformed biomass of each genus. Significant p values are highlighted in bold.

Abundance Biomass Genus χ2 p χ2 p 1 1 Ochetellus 2.865 0.091 2.739 0.106 Iridomyrmex 0.831 0.362 1.024 0.420

Nylanderia 0.006 0.938 0.088 0.775 Prolasius 0.081 0.776 0.590 0.487

Camponotus 0.272 0.602 0.668 0.516 Polyrhachis 1.975 0.160 0.493 0.606 Pheidole 0.016 0.900 0.063 0.812 Crematogaster 0.627 0.429 3.126 0.121 Myrmecia 0.125 0.724 0.297 0.638 Rhytidoponera 12.289 <0.001 5.417 0.020 TOTAL 3.519 0.061 0.476 0.490

111

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

TOTAL *

0 20 40 60 80 100 120 140 160 180

Ochetellus * Iridomyrmex

Nylanderia

Prolasius

Camponotus

Polyrhachis

Pheidole

Crematogaster

Myrmecia

Rhytidoponera ***

0 10 20 30 40 50 60 70

Abundance

Fig 6.1: The abundances – absolute (bars), and relative (%, points) – of the different ant genera

sampled in the different CO2 treatments, averaged across the six sampling periods. Ambient samples are shown in white; shaded bars/points denote elevated samples. Significance is shown by asterisks for 2 absolute abundance analyses (* p < 0.1, *** p < 0.01): Ochetellus χ 1 = 2.865, p = 0.091; 2 2 Rhytidoponera χ 1 = 12.289, p = <0.001; Total χ 1 = 3.519, p = 0.061. Error bars show ± SE of the mean.

In total, 18 genera were recorded under ambient conditions, 17 of which were also present in elevated rings (Table 6.1). Thus, rarefaction-based genera richness did not

differ significantly between the CO2 treatments (t (2) = –2.308, p = 0.137). However,

genus diversity was significantly increased under elevated CO2 conditions (SW mean ± 2 SD ambient1.17 ± 0.39 vs. elevated 1.56 ± 0.28, LME LRT χ 1 = 11.949, p = <0.001). 112

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

Assemblage composition in terms of genera was significantly altered under elevated CO2 conditions (Table 6.3, Fig 6.2). Under both CO2 treatments, ants from the genera Rhytidoponera, Pheidole and Nylanderia had the highest relative abundances (Fig 6.1). There was a tendency for communities under elevated CO2 to consist of relatively more ants from genera including Crematogaster, Nylanderia and Pheidole, while ambient communities contained proportionally more ants from the genera Ochetellus and Rhytidoponera (Figs 6.1 and 6.2, Table 6.2). Reductions in the abundance of Rhytidoponera ants under elevated

CO2 drove the alterations in assemblage composition between the two treatments (Table 6.3). Assemblage structure on a functional guild level, however, remained relatively consistent under elevated CO2 (Table 6.3, Fig 6.3).

Table 6.3: Results from multivariate permutational analysis (PERMANOVA) of the effect of CO2 treatment on ant assemblage, with abundance split by genera, both including and excluding the abundance of Rhytidoponera, and functional guild.

d.f. (treatment, Variable SS MS Pseudo-F P(perm) residual) Genera 1, 34 0.579 0.579 2.175 0.023

Genera - Rhytidoponera 1, 34 0.390 0.390 1.188 0.183 removed Functional guild 1, 34 0.315 0.315 1.410 0.226

113

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

Fig 6.2: NMDS plot of the genus-level ant communities under the two CO2 treatments (ambient CO2 shown in white ellipse/pale grey points, elevated CO2 in grey ellipse/black points). All sampling dates were included, with 36 samples in total (n=6x6); the points represent the assemblage composition of the individual samples. Data points closer together represent samples that are more compositionally similar to one another. Ellipses show the standard deviation of the centroid points for each treatment level. Genera labels are given by the first three letters of each generic name, listed in full in Table S1. All stress values remained below 0.2 with k=3 dimensions.

114

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

Fig 6.3: NMDS plot of the sampled ant communities split by functional guild and CO2 treatment.

Ambient CO2 samples are shown with the white ellipse/pale grey points, elevated CO2 by the grey ellipse/black points. Cold = Cold climate specialists, Cryptic = Cryptic genera, DomDoli = Dominant Dolichoderinae, GenMyrm = Generalised Myrmicinae, Hot = Hot climate specialists, Opp = Opportunists, Predators = Specialised predators, Subord = Subordinate Camponotini, Tropical = Tropical climate specialists. All stress values remained below 0.2 with k=2 dimensions.

115

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

6.5 Discussion This study investigated the abundance of foraging ants occurring in the understorey of a native Australian Eucalyptus woodland under current and future levels of CO2. Here, significant declines in the abundance of ants coupled with alterations in genera-based ant assemblage composition under elevated CO2 are reported for the first time. The decline in the overall abundance of ants in the woodland tracks the abundance of many other invertebrate groups sampled as part of a previous study, including beetles and mites, which also declined over the same period under elevated conditions (Facey et al., 2016). This finding therefore supports the use of ants as indicators of the wider community responses to environmental change. This finding is also consistent with work from other woodland FACE experiments in the USA, which reported declines in ant abundance under CO2 fumigation (Hillstrom & Lindroth 2008; Stiling et al. 2010).

So what could be driving the declines seen in ants at EucFACE? Many of the ant genera in this study will have at least some plant-based elements in their diets, including Rhytidoponera and Pheidole ants, which have extremely broad diets (Andersen 2000). If plant quality is negatively altered by changes in CO2, as found in other studies (Robinson et al. 2012), responses in the plant community could explain the changes seen in ant abundance through bottom-up processes. At this stage, however, other work from the EucFACE site conducted in parallel with this study has shown that the woodland plant quantity/quality is relatively resistant to alterations in CO2. There have been no detected changes in leaf area index, or canopy carbon to nitrogen ratios (Duursma et al., 2015; Gherlenda et al., 2015), at least within the timeframe of the experiment which has been under fumigation since 2012. Despite this, the declines seen in the abundances of certain ant genera, as well as ants overall and invertebrates in the wider community (Facey et al., 2016), potentially point to a broad deterioration in habitat quality occurring under elevated CO2; many small, hard-to-detect changes may be occurring in the ecosystem, contributing to the observed declines. It is possible that changes in other plant characteristics not currently being assessed may be occurring at the EucFACE site under elevated CO2, such as changes in plant structure, community composition, and understorey plant quality. More work would be needed to ascertain this, and potentially link any changes to the observed alterations in the ant community.

Ant populations are known to be sensitive to fluctuations in the availability of food resources (Andersen & Majer 2004). Given that many of the ant genera in this study

116

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

utilise invertebrate prey as a food resource (Shattuck 1999), the observed reduction in the abundance of ants could be explained by the previously reported reductions in invertebrates likely to constitute their food resources (Facey et al., 2016). Another element of the diet of many ant genera is Hemipteran honeydew, including but not limited to ants from the genera; Ochetellus, Rhytidoponera and Crematogaster (Shattuck 1999). Previous work found no difference in the abundance of (non-psyllid) Hemiptera under elevated CO2 (Facey et al., 2016). However, concurrent work by other researchers at the EucFACE site revealed significant declines in the abundances of three psyllid species

(Hemiptera) under elevated CO2, one of which (Glycaspis sp.) is tended by ants (Gray 1971; Gherlenda et al. 2016a). Further, a glasshouse study on green peach aphids found that the quantity of honeydew produced by aphids reared on tobacco plants declined under elevated CO2 conditions (Fu et al. 2010). Thus, reductions in the abundance (number of psyllids) and production (rate of excretion) of honeydew resources may be potential mechanisms explaining declines in the abundances of several of the ant genera observed in this study. Further work looking at arboreal ant activity and tending behaviour at the EucFACE site, as well as honeydew production, could provide evidence for such a mechanism.

This study has demonstrated for the first time the capacity of altered atmospheric concentrations of CO2 to have measurable indirect effects on ant assemblage composition. This alteration in structure was driven by reductions in the abundance of numerically dominant ants from the genus Rhytidoponera under elevated CO2. On a functional guild level, no evidence for shifts in assemblage guild structure was found. This is perhaps not surprising, given that the functional guild classifications are best suited to detecting differences across biogeographical scales (Andersen & Majer 2004). With multiple genera forming each of the functional guild classifications, the groupings are coarser than the genera-based analysis, making any effect more difficult to detect. This lack of alterations in functional composition suggests that changes in ant assemblage composition like the ones described here are unlikely to have major consequences for ecosystem functioning at large, for the levels of CO2 tested, in terms of altering the proportions of ants carrying out different roles within the ecosystem. However, the significant decline found in the total abundance of ants under elevated

CO2 suggests that future concentrations of the gas could have implications for the ecosystem processes these insects perform overall.

117

CHAPTER SIX ANT ASSEMBLAGES UNDER ELEVATED CO2

It was predicted that elevated CO2 would cause changes in genera richness and diversity. Declines in the abundance of certain numerically dominant ant genera, such as

Rhytidoponera, lead to an increase in assemblage evenness under elevated CO2, as evidenced by an increase in diversity with no accompanying decline in richness. This suggests that increases in CO2 could act as a moderate stress, causing shifts in dominant genera, reducing numerical dominance and increasing diversity. Given that competition between ant species is one of the primary forces structuring ant assemblages, with individual species exhibiting different levels of competitive ability (Cerdá, Arnan & Retana 2013), shifts in the dominant ant genera could lead to changes in the competitive dynamics and composition of ant communities. The lack of an effect of elevated CO2 on recorded ant generic richness suggests that ant responses to CO2 are not strong enough to cause the loss of genera from the assemblage altogether, for the levels of CO2 and timescale tested. Indeed, ant richness is known to be highly conserved and can be unresponsive to environmental change (Andersen 2010).

6.5.1 Conclusions Given that the broader-scale invertebrate community responded similarly to atmospheric change in previous work from the same experiment (Facey et al., 2016), this study suggests that ants can be valuable as indicators of wider community responses to environmental change. Elevated concentrations of CO2 could serve to modify ant assemblage composition by reducing numerical dominance and thus increase ant diversity. However, significant declines in the abundance of key ant genera and the ant assemblage at large suggest that elevated CO2 could eventually cause alterations in the important ecosystem functions these insects provide, including soil formation and nutrient cycling.

118

Chapter Seven: General Discussion

7.1 Main findings and synthesis The overall objective of this thesis was to investigate how predicted precipitation and atmospheric changes will impact invertebrate communities in the future, as our climate continues to change. Past work considering the effects of altered precipitation changes

(Chapter 2) and elevated concentrations of CO2 and O3 (Chapter 4) on invertebrates, and the communities they form, was reviewed to identify common themes and research gaps in the existing literature. Empirically, this thesis aimed to investigate; i) the effects of altered precipitation on the abundance and composition of invertebrates in a south eastern Australian grassland (Chapter 3); ii) the effects of elevated atmospheric concentrations of CO2 on the abundance and composition of invertebrates in a

Eucalyptus woodland (Chapter 5); and iii) the effects of elevated CO2 on the abundance, composition and diversity of the genus-level ant community occurring in the same woodland (Chapter 6). Of particular interest were the effects of the applied climatic and atmospheric changes on functional guilds (feeding groups).

Invertebrates are key faunal components in grassland and forest ecosystems, underpinning vital ecosystem processes including nutrient cycling, soil formation and pollination (Whiles & Charlton 2006; Meehan et al. 2014). Predicting the responses of invertebrate communities and the ecosystems they contribute to will require greater representation of invertebrate responses in field scale global change studies. Additionally, both review chapters (2 and 4) identified a need for greater geographical representation within such studies, across ecotypes. This will help to determine the generality of the observed responses, given that most studies to date have been carried out in the Northern Hemisphere, most notably within the USA, restricting our ability to make predictions about entire groups of ecosystems (Ainsworth & Long 2005; Beier et al. 2012; Mundim & Bruna 2016). In addition, Chapter 2 highlighted the need for greater representation across the full range of predicted rainfall scenarios. The majority of rainfall manipulation studies carried out to date have considered only the responses of plants to short-term rainfall alterations of limited scope – altered frequency scenarios, for instance, remain critically under-represented (Johnson et al. 2016b).

119

CHAPTER SEVEN GENERAL DISCUSSION

Arguably the most consistent finding emerging from the literature is the idiosyncratic nature of the responses of organisms to climatic and atmospheric changes (Chapters 2 and 4). Such species-specificity in the direction and/or strength of responses in the system potentially frustrates attempts to make broad predictions about responses at the ecosystem-level (Newman et al. 2011; Hillstrom et al. 2014). Viewed in a different light, the idiosyncratic nature of species responses to global change serves to highlight the importance of community level studies to provide realistic approximations of ecosystem responses. Extrapolating the findings from highly controlled studies on one or two focal species is likely to yield highly unrealistic conclusions at best, with the potential for misleading results (Chapter 4). One option is the use of feeding groups within community level studies to find trends, which allow some generalities to be made (Hillstrom & Lindroth 2008; Newman et al. 2011), with some patterns evident in the literature and within the empirical chapters of this thesis.

7.2 Effects of climatic and atmospheric change on invertebrate primary consumers

7.2.1 Altered precipitation Under altered precipitation, herbivores showed contrasting responses to altered water availability, which varied according to feeding mode (i.e. sap-sucking vs. folivorous taxa, Chapter 3). It was predicted that reductions in water availability would cause reductions in primary production in the grassland, leaving fewer resources available for herbivores. At the same time, reductions in water availability have been shown to alter plant chemistry in previous work (e.g. Tariq et al. 2012), which may have benefits for sap sucking herbivores in particular (plant stress and pulsed stress hypotheses (PSH and PuSH); White 1969; Huberty & Denno 2004). Thus, it was predicted that sap-sucking herbivores may respond differently (i.e. less negatively) than groups with other feeding methods such as chewing Orthopterans. Under increased rainfall, it was predicted that, given the naturally water limited state of the ecosystem, increased water availability would result in stimulations of primary productivity, promoting herbivore success (plant vigour hypothesis (PVH); Price 1991).

To some extent, these predictions were supported by the data from DRI-Grass (Chapter 3), with positive responses in the abundance of Hemipterans under reduced amounts of rainfall compared with all other scenarios, providing field-based evidence for the PSH. The mean abundance of Hemiptera in reduced frequency plots was comparable to that

120

CHAPTER SEVEN GENERAL DISCUSSION in reduced amount plots, but did not differ significantly from ambient levels. Thus, no support for the PuSH was found in this group, though with continued data collection over longer time scales, the differences in abundance between ambient and reduced frequency plots may reach significance. Under more extreme reductions in water availability – the summer drought treatment – Hemiptera abundance did not differ compared with ambient conditions, suggesting that past a certain level of drought-stress, any positive effects of plant water-stress on Hemiptera may be lost. This may be as a result of reductions in turgor pressure inhibiting feeding by sap-sucking insects (Huberty & Denno 2004). These findings are in line with those found in a glasshouse study by Tariq et al. (2012), who reported improved aphid performance on plants receiving moderate drought stress only; when plants were highly drought stressed, benefits to aphid performance were no longer apparent.

In contrast to Hemiptera, chewing herbivores showed more mixed responses to the applied rainfall regimes. Orthoptera responded positively to increases in water availability, as predicted by the PVH and supported by work in other systems (Lenhart et al. 2015). Omnivorous Diptera – many of which will have plant-based elements in their diets – also showed elevated abundances under higher water availability compared with reduced rainfall regimes and summer drought plots. This result supports similar findings for the same group in a previous study by Frampton, Van Den Brink & Gould (2000). Both Diptera and Orthoptera showed positive correlations with plant biomass, suggesting that resource abundance and apparency are important factors for these groups. Conversely, Lepidoptera caught by sticky traps had increased abundance under summer drought conditions compared with reduced frequency plots. Lepidoptera caught using this method showed no evidence of correlations with the assessed plant metrics (C3:C4 ratio, plant biomass); their response may instead suggest that plant quality is a more important factor for this group. Plant quality has been shown to be responsive to reductions in water availability and forms the basis of the PSH (White 1969). In a crucifer system, water-stress related alterations in plant quality (increased carbohydrate and amino acid concentrations) resulted in increases in herbivory by leaf chewing insects including moth larvae (Louda & Sharon 1992). Further work looking at plant primary and secondary chemistry would be needed at DRI-Grass to determine the mechanisms underpinning this observation.

121

CHAPTER SEVEN GENERAL DISCUSSION

Thus, the use of feeding groups to form generalisations about the responses of invertebrates to alterations in rainfall has some limitations. Whilst clearer differences can be seen between feeding groups (sap-suckers and folivorous herbivores), within the folivorous guild the picture becomes a little less clear cut. Part of this discrepancy will likely be due to differences between taxonomic groups in the factors determining and/or limiting their success within the environment (i.e. resource quantity or quality) as well as differences in the direct effects of altered precipitation for these groups (i.e. adaptive morphological and behavioural differences, discussed in Chapter 2). Our understanding of community responses to altered precipitation will be improved with the complementary use of more restricted studies exploring the physiological and mechanistic bases underpinning these findings in specific groups of invertebrates.

Though there was evidence of the responses discussed above occurring in the primary consumer community in response to altered rainfall, overall, watering treatments exerted weak influences on invertebrate abundance across sampling dates. Most effects were only apparent during the April sampling campaign when the differences between treatments will have been most extreme (at the end of the summer drought period). Some responses were inconsistent between sampling dates and methods, except Diptera which increased in both vacuum and sticky samples in response to increased rainfall. The short-term, transient nature of the observed effects suggests that any changes occurring in the plant community, in terms of resource abundance, composition and quality, are not great enough to i) maintain increased populations of invertebrates (in the case of elevated abundances of Diptera and Orthoptera under increased rainfall) or ii) inhibit the recolonization of plots subjected to summer drought upon the cessation of the drought period and subsequent regrowth of plant material.

7.2.2 Elevated carbon dioxide

In contrast to the effects of altered precipitation, elevated CO2 had a more consistent negative effect on the abundance of numerous primary consumer groups, including chewing herbivores, omnivores, Isopoda and Orthoptera (Chapter 5). This is in line with the predictions made for this study; it was anticipated that herbivores would be negatively affected by changes occurring in plant traits in response to elevated CO2, as found in similar studies in the Northern Hemisphere (e.g. Stiling, Rossi & Hungate 1999; Stiling et al. 2002; Hamilton et al. 2012). Similarly, scavenger and detritivorous primary consumer groups such as Acari and Collembola which are also likely to have

122

CHAPTER SEVEN GENERAL DISCUSSION

plant-based elements in their diets were negatively affected by increase in CO2. This is consistent with findings from Northern Hemisphere systems (Hansen et al. 2001; Loranger et al. 2004).

In contrast, it was predicted that sap-sucking Hemipterans would be less negatively affected by plant-mediated changes in their food quality, as found in other studies (Bezemer & Jones 1998), drawing a parallel with the positive responses expected for the same group under reduced water availability. No evidence of a decline in the abundance of non-psyllid Hemiptera was found under elevated CO2 – this finding, in combination with the positive responses seen in this group under reduced water availability in Chapter 3, suggests that Hemipterans could stand to benefit, or at the least be less negatively affected by future climatic and atmospheric changes. However, a concurrent study by other researchers at EucFACE found reductions in the abundance of three species of sap-sucking psyllids under elevated CO2 (Gherlenda et al. 2016a). This supports the idea that responses to altered atmospheric composition will also be species-specific, as noted by other researchers (e.g. Hillstrom et al. 2014).

Psocoptera were the only primary consumers to be positively affected by increases in

CO2, though their low abundance in the study requires consideration when interpreting this result. Psocids feed on algae and lichens, neither of which are well studied in terms of responses to elevated CO2, but which are present at EucFACE given that bark- adhering lichens are common on most Eucalyptus stems within the stand. One study in alpine grassland showed a trend towards increased lichen biomass under elevated conditions (Schäppi & Körner 1996); if such an increase occurred at the EucFACE site, this could be responsible for the increased abundance seen in this group under elevated

CO2, though we do not currently have data on whether or not this is the case.

The prediction that herbivore abundance would be reduced under elevated CO2 conditions was based on the widely reported decrease in plant resource quality observed in other FACE experiments and glasshouse studies (Robinson et al. 2012). However, work carried out at the EucFACE site has revealed no change in various plant quality metrics, including foliar C:N ratios (Gherlenda et al. 2015) and leaf area index (Duursma et al. 2015). This is not entirely unexpected; Hamilton et al. (2012) also observed changes in arthropod populations under elevated conditions with no accompanying alteration in C:N ratios of plant tissues in a northern hemisphere forested FACE site. One, as yet undetermined, plant-mediated mechanism for these declines could be alterations in

123

CHAPTER SEVEN GENERAL DISCUSSION

plant secondary chemistry occurring under elevated [CO2], which is known to affect invertebrate herbivores (Robinson et al. 2012), providing an avenue for further research.

Regardless of the specific mechanisms underpinning these observations, the declines seen across several groups at EucFACE suggest that elevated concentrations of CO2 may have implications for ecosystem processes. Many of the organisms captured in this study are likely to have roles in supporting ecosystem process like soil formation and nutrient cycling (such as Acari, Collembola and Isopoda) (Hansen et al. 2001), suggesting that the sustainability of such processes may be undermined as concentrations of CO2 continue to rise.

7.3 Effects of climatic and atmospheric change on higher trophic levels

7.3.1 Altered precipitation It was predicted that reduced water availability associated with summer drought and reduced amount treatments would cause reductions in the abundance of predatory groups such as spiders and parasitic wasps (Chapter 3), as a result of declines in the abundance of their primary consumer food resources. As with lower trophic levels, secondary consumers showed generally few responses, with only some groups showing evidence of season-dependent responses to alterations in water availability. However, the direction of the observed responses was the opposite of what was predicted – predators were found to be more abundant in the three water limited treatments, particularly the summer drought treatment.

In contrast, Buchholz et al. (2013) found reductions in semi-dry grassland spider diversity and abundance under water-limited conditions. However, at a similar site three years earlier, the same authors found no change in spider species richness, composition or abundance under precipitation manipulation (Buchholz et al. 2010). Zhu et al. (2014), on the other hand, reported a decline in herbivore abundance under drought conditions with stable populations of organisms at higher trophic levels, leading to an increasingly predator-dominated community in a Chinese steppe. In Chapter 3, such a pattern could be caused by increases in the abundances of spiders less reliant on complex vegetation structure for web-building, such as cursorial wandering spiders (e.g. Thomisidae, Lycosidae), given the reduction in plant biomass found in these plots. Greater taxonomic resolution would be needed to determine whether or not there is a shift occurring in the Family or species identity of spiders under the different rainfall regimes.

124

CHAPTER SEVEN GENERAL DISCUSSION

These spiders are known to feed on Collembola and Hemipterans, among other small, soft-bodied arthropods (Nyffeler & Breene 1990) which also increased in abundance under reduced water availability. This suggests that these predators may be tracking the availability of their prey. Another possibility is that reductions in vegetation complexity may also confer thermal benefits for these spiders, which experience high energy costs via activities like hunting. Like all invertebrates, spiders are reliant on their surroundings for thermal radiation, which is known to play an important role in spider habitat selection (Riechert & Tracy 1975; Porter, Parry & Carter 1991).

7.3.2 Elevated carbon dioxide In contrast to the potentially positive effects of altered precipitation on secondary consumers, increases in CO2 had predominantly negative effects on the abundances of higher trophic levels. Predatory groups, comprising spiders and other carnivorous beetle families (among others), showed no response to altered CO2. However, Neuroptera and non-ant Hymenoptera (parasitoid wasps) showed declines in abundance of 37.1 and

14.7%, respectively, under future levels of CO2. These reductions could have been driven by the declines seen in the abundances of organisms at lower trophic levels likely to constitute their prey/hosts. In addition, plant architectural complexity has been shown to negatively affect the search efficiency of parasitoid wasps (Gingras, Dutilleul & Boivin 2002). Increases in plant complexity, driven by carbon fertilisation occurring under elevated CO2, could also have contributed to the reductions in parasitoid abundance. Further work at the EucFACE site would be needed to determine if there are any differences in plant structure occurring under the two treatments.

Within the body of work from FACE studies looking at invertebrate responses, along with more controlled glasshouse studies, predictions about the responses of higher trophic levels are conflicting (Chapter 4). On the one hand, predators and parasitoids may stand to benefit from reductions in prey development rates, leaving them in vulnerable life stages for longer periods (Stiling et al. 2002), as well as increases in prey susceptibility making them easier to subdue (Coll & Hughes 2008). Indeed, previous FACE studies in the Northern Hemisphere have found evidence of increases in the numbers of parasitoids and parasitism rates under elevated CO2 (Percy et al. 2002; Stiling et al. 2002, 2003; Hillstrom & Lindroth 2008). However, at the same time, organisms at higher trophic levels will also be confronted with reductions in the quality and/or quantity of their food resources, as found at EucFACE. The reduction in the abundance

125

CHAPTER SEVEN GENERAL DISCUSSION of higher trophic levels present in Chapter 5 is consistent with findings from a sweetgum plantation in the US (Sanders et al. 2004); and suggests that elevated CO2, at the levels anticipated by the middle of this century, may have negative effects for parasitoids and some predatory groups.

In terms of implications for ecosystem functioning, taken together, the findings from Chapters 2 and 5 suggest that pest outbreaks of aphids and other Hemipterans could occur under altered climates and atmospheres, aided by reductions in pest control provided by forest insects such as parasitoid wasps. Further, such insects may provide pest suppression in other systems including adjacent agricultural land (Bianchi et al. 2006, 2008). In addition, all of the invertebrates in the system are likely to provide food resources for other vertebrate animals in the system, including insectivorous mammals and birds, many of which are already declining as a result of land use change and non- native species introductions (Ford et al. 2001; Newbold et al. 2016). The widespread declines in invertebrate abundance across trophic levels could thus have implications for the animals which rely on them as a food source as emissions of CO2 continue.

7.4 Climatic and atmospheric change impacts on invertebrate communities

7.4.1 Altered precipitation Altered precipitation regimes – specifically, summer drought conditions – caused short- term significant alterations in invertebrate composition. These changes were caused by alterations in the abundances of several taxonomic groups during the same period. These included increases in the abundance of predators, Coleoptera, Collembola, Lepidoptera and Hemiptera under reduced amount conditions, in contrast with the increase in Diptera under increased rainfall conditions. These changes show the potential for reductions in water availability to have implications for pest species from groups such as Hemiptera and Lepidoptera in agricultural systems with grassland components. However, the accompanying increase in the abundance of spiders could also indicate that any increases in the abundance of herbivorous groups will be kept in check by increased predation. Overall, however, no alterations in the total abundance of invertebrates were found and the alterations in abundance within individual groups were not strong enough to cause overall changes composition when data were considered across the full duration of the experiment. This suggests some resilience in the system at least over the short term and within habitats not uniformly affected by drought, because; i) there is rapid recolonization of plots with reduced abundances, for instance

126

CHAPTER SEVEN GENERAL DISCUSSION for Diptera and Orthoptera in reduced rainfall plots and ii) there is a lack of maintenance of greater populations, such as the comparably larger spider population occurring under reduced water availability. This resilience in the invertebrate community is likely related to the resilience of the underlying plant community which showed strong regrowth responses at the end of the summer drought period. This indicates that the applied drought was not strong enough to alter the plant community to the point of ‘community closure’, whereby the community can no longer support certain species after they have been lost, even if they are reintroduced (Lundberg et al. 2000). Such resilience may be a product of the selection pressures exerted by the strong climatic variability experienced over geological time in the area (Chiew et al. 2011).

But what about in reality, when such rainfall regimes will be experienced over greater geographical scales, for unpredictable periods of time? In Chapter 3, movement of animals between plots after the end of the summer drought could have contributed to the resilience seen in invertebrate community composition. In addition, long-term droughts – longer than the three month summer drought inflicted in Chapter 3 – are predicted to increase in frequency in the region (Chiew et al. 2011; Dai 2011). The invertebrate community at DRI-Grass showed the capacity to respond to reductions in water availability, under summer drought conditions. This could indicate that more extreme drought conditions, sustained over longer periods and across greater areas, could have the potential to alter grassland invertebrate communities, particularly if there is less potential for migration of animals to rebalance the community following such extreme events.

7.4.2 Elevated carbon dioxide

As for CO2, there was no evidence of alterations in community composition in response to elevated concentrations of the gas, despite the population declines across several groups. This is consistent with the findings of other studies in other forest systems which report findings ranging from weak, season-dependent impacts to non-detectable effects of CO2 fumigation on invertebrate composition (Sanders et al. 2004; Hillstrom & Lindroth 2008; Hillstrom et al. 2014). One potential explanation for this may be the idiosyncratic nature of the responses of different invertebrates to changes in atmospheric composition. With so many species in the forest, each responding in different directions, there may be a tendency for effects in one group to be offset by those of another responding in the opposite direction (Hamilton et al. 2012; Hillstrom et

127

CHAPTER SEVEN GENERAL DISCUSSION al. 2014). However, in Chapter 5, given the broadly consistent declines occurring across the community, coupled with significant reductions in total invertebrate abundance not seen in other systems (Hillstrom et al. 2014), the lack of compositional change is likely a result of a compositionally-similar community occurring under elevated CO2, with fewer total individuals. Communities comprised of fewer individuals may have implications for ecosystem processes, as discussed in section 7.3.2.

At finer taxonomic resolution in Chapter 6, the genus-level ant community showed evidence of compositional responses to elevated CO2, driven by changes in the relative abundances of key ant genera. Given the importance of intra-specific interactions in structuring ant community composition (Andersen 1995a), such changes in composition suggest that elevated CO2 may be altering the competitive dynamics between populations of ants in the Eucalyptus understorey. Such changes could be brought about by changes in food resources, to which ants are known to be sensitive (Andersen & Majer 2004), such as Hemipteran honeydew resources (Gherlenda et al. 2016a). Declines in the total abundance of ants under elevated CO2 suggest that the highly varied ecosystem processes these animals support – such as soil formation, decomposition and seed dispersal (Del Toro et al. 2012) – could be weakened as concentrations of the gas continue to rise.

7.5 Interactions with other climate factors In reality, the effects of atmospheric and precipitation change on invertebrate communities will occur within the context of wider global and climatic changes. The simultaneous effects of multiple climatic factors on a community may have three main outcomes. Firstly, the effect of change in both variables on the system is the sum of the effects of both factors applied in isolation (additive effect). Secondly, the net effect of both factors may be greater than the sum of both applied in isolation i.e. the effect of one or both of the factors is magnified by alterations in the other factor (synergistic effect). Thirdly, the effect of change in one factor may negate the impact of change in the other factor, as seen with the simultaneous elevation of O3 and CO2 (Chapter 4, antagonistic effect).

Predicted increases in temperature will have profound effects on invertebrates via a variety of mechanisms. Warmer average temperatures will alter microclimate/habitat suitability, affect invertebrate physiology and development, and cause changes in the timing of biological events (phenology) (reviewed in Bale et al. 2002). To date, there is

128

CHAPTER SEVEN GENERAL DISCUSSION

evidence that the simultaneous elevation of temperature and atmospheric CO2 concentrations will have complex outcomes for natural systems. In particular, at the plant-level, there is evidence to support antagonistic effects of CO2 and increased temperatures on plant chemistry and defence variables, such that the combined effects of the two factors in reality may be less severe than those predicted by single-factor studies (Robinson et al. 2014). The same also appears to be true for invertebrate herbivores and consumers, with the potential for work considering both factors to reveal complex and unexpected outcomes relative to single-factor studies (Facey et al. 2014). However, to date, very little work has considered the effects of alterations in multiple climatic factors at scales greater than simple pairwise comparisons in glasshouse settings (Villalpando et al. 2009; Jamieson et al. 2012). There is a pressing need for community-level field studies looking at the effects of multiple climatic factors on invertebrate community structure and dynamics if we are to gain realistic insight into the responses of the biosphere to predicted change (Facey et al. 2014).

7.6 Constraints, caveats and further study Large-scale climatic and atmospheric change experiments – like the ones used here – exist at the cutting edge of community and global change ecology, though they are not free from limitations and necessary compromises. In terms of benefits, such experiments enable the testing of hypotheses at scales previously not possible, allowing the alteration of climatic and atmospheric factors at the community-level and in realistic settings. As such, the responses of the community, as well as those of the individual species and functional groups within the ecosystem, can be assessed with the full integration of direct and indirect responses to the imposed alteration (Fig 1.1). However, such experiments also incur high running costs and suffer from low levels of replication. This lack of replication presents a recurrent problem within the field, whereby the results from such experiments may be more accurate over glasshouse studies and other work in more controlled environments, but at the same time are treated with higher circumspection because of the statistical issues involved with analysing data from large-scale experiments (Johnson & Jones 2016). Continued improvements in the robustness of statistical methods, along with a balanced approach towards the results of such studies (i.e. one that is both rigorous and realistic), will be needed to fully recognise the potential value of such work.

129

CHAPTER SEVEN GENERAL DISCUSSION

An additional problem with open air field experiments – inherent to all plot-level work – is the potential for the movement of organisms between plots, confounding treatment effects, as discussed by Moise and Henry (2010) and in section 7.4.1. Particularly, there is the potential for organisms to be attracted (or repelled) by certain imposed treatments, leading to artificially high (or low) densities within plots, exaggerating treatment effects. On the flipside, animals may also move between plots, swamping out treatment effects and making them harder to detect. Nevertheless, differences in invertebrate composition have been recorded over short distances (Muff et al. 2009), with treatment differences detected in Chapter 3 with two metre separations between plots. As with replication issues, such issues are an acceptable compromise necessary for making progress in the field, requiring careful consideration when interpreting the results from such studies.

In terms of issues of community-level studies, the widely reported idiosyncratic nature of invertebrate responses to climatic and atmospheric change presents a theoretical quandary. It could be argued that the species-specific nature of such responses makes their study futile as no generalisations can be drawn (Newman et al. 2011). However, the empirical chapters of this thesis demonstrate the potential use of broader taxonomic classifications and feeding guilds as tools to describe the responses of invertebrates to future change. Across Chapters 3 and 5, several groups showed predictable responses to alterations in water availability and atmospheric change that were broadly consistent with current theory and findings in the literature. Such approaches will be a necessary compromise if we are to make predictions which are at once both realistic (i.e. not based on the responses of a single species studied in a glasshouse) and feasible, given restricted access to funding, time and human resources.

Chapter 6 also demonstrated the potential for the use of indicator taxa to stand as a proxy for the responses of the broader invertebrate community, with the ant population showing declines which were mirrored by other taxa in the system. Although, the genus- level assemblage changes occurring in the ant community were not reflected in the general invertebrate community, this could be an artefact of the lower taxonomic resolution used in Chapter 5. Indeed, specific to Chapter 3 and 5, the coarse taxonomic resolution used in these studies will have limited the ability to detect compositional changes occurring in response to the imposed changes. Further, it was unfeasible to assign more detailed functional guilds, such as intraguild predators and so on, limiting

130

CHAPTER SEVEN GENERAL DISCUSSION the ability to make predictions about the responses of such groups. The use of Order/Family level resolution was necessary in order to process the significant number of individuals captured across both studies. Future work could overcome this problem by adopting novel molecular techniques such as DNA-based metabarcoding, as DNA reference databases become more comprehensive. Such techniques have previously been used for microbial and fungal studies, but are increasingly being adapted for use in plant and animal studies (Coissac, Riaz & Puillandre 2012). Metabarcoding allows for morpho-species-equivalent identities to be determined within community-level samples (Baselga et al. 2013), overcoming the need for time-consuming and expensive taxonomic expertise.

Looking forward, avenues for research are proposed as priorities across climatic and atmospheric studies in Chapter 2 and 4. Firstly, given that the timescale over which studies are carried out may be important, studies over longer durations are needed. For instance, work by Stiling et al. initially revealed potential benefits of elevated CO2 for parasitoids due to increased host susceptibility (Stiling et al. 1999, 2002, 2003). Work from the same site roughly seven years later revealed that this effect had not persisted, with higher trophic levels no longer responsive to alterations in CO2 (Stiling et al. 2010). This highlights the need for such studies in order to take into account any resilience or lag effects in the system.

Perhaps most importantly if we are to gain realistic insights into the ecosystems of the future, studies incorporating multiple climate change factors will be necessary (Robinson et al. 2012; Facey et al. 2014). Whilst this is true, it is also the case that this should not dismiss the importance of large scale field manipulative studies such as those presented in this thesis with one central variable – these studies in themselves are still underrepresented in the literature and will provide key information for the generation of hypotheses. Nevertheless, there is the potential for interactive effects between variables, whereby the elevation of one variable serves to enhance or mitigate the effects of the other, as discussed in Chapter 4 between O3 and CO2. Thus, such studies will be crucial moving forward, and will require innovation to overcome the considerable logistical difficulties invited by more complex experimental designs (Johnson et al. 2016a).

Such work – over greater timescales and incorporating the complexity of climatic and atmospheric change – in combination with improvements in statistical tests and reductions in experimental infrastructure-related costs, will help to inform theoretical

131

CHAPTER SEVEN GENERAL DISCUSSION approaches and devise management strategies to ensure the sustainability of the important roles carried out by invertebrates in the world’s ecosystems.

7.7 Conclusions and future outlook This thesis has highlighted the responsiveness of invertebrates to climatic and atmospheric change in two disparate ecosystems, as well as the need for community- level studies to determine how invertebrate assemblages will respond to our changing climate. Two state-of-the-art field-scale experimental platforms were used to address the question of how climatic and atmospheric changes will impact invertebrate communities, the results of which are summarised in Fig 7.1. Overall, altered precipitation regimes caused highly variable responses in the abundance of invertebrates across the community, which were strongly seasonal and only weakly related to changes in the underlying plant community. The short-term, transient nature of the observed responses suggests that the invertebrate community – which has evolved against a background of strong precipitation variability – will be resilient to changes in rainfall, at least in the short term. The invertebrate community of an endangered Eucalyptus woodland showed resistance to change within both the plant and invertebrate community, in terms of composition and plant quality. Despite this resistance, elevated concentrations of CO2 also caused widespread declines in the populations of various arthropods across the community, including herbivore and parasitoid functional groups. Further, for the ant community, shifts in the dominant genus-level ant populations occurring under elevated CO2 drove changes in ant assemblage structure. This, coupled with the general declines witnessed in the ant and broader invertebrate community, support the notion that elevated CO2 could lead to changes in the ecosystem processes these organisms support.

Taken together, these results present contrasting evidence for invertebrate community- level responses to climatic and atmospheric change. On the one hand, communities may be able to cope with future increases in precipitation variability, suggesting that the ecosystem processes underpinned by invertebrates may remain stable in this system. On the other hand, exposure to levels of CO2 not recently experienced within evolutionary time could result in declines in the abundance of organisms likely to play important roles in processes like soil formation and nutrient cycling.

132

CHAPTER SEVEN GENERAL DISCUSSION

A continued focus on the responses of invertebrate communities to climatic and atmospheric change will be a challenging yet necessary endeavour to ensure that the roles of these diverse and fascinating organisms in ‘running the world’ are maintained.

133

CHAPTER SEVEN GENERAL DISCUSSION

Fig 7.1: A conceptual diagram linking together the different chapters of this thesis, highlighting the main findings from the empirical research chapters.

134

Acknowledgements

I thank V. Kumar, S. Wohl, C. McNamara, D. Fidler, C. Barton, G. Lopaticki, B. May, S. Pulumbarit, B. Amiji, C. Maier, E. Gibson-Forty and J. Vogelzang for their assistance in the field and laboratory. I thank K. Barnett, J. Powell and R. Duursma for statistical advice. H. Campbell provided helpful comments on early versions of Chapter 6. DRI- Grass (Chapter 3) was constructed with funds from Western Sydney University. Research activity was supported by a project grant from the Hermon Slade Foundation (P00021516) and funding provided by Western Sydney University. EucFACE (Chapters 5 & 6) is supported by the Australian Commonwealth Government through the Education Investment Fund, the Department of Industry and Science, and the Australian Research Council in partnership with Western Sydney University. The facilities at EucFACE were built as part of the Australian Government’s Nation- building Economic Stimulus Package. The work forming this thesis was supported by a Hawkesbury Institute for the Environment Higher Degree Research Award.

135

Bibliography

Addo-Bediako, A., Chown, S.L. & Gaston, K.J. (2001) Revisiting water loss in insects: a large scale view. Journal of Insect Physiology, 47, 1377–1388.

Agrell, J., Kopper, B., McDonald, E.P. & Lindroth, R.L. (2005) CO2 and O3 effects on host plant preferences of the forest tent caterpillar (Malacosoma disstria). Global Change Biology, 11, 588–599.

Ainsworth, E.A. & Long, S.P. (2005) What have we learned from 15 years of free-air

CO2 enrichment (FACE)? A meta-analytic review of the responses of

photosynthesis, canopy properties and plant production to rising CO2. New Phytologist, 165, 351–372.

Allen, L.H. (1992) Free-air CO2 enrichment field experiments: An historical overview. Critical Reviews in Plant Sciences, 11, 121–134.

Altermatt, F. (2003) Potential negative effects of atmospheric CO2 enrichment on insect communities in the canopy of a mature deciduous forest in Switzerland. Mitteilungen der Schweizerischen Entomologischen Gesellschaft, 76, 191–199.

Andersen, A.N. (1991a) Sampling communities of ground-foraging ants: Pitfall catches compared with quadrat counts in an Australian tropical savanna. Australian Journal of Ecology, 16, 273–279.

Andersen, A.N. (1991b) The Ants of Southern Australia: A Guide to the Bassian Fauna. CSIRO Publishing, Melbourne, Australia.

Andersen, A.N. (1995a) A classification of Australian ant communities, based on functional groups which parallel plant life-forms in relation to stress and disturbance. Journal of Biogeography, 22, 15–29.

Andersen, A.N. (1995b) Measuring more of biodiversity: genus richness as a surrogate for species richness in Australian ant faunas. Biological Conservation, 73, 39–43.

136

BIBLIOGRAPHY

Andersen, A.N. (1997) Using ants as bioindicators: multiscale issues in ant community ecology. Conservation Ecology, 1, 8.

Andersen, A.N. (2000) The Ants of Northern Australia: A Guide to the Monsoonal Fauna. CSIRO Publishing, Melbourne, Australia.

Andersen, A.N. (2010) Using ants as indicators of ecosystem change. In Ant Ecology (eds L. Lach, C.L. Parr & K.L. Abbott), pp. 138–141. Oxford University Press, Oxford, UK.

Andersen, A.N., Fisher, A., Hoffmann, B.D., Read, J.L. & Richards, R. (2004) Use of terrestrial invertebrates for biodiversity monitoring in Australian rangelands, with particular reference to ants. Austral Ecology, 29, 87–92.

Andersen, A.N. & Majer, J.D. (2004) Ants show the way down under: invertebrates as bioindicators in land management. Frontiers in Ecology and Environment, 2, 291–298.

Andersen, A.N., Ribbons, R.R., Pettit, M. & Parr, C.L. (2014) Burning for biodiversity: highly resilient ant communities respond only to strongly contrasting fire regimes in Australia’s seasonal tropics. Journal of Applied Ecology, 51, 1406–1413.

Andersen, A.N., Del Toro, I. & Parr, C.L. (2015) Savanna ant species richness is maintained along a bioclimatic gradient of increasing latitude and decreasing rainfall in northern Australia. Journal of Biogeography, 42, 2313–2322.

Andrey, A., Humbert, J.-Y., Pernollet, C. & Arlettaz, R. (2014) Experimental evidence for the immediate impact of fertilization and irrigation upon the plant and invertebrate communities of mountain grasslands. Ecology and Evolution, 4, 2610– 2623.

Andrew, N.R., Hill, S.J., Binns, M., Bahar, M.H., Ridley, E.V., Jung, M.-P., Fyfe, C., Yates, M. & Khusro, M. (2013) Assessing insect responses to climate change: what are we testing for? Where should we be heading? PeerJ, 1, e11.

Aslam, T.J., Johnson, S.N. & Karley, A.J. (2013) Plant-mediated effects of drought on aphid population structure and parasitoid attack. Journal of Applied Entomology, 137, 136–145.

137

BIBLIOGRAPHY

Aukema, J.E., Leung, B., Kovacs, K., Chivers, C., Britton, K.O., Englin, J., Frankel, S.J., Haight, R.G., Holmes, T.P., Liebhold, A.M., McCullough, D.G. & von Holle, B. (2011) Economic impacts of non-native forest insects in the continental United States. PLoS ONE, 6, 1–7.

Awmack, C.S., Harrington, R. & Lindroth, R.L. (2004) Aphid individual performance

may not predict population responses to elevated CO2 or O3. Global Change Biology, 10, 1414–1423.

Awmack, C. & Leather, S. (2002) Host plant quality and fecundity in herbivorous insects. Annual Review of Entomology, 47, 817–844.

Awmack, C., Woodcock, C. & Harrington, R. (1997) Climate change may increase vulnerability of aphids to natural enemies. Ecological Entomology, 22, 366–368.

Bale, J.S., Masters, G.J., Hodkinson, I.D., Awmack, C., Bezemer, T.M., Brown, V.K., Butterfield, J., Buse, A., Coulson, J.C., Farrar, J., Good, J.E.G., Harrington, R., Hartley, S., Jones, T.H., Lindroth, R.L., Press, M.C., Symrnioudis, I., Watt, A.D. & Whittaker, J.B. (2002) Herbivory in global climate change research: direct effects of rising temperature on insect herbivores. Global Change Biology, 8, 1–16.

Ballinger, A., Mac Nally, R. & Lake, P.S. (2005) Immediate and longer-term effects of managed flooding on floodplain invertebrate assemblages in south-eastern Australia: generation and maintenance of a mosaic landscape. Freshwater Biology, 50, 1190–1205.

Barbehenn, R.V., Chen, Z., Karowe, D.N. & Spickard, A. (2004) C3 grasses have higher

nutritional quality than C4 grasses under ambient and elevated atmospheric CO2. Global Change Biology, 10, 1565–1575.

Barbosa, P., Saunders, J. & Waldvogel, M. (1982) Plant-mediated variation in herbivore suitability and parasitoid fitness. Proceedings of the 5th International Symposium on Insect- Plant Relationships, pp. 63–71.

Barker, G.M. (2004) Millipedes (Diplopoda) and centipedes (Chilopoda)(Myriapoda) as predators of terrestrial gastropods. In Natural Enemies of Terrestrial Molluscs (ed. G. M. Barker), pp. 405–426. CABI Publishing, Wallingford, UK.

138

BIBLIOGRAPHY

Barnett, K.L. & Facey, S.L. (2015) Grasslands, invertebrates and precipitation: a review of the effects of climate change. Frontiers in Plant Science, 7, 01196.

Barnett, K. & Johnson, S.N. (2013) Living in the soil matrix: abiotic factors affecting root herbivores. In Behaviour and Physiology of Root Herbivores (eds S. Johnson, I. Hiltpold & T.C.J. Turlings), pp. 1–52. Elsevier, Oxford, UK.

Barton, B.T., Beckerman, A.P. & Schmitz, O.J. (2009) Climate warming strengthens indirect interactions in an old-field food web. Ecology, 90, 2346–2351.

Barton, P., Sato, C., Kay, G., Florence, D. & Lindenmayer, D. (2016) Effects of environmental variation and livestock grazing on ant community structure in temperate eucalypt woodland. Insect Conservation and Diversity, 9, 124–134.

Baselga, A., Fujisawa, T., Crampton-Platt, A., Bergsten, J., Foster, P.G., Monaghan, M.T. & Vogler, A.P. (2013) Whole-community DNA barcoding reveals a spatio- temporal continuum of biodiversity at species and genetic levels. Nature Communications, 4, 1892.

Bates, D., Maechler, M., Bolker, B. & Walker, S. (2014) lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1-7.

Bazzaz, F. (1990) The response of natural ecosystems to the rising global CO2 levels. Annual Review of Ecology and Systematics, 21, 167–196.

Begon, M., Townsend, C. & Harper, J. (2006) Ecology: From Individuals to Ecosystems, 4th ed. Blackwell Publishing, Oxford, UK.

Beier, C., Beierkuhnlein, C., Wohlgemuth, T., Peñuelas, J., Emmett, B., Körner, C., Boeck, H., Christensen, J.H., Leuzinger, S., Janssens, I.A. & Hansen, K. (2012) Precipitation manipulation experiments – challenges and recommendations for the future. Ecology Letters, 15, 899–911.

Bell, J.R., Wheater, C.P., Henderson, R. & Cullen, W.R. (2000) Testing the efficiency of suction samplers (G-vacs) on spiders: the effect of increasing nozzle size and suction time. Proceedings of the 19th European Colloquium of Arachnology, pp. 285–290.

Benoit, J.B. (2010) Water management by dormant insects: comparisons between dehydration resistance during summer aestivation and winter diapause. In: Navas,

139

BIBLIOGRAPHY

C.A. & Carvalho, J.E. (Eds.), Progress in Molecular and Subcellular Biology Aestivation: Molecular and Physiological Aspects, vol. 49. Springer, Berlin, pp. 209–230.

Bernays, E. (1986) Diet-induced head allometry among foliage-chewing insects and its importance for graminivores. Science, 231, 495–497.

Berridge, M. (2012) Osmoregulation in terrestrial arthropods. In Chemical Zoology V5: Arthropoda, Part 1 (ed. M. Florkin), pp. 287–320. Academic Press, New York.

Bestelmeyer, B.T. & Wiens, J.A. (2001) Ant biodiversity in semiarid landscape mosaics: the consequences of grazing vs. natural heterogeneity. Ecological Applications, 11, 1123–1140.

Bezemer, T. & Jones, T. (1998) Plant–insect herbivore interactions in elevated

atmospheric CO2: quantitative analyses and guild effects. Oikos, 82, 212–222.

Bezemer, T.M., Jones, T.H. & Knight, K.K.J. (1998) Long-term effects of elevated CO2 and temperature on populations of the peach potato aphid Myzus persicae and its parasitoid Aphidius matricariae. Oecologia, 116, 128–135.

Bianchi, F.J.J.A., Booij, C.J.H. & Tscharntke, T. (2006) Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proceedings of the Royal Society B: Biological Sciences, 273, 1715– 1727.

Bianchi, F.J.J.A., Goedhart, P.W. & Baveco, J.M. (2008) Enhanced pest control in cabbage crops near forest in the Netherlands. Landscape Ecology, 23, 595–602.

Borer, E.T., Seabloom, E.W. & Tilman, D. (2012) Plant diversity controls arthropod biomass and temporal stability. Ecology Letters, 15, 1457–1464.

Boutton, T.W., Cameron, G.N. & Smith, B.N. (1978) Insect herbivory on C3 and C4 grasses. Oecologia, 36, 21–32.

Brodbeck, B. & Strong, D. (1987) Amino acid nutrition of herbivorous insects and stress to host plants. In Insect Outbreaks (eds P. Barbosa & J.C. Schultz), pp. 347– 364. Academic Press, Inc., New York, New York, USA.

140

BIBLIOGRAPHY

Brook, A.J., Woodcock, B.A., Sinka, M. & Vanbergen, A.J. (2008) Experimental verification of suction sampler capture efficiency in grasslands of differing vegetation height and structure. Journal of Applied Ecology, 45, 1357–1363.

Brown, J.H., Gillooly, J., Allen, A., Savage, V. & West, G. (2004) Toward a metabolic theory of ecology. Ecology, 85, 1771–1789.

Brust, G.E. & House, G.J. (1990) Effects of soil moisture, no-tillage and predators on southern corn rootworm (Diabrotica undecimpunctata howardi) survival in corn agroecosystems. Agriculture, Ecosystems & Environment, 31, 199–215.

Buchholz, S., Hannig, K. & Schirmel, J. (2013) Losing uniqueness – shifts in carabid species composition during dry grassland and heathland succession. Animal Conservation, 16, 661–670.

Buchholz, S., Jess, A.-M., Hertenstein, F. & Schirmel, J. (2010) Effect of the colour of pitfall traps on their capture efficiency of carabid beetles (Coleoptera: Carabidae), spiders (Araneae) and other arthropods. European Journal of Entomology, 107, 277– 280.

Butler, C.D., Beckage, N.E. & Trumble, J.T. (2009) Effects of terrestrial pollutants on insect parasitoids. Environmental Toxicology and Chemistry, 28, 1111–1119.

Byrne, K.M., Lauenroth, W.K. & Adler, P.B. (2013) Contrasting effects of precipitation manipulations on production in two sites within the central grassland region, USA. Ecosystems, 16, 1039–1051.

Caswell, H. & Reed, F.C. (1976) Plant–herbivore interactions – The indigestibility of C4 bundle sheath cells by grasshoppers. Oecologia, 26, 151–156.

Caswell, H., Reed, F., Stephenson, S.N. & Werner, P.A. (1973) Photosynthetic pathways and selective herbivory: a hypothesis. The American Naturalist, 107, 465–480.

Cerdá, X., Arnan, X. & Retana, J. (2013) Is competition a significant hallmark of ant (Hymenoptera: Formicidae) ecology? Myrmecological News, 18, 131–147.

Chen, F., Ge, F. & Parajulee, M. (2005) Impact of elevated CO2 on tri-trophic interaction of Gossypium hirsutum, Aphis gossypii, and Leis axyridis. Environmental Entomology, 34, 37–46.

141

BIBLIOGRAPHY

Chiew, F.H.S., Young, W.J., Cai, W. & Teng, J. (2011) Current drought and future hydroclimate projections in southeast Australia and implications for water resources management. Stochastic Environmental Research and Risk Assessment, 25, 601–612.

Coissac, E., Riaz, T. & Puillandre, N. (2012) Bioinformatic challenges for DNA metabarcoding of plants and animals. Molecular Ecology, 21, 1834–1847.

Coleman, J.S. & Jones, C.G. (1988a) Plant stress and insect performance: cottonwood, ozone and a leaf beetle. Oecologia, 76, 57–61.

Coleman, J.S. & Jones, C.G. (1988b) Acute ozone stress on eastern cottonwood (Populus deltoides Bartr.) and the pest potential of the aphid, Chaitophorus populicola Thomas (Homoptera: Aphididae). Environmental Entomology, 17, 207–212.

Coll, M. & Hughes, L. (2008) Effects of elevated CO2 on an insect omnivore: a test for nutritional effects mediated by host plants and prey. Agriculture, Ecosystems & Environment, 123, 271–279.

Collins, S.L., Koerner, S.E., Plaut, J.A., Okie, J.G., Breese, D., Calabrese, L.B., Carvajal, A., Evansen, R.J. & Nonaka, E. (2012) Stability of tallgrass prairie during a 19-year increase in growing season precipitation. Functional Ecology, 26, 1450–1459.

Coupe, M.D., Stacey, J. & Cahill, J.F. (2009) Limited effects of above-and belowground insects on community structure and function in a species-rich grassland. Journal of Vegetation Science, 20, 121–129.

Couture, J.J. & Lindroth, R.L. (2012) Atmospheric change alters performance of an invasive forest insect. Global Change Biology, 18, 3543–3557.

Couture, J.J. & Lindroth, R.L. (2013) Impacts of atmospheric change on tree–arthropod interactions. In Climate Change, Air Pollution and Global Challenges: Understanding and Perspectives from Forest Research, 1st ed. (eds R. Matyssek, N. Clarke, P. Cudlin, T.N. Mikkelsen, J.-P. Tuovinen, G. Wieser & E. Paoletti), pp. 227–248. Elsevier Ltd., Oxford, UK.

Couture, J.J., Mason, C.J., Habeck, C.W. & Lindroth, R.L. (2016) Behavioral and morphological responses of an insect herbivore to low nutrient quality are inhibited by plant chemical defenses. Arthropod–Plant Interactions, 10, 1–9.

142

BIBLIOGRAPHY

Couture, J.J., Meehan, T.D., Kruger, E.L. & Lindroth, R.L. (2015) Insect herbivory alters impact of atmospheric change on northern temperate forests. Nature Plants, 1, 15016.

Couture, J.J., Meehan, T.D. & Lindroth, R.L. (2012) Atmospheric change alters foliar quality of host trees and performance of two outbreak insect species. Oecologia, 168, 863–876.

Coviella, C.E. & Trumble, J.T. (1999) Effects of elevated atmospheric carbon dioxide on insect–plant interactions. Conservation Biology, 13, 700–712.

Crawley, M.J. (1989) Insect herbivores and plant population dynamics. Annual Review of Entomology, 34, 531–564.

Cristescu, M.E. (2014) From barcoding single individuals to metabarcoding biological communities: towards an integrative approach to the study of global biodiversity. Trends in Ecology & Evolution, 29, 566–571.

Crous, K.Y., Ósvaldsson, A. & Ellsworth, D.S. (2015) Is phosphorus limiting in a mature Eucalyptus woodland? Phosphorus fertilisation stimulates stem growth. Plant and Soil, 391, 293–305.

CSIRO (1991) The Insects of Australia, 2nd ed. (ed. I. Naumann). Melbourne University Press, Melbourne, Australia.

Curtis, P.S. & Wang, X. (1998) A meta-analysis of elevated CO2 effects on woody plant mass, form, and physiology. Oecologia, 113, 299–313.

Dai, A. (2011) Drought under global warming: a review. Wiley Interdisciplinary Reviews: Climate Change, 2, 45–65.

Danks, H. (2000) Dehydration in dormant insects. Journal of Insect Physiology, 46, 837–852.

Davis, C.A., Austin, J.E. & Buhl, D.A. (2006) Factors influencing soil invertebrate communities in riparian grasslands of the Central Platte river floodplain. Wetlands, 26, 438–454.

Del Toro, I., Ribbons, R.R. & Pelini, S.L. (2012) The little things that run the world revisited: a review of ant-mediated ecosystem services and disservices (Hymenoptera: Formicidae). Myrmecological News, 17, 133–146.

143

BIBLIOGRAPHY

Del Toro, I., Silva, R.R. & Ellison, A.M. (2015) Predicted impacts of climatic change on ant functional diversity and distributions in eastern North American forests. Diversity and Distributions, 21, 781–791.

DeLucia, E.H., Nabity, P.D., Zavala, J.A. & Berenbaum, M.R. (2012) Climate change: resetting plant–insect interactions. Plant Physiology, 160, 1677–1685.

Department of the Environment and Water Resources. (2007a) Australia’s Native Vegetation: A Summary of Australia’s Major Vegetation Groups.

Department of the Environment and Water Resources. (2007b) Vegetation Fact Sheet: MVG 5 - Eucalypt Woodlands.

Dermody, O., O’Neill, B.F., Zangerl, A.R., Berenbaum, M.R. & DeLucia, E.H. (2008)

Effects of elevated CO2 and O3 on leaf damage and insect abundance in a soybean agroecosystem. Arthropod–Plant Interactions, 2, 125–135.

Diamond, S.E., Sorger, D.M., Hulcr, J., Pelini, S.L., Del Toro, I., Hirsch, C., Oberg, E. & Dunn, R.R. (2012) Who likes it hot? A global analysis of the climatic, ecological, and evolutionary determinants of warming tolerance in ants. Global Change Biology, 18, 448–456.

Docherty, M. & Hurst, D. (1996) Carbon dioxide-induced changes in beech foliage cause female beech weevil larvae to feed in a compensatory manner. Global Change Biology, 2, 335–341.

Drake, B.G., Gonzalez-Meler, M.A. & Long, S.P. (1997) More efficient plants: a

consequence of rising atmospheric CO2? Annual Review of Plant Physiology, 48, 609– 639.

Drake, V. (1994) The influence of weather and climate on agriculturally important insects: an Australian perspective. Australian Journal of Agricultural Research, 45, 487– 509.

Duursma, R.A., Gimeno, T.E., Boer, M.M., Crous, K.Y., Mark, G. & Ellsworth, D.S. (2015) Canopy leaf area of a mature evergreen Eucalyptus woodland does not

respond to elevated atmospheric [CO2] but tracks water availability. Global Change Biology, 22, 1666–1676.

144

BIBLIOGRAPHY

Eatough Jones, M. & Paine, T.D. (2006) Detecting changes in insect herbivore communities along a pollution gradient. Environmental Pollution, 143, 377–387.

Endara, M.J. & Coley, P.D. (2011) The resource availability hypothesis revisited: a meta- analysis. Functional Ecology, 25, 389–398.

Facey, S.L., Ellsworth, D.S., Staley, J.T., Wright, D.J. & Johnson, S.N. (2014) Upsetting the order: how climate and atmospheric change affects herbivore–enemy interactions. Current Opinion in Insect Science, 5, 66–74.

Facey, S.L., Fidler, D.B., Rowe, R.C., Bromfield, L.M., Nooten, S.S., Staley, J.T., Ellsworth, D.S. & Johnson, S.N. (2016) Atmospheric change causes declines in woodland arthropods and impacts specific trophic groups. Agricultural & Forest Entomology (accepted).

Facey, S.L. & Gherlenda, A.N. (2016) Forest invertebrate communities and atmospheric change. In Global Climate Change and Terrestrial Invertebrates, 1st ed. (eds S.N. Johnson & T.H. Jones) (in press).

Facey, S.L. & Torode, M.D. (2016) An assessment of the effect of sward height on suction sampling efficiency for the capture of grassland invertebrates using a G- Vac device. Proceedings of the 9th Australasian Conference of Grassland Invertebrate Ecology, Sydney, Australia.

Fajer, E.D., Bowers, M.D. & Bazzaz, F.A. (1989) The effects of enriched carbon dioxide atmospheres on plant–insect herbivore interactions. Science, 243, 1198–1200.

Faraway, J.J. (2006) Extending the Linear Model with R: Generalized Linear, Mixed Effects and Nonparametric Regression Models. CRC Press, Boca Raton, Florida.

Fay, P.A., Carlisle, J.D., Danner, B.N., Lett, M.S., McCarron, J.K., Stewart, C., Knapp, A.K., Blair, J.M. & Collins, S.L. (2002) Altered rainfall patterns, gas exchange, and growth in grasses and forbs. International Journal of Plant Sciences, 163, 549–557.

Fay, P.A., Carlisle, J.D., Knapp, A.K., Blair, J.M. & Collins, S.L. (2003) Productivity

responses to altered rainfall patterns in a C4-dominated grassland. Oecologia, 137, 245–251.

145

BIBLIOGRAPHY

Fischer, E.M., Beyerle, U. & Knutti, R. (2013) Robust spatially aggregated projections of climate extremes. Nature Climate Change, 3, 1033–1038.

Folgarait, P. (1998) Ant biodiversity to ecosystem functioning: a review. Biodiversity and Conservation, 7, 1121–1244.

Ford, H.A., Barrett, G.W., Saunders, D.A. & Recher, H.F. (2001) Why have birds in the woodlands of southern Australia declined? Biological Conservation, 97, 71–88.

Fowler, D., Cape, J.N., Coyle, M., Flechard, C., Kuylenstierna, J., Hicks, K., Derwent, D., Johnson, C. & Stevenson, D. (1999) The global exposure of forests to air pollutants. Water, Air, and Soil Pollution, 116, 5–32.

Fox, J. & Weisberg, S. (2011) An R Companion to Applied Regression, 2nd ed. Sage, Thousand Oaks, CA, USA. URL: http://socserv.socsci.mcmaster.ca/jfox/Books/Companion.

Frampton, G., Van Den Brink, P. & Gould, P. (2000) Effects of spring drought and irrigation on farmland arthropods in southern Britain. Journal of Applied Ecology, 37, 865–883.

Frew, A., Nielsen, U.N., Riegler, M. & Johnson, S.N. (2013) Do eucalypt plantation management practices create understory reservoirs of scarab beetle pests in the soil? Forest Ecology and Management, 306, 275–280.

Fry, E.L., Power, S.A. & Manning, P. (2014) Trait-based classification and manipulation of plant functional groups for biodiversity–ecosystem function experiments. Journal of Vegetation Science, 25, 248–261

Fu, X., Ye, L., Kang, L. & Ge, F. (2010) Elevated CO2 shifts the focus of tobacco plant defences from cucumber mosaic virus to the green peach aphid. Plant, Cell and Environment, 33, 2056–2064.

Gao, Q. & Reynolds, J.F. (2003) Historical shrub-grass transitions in the northern Chihuahuan Desert: modeling the effects of shifting rainfall seasonality and event size over a landscape gradient. Global Change Biology, 9, 1475–1493.

Gao, F., Zhu, S.-R., Sun, Y.-C., Du, L., Parajulee, M., Kang, L. & Ge, F. (2008)

Interactive effects of elevated CO2 and cotton cultivar on tri-trophic interaction of

146

BIBLIOGRAPHY

Gossypium hirsutum, Aphis gossyppii, and Propylaea japonica. Environmental Entomology, 37, 29–37.

Gardner, T.A., Barlow, J., Araujo, I.S., Ávila-Pires, T.C., Bonaldo, A.B., Costa, J.E., Esposito, M.C., Ferreira, L.V., Hawes, J., Hernandez, M.I.M., Hoogmoed, M.S., Leite, R.N., Lo-Man-Hung, N.F., Malcolm, J.R., Martins, M.B., Mestre, L.A.M., Miranda-Santos, R., Overal, W.L., Parry, L., Peters, S.L., Ribeiro, M.A., Da Silva, M.N.F., Da Silva Motta, C. & Peres, C.A. (2008) The cost-effectiveness of biodiversity surveys in tropical forests. Ecology Letters, 11, 139–150.

Gate, I., McNeill, S. & Ashmore, M. (1995) Effects of air pollution on the searching behaviour of an insect parasitoid. Water, Air, and Soil Pollution, 85, 1425–1430.

Gherlenda, A.N., Crous, K.Y., Moore, B.D., Haigh, A.M., Johnson, S.N. & Riegler, M.

(2015a) Precipitation, not CO2 enrichment, drives insect herbivore frass deposition and subsequent nutrient dynamics in a mature Eucalyptus woodland. Plant and Soil, 399, 29–39.

Gherlenda, A.N., Esveld, J.L., Hall, A.G., Duursma, R.A. & Riegler, M. (2016a) Boom and bust: rapid feedback responses between insect outbreak dynamics and canopy

leaf area impacted by rainfall and CO2. Global Change Biology, doi: 10.1111/gcb.13334.

Gherlenda, A.N., Haigh, A.M., Moore, B.D., Johnson, S.N. & Riegler, M. (2015b)

Responses of leaf beetle larvae to elevated [CO2] and temperature depend on Eucalyptus species. Oecologia, 177, 607–617.

Gherlenda, A.N., Moore, B.D., Haigh, A.M., Johnson, S.N. & Riegler, M. (2016b) Insect herbivory in a mature Eucalyptus woodland depends on rainfall and leaf

phenology but not CO2 enrichment. BMC Ecology (accepted).

Gibb, H., Sanders, N.J., Dunn, R.R., Watson, S., Photakis, M., Abril, S., Andersen, A.N., Angulo, E., Armbrecht, I., Arnan, X., Baccaro, F.B., Bishop, T.R., Boulay, R., Castracani, C., Del Toro, I., Delsinne, T., Diaz, M., Donoso, D.A., Enríquez, M.L., Fayle, T.M., Feener, D.H., Fitzpatrick, M.C., Gómez, C., Grasso, D.A., Groc, S., Heterick, B., Hoffmann, B.D., Lach, L., Lattke, J., Leponce, M., Lessard, J.-P., Longino, J., Lucky, A., Majer, J., Menke, S.B., Mezger, D., Mori, A., Munyai, T.C., Paknia, O., Pearce-Duvet, J., Pfeiffer, M., Philpott, S.M., de Souza, J.L.P.,

147

BIBLIOGRAPHY

Tista, M., Vasconcelos, H.L., Vonshak, M. & Parr, C.L. (2015) Climate mediates the effects of disturbance on ant assemblage structure. Proceedings of the Royal Society of London B: Biological Sciences, 282, 20150418.

Gilman, S.E., Urban, M.C., Tewksbury, J., Gilchrist, G.W. & Holt, R.D. (2010) A framework for community interactions under climate change. Trends in Ecology & Evolution, 25, 325–331.

Gingras, D., Dutilleul, P. & Boivin, G. (2002) Modeling the impact of plant structure on host-finding behavior of parasitoids. Oecologia, 130, 396–402.

Girousse, C., Bournoville, R. & Bonnemain, J.L. (1996) Water deficit-induced changes in concentrations in proline and some other amino acids in the phloem sap of alfalfa. Plant Physiology, 111, 109–113.

Gómez, C., Casellas, D., Oliveras, J. & Bas, J.M. (2003) Structure of ground-foraging ant assemblages in relation to land-use change in the northwestern Mediterranean region. Biodiversity and Conservation, 12, 2135–2146.

González-Megías, A. & Menéndez, R. (2012) Climate change effects on above- and below-ground interactions in a dryland ecosystem. Philosophical Transactions of the Royal Society of London, Series B, Biological Sciences, 367, 3115–3124.

Grant, J.A. & Villani, M.G. (2003) Soil moisture effects on entomopathogenic nematodes. Environmental Entomology, 32, 80–87.

Gray, B. (1971) Notes on the field behaviour of two ant species Myrmecia desertorum Wheeler and Myrmecia dispar (Clark) (Hymenoptera: Formicidæ). Insectes Sociaux, 18, 81–94.

Greaver, T.L., Sullivan, T.J., Herrick, J.D., Barber, M.C., Baron, J.S., Cosby, B.J., Deerhake, M.E., Dennis, R.L., Dubois, J.J.B., Goodale, C.L., Herlihy, A.T., Lawrence, G.B., Liu, L., Lynch, J.A. & Novak, K.J. (2012) Ecological effects of nitrogen and sulfur air pollution in the US: What do we know? Frontiers in Ecology and the Environment, 10, 365–372.

Guo, K., Hao, S.G., Sun, O.J. & Kang, L. (2009) Differential responses to warming and increased precipitation among three contrasting grasshopper species. Global Change Biology, 15, 2539–2548.

148

BIBLIOGRAPHY

Haile, F.J. (2001) Drought stress, insects, and yield loss. In Biotic Stress and yield Loss, (eds R.K.D. Peterson & L.G. Higley), pp 117–134. CRC Press, Boca Raton, US.

Hamilton, J.J.G., Dermody, O., Aldea, M., Zangerl, A.R., Rogers, A., Berenbaum, M.R. & DeLucia, E.H. (2005) Anthropogenic changes in tropospheric composition increase susceptibility of soybean to insect herbivory. Environmental Entomology, 34, 479–485.

Hamilton, J.G., Zangerl, A.R., Berenbaum, M.R., Pippen, J., Aldea, M. & DeLucia, E.H.

(2004) Insect herbivory in an intact forest understory under experimental CO2 enrichment. Oecologia, 138, 566–573.

Hamilton, J., Zangerl, A.R., Berenbaum, M.R., Sparks, J.P., Elich, L., Eisenstein, A. &

DeLucia, E.H. (2012) Elevated atmospheric CO2 alters the arthropod community in a forest understory. Acta Oecologica, 43, 80–85.

Hance, T., Baaren, J. Van, Vernon, P., Boivin, G. & van Baaren, J. (2007) Impact of extreme temperatures on parasitoids in a climate change perspective. Annual Review of Entomology, 52, 107–126.

Hansen, R.A., Williams, R.S., Degenhardt, D.C. & Lincoln, D.E. (2001) Non-litter

effects of elevated CO2 on forest floor microarthropod abundances. Plant and Soil, 236, 139–144.

Harrington, R., Woiwod, I. & Sparks, T. (1999) Climate change and trophic interactions. Trends in Ecology & Evolution, 14, 146–150.

Harrison, X.A. (2014) Using observation-level random effects to model overdispersion in count data in ecology and evolution. PeerJ, 2, e616.

Hartley, M. (2002) Rationale and methods for conservating biodiversity in plantation forests. Forest Ecology and Management, 155, 81–95.

Hättenschwiler, S. & Schafellner, C. (2004) Gypsy moth feeding in the canopy of a CO2- enriched mature forest. Global Change Biology, 10, 1899–1908.

Hättenschwiler, S. & Schafellner, C. (1999) Opposing effects of elevated CO2 and N deposition on Lymantria monacha larvae feeding on spruce trees. Oecologia, 118, 210– 217.

149

BIBLIOGRAPHY

Hawes, C., Haughton, A.J., Osborne, J.L., Roy, D.B., Clark, S.J., Perry, J.N., Rothery, P., Bohan, D.A., Brooks, D.R., Champion, G.T., Dewar, A.M., Heard, M.S., Woiwod, I.P., Daniels, R.E., Young, M.W., Parish, A.M., Scott, R.J., Firbank, L.G. & Squire, G.R. (2003) Responses of plants and invertebrate trophic groups to contrasting herbicide regimes in the Farm Scale Evaluations of genetically modified herbicide-tolerant crops. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 358, 1899–1913.

Heisler-White, J.L., Blair, J.M., Kelly, E.F., Harmoney, K. & Knapp, A.K. (2009) Contingent productivity responses to more extreme rainfall regimes across a grassland biome. Global Change Biology, 15, 2894–2904.

Helms, K.R. & Vinson, S.B. (2008) Plant resources and colony growth in an invasive ant: the importance of honeydew-producing Hemiptera in carbohydrate transfer across trophic levels. Environmental Entomology, 37, 487–93.

Hentley, W.T., Hails, R.S., Johnson, S.N., Hefin Jones, T. & Vanbergen, A.J. (2014) Top-down control by Harmonia axyridis mitigates the impact of elevated

atmospheric CO2 on a plant–aphid interaction. Agricultural and Forest Entomology, 16, 350–358.

Hertzog, L.R., Meyer, S.T., Weisser, W.W. & Ebeling, A. (2016) Experimental manipulation of grassland plant diversity induces complex shifts in aboveground arthropod diversity. PLoS ONE, 11, e0148768.

Hillstrom, M.L., Couture, J.J. & Lindroth, R.L. (2014) Elevated carbon dioxide and ozone have weak, idiosyncratic effects on herbivorous forest insect abundance, species richness, and community composition. Insect Conservation and Diversity, 7, 553–562.

Hillstrom, M.L. & Lindroth, R.L. (2008) Elevated atmospheric carbon dioxide and ozone alter forest insect abundance and community composition. Insect Conservation and Diversity, 1, 233–241.

Hillstrom, M., Meehan, T.D., Kelly, K. & Lindroth, R.L. (2010) Soil carbon and nitrogen mineralization following deposition of insect frass and greenfall from

forests under elevated CO2 and O3. Plant and Soil, 336, 75–85.

150

BIBLIOGRAPHY

Himanen, S., Nerg, A., Nissinen, A., Pinto, D., Stewart Jr, C., Poppy, G. & Holopainen, J.K. (2009) Effects of elevated carbon dioxide and ozone on volatile terpenoid emissions and multitrophic communication of transgenic insecticidal oilseed rape (Brassica napus). New Phytologist, 181, 174–186.

Hoffmann, B.D. & Andersen, A.N. (2003) Responses of ants to disturbances in Australia, with particular reference to functional groups. Austral Ecology, 28, 444– 464.

Hölldobler, B. & Wilson, E. (1990) The Ants. Harvard University Press, Cambridge, Massachusetts.

Holton, M.K., Lindroth, R.L. & Nordheim, E. V. (2003) Foliar quality influences tree–

herbivore–parasitoid interactions: effects of elevated CO2, O3, and plant genotype. Oecologia, 137, 233–244.

Horne, F.R. (1968) Nitrogen excretion in crustacea – I. The herbivorous land crab Cardisoma guanhumi latreille. Comparative Biochemistry and Physiology, 26, 687–695.

Hoover, D.L., Knapp, A.K. & Smith, M.D. (2014) Resistance and resilience of a grassland ecosystem to climate extremes. Ecology, 95, 2646–2656.

Hoover, S.E.R., Ladley, J.J., Shchepetkina, A.A., Tisch, M., Gieseg, S.P. & Tylianakis,

J.M. (2012) Warming, CO2, and nitrogen deposition interactively affect a plant– pollinator mutualism. Ecology Letters, 15, 227–234.

Hothorn, T., Bretz, F. & Westfall, P. (2008) Simultaneous inference in general parametric models. Biometrical Journal, 50, 346–363.

Hsiao, T. (1973) Plant responses to water stress. Annual Review of Plant Physiology, 24, 519–570.

Huberty, A. & Denno, R. (2004) Plant water stress and its consequences for herbivorous insects: a new synthesis. Ecology, 85, 1383–1398.

Hunter, M.D. (2001) Insect population dynamics meets ecosystem ecology: effects of herbivory on soil nutrient dynamics. Agricultural and Forest Entomology, 3, 77–84.

Huxman, T.E., Cable, J.M., Ignace, D.D., Eilts, J.A., English, N.B., Weltzin, J. & Williams, D.G. (2003) Response of net ecosystem gas exchange to a simulated

151

BIBLIOGRAPHY

precipitation pulse in a semi-arid grassland: the role of native versus non-native grasses and soil texture. Oecologia, 141, 295–305.

IPCC (2007) Climate Change 2007 – The Physical Science Basis: Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (eds S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, and H.L. Miller). Cambridge University Press, Cambridge, UK.

IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds T. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P.M. Midgley). Cambridge University Press, Cambridge, UK.

Jamieson, M.A., Trowbridge, A.M., Raffa, K.F. & Lindroth, R.L. (2012) Consequences of climate warming and altered precipitation patterns for plant–insect and multitrophic interactions. Plant Physiology, 160, 1719–1727.

Johnson, S.N., Clark, K.E., Hartley, S.E., Jones, T.H., McKenzie, S.W. & Koricheva, J. (2012) Aboveground-belowground herbivore interactions: a meta-analysis. Ecology, 93, 2208–2215.

Johnson, S.N., Gherlenda, A.N., Frew, A. & Ryalls, J.M.W. (2016a) The importance of testing multiple environmental factors in legume–insect research: replication, reviewers, and rebuttal. Frontiers in Plant Science, 7, 2013–2016.

Johnson, S. & Jones, T. (2016) Emerging issues and future perspectives for global climate change and invertebrates. In Global Climate Change and Terrestrial Invertebrates, 1st ed. (eds S.N. Johnson & T.H. Jones) (in press).

Johnson, S.N. & McNicol, J.W. (2010) Elevated CO2 and aboveground–belowground herbivory by the clover root weevil. Oecologia, 162, 209–216.

Johnson, S.N., Ryalls, J.M.W. & Staley, J. (2016b) Impacts of climate and atmospheric change on aboveground–belowground invertebrate interactions. In Global Climate Change and Terrestrial Invertebrates, 1st ed. (eds S.N. Johnson & T.H. Jones) (in press).

152

BIBLIOGRAPHY

Johnson, S.N., Staley, J.T., McLeod, F.A.L. & Hartley, S.E. (2011) Plant-mediated effects of soil invertebrates and summer drought on above-ground multitrophic interactions. Journal of Ecology, 99, 57–65.

Jones, C.G. & Coleman, J.S. (1988) Plant stress and insect behaviour: cottonwood, ozone and the feeding and oviposition preference of a beetle. Oecologia, 76, 51–56.

Karl, T.R. & Trenberth, K.E. (2003) Modern global climate change. Science, 302, 1719– 1723.

Kasper, M.L., Reeson, A.F., Mackay, D.A. & Austin, A.D. (2008) Environmental factors influencing daily foraging activity of Vespula germanica (Hymenoptera, Vespidae) in Mediterranean Australia. Insectes Sociaux, 55, 288–295.

Karowe, D.N. & Schoonhoven, L.M. (1992) Interactions among three trophic levels – the influence of host plant on performance of Pieris brassicae and its parasitoid, Cotesia glomerata. Entomologia Experimentalis et Applicata, 62, 241–251.

Kemp, W.P. & Cigliano, M.M. (1994) Drought and rangeland grasshopper species diversity. The Canadian Entomologist, 126, 1075–1092.

Kimura, K., Shimozawa, T. & Tanimura, T. (1985) Water loss through the integument in the desiccation-sensitive mutant, Parched, of Drosophila melanogaster. Journal of Insect Physiology, 31, 573–580.

King, P., Pugh, P., Fordy, M.R., Love, N. & Wheeler, S.A. (1990) A comparison of some environmental adaptations of the littoral collembolans Anuridella marina (Willem) and Anurida maritima (Guérin). Journal of Natural History, 24, 673–688.

Klaiber, J., Najar-Rodriguez, A.J., Dialer, E. & Dorn, S. (2013) Elevated carbon dioxide impairs the performance of a specialized parasitoid of an aphid host feeding on Brassica plants. Biological Control, 66, 49–55.

Knapp, A.K., Fay, P.A., Blair, J.M., Collins, S.L., Smith, M.D., Carlisle, J.D., Harper, C.W., Danner, B.T., Lett, M.S. & McCarron, J.K. (2002) Rainfall variability, carbon cycling, and plant species diversity in a mesic grassland. Science, 298, 2202– 2205.

153

BIBLIOGRAPHY

Knepp, R.G., Hamilton, J.G., Mohan, J.E., Zangerl, A.R., Berenbaum, M.R. & DeLucia,

E.H. (2005) Elevated CO2 reduces leaf damage by insect herbivores in a forest community. New Phytologist, 167, 207–218.

Knepp, R.G., Hamilton, J.G., Zangerl, A.R., Berenbaum, M.R. & DeLucia, E.H. (2007)

Foliage of oaks grown under elevated CO2 reduces performance of Antheraea polyphemus (Lepidoptera: Saturniidae). Environmental Entomology, 36, 609–617.

Kopper, B.J. & Lindroth, R.L. (2003) Effects of elevated carbon dioxide and ozone on the phytochemistry of aspen and performance of an herbivore. Oecologia, 134, 95– 103.

Kopper, B.J., Lindroth, R.L. & Nordheim, E.V. (2001) CO2 and O3 effects on paper birch (Betulaceae: Betula papyrifera) phytochemistry and whitemarked tussock moth (Lymantriidae: Orgyia leucostigma) performance. Environmental Entomology, 30, 1119– 1126.

Koricheva, J., Mulder, C.P.H., Schmid, B., Joshi, J. & Huss-Danell, K. (2000) Numerical responses of different trophic groups of invertebrates to manipulations of plant diversity in grasslands. Oecologia, 125, 271–282.

Kremen, A.C., Colwell, R.K., Erwin, T.L., Murphy, D.D., Noss, R.F. & Sanjayan, M.A. (1993) Terrestrial arthropod assemblages: their use in conservation planning. Conservation Biology, 7, 796–808.

Lambeets, K., Maelfait, J.-P. & Bonte, D. (2008) Plasticity in flood-avoiding behaviour in two congeneric riparian wolf spiders. Animal Biology, 58, 389–400.

Lamarche, V.C., Graybill, D.A., Fritts, H.C. & Rose, M.R. (1984) Increasing atmospheric carbon dioxide: tree ring evidence for growth enhancement in natural vegetation. Science, 225, 1019–1021.

Lawler, I., Foley, W., Woodrow, I. & Cork, S. (1996) The effects of elevated CO2 atmospheres on the nutritional quality of Eucalyptus foliage and its interaction with soil nutrient and light availability. Oecologia, 109, 59–68.

Lee, M.A., Manning, P., Walker, C.S. & Power, S.A. (2014) Plant and arthropod community sensitivity to rainfall manipulation but not nitrogen enrichment in a successional grassland ecosystem. Oecologia, 176, 1173–1185.

154

BIBLIOGRAPHY

Lees, A.D. (1943) On the behaviour of wireworms of the genus Agriotes Esch. (Coleoptera, Elateridae): II. Reactions to moisture. Journal of Experimental Biology, 20, 54–60.

Lenhart, P.A., Eubanks, M.D. & Behmer, S.T. (2015) Water stress in grasslands: dynamic responses of plants and insect herbivores. Oikos, 124, 381–390.

LePage, M. (2016) Highest ever annual rise in carbon dioxide levels recorded. New Scientist: https://www.newscientist.com/article/2079995-highest-ever-annual-rise- in-carbon-dioxide-levels-recorded/.

Lincoln, D., Couvet, D. & Sionit, N. (1986) Response of an insect herbivore to host plants grown in carbon dioxide enriched atmospheres. Oecologia, 69, 556–560.

Lincoln, D., Sionit, N. & Strain, B. (1984) Growth and feeding response of Pseudoplusia includens (Lepidoptera: Noctuidae) to host plants grown in controlled carbon dioxide atmospheres. Environmental Entomology, 13, 1527–1530.

Lindroth, R.L. (2010) Impacts of elevated atmospheric CO2 and O3 on forests: phytochemistry, trophic interactions, and ecosystem dynamics. Journal of Chemical Ecology, 36, 2–21.

Lindroth, R.L. & Kinney, K.K. (1998) Consequences of enriched atmospheric CO2 and defoliation for foliar chemistry and gypsy moth performance. Journal of Chemical Ecology, 24, 1677–1695.

Lindroth, R.L. & Raffa, K.F. (2016) Experimental approaches for assessing invertebrate responses to global change factors. In Global Climate Change and Terrestrial Invertebrates, 1st ed. (eds S.N. Johnson & T.H. Jones) (in press).

Lindroth, R.L., Wood, S.A. & Kopper, B.J. (2002) Response of quaking aspen genotypes

to enriched CO2: foliar chemistry and tussock moth performance. Agricultural and Forest Entomology, 4, 315–323.

Loranger, G.I., Pregitzer, K.S. & King, J.S. (2004) Elevated CO2 and O3t concentrations differentially affect selected groups of the fauna in temperate forest soils. Soil Biology and Biochemistry, 36, 1521–1524.

155

BIBLIOGRAPHY

Louda, S.M. & Sharon, K. (1992) Plant resistance to insect herbivores: a field test of the environmental stress hypothesis. Ecology, 73, 153–169.

Lundberg, P., Ranta, E. & Kaitala, V. (2000) Species loss leads to community closure. Ecology Letters, 3, 465–468.

Majer, J.D. (1983) Ants: bio-indicators of minesite rehabilitation, land-use, and land conservation. Environmental Management, 7, 375–383.

Mariotte, P., Le Bayon, R.-C., Eisenhauer, N., Guenat, C. & Buttler, A. (2016) Subordinate plant species moderate drought effects on earthworm communities in grasslands. Soil Biology and Biochemistry, 96, 119–127.

Masters, G.J., Brown, V.K., Clarke, I.P., Whittaker, J.B. & Hollier, J.A. (1998) Direct and indirect effects of climate change on insect herbivores: Auchenorrhyncha (Homoptera). Ecological Entomology, 23, 45–52.

Mattson, W.J. (1980) Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics, 11, 119–161.

Meehan, T.D., Couture, J.J., Bennett, A.E. & Lindroth, R.L. (2014) Herbivore-mediated material fluxes in a northern deciduous forest under elevated carbon dioxide and ozone concentrations. New Phytologist, 204, 397–407.

Moise, E.R.D. & Henry, H.A.L. (2010) Like moths to a street lamp: Exaggerated animal densities in plot-level global change field experiments. Oikos, 119, 791–795.

Mondor, E.B., Tremblay, M.N., Awmack, C.S. & Lindroth, R.L. (2004) Divergent pheromone-mediated insect behaviour under global atmospheric change. Global Change Biology, 10, 1820–1824.

Mora, C., Tittensor, D.P., Adl, S., Simpson, A.G.B. & Worm, B. (2011) How many species are there on earth and in the ocean? PLoS Biology, 9, 1–8.

Moran, V.C. & Southwood, T.R.E. (1982) The guild composition of arthropod communities in trees. Journal of Animal Ecology, 51, 289–306.

Morgan, J.A., LeCain, D.R., Pendall, E., Blumenthal, D.M., Kimball, B.A., Carrillo, Y.,

Williams, D.G., Heisler-White, J., Dijkstra, F.A & West, M. (2011) C4 grasses

156

BIBLIOGRAPHY

prosper as carbon dioxide eliminates desiccation in warmed semi-arid grassland. Nature, 476, 202–205.

Morris, E.K., Caruso, T., Buscot, F., Fischer, M., Hancock, C., Maier, T.S., Meiners, T., Müller, C., Obermaier, E., Prati, D., Socher, S.A., Sonnemann, I., Wäschke, N., Wubet, T., Wurst, S. & Rillig, M.C. (2014). Choosing and using diversity indices: insights for ecological applications from the German Biodiversity Exploratories. Ecology and Evolution, 4, 3514–3524.

Muff, P., Kropf, C., Frick, H., Nentwig, W. & Schmidt-Entling, M.H. (2009) Co- existence of divergent communities at natural boundaries: spider (Arachnida: Araneae) diversity across an alpine timberline. Insect Conservation and Diversity, 2, 36–44.

Mundim, F. & Bruna, E. (2016) Is there a temperate bias in our understanding of how climate change will alter plant–herbivore interactions? A meta-analysis of experimental studies. The American Naturalist, 188, 74-89.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature, 403, 853–858.

Myers, J. & Post, B. (1981) Plant nitrogen and fluctuations of insect populations: a test with the cinnabar moth—tansy ragwort system. Oecologia, 48, 151–156.

Nakamura, A., Burwell, C.J., Ashton, L.A., Laidlaw, M.J., Katabuchi, M. & Kitching, R.L. (2015) Identifying indicator species of elevation: comparing the utility of woody plants, ants and moths for long-term monitoring. Austral Ecology, 41, 179– 188.

Nielsen, U.N. & Ball, B.A. (2015) Impacts of altered precipitation regimes on soil communities and biogeochemistry in arid and semi-arid ecosystems. Global Change Biology, 21, 1407–1421.

Newbold, T., Hudson, L.N., Arnell, A.P., Contu, S., Palma, A. De, Ferrier, S., Hill, S.L.L., Hoskins, A.J., Lysenko, I., Phillips, H.R.P., Burton, V.J., Chng, C.W.T., Emerson, S., Gao, D., Pask-Hale, G., Hutton, J., Jung, M., Sanchez-Ortiz, K., Simmons, B.I., Whitmee, S., Zhang, H., Scharlemann, J. & Purvis, A. (2016) Has

157

BIBLIOGRAPHY

land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353, 288–291.

Newman, J.A., Anand, M., Henry, H., Hunt, S. & Gedalof, Z. (2011) Climate Change Biology. CABI Publishing, Wallingford, UK.

Niu, S., Liu, W. & Wan, S. (2008) Different growth responses of C3 and C4 grasses to seasonal water and nitrogen regimes and competition in a pot experiment. Journal of Experimental Botany, 59, 1431–1439.

NOAA. (2013) Trends in Atmospheric Carbon Dioxide. URL http://www.esrl.noaa.gov/gmd/ccgg/trends/weekly.html [accessed 23 December 2013]

Nokelainen, O., Ripley, B.S., van Bergen, E., Osborne, C.P. & Brakefield, P.M. (2016)

Preference for C4 shade grasses increases hatchling performance in the butterfly, Bicyclus safitza. Ecology and Evolution, 6, 5246–5255.

Novotny, V., Basset, Y., Miller, S.E., Weiblen, G.D., Bremer, B., Cizek, L. & Drozd, P. (2002) Low host specificity of herbivorous insects in a tropical forest. Nature, 416, 841–844.

Nyffeler, M. & Breene, R.G. (1990) Evidence of low daily food consumption by wolf spiders in meadowland and comparison with other cursorial hunters. Journal of Applied Entomology, 110, 73–81.

Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H. & Wagner, H. (2015) vegan: Community Ecology Package. R package version 2.2-1. http://CRAN.R- project.org/package=vegan.

Olff, H. & Ritchie, M.E.M. (1998) Effects of herbivores on grassland plant diversity. Trends in Ecology & Evolution, 13, 261–265.

Osier, T.L., Hwang, S.Y. & Lindroth, R.L. (2000) Effects of phytochemical variation in quaking aspen Populus tremuloides clones on gypsy moth Lymantria dispar performance in the field and laboratory. Ecological Entomology, 25, 197–207.

158

BIBLIOGRAPHY

Owensby, C., Ham, J., Knapp, A. & Auen, L. (1999) Biomass production and species composition change in a tallgrass prairie ecosystem after long‐term exposure to

elevated atmospheric CO2. Global Change Biology, 5, 497–506.

Paine, T.D., Steinbauer, M.J. & Lawson, S.A. (2011) Native and exotic pests of eucalyptus: a worldwide perspective. Annual Review of Entomology, 56, 181–201.

Palmer, C.M. (2010) Chronological changes in terrestrial insect assemblages in the arid zone of Australia. Environmental Entomology, 39, 1775–1787.

Pan, Y., Birdsey, R.A., Fang, J., Houghton, R., Kauppi, P.E., Kurz, W.A., Phillips, O.L., Shvidenko, A., Lewis, S.L., Canadell, J.G., Ciais, P., Jackson, R.B., Pacala, S.W., McGuire, A.D., Piao, S., Rautiainen, A., Sitch, S. & Hayes, D. (2011) A large and persistent carbon sink in the world’s forests. Science, 333, 988–993.

Park, C.C. (2014) Acid Rain: Rhetoric and Reality. Reprint ed. Routledge, Abingdon, UK.

Pedigo, L.P. & Buntin, G.D. (eds). (1993) Handbook of Sampling Methods for Arthropods in Agriculture. CRC Press, Boca Raton, US.

Peng, R.K., Fletcher, C.R. & Sutton, S.L. (1992) The effect of microclimate on flying dipterans. International Journal of Biometeorology, 36, 69–76.

Peng, S., Piao, S., Shen, Z., Ciais, P., Sun, Z., Chen, S., Bacour, C., Peylin, P. & Chen, A. (2013) Precipitation amount, seasonality and frequency regulate carbon cycling of a semi-arid grassland ecosystem in Inner Mongolia, China: a modeling analysis. Agricultural and Forest Meteorology, 178, 46–55.

Percy, K., Awmack, C. & Lindroth, R. (2002) Altered performance of forest pests under

atmospheres enriched by CO2 and O3. Nature, 420, 403–408.

Perfecto, I., Vandermeer, J., Hanson, P. & Carti, V. (1997) Arthropod biodiversity loss and the transformation of a tropical agro-ecosystem. Biodiversity and Conservation, 6, 935–945.

Perner, J., Wytrykush, C., Kahmen, A., Buchmann, N., Egerer, I., Creutzburg, S., Odat, N., Audorff, V. & Weisser, W.W. (2005) Effects of plant diversity, plant productivity and habitat parameters on arthropod abundance in montane European grasslands. Ecography, 28, 429–442.

159

BIBLIOGRAPHY

Pik, A.J., Dangerfield, J.M., Bramble, R.A., Angus, C. & Nipperess, D.A. (2002) The use of invertebrates to detect small-scale habitat heterogeneity and its application to restoration practices. Environmental Monitoring and Assessment, 75, 179–199.

Pinder, J.E. & Kroh, G.C. (1987) Insect herbivory and photosynthetic pathways in old- field ecosystems. Ecology, 68, 254–259.

Pinto, D.M., Blande, J.D., Nykänen, R., Dong, W.X., Nerg, A.M. & Holopainen, J.K. (2007) Ozone degrades common herbivore-induced plant volatiles: does this affect herbivore prey location by predators and parasitoids? Journal of Chemical Ecology, 33, 683–694.

Pinto, D.M., Blande, J.D., Souza, S.R., Nerg, A.-M. & Holopainen, J.K. (2010) Plant

volatile organic compounds (VOCs) in ozone (O3) polluted atmospheres: the ecological effects. Journal of Chemical Ecology, 36, 22–34.

Pinto, D.M., Himanen, S.J., Nissinen, A., Nerg, A.-M. & Holopainen, J.K. (2008) Host location behavior of Cotesia plutellae Kurdjumov (Hymenoptera: Braconidae) in ambient and moderately elevated ozone in field conditions. Environmental Pollution, 156, 227–231.

Piper, S.D., Catterall, C.P., Kanowski, J.J. & Proctor, H.C. (2009) Biodiversity recovery during rainforest reforestation as indicated by rapid assessment of epigaeic ants in tropical and subtropical Australia. Austral Ecology, 34, 422–434.

Plum, N. (2005) Terrestrial invertebrates in flooded grassland: a literature review. Wetlands, 25, 721–737.

Pocock, M.J.O., Evans, D.M. & Memmott, J. (2012) The robustness and restoration of a network of ecological networks. Science, 335, 973–977.

Polis, G. (1998) Ecology: Stability is woven by complex webs. Nature, 395, 744–745.

Porter, J.H., Parry, M.L. & Carter, T.R. (1991) The potential effects of climatic change on agricultural insect pests. Agricultural and Forest Meteorology, 57, 221–240.

Potts, S.G., Biesmeijer, J.C., Kremen, C., Neumann, P., Schweiger, O. & Kunin, W.E. (2010) Global pollinator declines: trends, impacts and drivers. Trends in Ecology and Evolution, 25, 345–353.

160

BIBLIOGRAPHY

Power, S.A., Barnett, K.L., Ochoa-Hueso, R., Facey, S.L., Gibson-Forty, E.V-J., Hartley, S.E., Nielsen, U.N., Tissue, D.T. & Johnson, S.N. (2016) DRI-Grass: a new experimental platform for addressing grassland ecosystem responses to future precipitation scenarios in south-east Australia. Frontiers in Plant Science (accepted).

Price, P.W. (1991) The plant vigor hypothesis and herbivore attack. Oikos, 62, 244–251.

Price, P.W. (2002) Resource-driven terrestrial interaction webs. Ecological Research, 17, 241–247.

Pritchard, S.G., Rogers, H.H., Prior, S.A. & Peterson, C.M. (1999) Elevated CO2 and plant structure: a review. Global Change Biology, 5, 807–837. van der Putten, W.H., de Ruiter, P.C., Martijn Bezemer, T., Harvey, J.A., Wassen, M. & Wolters, V. (2004) Trophic interactions in a changing world. Basic and Applied Ecology, 5, 487–494.

R Core Team. (2015) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Raffaelli, D. (2004) How extinction patterns affect ecosystems. Science, 306, 1141–1142.

Read, J.L. & Andersen, A.N. (2000) The value of ants as early warning bioindicators: responses to pulsed cattle grazing at an Australian arid zone locality. Journal of Arid Environments, 45, 231–251.

Riechert, S. & Tracy, C. (1975) Thermal balance and prey availability: bases for a model relating web-site characteristics to spider reproductive success. Ecology, 56, 265– 284.

Robinson, E.A., Ryan, G.D. & Newman, J.A. (2012) A meta-analytical review of the

effects of elevated CO2 on plant–arthropod interactions highlights the importance of interacting environmental and biological variables. New Phytologist, 194, 321– 336.

Romo, C. & Tylianakis, J. (2013) Elevated temperature and drought interact to reduce parasitoid effectiveness in suppressing hosts. PLoS ONE, 8, e58136.

161

BIBLIOGRAPHY

Roth, S.K. & Lindroth, R.L. (1995) Elevated atmospheric CO2: effects on phytochemistry, insect performance and insect–parasitoid interactions. Global Change Biology, 1, 173–182.

Roth, S., Lindroth, R.L., Volin, J.C. & Kruger, E.L. (1998) Enriched atmospheric CO2 and defoliation: effects on tree chemistry and insect performance. Global Change Biology, 4, 419–430.

Ryan, G.D., Rasmussen, S. & Newman, J.A. (2010) Plant Communication from an Ecological Perspective (eds F. Baluška & V. Ninkovic). Springer, Berlin, Heidelberg.

Rypstra, A.A.L., Carter, P.P.E., Balfour, R.R.A. & Marshall, S.S.D. (1999) Architectural features of agricultural habitats and their impact on the spider inhabitants. Journal of Arachnology, 27, 371–377.

Saint-Germain, M., Buddle, C.M., Larrivée, M., Mercado, A., Motchula, T., Reichert, E., Sackett, T.E., Sylvain, Z. & Webb, A. (2007) Should biomass be considered more frequently as a currency in terrestrial arthropod community analyses? Journal of Applied Ecology, 44, 330–339.

Sanders, N.J., Belote, R.T. & Weltzin, J.F. (2004) Multitrophic effects of elevated

atmospheric CO2 on understory plant and arthropod communities. Environmental Entomology, 33, 1609–1616.

Sanders, N.J., Gotelli, N.J., Heller, N.E. & Gordon, D.M. (2003) Community disassembly by an invasive species. Proceedings of the National Academy of Sciences of the United States of America, 100, 2474–2477.

Savolainen, R. & Vepsäläinen, K. (1988) A competition hierarchy among boreal ants: impact on resource partitioning and community structure. Oikos, 51, 135–155.

Schäppi, B. & Körner, C. (1996) Growth responses of an alpine grassland to elevated

CO2. Oecologia, 105, 43–52.

Scheffers, B.R., Joppa, L.N., Pimm, S.L. & Laurance, W.F. (2012) What we know and don't know about Earth's missing biodiversity. Trends in Ecology & Evolution, 27, 501–510.

162

BIBLIOGRAPHY

Schnell, M.R., Pik, A.J. & Dangerfield, J.M. (2003) Ant community succession within eucalypt plantations on used pasture and implications for taxonomic sufficiency in biomonitoring. Austral Ecology, 28, 553–565.

Schowalter, T.D., Lightfoot, D.C. & Whitford, W.G. (1999) Diversity of arthropod responses to host-plant water stress in a desert ecosystem in southern New Mexico. The American Midland Naturalist, 142, 281–290.

Schowalter, T. (2006) Insect Ecology: An Ecosystem Approach. 2nd ed.. Academic Press, London, UK.

Schweitzer, J.A., Madritch, M.D., Bailey, J.K., Leroy, C.J., Fischer, D.G., Rehill, B.J., Lindroth, R.L., Hagerman, A.E., Wooley, S.C., Hart, S.C. & Whitham, T.G. (2008) From genes to ecosystems: the genetic basis of condensed tannins and their role in nutrient regulation in a Populus model system. Ecosystems, 11, 1005– 1020.

Shattuck, S.O. (1999) Australian Ants: Their Biology and Identification. CSIRO Publishing, Collingwood, Australia.

Slansky, F. & Wheeler, G. (1992) Caterpillars’ compensatory feeding response to diluted nutrients leads to toxic allelochemical dose. Entomologia Experimentalis et Applicata, 65, 171–186.

Slosser, J.E., Phillips, J.R., Herzog, G.A. & Reynolds, C.R. (1975) Overwinter survival and spring emergence of the bollworm in Arkansas. Environmental Entomology, 4, 1015–1024.

Speakman, J.R. (2005) Body size, energy metabolism and lifespan. Journal of Experimental Biology, 208, 1717–1730.

Speight, M.R., Hunter, M.D. & Watt, A.D. (1999) The Ecology of Insects: Concepts and Applications. Blackwell Science, Oxford, UK.

Staley, J.T., Mortimer, S.R., Morecroft, M.D., Brown, V.K. & Masters, G.J. (2007) Summer drought alters plant-mediated competition between foliar- and root- feeding insects. Global Change Biology, 13, 866–877.

163

BIBLIOGRAPHY

Staley, J.T., Mortimer, S.R. & Morecroft, M.D. (2008) Drought impacts on above– belowground interactions: do effects differ between annual and perennial host species? Basic and Applied Ecology, 9, 673–681.

Stange, G. (1997) Effects of changes in atmospheric carbon dioxide on the location of hosts by the moth, Cactoblastis cactorum. Oecologia, 110, 539–545.

Sternberg, M. (2000) Terrestrial gastropods and experimental climate change: a field study in a calcareous grassland. Ecological Research, 15, 73–81.

Stiling, P., Cattell, M., Moon, D.C., Rossi, A., Hungate, B.A., Hymus, G. & Drake, B.

(2002) Elevated atmospheric CO2 lowers herbivore abundance, but increases leaf abscission rates. Global Change Biology, 8, 658–667.

Stiling, P. & Cornelissen, T. (2007) How does elevated carbon dioxide (CO2) affect

plant–herbivore interactions? A field experiment and meta-analysis of CO2- mediated changes on plant chemistry and herbivore performance. Global Change Biology, 13, 1823–1842.

Stiling, P., Forkner, R. & Drake, B. (2010) Long-term exposure to elevated CO2 in a Florida scrub-oak forest increases herbivore densities but has no effect on other arthropod guilds. Insect Conservation and Diversity, 3, 152–156.

Stiling, P., Moon, D.C., Hunter, M.D., Colson, J., Rossi, A.M., Hymus, G.J. & Drake,

B.G. (2003) Elevated CO2 lowers relative and absolute herbivore density across all species of a scrub-oak forest. Oecologia, 134, 82–87.

Stiling, P., Moon, D., Rossi, A., Hungate, B.A. & Drake, B. (2009) Seeing the forest for

the trees: Long-term exposure to elevated CO2 increases some herbivore densities. Global Change Biology, 15, 1895–1902.

Stiling, P., Rossi, A. & Hungate, B. (1999) Decreased leaf-miner abundance in elevated

CO2: reduced leaf quality and increased parasitoid attack. Ecological Applications, 9, 240–244.

Stireman, J.O., Dyer, L.A., Janzen, D.H., Singer, M.S., Lill, J.T., Marquis, R.J., Ricklefs, R.E., Gentry, G.L., Hallwachs, W., Coley, P.D., Barone, J.A., Greeney, H.F., Connahs, H., Barbosa, P., Morais, H.C. & Diniz, I.R. (2005) Climatic unpredictability and parasitism of caterpillars: implications of global warming.

164

BIBLIOGRAPHY

Proceedings of the National Academy of Sciences of the United States of America, 102, 17384–17387.

Stork, N.E., McBroom, J., Gely, C. & Hamilton, A.J. (2015) New approaches narrow global species estimates for beetles, insects, and terrestrial arthropods. Proceedings of the National Academy of Sciences of the United States of America, 112, 7519–7523.

Suttle, K.B., Thomsen, M.A. & Power, M.E. (2007) Species interactions reverse grassland responses to changing climate. Science, 315, 640–642.

Sylvain, Z.A., Wall, D.H., Cherwin, K.L., Peters, D.P., Reichmann, L.G. & Sala, O.E. (2014) Soil animal responses to moisture availability are largely scale, not ecosystem dependent: insight from a cross-site study. Global Change Biology, 20, 2631–2643.

Tariq, M., Wright, D.J., Bruce, T.J.A & Staley, J.T. (2013) Drought and root herbivory interact to alter the response of above-ground parasitoids to aphid infested plants and associated plant volatile signals. PLoS ONE, 8, e69013.

Tariq, M., Wright, D., Rossiter, J.T. & Staley, J.T. (2012) Aphids in a changing world: testing the plant stress, plant vigour and pulsed stress hypotheses. Agricultural and Forest Entomology, 14, 177–185.

Taylor, S.H., Ripley, B.S., Martin, T., De-Wet, L.A., Woodward, F.I. & Osborne, C.P.

(2014) Physiological advantages of C4 grasses in the field: A comparative experiment demonstrating the importance of drought. Global Change Biology, 20, 1992–2003.

Teeri, J.A. & Stowe, L.G. (1976) Climatic patterns and the distribution of C4 grasses in North America. Oecologia, 23, 1–12.

Tilman, D. & El Haddi, A. (1992) Drought and biodiversity in grasslands. Oecologia, 89, 257–264.

Tobin, P.C., Parry, D. & Aukema, B.H. (2014) The influence of climate change on insect invasions in temperate forest ecosystems. In Challenges and Opportunities for the World’s Forests in the 21st Century, Forestry Sciences (ed. T. Fenning), pp. 267– 293. Springer, Dordrecht, Netherlands.

165

BIBLIOGRAPHY

Torode, M.D., Barnett, K.L., Facey, S.L., Nielsen, U.N., Tissue, D.T., Power, S.A. & Johnson, S.N. (2016) Altered precipitation impacts on above- and belowground grassland invertebrates: summer drought leads to outbreaks in spring. Frontiers in Plant Science (accepted).

Tscharntke, T. & Greiler, H.-J. (1995) Insect communities, grasses and grasslands. Antenna, 40, 535–558.

Tylianakis, J.M., Didham, R.K., Bascompte, J. & Wardle, D.A. (2008) Global change and species interactions in terrestrial ecosystems. Ecology Letters, 11, 1351–1363.

Valkama, E., Koricheva, J. & Oksanen, E. (2007) Effects of elevated O3, alone and in

combination with elevated CO2, on tree leaf chemistry and insect herbivore performance: a meta-analysis. Global Change Biology, 13, 184–201.

Vanbergen, A.J., Woodcock, B.A., Koivula, M., Niemelä, J., Kotze, D.J., Bolger, T., Golden, V., Dubs, F., Boulanger, G., Serrano, J., Lencina, J.L., Serrano, A., Aguiar, C., Grandchamp, A.C., Stofer, S., Szél, G., Ivits, E., Adler, P., Markus, J. & Watt, A.D. (2010) Trophic level modulates carabid beetle responses to habitat and landscape structure: a pan-European study. Ecological Entomology, 35, 226–235.

Vasconcelos, H.L., Leite, M.F., Vilhena, J.M.S., Lima, A.P. & Magnusson, W.E. (2008) Ant diversity in an Amazonian savanna: relationship with vegetation structure, disturbance by fire, and dominant ants. Austral Ecology, 33, 221–231.

Vasseur, L. & Potvin, C. (1998) Natural pasture community response to enriched carbon dioxide atmosphere. Plant Ecology, 135, 31–41.

Verhoef, H.A. & Witteveen, J. (1980) Water balance in Collembola and its relation to habitat selection; cuticular water loss and water uptake. Journal of Insect Physiology, 26, 201–208.

Villalpando, S.N., Williams, R.S. & Norby, R.J. (2009) Elevated air temperature alters an old-field insect community in a multifactor climate change experiment. Global Change Biology, 15, 930–942.

Villani, M.G., Allee, L.L., Diaz, A. & Robbins, P.S. (1999) Adaptive strategies of edaphic arthropods. Annual Review of Entomology, 44, 233–256.

166

BIBLIOGRAPHY

Vingarzan, R. (2004) A review of surface ozone background levels and trends. Atmospheric Environment, 38, 3431–3442.

Voigt, W., Perner, J., Davis, A.J., Eggers, T., Schumacher, J., Bährmann, R., Fabian, B., Heinrich, W., Köhler, G., Lichter, D., Marstaller, R. & Sander, F.W. (2003) Trophic levels are differentially sensitive to climate. Ecology, 84, 2444–2453.

Vuorinen, T., Nerg, A.M. & Holopainen, J.K. (2004) Ozone exposure triggers the emission of herbivore-induced plant volatiles, but does not disturb tritrophic signalling. Environmental Pollution, 131, 305–311.

Wagner, T., Neinhuis, C. & Barthlott, W. (1996) Wettability and contaminability of insect wings as a function of their surface sculptures. Acta Zoologica, 77, 213–225.

Walter, J., Hein, R., Auge, H., Beierkuhnlein, C., Löffler, S., Reifenrath, K., Schädler, M., Weber, M. & Jentsch, A. (2012) How do extreme drought and plant community composition affect host plant metabolites and herbivore performance? Arthropod– Plant Interactions, 6, 15–25.

Walther, G.-R. (2010) Community and ecosystem responses to recent climate change. Philosophical Transactions of the Royal Society of London, Series B, Biological Sciences, 365, 2019–2024.

Wang, W. & Fang, J. (2009) Soil respiration and human effects on global grasslands. Global and Planetary Change, 67, 20–28.

Ward, J.K., Tissue, D.T., Thomas, R.B. & Strain, B.R. (1999) Comparative responses of

model C3 and C4 plants to drought in low and elevated CO2. Global Change Biology, 5, 857–867.

Warne, R.W., Pershall, A.D. & Wolf, B.O. (2010) Linking precipitation and C3–C4 plant production to resource dynamics in higher-trophic-level consumers. Ecology, 91, 1628–1638.

Werling, B.P., Dickson, T.L., Isaacs, R., Gaines, H., Gratton, C., Gross, K.L., Liere, H., Malmstrom, C.M., Meehan, T.D., Ruan, L., Robertson, B.A., Robertson, G.P., Schmidt, T.M., Schrotenboer, A.C., Teal, T.K., Wilson, J.K. & Landis, D.A. (2014) Perennial grasslands enhance biodiversity and multiple ecosystem services

167

BIBLIOGRAPHY

in bioenergy landscapes. Proceedings of the National Academy of Sciences of the United States of America, 111, 1652–1657.

Whiles, M.R. & Charlton, R.E. (2006) The ecological significance of tallgrass prairie arthropods. Annual Review of Entomology, 51, 387–412.

White, T. (1969) An index to measure weather-induced stress of trees associated with outbreaks of psyllids in Australia. Ecology, 50, 905–909.

White, T.C.R. (1974) A hypothesis to explain outbreaks of looper caterpillars, with special reference to populations of Selidosema suavis in a plantation of Pinus radiata in New Zealand. Oecologia, 16, 279–301.

White, T.C.R. (1984) The abundance of invertebrate herbivores in relation to the availability of nitrogen in stressed food plants. Oecologia, 63, 90–105.

White, S.R., Carlyle, C.N., Fraser, L.H. & Cahill, J.F. (2012) Climate change experiments in temperate grasslands: Synthesis and future directions. Biology Letters, 8, 484–487.

Wilcox, K.R., Blair, J.M., Smith, M.D. & Knapp, A.K. (2016) Does ecosystem sensitivity to precipitation at the site-level conform to regional-scale predictions? Ecology, 97, 561–568.

Wilcox, K.R., von Fischer, J.C., Muscha, J.M., Petersen, M.K. & Knapp, A.K. (2015) Contrasting above- and belowground sensitivity of three Great Plains grasslands to altered rainfall regimes. Global Change Biology, 21, 335–344.

Williams, R.S., Lincoln, D.E. & Norby, R.J. (1998) Leaf age effects of elevated CO2- grown white oak leaves on spring-feeding lepidopterans. Global Change Biology, 4, 235–246.

Willmer, P.G. (1982) Microclimate and the environmental physiology of insects. Advances in Insect Physiology, 16, 1–57.

Wilson, E.O. (1987) The little things that run the world (the importance and conservation of invertebrates). Conservation Biology, 1, 344–346.

Wittig, V.E., Ainsworth, E.A., Naidu, S.L., Karnosky, D.F. & Long, S.P. (2009) Quantifying the impact of current and future tropospheric ozone on tree biomass,

168

BIBLIOGRAPHY

growth, physiology and biochemistry: a quantitative meta-analysis. Global Change Biology, 15, 396–424.

Woodcock, B.A. (2005) Pitfall trapping in ecological studies. In Insect Sampling in Forest Ecosystems (ed. S Leather). Blackwell Science Ltd, Oxford, UK.

Wu, Z., Dijkstra, P., Koch, G.W., Peñuelas, J. & Hungate, B.A. (2011) Responses of terrestrial ecosystems to temperature and precipitation change: a meta-analysis of experimental manipulation. Global Change Biology, 17, 927–942.

Xu, G.L., Fu, S.L., Schleppi, P. & Li, M.H. (2013) Responses of soil Collembola to long-

term atmospheric CO2 enrichment in a mature temperate forest. Environmental Pollution, 173, 23–28.

Zaller, J. & Arnone, J, (1999) Earthworm and soil moisture effects on the productivity and structure of grassland communities. Soil Biology and Biochemistry, 31, 517–523.

Zalucki, M.P., Clarke, A.R. & Malcolm, S.B. (2002) Ecology and behavior of first instar larval lepidoptera. Annual Review of Entomology, 47, 361–393.

Zhu, Z., Piao, S., Myneni, R.B., Huang, M., Zeng, Z., Canadell, J.G., Ciais, P., Sitch, S., Friedlingstein, P., Arneth, A., Liu, R., Mao, J., Pan, Y., Peng, S., Peñuelas, J., Poulter, B., Pugh, T.A.M., Stocker, B.D., Viovy, N., Wang, X., Wang, Y., Xiao, Z., Yang, H., Zaehle, S. & Zeng, N. (2016) Greening of the Earth and its drivers. Nature Climate Change, 6, 791–795.

Zhu, H., Wang, D., Wang, L., Fang, J., Sun, W. & Ren, B. (2014) Effects of altered precipitation on insect community composition and structure in a meadow steppe. Ecological Entomology, 39, 453–461.

Zimmer, M. (2002) Nutrition in terrestrial isopods (Isopoda: Oniscidea): an evolutionary-ecological approach. Biological Reviews of the Cambridge Philosophical Society, 77, 455–493.

Zvereva, E.L. & Kozlov, M. V. (2006) Consequences of simultaneous elevation of carbon dioxide and temperature for plant–herbivore interactions: a metaanalysis. Global Change Biology, 12, 27–41.

169

BIBLIOGRAPHY

Zvereva, E.L. & Kozlov, M. V. (2012) Sources of variation in plant responses to belowground insect herbivory: a meta-analysis. Oecologia, 169, 441–452.

170

Appendices

Appendix I – Chapter Three Supplementary Material

Table S3.1: a list of the total abundances of the groups identified in this study and their functional guild classifications (from Barker, 2004; CSIRO, 1991; Moran & Southwood, 1982; Zimmer, 2002). Those groups marked with an asterisk were identified to Subclass level only.

Sampling method Functional guild Groups identified Sticky trap Vacuum classification Diptera: (26569) (2904) - Sciaridae 10104 143 Detritivores Culicidae 5 1 Sanguivores Psychodidae 7847 182 Detritivores Syrphidae 2 0 Predators Asilidae 0 6 Predators Other Diptera 8610 2572 Omnivores Tipulidae 1 0 Detritivores Coleoptera: (1171) (2216) - Elateridae 84 2 Omnivores Carabidae 0 1 Predators Stapylinidae 162 15 Predators Cerambycidae 1 0 Detritivores Curculionidae 7 55 Chewing herbivores Cantharidae 3 3 Omnivores Coccinellidae 51 10 Predators Other Coleoptera 863 2130 Omnivores Hymenoptera: (6504) (4280) - Non-ant, non-bee Hymenoptera 6365 2930 Parasitoids Formicidae 139 1350 Scavengers Neuroptera 81 17 Predators Lepidoptera 69 265 Chewing herbivores Hemiptera: (1435) (3190) - Predatory Hemiptera (e.g. Reduviidae) 0 76 Predators Other Hemiptera 1435 3114 Sucking herbivores Thysanoptera 10985 239 Omnivores Orthoptera 2 741 Chewing herbivores Psocoptera 178 389 Detritivores Blattodea 7 2 Scavengers Mantodea 0 5 Predators Araneae 311 1293 Predators Acari* 9 6201 Scavengers Collembola* 0 4227 Detritivores TOTAL: 47321 25969 (73290) 171

Fig S3.1 (overleaf): soil water content and rain applied under the four watering regimes (increased amount, reduced amount, summer drought and reduced frequency) relative to ambient, over the duration of the experiment.

172

Fig S3.2: the mean abundance of invertebrates captured by sticky trapping, divided into Order groups, over the course of the experiment. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD. O = October, Ja = January, A = April, Ju = July.

174

Fig S3.3: the mean abundance of invertebrates captured by sticky trapping, divided into feeding groups, over the course of the experiment. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD. O = October, Ja = January, A = April, Ju = July.

175

Fig S3.4: the mean abundance of invertebrates captured by vacuum sampling, divided into Order groups, over the course of the experiment, scaled to 1 m2. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

176

Fig S3.5: the mean abundance of invertebrates captured by vacuum sampling, divided into feeding groups, over the course of the experiment, scaled to 1 m2. Significance stars denote significant watering treatment p-values from individual time point LM analyses (Tables 2 and 4): * p<0.05, ** p <0.01, *** p<0.001. Error bars show ± SE of the mean. AMB ambient rainfall; Increased Amount IA; Reduced Amount RA; Reduced Frequency RF; Summer Drought SD.

177

Appendix II – Chapter Five Supplementary Material

Table S5.1: a list of the total abundances of the groups identified in this study and their functional guild classifications (from Barker, 2004; CSIRO, 1991; Moran & Southwood, 1982; Zimmer, 2002). Those groups marked with an asterisk were identified to Subclass level only.

Total abundance Functional guild Groups identified ground- understorey aerial classification dwelling Diptera: (520 1576 14605) - Sciaridae 179 368 8144 Detritivores Mosquitoes 0 9 0 Sanguivores Psychodidae 1 7 38 Detritivores Syrphidae 0 0 29 Predators Asilidae 0 0 1 Predators Other Diptera 340 1192 6393 Omnivores Coleoptera: (342 224 1314) - Elateridae 8 0 12 Omnivores Carabidae 157 7 0 Predators Tenebrionidae 2 0 0 Scavengers Stapylinidae 45 32 109 Predators Scarabaeidae 23 0 0 Detritivores Curculionidae 66 32 66 Chewing herbivores Cantharidae 0 1 12 Omnivores Coccinellidae 0 0 72 Predators Other Coleoptera 41 152 1043 Omnivores Hymenoptera: (4833 1966 21711) - Non-ant Hymenoptera 290 637 21540 Parasitoids Formicidae 4543 1329 171 Scavengers Neuroptera 3 0 114 Predators Lepidoptera 9 22 54 Chewing herbivores Hemiptera: (209 673 2784) - Predatory Hemiptera (e.g. Reduviidae) 26 14 0 Predators Other (non-psyllid) Hemiptera 183 659 2784 Sucking herbivores Thysanoptera 9 715 7317 Omnivores Orthoptera 57 45 5 Chewing herbivores Psocoptera 4 23 51 Detritivores Blattodea 32 57 21 Scavengers Mantodea 0 1 4 Predators Lithobiomorpha 25 1 0 Predators Scolopendromorpha 15 0 0 Predators Chilognatha* 1 0 0 Detritivores Araneae 1644 598 1028 Predators Pseudoscorpiones 8 21 0 Predators Acari* 2191 10531 346 Scavengers Isopoda 1788 85 0 Detritivores Collembola* 2769 2615 562 Detritivores GRAND TOTAL: 83,528

178

Table S5.2: Results from multivariate permutational analysis (PERMANOVA) of the effect of CO2 treatment on community data from the three different niche-types, partitioned by Order identity and functional feeding guild classification (FG).

Niche type Community d.f. (treatment, SS MS Pseudo-F R2 P (perm) tested residual) Ground- dwelling invertebrates Order 1, 34 0.204 0.204 1.119 0.032 0.326 FG 1, 34 0.124 0.124 0.837 0.024 0.479 Understorey invertebrates Order 1, 28 0.068 0.068 0.372 0.013 0.823 FG 1, 28 0.055 0.055 0.323 0.011 0.812 Aerial invertebrates Order 1, 34 0.064 0.064 0.882 0.025 0.421 FG 1, 34 0.066 0.066 1.076 0.031 0.338

Fig S5.1 (overleaf): NMDS plots of arthropod community data in each of the three niche types, partitioned by functional guild classification and Order identity. Ambient CO2 samples are shown in white, with elevated CO2 in dark grey/black. Ellipses show the standard deviation around each community centroid. All sampling dates were included in the analysis. Pr Predators, Pa Parasitoids, Sc Scavengers, De Detritivores, Om Omnivores, Su Suckers, Ch Chewers, Sa Sanguivores; Di Diptera, Co Coleoptera, Ar Araneae, Ac Acarina, He Hemiptera, Th Thysanoptera, Bl Blattodea, Is Isopoda, Col Collembola, Hy Hymenoptera, Pse Pseudoscorpiones, Or Orthoptera, Le Lepidoptera, Li Lithobiomorpha, Ma Mantodea, Pso Psocoptera, Ne Neuroptera, Mi Millipedes (Chilognatha), Sc Scolopendromorpha. Stress values remained below 0.2 for all analyses, with k=3 dimensions.

179

Ground-dwelling Understorey Aerial

level

-

Guild

level level

-

Order

List of Publications

Peer-reviewed journal articles

Facey, S.L., Fidler, D.B., Rowe, R.C., Bromfield, L.M., Nooten, S.S., Staley, J.T., Ellsworth, D.S. & Johnson, S.N. (2017) Atmospheric change causes declines in woodland arthropods and impacts specific trophic groups. Agricultural & Forest Entomology, 19, 101–112.

Torode, M.D., Barnett, K.L., Facey, S.L., Nielsen, U.N., Tissue, D.T., Power, S.A. & Johnson, S.N. (2016) Altered precipitation impacts on above- and belowground grassland invertebrates: summer drought leads to outbreaks in spring. Frontiers in Plant Science, 7, 1468.

Power, S.A., Barnett, K.L., Ochoa-Hueso, R., Facey, S.L., Gibson-Forty, E.V-J., Hartley, S.E., Nielsen, U.N., Tissue, D.T. & Johnson, S.N. (2016) DRI-Grass: a new experimental platform for addressing grassland ecosystem responses to future precipitation scenarios in south-east Australia. Frontiers in Plant Science , 7, 1373.

Barnett, K.L. & Facey, S.L.* (2016) Grasslands, invertebrates and precipitation: a review of the effects of climate change. Frontiers in Plant Science, 7, 1196.

Johnson, S., Lopaticki, G., Barnett, K., Facey, S. & Hartley, S. (2015) An insect ecosystem engineer alleviates drought stress in plants without increasing plant susceptibility to an aboveground herbivore. Functional Ecology, 30, 894–902.

Facey, S.L., Ellsworth, D.S., Staley, J.T., Wright, D. & Johnson, S.N. (2014) Upsetting the order: how climate and atmospheric change affects herbivore-enemy interactions. Current Opinion in Insect Science, 5, 66–74.

Facey, S., Botham, M., Heard, M., Pywell, R. & Staley, J. (2014) Lepidoptera communities and agri-environment schemes; examining the effects of hedgerow cutting regime on diversity, abundance and parasitism. Insect Conservation and Diversity, 7, 543–552.

*denotes joint-first author contribution 181

Peer-reviewed book chapter

Facey, S.L. & Gherlenda, A.N. (2017) Forest invertebrate communities and atmospheric change. In: Global Climate Change and Terrestrial Invertebrates (eds Johnson, S.N. & Jones, T.H.). John Wiley & Sons, Chichester, UK.

Peer-reviewed conference proceedings

Barnett, K.B. & Facey, S.L.* (2016) 20 years of rain manipulation experiments with grassland invertebrates – what do we know? Proceedings of the Ninth Australian Conference on Grassland Invertebrate Ecology

Facey, S.L. & Torode, M.D. (2016) An assessment of the effect of sward height on suction sampling efficiency for the capture of grassland invertebrates using a G-Vac device. Proceedings of the Ninth Australian Conference on Grassland Invertebrate Ecology

*denotes joint-first author contribution 182