HERBIVORY OF THE COMMON BRUSHTAIL POSSUM

(TRICHOSURUS VULPECULA, MARSUPIALIA: PHALANGERIDAE)

AT DIFFERENT SCALES OF RESOURCE HETEROGENEITY

KAROLINA PETROVIĆ

Master of Science (Biology)

A thesis submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

September 2014

Faculty of Science

School of Environmental Sciences

Table of Contents

Certificate of Authorship…………………………………………………………………1

Acknowledgments………………………………………………………………………...2

Abstract…………………………………………………………………………………...5

Chapter 1 A world on a plate: Heterogeneity in ecology and animal nutrition……………………...7

General introduction and thesis overview………………………………………………...8

References……………………………………………………………………………….22

Chapter 2 Dietary composition of the common brushtail possum (Trichosurus vulpecula) in a

heterogeneous landscape: Choice determinants at an “all-you-can-eat buffet”………………………. .27

Introduction……………………………………………………………………………..28

Aims and hypotheses…………………………………………………………………….32

Study sites……………………………………………………………………………….33

Methods…………………………………………………………………………………39

Data analysis……………………………………………………………………………..48

Results…………………………………………………………………………………...56

Discussion……………………………………………………………………………….66

References……………………………………………………………………………….78

Chapter 3 Home range attributes of the common brushtail possum (Trichosurus vulpecula):

How does individual variability in resource use, food, and shelter influence spatial patterns? ………….87

Introduction……………………………………………………………………………..88

Aims and hypotheses…………………………………………………………………….92

Study sites……………………………………………………………………………….93

Methods…………………………………………………………………………………95

Data analysis…………………………………………………………………………....102

Results………………………………………………………………………………….106

Discussion……………………………………………………………………………...128

References……………………………………………………………………………...133

Chapter 4 Dietary choices of the common brushtail possum (Trichosurus vulpecula) in a chemically

complex environment: When not to take a green smoothie? ….………………………………….138

Introduction……………………………………………………………………………139

Aims and hypotheses…………………………………………………………………...144

Methods………………………………………………………………………………..145

Data analysis…………………………………………………………………………....159

Results………………………………………………………………………………….160

Discussion……………………………………………………………………………...180

References……………………………………………………………………………...187

Chapter 5 Herbivory in the heterogeneous world……………………………………………..194

Synthesis of results and conclusions……………………………………………………195

References……………………………………………………………………………...202

1

Certificate of Authorship

I hereby declare that this submission is my own work and that, to the best of my

knowledge and belief, it contains no material previously published or written by another

person nor material which to a substantial extent has been accepted for the award of any

other degree or diploma at Charles Sturt University or any other educational institution,

except where due acknowledgment is made in the thesis. Any contribution made to the

research by colleagues with whom I have worked at Charles Sturt University or elsewhere

during my candidature is fully acknowledged. I agree that this thesis be accessible for the

purpose of study and research in accordance with the normal conditions established by the

Executive Director, Library Services or nominee, for the care, loan and reproduction of

theses.

……………………………..

September 2014

2

Acknowledgments

In 2009, Gardner in her seminal work, “The Devolvement of Doctoral Students: Phases of

Challenge and Support,” proposed a conceptual framework for doctoral student identity

development, which is grounded in the idea of growth through challenges and the

provision of support. While my growth through challenges still awaits some empirical

testing, the endless support of my supervisors, colleagues, family, doctors, and friends is an

unquestionable fact.

I am deeply grateful to my supervisors, Assoc. Prof. David M. Watson and Assoc. Prof. Ian

Lunt, for assisting me on my rocky road to academic excellence and professional

fulfilment. You have oriented and supported me with promptness and care and have

always been patient and encouraging in times of new ideas and personal difficulties.

Dave, thank you for giving me this amazing opportunity of living and studying in Australia,

teaching me along the way how to be an independent researcher, critical thinker, and

writer. Ian, thank you for teaching me how to be a logical and organized scientist,

supporting me in all my moments of self-doubt and despair. Without you guys I would not

be here and that is an unquestionable fact!

I am very grateful to my dear colleagues, Simon McDonald and Deanna Duffy, from

Charles Sturt University’s Spatial Analysis Unit, who assisted me in the gruelling task of

data analysis allowing me to see ecological patterns in numbers and maps that my artsy

brain and a short-sighted naturalist eye could not see. I am also sincerely grateful to

Dr Min An from Charles Sturt University’s Environmental and Analytical Laboratories for

helping me get a Charles Sturt University research grant and bringing me all these sweet,

little gifts from China. Also, I am thankful to a talented, young Polish statistician Grzegorz

Golonka, who came up with a novel way of measuring mistletoe biomass.

I have been very privileged to get to know and collaborate with many great scientists

around Australia and learn from their research and experience. I am immensely grateful to

Dr Katherine Handasyde from the University of Melbourne for generously sharing her

practical knowledge of Aussie wildlife and letting me stay in her research station, where

I had a chance to hang around a bunch of awesome possum researchers.

Moreover, it has been a great honour to work with marsupial–eucalypt chemistry pioneers,

Prof. William Foley and Dr Ian Wallis, from the Australian National University.

3

Their technical excellence and tremendous grasp of experimental and theoretical problems

had a great impact on my thinking of herbivory.

Also, I have to acknowledge a great job done by Hugh McGregor, who analysed my

possum faecal samples, and the incredible tree climbing Kenyon brothers from Tree

Tactics, who collected leaf samples for me. Also, I am very grateful to all Charles Sturt

University undergrad and postgrad volunteers, who made my research easier and kept me

company during those long night hours spent chasing possums in the bush. I would also

like to thank landholders in the Strathbogie Ranges that kindly provided access to their

properties, shared their possum stories, and invited me into their homes. You guys really

made me feel a part of the community once when we got to the subject of local gossip.

Finally, I would like to thank my dear family and friends for their never ending faith in me,

constant encouragement, and support. I am incredibly grateful to my beautiful and smart

Polish mother, Katarzyna Piórkowska, for all her acts of kindness, generosity, and love for

me, standing beside me in my darkest hours, and for letting me hear true words of wisdom

from her partner, Salomon Cymerman. I am also immensely grateful to my loving Serbian

father, Ivan Petrovic, who came all way from Europe to help me survive in the Aussie bush

and drove me crazy with his fear of snakes and my driving.

Naming here all my beautiful Aussie, Kiwi, Indian, Indonesian, Polish, and Serbian friends

that supported and entertained me through all these years would take at least a page, but

you guys already know you are in my heart forever. I would especially like to thank here my

dear friends, Anna and Tom, who generously shared their home during the last stage of my

PhD, cooked me amazing food, and hooked me on Taiwanese pineapple cakes.

Last but not the least, I would like to thank all the staff from Karinya Clinic in Albury,

especially Dr. Paresh Ramjee, for giving me personal wisdom, strength, and hope that there

is always light at the end of the tunnel.

During my candidature I was supported by the Australian Government Endeavour Europe

Award and the International Postgraduate Research Scholarship, as well as the Charles

Sturt University Postgraduate Research Scholarship. I wish to express here my immense

gratitude to my supervisor, Assoc. Prof. David M. Watson, and Charles Sturt University’s

Research Office for their ongoing technical and material support during periods when I

experienced health problems.

4

Operational funding and resources for this research were generously provided by the

following organizations, to which I express my gratitude: Charles Sturt University’s

Postgraduate Operating Funds, Charles Sturt University Competitive Grant, and

Holsworth Wildlife Research Endowment. I am also very thankful to my generous Polish

grandmother, Helena Piórkowska, for leaving me an inheritance so I could support myself

during my candidature and concentrate fully on my PhD. The list could go on and on . . .

I love you all and, once again, thank you for your help and support through my phases of

challenge and personal growth.

5

Abstract

Resource heterogeneity has been a broadly discussed concept in ecology used to explain

adaptive responses of herbivores to the distribution, abundance, and nutritional quality of

forage, as well as the patchy distribution of other key resources (i.e., shelter) and

interactions with competitors and predators. However, few studies have considered that

generalist herbivores exhibit resource selectivity at different scales of environmental

heterogeneity. Australian arboreal marsupials represent a unique opportunity for exploring

herbivore nutrition as they represent a continuum from specialist to generalist herbivores,

live in habitats dominated by a single tree genus (Eucalyptus) of varying nutritional quality,

and rely on trees as a shelter and platform for facilitating interactions with conspecifics,

competitors, and predators.

In this study, resource selectivity of a generalist arboreal marsupial, the common brushtail

possum (Trichosurus vulpecula, Marsupialia: Phalangeridae), was examined in the context of

forage availability and nutritional quality at three different scales, landscape, home range,

and individual tree, with tree-dependant parasitic plants representing a special case of small-

scale resource heterogeneity. Furthermore, other factors associated with tree use by

common brushtail possums were explored, including availability of shelter (tree hollows)

and presence of competitors, the mountain brushtail possum (Trichosurus cunninghami).

Integrative field and laboratory techniques were employed, including spotlighting, radio-

tracking, and faecal content analysis. The chemical variability of trees and parasitic plants

was studied using a recently developed method for estimating the nutritional quality of

forage for herbivores or foliar concentrations of available nitrogen.

The present study has demonstrated that common brushtail possums had a broad diet,

including the most common, but chemically defended and hard-to-digest Eucalyptus species;

nitrogen-rich Acacia species; succulent parasitic plants, drooping mistletoe (Amyema pendula)

and cherry ballart (Exocarpos cupressiformis); and supplemented their diet with other plant

groups. However, animals were selective over multiple scales of resource heterogeneity,

ranging from across a landscape to individual home ranges to specific species of trees,

individual trees, and discrete patches of mistletoes within tree canopies. They preferred

parasitic plants over their tree hosts and selectively foraged on trees parasitized by

mistletoes. They used Eucalyptus species less than expected from their availability in the

habitat, and were found foraging on both Monocalyptus and Symphyomyrtus species of

differing nutritional quality.

6

Moreover, tree choices and space-use patterns of common brushtail possum were shaped

by the availability of hollows and the presence of mountain brushtail possums.

The current research has shown that studying herbivory over different spatial scales,

landscape, home range, and individual plants, allows general conclusions to be made about

the nature of herbivore nutrition, focusing not only on food availability and nutritional

quality, but also on a broader context of ecological interactions with conspecifics and

competitors. Moreover, this study has shown that focusing on a single measure of the

nutritional quality of forage for herbivores, available nitrogen, simplifies complex

interactions between multiple nutrients and herbivore-deterrents. Therefore, future studies

should encourage a more holistic approach when analysing availability and nutritional

quality of forage for herbivores.

7

CH

AP

TE

R

1

A world on a plate: Heterogeneity in ecology and animal nutrition

8

CHAPTER 1

A world on a plate: Heterogeneity in ecology and animal nutrition

Introduction

Plant–animal interactions

Antagonistic relationships between plants and animal consumers (herbivores) are one of

the most important ecological interactions in nature (Crawley, 1983; Huntly, 1991; Karban

& Baldwin, 1997). Plants and herbivores represent three-quarters of earth’s biodiversity and

biomass, ranging from microscopic phytoplankton and zooplankton to some of the largest

organisms on earth, the giant sequoia (Sequoiadendron giganteum) and bush elephant

(Loxodonta africana). Herbivores consume around 15% of the plant biomass produced

annually in temperate and tropical ecosystems, making herbivory the most common mode

of animal foraging and the major agent by which energy enters food webs (Agrawal, 2011;

Cyr & Face, 1993).

During approximately 410 million years of shared evolutionary history, plants have

developed an arsenal of physical and chemical defences against herbivores. Consequently,

plant defences have evolved due to long-term selection pressure by herbivores (Agrawal &

Fishbein, 2006; Johnson, 2011). Furthermore, plants respond to short-term threats

imposed by herbivores and can change their metabolic pathways by investing in the

production of chemical defences at the time of attack (Karban & Myers, 1989; Ohgushi,

2005). These induced defences can increase the resistance of the plant to herbivore attack

by reducing the preference for, or effect of, herbivores on the damaged plant. Some plants

harbor many natural enemies, whereas others just a few. Hence, the intensity of selection—

pairwise or diffuse—will depend upon the reliability of attack from a diverse array of

potential herbivores (Ehrlich & Raven, 1964; Janzen, 1980).

For herbivores consumption of plants is a great ecological advantage as plants are the most

abundant food in terrestrial and aquatic ecosystems (Cyr & Face, 1993; Wiegert & Owen,

1971). However, plants are a challenging food source for herbivores as reviewed by Moles

and colleagues (2013). Specifically, plants are considered to be of poor nutritional quality

(low in protein, carbohydrates, and other essential nutrients).

9

Moreover, plants possess physical defences that limit herbivore access and foliage

processing (e.g., hard cuticles, spines, hairs, and latex). Finally, plants have developed a

range of chemical defences that reduce the nutritional quality of the foliage, are hard to

digest (e.g., tannins and fibre), and can be potentially toxic to herbivores (e.g., different

groups of plant secondary metabolites).

Plant secondary metabolites, commonly known as toxins, impose high metabolic costs on

herbivores (Freeland & Janzen, 1974). Upon ingestion of food, plant toxins enter the

animal via absorption from gastrointestinal tract and exert their effects via lipophilicity.

Toxins can cause disruption of cell membrane functions, which lead to the induction of

cysts and tissue damage and can lead ultimately to animal death (Iason, 2005; Scheline,

1991). As a result, herbivore’s ingestion of a particular food type is limited by its ability to

detoxify and excrete specific plant secondary metabolites (Freeland & Janzen, 1974).

Furthermore, plant toxicity is closely linked to the potency of plant secondary metabolites

or the amount and type of toxins that are required to produce a given intoxicating effect on

herbivores. According to Iason and Villalba (2006), all plant secondary metabolites act on

herbivores in a dose-dependent fashion, although their functions, thresholds, and

maximum tolerable doses vary greatly depending on the species of animal.

Some plant secondary metabolites can act as both toxins and digestibility reducers and, in

some instances, as pharmaceuticals exerting beneficial effects on herbivore digestion

(Barbehenn & Constabel, 2011). For instance, tannins—the most ubiquitous plant

secondary metabolites in nature—are known to be toxic if present in high concentrations

for both invertebrate and vertebrate species (Aerts, Barry, & McNabb, 1999). Furthermore,

tannins can reduce the digestibility of forage for herbivores through their protein-binding

capacity or inhibition of enzymatic degradation (Iason, 2005). Nevertheless, the same

process that inhibits digestive protein degradation in ruminants can help ruminants obtain

more essential amino acids through the formation of protein complexes limiting

degradation of protein by rumen bacteria (Min et al., 2003). A new discipline of

Pharm-ecology recognizes the importance of both nutrients and plant secondary

metabolites (toxins and pharmaceuticals) in plant–herbivore interactions (Raubenheimer &

Simpson, 2009; Forbey et al., 2013).

10

To overcome chemical defences imposed by plants, herbivores have developed a series of

anatomical, physiological, and behavioural counter-adaptations in an evolutionary “arms

race” (Brodie III & Brodie Jr, 1999; Mello & Silva-Filho, 2002; Rhoades, 1985). For

instance, to sustain themselves in nutritionally poor environments, herbivores have

increased their body size and gut capacity, lowered their nutrient requirements in

comparison to predatory animals, and developed specialized dentition and detoxification

mechanisms to deal with tough and toxic foliage (Freeland & Janzen, 1974; Karasov &

Diamond, 1988; Van Soest, 1994). Furthermore, herbivores have developed broad diets to

avoid over-ingestion of specific toxins and to prevent overloading of individual

detoxification pathways (Dearing, Mangione, & Karasov, 2000; Marsh et al., 2006). Finally,

to maximize their nutrient acquisition herbivores have developed food preferences by

selecting foliage of particular groups (monocots vs. dicots), favouring certain species or

individuals of the same species, and even choosing individual plant parts that are of higher

nutritional quality (Crawley, 1983; Marquis, 1992).

Herbivory at different scales of resource heterogeneity

Herbivores are known to be selective and alter their movements and foraging behaviour

across different scales of resource heterogeneity, ranging from geographic regions and

landscapes to plant communities, species and individuals (Kotliar & Wiens, 1990; Senft et

al., 1987; Fig. 1). According to Senft et al. (1987), herbivores at a regional scale select

different landscapes based on their geophysical properties (i.e., geomorphology, proximity

to waterways, regional climate), as well as availability and distribution of resources (i.e.,

food and shelter), and adjust their life strategies accordingly from sedentary (home ranging)

to migratory. At a landscape scale, herbivores select foraging areas or plant communities

based on their availability, habitat properties (substrate, topography, and microclimate), and

presence of predators, competitors, and conspecifics. Finally, at a community scale,

herbivores select species, individuals, and plant parts based on their availability or

nutritional quality or both.

The concept of resource selectivity was first introduced by Johnson (1980), who focused

on comparing resource availability with resource usage by animal consumers and

introduced a classification system for defining resource selectivity. He described resource

abundance as the resource quantity in the environment measured independently of the

consumer and distinguished resource availability as the accessibility of the resource to the

consumer.

11

Resource availability is compared with resource usage, or the amount consumed by an

animal in a fixed period of time, to generate an overall understanding of the degree of

resource selectivity, by which the consumer selects a resource disproportionately to its

availability. Based on the concept of resource selectivity, the foraging preferences of

herbivores and other groups of consumers can be established on the premise that all foods

are offered on equal basis within a spatially defined area. The concept of resource selectivity

was further extended to incorporate foraging on different spatial scales (i.e., within

different habitats and animals’ home ranges) and across landscapes and regions (Searle et

al., 2005; Senft et al., 1987).

Furthermore, to recognize the hierarchical nature of resource selectivity, Johnson (1980)

introduced the concept of selection order. First-order selection is defined as the selection

of the physical and geographical range of a species. Within that distributional range,

second-order selection determines the home ranges of individuals and the location of social

groups. Third-order selection refers to the usage of different habitat components by

animals within their home ranges, of which fourth-order selection incorporates differential

usage of food items from those available. Hence, resource selectivity is a hierarchically

structured process (see O'Neill, Johnson, & King, 1989), where processes at different

spatial scales can influence the selection of a resource at one spatial scale. Therefore, it is

critical to test for selection at the appropriate scale(s) at which organisms select resources

and to identify how selection processes vary across scales.

12

Fig. 1 Different scales of resource selectivity by herbivores while foraging

13

Herbivory theories: Past, present and future

The effect of food availability on resource selection by herbivores (known as frequency

dependence sensu Johnson, 1980) incorporates at least two inconsistent theoretical

assumptions as recognized by Di Stefano and Newell (2008). The first assumption is

quantitative and implies the existence of a positive relationship between selection and

relative availability of resources in the habitat (positive frequency dependence). This

assumption is consistent with diet specialization as predicted by the optimal foraging theory

(Pyke, 1984; Pyke, Pulliam, & Charnov, 1977; Table 1). The second assumption is

qualitative and proposes that mixed diets are to be expected if consumption is constrained

by nutrient availability (the partial preference theory; Pulliam, 1975; Westoby, 1978) or if

eating a range of foods reduces the negative effects of plant toxins and other hard-to-digest

compounds (the detoxification limitation theory; Freeland & Janzen, 1974; Marsh et al.,

2006). In this case, a negative relationship between selection and relative availability is to be

expected (negative frequency dependence).The observed negative frequency dependence

can be linked to dietary constraints of herbivores.

Namely, to maximize nutrient intake, dietary preferences of herbivores are non-linearly

related to food availability and nutritional quality and rely also on animal sensory cues

(Forbes, 1998). Forage palatability can be defined as the outcome of herbivores’ specific

foraging behavior and involves complex interactions between the current nutritional state

of herbivores, their past experience with forage, and their perception of the trade-offs

between detrimental aspects of forage (i.e., toughness and toxicity of plant secondary

metabolites) against its nutritional quality (macro- and micro-nutrients). According to Iason

and Villalba (2006), the extent to which these processes determine foraging choices of

herbivores is limited by the ability of animals to experience consequences of their behaviors

and associate particular cues in foods with their specific effects in the body.

For example, Provenza (1996) argues that ruminants possess a degree of nutritional

wisdom in the sense that they generally select foods that meet their nutritional needs and

avoid foods that cause toxicosis. The author postulates that neurally mediated interactions

between the senses (i.e., taste and smell) and the viscera enable ruminants to sense the

consequences of food ingestion. Satiety occurs when animals ingest adequate kinds and

amounts of nutritious foods and acquire preferences (mild to strong) for foods that cause

satiety.

14

Unpleasant feelings of physical discomfort (i.e., malaise) are caused by excesses of nutrients

and toxins and by nutrient deficits in food; animals can acquire aversions to those types of

foods that cause this discomfort (the satiety hypothesis; Provenza et al., 2003).

Recently, Simpson and Raubenheimer (2001) have developed the Geometric Framework

for nutrition to deal with the complexity of nutrient–toxin interactions and their impacts

on herbivore foraging behavior and post-ingestive responses. In this framework both

nutrients and toxins can be accommodated as dimensions in a simple geometrical model

with differing animal intake targets and metabolic costs enabling a relatively simple

comparison between observed and predicted patterns of food consumption. The focus of

this framework is on observing animal free food choices when faced with diverse dietary

options scattered throughout a multi-dimensional space and assessing their rule of

compromise when inhibited from reaching the preferred intake target.

Finally, along with food availability and nutritional or toxic constraints, the intraspecific

and interspecific competition over limited food resources and the risk of predation have

been recognized as important factors shaping foraging patterns of herbivores (Lima & Dill,

1990; Chase et al., 2002). Generally, herbivores with overlapping diets exhibit greater

competitive interactions than species that partition resources due to competitive

displacement in the past and niche separation (Stewart et al., 2002). Predation can reduce

the intensity of competitive interactions among herbivores (Gurevitch, Morrison, &

Hedges, 2000), but can also represent an additional cost for foraging animals (Nersesian,

Banks, & McArthur, 2011). Namely, the predation risk alters how long herbivores are

prepared to forage in certain areas and how long animals spend in risky areas if forage is of

better nutritional quality. Hence, it has been recognized that herbivores are able to quantify,

compare, and balance these different, but proximate costs, altering their foraging behaviour

in the process.

15

Table 1 Key theories explaining foraging behaviour of herbivores

Theory description

Foraging consequences

Evolutionary consequences

Key references

Optimal foraging theory

The optimal diet maximizes the net

rate of energy or mass intake while

minimising food searching and

handling time.

Behavioural and physiological adaptations

to forage of low nutritional quality, ability

to utilize nutrients from a single species

more efficiently.

Increasing availability of a preferred

food in a habitat should lead to diet

specialisation as less preferred foods are

dropped out of the diet.

(Pyke, 1984; Pyke, Pulliam,

& Charnov, 1977)

Partial preference theory

The optimal diet is the outcome of

nutritional balancing of foods with

different nutritional content.

Behavioural and physiological adaptations

to forage of variable nutritional quality, less

efficient in utilizing nutrients from a single

species.

Meeting nutritional needs on low quality

food should lead to a generalist foraging

strategy irrespective of availability of

specific foods.

(Pulliam, 1975; Westoby,

1978)

Detoxification limitation theory

The optimal diet is the outcome of

detoxification processes of different

plant secondary metabolites or

toxins.

Behavioural and physiological adaptations

to forage containing different toxins, less

efficient in detoxifying plant secondary

metabolites from a single species.

Higher detoxification efficiency of

generalist herbivores is realized through

different detoxification pathways for

different plant secondary metabolites.

(Freeland & Janzen, 1974;

Marsh et al., 2006)

16

Herbivory in the Australian context

The concepts generally associated with herbivory, such as resource heterogeneity and

selectivity, have been extensively studied in the Australian context. Australian ecosystems

are dominated by nutritionally poor, highly toxic, and hard to digest Eucalyptus species with

associated arboreal marsupial fauna, including both specialist and generalist species

(DeGabriel, Moore, Marsh, & Foley, 2010; Moore, Wallis, Marsh, & Foley, 2004). Past

research has investigated how variability of abiotic factors (i.e., geological parent material,

climate, elevation, and topography) and biotic factors (i.e., tree species composition,

structure, and abundance) influenced populations of folivorous marsupials over large

spatial scales (i.e., first- and second-order selection; Braithwaite, Binns, & Nowlan, 1988;

Braithwaite, Dudzinski, & Turner, 1983; Braithwaite, Turner, & Kelly, 1984; Pausas,

Braithwaite, & Austin, 1995). Moreover, many studies have examined relationships

between the spatio-temporal variability of biotic factors (including tree availability and

nutritional quality) and marsupial habitat preferences and tree use within their home ranges

(i.e., third- and fourth-order selection DeGabriel, 2008; Moore & Foley, 2005; Moore,

Lawler, Wallis, Beale, & Foley, 2010).

Nutritional studies have found that the extent and structure of resource heterogeneity

determine how and when arboreal marsupials encounter different types of food as they

move through superficially “uniform” Eucalyptus stands (summarized by DeGabriel et al.,

2010). For instance, the strongest influence of resource heterogeneity on arboreal

marsupials is expressed through the chemical variability of Eucalyptus tree assemblages,

because closely related tree species can differ in their palatability (Moore & DeGabriel,

2012). Also, individual trees of the same species can vary greatly in nutrients and plant

secondary metabolites as can different tree parts (i.e., mature vs. young foliage; Edwards et

al., 1993; Lawler et al., 1998). The observed chemical variability may limit or enhance the

opportunity of arboreal marsupials to avoid intoxication by plant secondary metabolites

and their ability to assemble a balanced and nutritious diet (Moore & Foley, 2005; Scrivener

et al., 2004; Youngentob et al., 2011).

Nevertheless, many studies have identified the dietary choices of arboreal marsupials only

in controlled feeding trials in which captive animals were offered Eucalyptus foliage of

varying palatability (e.g., Cork, Hume, & Dawson, 1983; Marsh, Wallis, & Foley, 2003;

Wiggins, McArthur, McLean, & Boyle, 2003).

17

However, the factors that govern the foraging decisions of wild populations have received

less attention, due to difficulties in applying laboratory findings to the complex

environments of wild populations. Almost all studies known to the author, performed both

in laboratory and field conditions, have taken a “eucalypt-centric” approach by looking

only at the eucalypt component of diets of arboreal marsupials (summarized by DeGabriel

et al., 2010; Moore et al., 2004). However, Eucalyptus stands are usually intermixed with

Acacia species, forming a dense understorey and mid-stratum, and have well developed a

grassy and herbaceous ground layer. Both Acacia and understorey species have been

identified as a valuable food resource for arboreal marsupials (McArthur & Turner, 1997;

Shepherd, 1997; Irlbeck & Hume, 2003).

Moreover, what has been largely overlooked in previous studies of the foraging decisions

of arboreal marsupials is the occurrence of parasitic plants within their reach at a tree level.

Aerial parasites, such as native mistletoes (from Loranthaceae and Viscaceae families) and

root parasites (native cherries, Exocarpos spp.; Santalaceae), introduce another level of

resource heterogeneity for arboreal marsupials as they provide additional nutrients without

incurring costs, such as extra travelling or risk of predation. Furthermore, individual

variability and plasticity in foraging choices of Australian arboreal marsupials observed in

other marsupial species, such as the gracile mouse opossum from Brazil (Gracilinanus

microtarsus, Araújo et al., 2009), have been largely overlooked in previous studies.

Arboreal marsupials are dependent on trees not only for food, but also for shelter, with

tree limbs, bark, and hollows providing concealment from predators and protection from

adverse weather conditions (Gibbons & Lindenmayer, 2002). Furthermore, trees facilitate

social interactions and provide a platform on which competition and predation processes

operate (Lindenmayer, 1997; Nersesian, Banks, & McArthur, 2011). Rather than simple

pairwise interactions, individual trees can support multiple individuals of different arboreal

marsupials and their natural enemies, allowing entire food webs to play out within their

canopies. Therefore, the present study takes an integrative approach to explore a basis of

nutrition by looking at the foraging behaviour and tree use of a generalist herbivore, the

common brushtail possum Trichosurus vulpecula (Marsupialia: Phalangeridae; Fig. 2), in

heterogeneous environments of Australian eucalypt forests and woodlands.

18

Fig. 2 The common brushtail possum (Trichosurus vulpecula), photo by H. McGregor

19

The common brushtail possum is a model organism for studying foraging decisions of

herbivores as it represents the epitome of a generalist herbivore in its native range in

Australia and as an introduced species in New Zealand (summarized by Kerle, 2001).

Regardless of a generalist diet, common brushtail possums show clear foraging preferences

at multiple scales of resource heterogeneity and selectively use specific forest types, trees

and understorey species, individual plants and parts, and discrete patches of mistletoes

within tree canopies depending on their availability and nutritional quality (Loney et al.,

2006; McArthur, Goodwin, & Turner, 2000; Scrivener et al., 2004; Sessions & Kelly, 2002).

Furthermore, the current study moves away from nutritional studies done in the past that

attempted to explain foraging decisions of common brushtail possums by deterring effects

of different groups of plant secondary metabolites found in Eucalyptus foliage (e.g. terpenes,

cyanogenic glycosides, formylated phloroglucinol compounds (FPCs); summarized by

Moore et al., 2004). Instead this study focuses on the amount of nitrogen in foliage that is

readily available to animals and not bound with tannins (DeGabriel et al., 2008). Tannins

have been found previously to affect foraging decisions, physiological processes, and

reproductive success in common brushtail possums (Marsh, Wallis, & Foley, 2003;

DeGabriel et al., 2009). Studying the effects of tannins on food nutritional quality instead

of a variety of plant chemical defences against herbivores represents a somewhat a

restrictive approach; but, this approach allows easy cross-taxa comparisons (Eucalyptus vs.

Acacia vs. parasitic plants). This approach also reduces chemical noise resulting from

differences in the amount and type of plant secondary metabolites found in different

species and allows for a clear perception of food as a source of protein for herbivores.

20

Therefore, the overall research objectives of this study are as follows:

1) To determine food preferences, resource selection, and space use patterns of

common brushtail possums at three different scales of resource heterogeneity,

landscape, home range and individual tree-scale, with a special consideration of the

contribution of parasitic plants to small-scale resource heterogeneity.

2) To investigate between and within plant chemical heterogeneity by using a

common currency, available nitrogen, to measure the nutritional value of different

tree species and co-occurring parasitic plants for common brushtail possums.

3) To explore and compare underlying causes of tree use by common brushtail

possums across different spatial scales, taking into account food availability and

nutritional quality, as well as the availability of parasitic plants, tree hollows, and the

presence of competitors.

In the present study, the specific research aims are addressed in Chapters 2–4 while

Chapter 5 summarises and integrates the findings and suggests future research directions.

Chapters 2–4 outline the results of field investigations and laboratory analyses and have

been written as self-contained manuscripts to facilitate subsequent publication.

Consequently, each chapter contains a separate introduction, methods, results, discussion,

and reference list.

In Chapter 2, a landscape-scale study looked at resource selectivity by common brushtail

possums in the context of food availability (species composition of trees and parasitic

plants), hollow availability (used as shelter), as well as potential competition with a

sympatric species, the mountain brushtail possum (Trichosurus cunninghami). This study

focused on the overall coarse-grain patterns of resource selection by establishing the diet

breadth, food preferences, and tree use of the entire population of common brushtail

possums occurring in the study area. In this chapter, standard dietary techniques were used,

direct observations of possum foraging activity and faecal content analysis. This integrative

approach provided a broader nutritional context for the following chapters and enabled

comparisons of resource selectivity on finer scales, home range, and individual tree.

21

In Chapter 3, resource selection (i.e., food and shelter) and space use patterns of common

brushtail possums were explored to establish the home range size and foraging and

denning behavior of individual animals. Both nightly and daily tree sightings of nine

randomly selected individuals (3 males and 6 females) were recorded in a radio-tracking

survey to establish the total home range size for common brushtail possums. This

approach allowed identification of both trees used for foraging and trees used as denning

sites. This information was then depicted in a spatially explicit context to estimate foraging

and denning ranges for each individual possum. Possum food preferences were established

in a more direct way by observing animals foraging within their home ranges. This study

also provided background data for the following chapter by identifying trees used and

avoided by possums while foraging.

In Chapter 4, inter- and intra-specific chemical variability of foliage of the most common

native trees and parasitic plants available to possums within their home ranges was

measured. This approach allowed exploration of the finest scale of resource heterogeneity

affecting foraging decisions of arboreal herbivores. In this study, a recently developed

method by DeGabriel and colleagues (2007, 2009) was used for measuring nutritional value

of browse. It combines the effects of total nitrogen (a proxy of plant protein content),

tannins (protein-binding agents) and fibre (digestibility reducing component of cell walls)

into a single measure of available nitrogen to herbivores. This approach allowed for the use

of a single currency to measure quality of browse for herbivores across different species

and taxa with otherwise unique chemical signatures.

In summary, to understand dietary choices of herbivores in chemically complex

environments, it is important to explore the variability in quantity and quality of forage

available to animals. By knowing what represents a “good” and “easily accessible” food for

herbivores and coupling this knowledge with information gathered from faecal content

analysis and direct observations of animal foraging behavior, we are able to establish the

dietary breadth and foraging preferences of herbivores. Furthermore, studying herbivory

over different spatial scales, landscape, home range, and individual plant, allows for general

conclusions to be made about the nature of herbivory. Finally, it is possible to understand a

broader context of animal nutrition by focusing not only on direct measures of food

availability and nutritional quality, but also on indirect measures of ecological interactions

of herbivores with their conspecifics, predators, and competitors.

22

References

Aerts, R. J., Barry, T. N., & McNabb, W. C. (1999). Polyphenols and agriculture: Beneficial effects of proanthocyanidins in forages. Agriculture, Ecosystems & Environment, 75(1), 1–12.

Agrawal, A. A. (2011). Current trends in the evolutionary ecology of plant defence. Functional Ecology, 25(2), 420–432.

Agrawal, A. A., & Fishbein, M. (2006). Plant defense syndromes. Ecology, 87(7), 132–149. Araújo, M. S., Martins, E. G., Cruz, L. D., Fernandes, F. R., Linhares, A. X., Dos Reis, S.

F., & Guimarães, P. R. (2009). Nested diets: A novel pattern of individual‐level resource use. Oikos, 119(1), 81–88.

Armati, P. J., Dickman, C. R., & Hume, I. D. (Eds.) (2006). Marsupials. Cambridge: Cambridge University Press.

Bailey, D. W., Gross, J. E., Laca, E. A., Rittenhouse, L. R., Coughenour, M. B., Swift, D. M., & Sims, P. L. (2006). Mechanisms that result in large herbivore grazing distribution patterns. Journal of Range Management Archives, 49(5), 386–400.

Barbehenn, R. V., & Peter Constabel, C. (2011). Tannins in plant–herbivore interactions. Phytochemistry, 72(13), 1551–1565.

Belovsky, G. E. (1986). Generalist herbivore foraging and its role in competitive interactions. American Zoologist, 26(1), 51–69.

Braithwaite, L. W., Binns, D. L., & Nowlan, R. D. (1988). The distribution of arboreal marsupials in relation to eucalypt forest types in the Eden (NSW) woodchip concession area. Australian Wildlife Research, 15(4), 363–373.

Braithwaite, L. W., Dudzinski, M. L., & Turner, J. (1983). Studies on the arboreal marsupial fauna of eucalypt forests being harvested for woodpulp at Eden, N.S.W. II. Relationship between the fauna density, richness and diversity, and measured variables of the habitat. Australian Wildlife Research, 10(2), 231–247.

Braithwaite, L. W., Turner, J., & Kelly, J. (1984). Studies on the Arboreal Marsupial Fauna of Eucalypt Forests Being Harvested for Wood Pulp at Eden, N.s.w. Iii. Relationships Between Faunal Densities, Eucalypt Occurrence and Foliage Nutrients, and Soil Parent Materials. Australian Wildlife Research, 11(1), 41–48.

Brodie III, E. D., & Brodie Jr, E. D. (1999). Predator-prey arms races. BioScience, 49(7), 557–568.

Brophy, J. J., & Southwell, I. A. (2002). Eucalyptus chemistry. In J. J. W. Coppen (Ed.), Eucalyptus: The Genus Eucalyptus (pp. 102-161). London: Taylor & Francis.

Brown, K., Innes, J., & Shorten, R. (1993). Evidence that possums prey on and scavenge birds’ eggs, birds and mammals. Notornis, 40(3), 169–177.

Chase, J. M., Abrams, P. A., Grover, J. P., Diehl, S., Chesson, P., Holt, R. D., Richards, S. A., Nisbet, R. M., Case, T. J. (2002). The interaction between predation and competition: A review and synthesis. Ecology Letters, 5(2), 302–315.

Chilcott, M., & Hume, I. (1985). Coprophagy and selective retention of fluid digesta: Their role in the nutrition of the common ringtail possum, Pseudocheirus peregrinus. Australian Journal of Zoology, 33(1), 1–15.

Cork, S., Hume, I., & Dawson, T. (1983). Digestion and metabolism of a natural foliar diet (Eucalyptus punctata) by an arboreal marsupial, the koala (Phascolarctos cinereus). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology, 153(2), 181–190.

Cowan, P. (1990). Fruits, seeds, and flowers in the diet of brushtail possums, Trichosurus vulpecula, in lowland podocarp/mixed hardwood forest, Orongorongo Valley, New Zealand. New Zealand Journal of Zoology, 17(4), 549–566.

Cowan, P. E. (1989). A vesicular-arbuscular fungus in the diet of brushtail possums, Trichosurus vulpecula. New Zealand Journal of Botany, 27(1), 129–131.

23

Cowan, P. E., & Moeed, A. (1987). Invertebrates in the diet of brushtail possums, Trichosurus vulpecula, in lowland podocarp/broadleaf forest, Orongorongo Valley, Wellington, New Zealand. New Zealand Journal of Zoology, 14(2), 163–177.

Crawley, M. J. (1983). Herbivory. The dynamics of animal–plant interactions. Oxford: Blackwell Scientific Publications.

Cyr, H., & Face, M. L. (1993). Magnitude and patterns of herbivory in aquatic and terrestrial ecosystems. Nature, 361(6408), 148–150.

Dearing, M. D., Mangione, A. M., & Karasov, W. H. (2000). Diet breadth of mammalian herbivores: Nutrient versus detoxification constraints. Oecologia, 123(3), 397-405.

DeGabriel, J. L. (2008). Demographic constraints imposed by plant nutrients and secondary chemistry on the common brushtail possum (Trichosurus vulpecula Kerr.). (Ph. D. thesis), the Australian National University, Canberra.

DeGabriel, J. L., Wallis, I. R., Moore, B. D., & Foley, W. J. (2008). A simple, integrative assay to quantify nutritional quality of browses for herbivores. Oecologia, 156(1), 107–116.

DeGabriel, J. L., Moore, B. D., Foley, W. J., & Johnson, C. N. (2009). The effects of plant defensive chemistry on nutrient availability predict reproductive success in a mammal. Ecology, 90(3), 711–719.

DeGabriel, J., Moore, B. D., Marsh, K. J., & Foley, W. J. (2010). The effect of plant secondary metabolites on the interplay between the internal and external environments of marsupial folivores. Chemoecology, 20(2), 97–108.

Di Stefano, J., & Newell, G. R. (2008). Diet selection by the swamp wallaby (Wallabia bicolor): Feeding strategies under conditions of changed food availability. Journal of Mammalogy, 89(6), 1540–1549.

Edwards, P. B., Wanjura, W., & Brown, W. (1993). Selective herbivory by Christmas beetles in response to intraspecific variation in Eucalyptus terpenoids. Oecologia, 95(4), 551–557.

Ehrlich, P. R., & Raven, P. H. (1969). Differentiation of populations. Science, 165(3899), 1228–1232.

Eschler, B. M., Pass, D. M., Willis, R., & Foley, W. J. (2000). Distribution of foliar formylated phloroglucinol derivatives amongst Eucalyptus species. Biochemical Systematics and Ecology, 28(9), 813–824.

Evans, M. C. (1992). Diet of the brushtail possum Trichosurus vulpecula (Marsupialia: Phalangeridae) in central Australia. Australian Mammalogy, 15, 25–30.

Foley, W., Iason, G., & McArthur, C. (1999). Role of plant secondary metabolites in the nutritional ecology of mammalian herbivores: How far have we come in 25 years? In H.-J. Jung & G. C. Fahey (Eds.), Nutritional Ecology of Herbivores (pp. 130–209). Savoy, IL: American Society of Animal Science.

Forbes, J. M. (1998). Dietary awareness. Applied Animal Behavior Science, 57(3), 287–297. Forbey, J. S., Dearing, M. D., Gross, E. M., Orians, C. M., Sotka, E. E., & Foley, W. J.

(2013). A Pharm-Ecological Perspective of Terrestrial and Aquatic Plant-Herbivore Interactions. Journal of Chemical Ecology, 39(4), 1–16.

Fox, L. R., & Macauley, B. J. (1977). Insect grazing on Eucalyptus in response to variation in leaf tannins and nitrogen. Oecologia, 29(2), 145–162.

Freeland, W. J., & Janzen, D. H. (1974). Strategies in herbivory by mammals: The role of plant secondary compounds. The American Naturalist, 108(961), 269–287.

Freeland, W. J., & Winter, J. W. (1975). Evolutionary consequences of eating:Trichosurus vulpecula (Marsupialia) and the genus Eucalyptus. Journal of Chemical Ecology, 1(4), 439–455.

Gibbons, P., & Lindenmayer, D. (2002). Tree hollows and wildlife conservation in Australia. Collingwood: CSIRO.

24

Gurevitch, J., Morrison, J. A., & Hedges, L. V. (2000). The interaction between

competition and predation: A meta‐analysis of field experiments. The American Naturalist, 155(4), 435–453.

Hume, I. D. (1999). Marsupial nutrition. Cambridge: Cambridge University Press. Huntly, N. (1991). Herbivores and the dynamics of communities and ecosystems. Annual

Review of Ecology and Systematics, 22(1), 477–503. Iason, G. (2005). The role of plant secondary metabolites in mammalian herbivory:

Ecological perspectives. Proceedings of the Nutrition Society, 64(01), 123–131. Iason, G., & Villalba, J. (2006). Behavioral strategies of mammal herbivores against plant

secondary metabolites: The avoidance-tolerance continuum. Journal of Chemical Ecology, 32(6), 1115–1132.

Irlbeck, N. A., & Hume, I. D. (2003). The role of Acacia in the diets of Australian marsupials? A review. Australian Mammalogy, 25(2), 121–134.

Janzen, D. H. (1980). When is it coevolution. Evolution, 34(3), 611–612. Johnson, D. H. (1980). The comparison of usage and availability measurements for

evaluating resource preference. Ecology, 61(1), 65–71. Johnson, M. T. (2011). Evolutionary ecology of plant defences against herbivores.

Functional Ecology, 25(2), 305–311. Karasov, W. H., & Diamond, J. M. (1988). Interplay between physiology and ecology in

digestion. BioScience, 38(9), 602–611. Karban, R., & Myers, J. H. (1989). Induced plant responses to herbivory. Annual Review of

Ecology and Systematics, 20(1), 331–348. Karban, R., & Baldwin, I. T. (1997). Induced responses to herbivory. Chicago: University of

Chicago Press. Kavanagh, R. P., & Lambert, M. J. (1990). Food selection by the greater glider, Petauroides

volans: Is foliar nitrogen a determinant of habitat quality? Wildlife Research, 17(3), 285–299.

Kerle, A. (2001). Possums: The brushtails, ringtails and greater glider. Sydney: UNSW Press. Kotliar, N. B., & Wiens, J. A. (1990). Multiple scales of patchiness and patch structure:

A hierarchical framework for the study of heterogeneity. Oikos, 59(2), 253–260. Lawler, I. R., Foley, W. J., Eschler, B. M., Pass, D. M., & Handasyde, K. (1998).

Intraspecific variation in Eucalyptus secondary metabolites determines food intake by folivorous marsupials. Oecologia, 116(1), 160–169.

Lima, S. L., & Dill, L. M. (1990). Behavioral decisions made under the risk of predation: A review and prospectus. Canadian Journal of Zoology, 68(4), 619–640.

Lindenmayer, D., Cunningham, R., Tanton, M., Smith, A., & Nix, H. (1990). Habitat requirements of the mountain brushtail possum and the greater glider in the montane ash-type eucalypt forests of the central highlands of Victoria. Wildlife Research, 17(5), 467–478.

Lindenmayer, D. B. (1997). Differences in the biology and ecology of arboreal marsupials in forests of southeastern Australia. Journal of Mammalogy, 78(4), 1117–1127.

Logan, M. (2001). Evidence for the occurrence of rumination‐like behaviour, or merycism, in the koala (Phascolarctos cinereus, Goldfuss). Journal of Zoology, 255(1), 83–87.

Marquis, R. J. (1992). The selective impact of herbivores. In R. S. Fritz & E. L. Simms (Eds.), Plant Resistance to Herbivores and Pathogens: Ecology, Evolution, and Genetics (pp. 301–325). Chicago: University of Chicago Press.

Marsh, K. J., Wallis, I. R., & Foley, W. J. (2003). The effect of inactivating tannins on the intake of Eucalyptus foliage by a specialist Eucalyptus folivore (Pseudocheirus peregrinus) and a generalist herbivore (Trichosurus vulpecula). Australian Journal of Zoology, 51(1), 31–42.

Marsh, K. J., Wallis, I. R., Andrew, R. L., & Foley, W. J. (2006). The detoxification limitation hypothesis: Where did it come from and where is it going? Journal of Chemical Ecology, 32(6), 1247–1266.

25

McArthur, C., & Turner, S. (1997). Feeding preferences of captive brushtail possums for eucalypt and acacia foliage. Tasmanian Forests, 9, 155–162.

Mello, M. O., & Silva-Filho, M. C. (2002). Plant-insect interactions: An evolutionary arms race between two distinct defense mechanisms. Brazilian Journal of Plant Physiology, 14(2), 71–81.

Min, B., Barry, T., Attwood, G., & McNabb, W. (2003). The effect of condensed tannins on the nutrition and health of ruminants fed fresh temperate forages: A review. Animal Feed Science and Technology, 106(1), 3–19.

Moles, A. T., Peco, B., Wallis, I. R., Foley, W. J., Poore, A. G., Seabloom, E. W., Vesk, P. A., Bisigato, A. J., Cella-Pizarro, L., Clark, C. J., Cohen, P. S., Cornwell, W. K., Edwards, W., Ejrnæs, R., Gonzales-Ojeda, T., Graae, B. J., Hay, G., Lumbwe, F. C., Magana-Rodrıguez, B., Moore, B. D., Peri, P. L., Poulsen, J. R., Stegen, J. C., Veldtman, R., Zeipel, H., Andrew, N. R., Boulter, S. L., Borer, E. T., Cornelissen, J. H. C., Farji-Brener, A. G., DeGabriel, J. L., Jurado, E., Kyhn, L. A., Low, B., Mulder, C. P. H., Reardon-Smith K., Rodrıguez-Velazquez, J., De Fortier, A., Zheng, Z., Blendinger, P. G., Enquist, B. J., Facelli, J. M., Knight, T., Majer, J. D., Martınez-Ramos, M., McQuillan, P., & Hui, F. K. C. (2013). Correlations between physical and chemical defences in plants: Tradeoffs, syndromes, or just many different ways to skin a herbivorous cat? New Phytologist, 198(1), 252–263.

Moore, B. D., & DeGabriel, J. L. (2012). Integrating the effects of PSMs on vertebrate herbivores across spatial and temporal scales. In G. R. Iason, M. Dicke & S. E.

Hartley (Eds.), The Ecology of Plant Secondary Metabolites: From Genes to Global Processes (pp. 226–246). Cambridge: Cambridge University Press.

Moore, B. D., & Foley, W. J. (2000). A review of feeding and diet selection in koalas (Phascolarctos cinereus). Australian Journal of Zoology, 48(3), 317–333.

Moore, B. D., & Foley, W. J. (2005). Tree use by koalas in a chemically complex landscape. Nature, 435(7041), 488–490.

Moore, B. D., Lawler, I. R., Wallis, I. R., Beale, C. M., & Foley, W. J. (2010). Palatability mapping: A koala's eye view of spatial variation in habitat quality. Ecology, 91(11), 3165–3176.

Moore, B. D., Wallis, I. R., Marsh, K. J., & Foley, W. J. (2004). The role of nutrition in the conservation of the marsupial folivores of eucalypt forests. In D. Lunney (Ed.), Conservation of Australia’s forest fauna (2 ed., pp. 549–575). Mosman: Royal Zoological Society of New South Wales.

Nersesian, C. L., Banks, P. B., & McArthur, C. (2011). Titrating the cost of plant toxins against predators: Determining the tipping point for foraging herbivores. Journal of Animal Ecology, 80(4), 753–760.

Ohgushi, T. (2005). Indirect interaction webs: Herbivore-induced effects through trait change in plants. Annual Review of Ecology, Evolution, and Systematics, 36, 81–105.

O'Neill, R. V., Johnson, A., & King, A. (1989). A hierarchical framework for the analysis of scale. Landscape Ecology, 3(3), 193–205.

Pahl, L. (1987). Feeding-Behavior and Diet of the Common Ringtail Possum, Pseudocheirus-Peregrinus, in Eucalyptus Woodlands and Leptospermum Thickets in Southern Victoria. Australian Journal of Zoology, 35(5), 487–506.

Pausas, J. G., Braithwaite, L. W., & Austin, M. P. (1995). Modelling habitat quality for arboreal marsupials in the south coastal forests of New South Wales, Australia. Forest Ecology and Management, 78(1–3), 39–49.

Provenza, F. D. (1996). Acquired aversions as the basis for varied diets of ruminants foraging on rangelands. Journal of Animal Science, 74(8), 2010–2020.

Provenza, F. D., Villalba, J. J., Dziba, L. E., Atwood, S. B., & Banner, R. E. (2003). Linking herbivore experience, varied diets, and plant biochemical diversity. Small Ruminant Research, 49(3), 257–274.

26

Pulliam, H. R. (1975). Diet optimization with nutrient constraints. The American Naturalist, 109(970), 765–768.

Pyke, G. H. (1984). Optimal foraging theory: A critical review. Annual Review of Ecology and Systematics, 15(1), 523–575.

Pyke, G. H., Pulliam, H. R., & Charnov, E. (1977). Optimal foraging: A selective review of theory and tests. Quarterly Review of Biology, 52(2), 137–154.

Raubenheimer, D., & Simpson, S. J. (2009). Nutritional PharmEcology: Doses, nutrients, toxins, and medicines. Integrative and Comparative Biology, 49(3), 329–337.

Rhoades, D. F. (1985). Offensive-defensive interactions between herbivores and plants: Their relevance in herbivore population dynamics and ecological theory. The American Naturalist, 125(2), 205–238.

Rothman, J. M., Raubenheimer, D., & Chapman, C. A. (2011). Nutritional geometry: Gorillas prioritize non-protein energy while consuming surplus protein. Biology Letters, 7(6), 847–849.

Scheline, R. P. (1991). Handbook of Mammalian metabolism of Plant Compounds. Boca Raton: CRC Press.

Scrivener, N. J., Johnson, C. N., Wallis, I. R., Takasaki, M., Foley, W. J., & Krockenberger, A. K. (2004). Which trees do wild common brushtail possums (Trichosurus vulpecula) prefer? Problems and solutions in scaling laboratory findings to diet selection in the field. Evolutionary Ecology Research, 6(1), 77–87.

Searle, K. R., Thompson Hobbs, N., & Shipley, L. A. (2005). Should I stay or should I go? Patch departure decisions by herbivores at multiple scales. Oikos, 111(3), 417–424.

Senft, R. L., Coughenour, M. B., Bailey, D. W., Rittenhouse, L. R., Sala, O. E., & Swift, D. M. (1987). Large herbivore foraging and ecological hierarchies. BioScience, 37(11), 789–799.

Shepherd, K. A., Wardell-Johnson, G. W., Loneragan, W. A., & Bell, D. T. (1997). Diet of herbivorous marsupials in a Eucalyptus marginata forest and their impact on the understorey vegetation. Journal of the Royal Society of Western Australia, 80, 47–54.

Simpson, S. J., Raubenheimer, D. (2001). The geometric analysis of nutrient-allelochemical interactions: A case study using locusts. Ecology, 82(2), 422–39.

Stewart, K. M., Bowyer, R. T., Kie, J. G., Cimon, N. J., & Johnson, B. K. (2002). Temporospatial distributions of elk, mule deer, and cattle: Resource partitioning and competitive displacement. Journal of Mammalogy, 83(1), 229–244.

Van Soest, P. J. (1994). Nutritional ecology of the ruminant. Ithaca: Cornell University Press. Westoby, M. (1978). What are the biological bases of varied diets? American Naturalist,

112(985), 627–631. Wiegert, R., & Owen, D. (1971). Trophic structure, available resources and population

density in terrestrial vs. aquatic ecosystems. Journal of Theoretical Biology, 30(1), 69–81. Wiggins, N. L., McArthur, C., McLean, S., & Boyle, R. (2003). Effects of two plant

secondary metabolites, cineole and gallic acid, on nightly feeding patterns of the common brushtail possum. Journal of Chemical Ecology, 29(6), 1447–1464.

Youngentob, K., Wallis, I. R., Lindenmayer, D. B., Wood, J., Pope, M., & Foley, W. J. (2011). Foliage chemistry influences tree choice and landscape use of a gliding marsupial folivore. Journal of Chemical Ecology, 37(1), 71–84.

27

CH

AP

TE

R

2

Dietary composition of the common brushtail possum (Trichosurus vulpecula)

in a heterogeneous landscape: Choice determinants at an “all-you-can-eat buffet”

28

CHAPTER 2

Dietary composition of the common brushtail possum (Trichosurus vulpecula)

in a heterogeneous landscape: Choice determinants at an “all-you-can-eat buffet”

Introduction

Knowledge of what drives foraging choices of herbivores in the wild is essential for our

understanding of the multifaceted dynamics of animal–plant systems. Nutrition is

recognized as a major mechanism regulating the fitness and reproduction of individual

animals, as well as the growth and maintenance of entire populations of herbivores

(Crawley, 1983). The majority of herbivores are generalist foragers satisfying their

nutritional needs from a range of plant species available in the habitat and include such

diverse species as grasshoppers (Unsicker et al., 2008), North American pika (Ochotona

princeps; Dearing, 1996), moose (Alces alces; Belovsky, 1978), and mountain gorilla (Gorilla

gorilla beringei; Watts, 1984). However, plants vary spatially and temporally in concentrations

of nutrients and toxins forcing herbivores to make complex foraging decisions and limiting

their foraging choices to a subset of species available.

Recent studies with captive and domestic animals have revealed the strong effects of plant

nutrients and toxins on diet selection, digestive physiology, and reproductive success of

herbivores (summarized by DeGabriel et al., 2013; Provenza et al., 2009). Nevertheless,

extending our knowledge from the laboratory to the field has proved challenging, and the

existing field studies suffer from several inconsistent theoretical expectations. On one side,

the optimal foraging theory predicts a positive relationship between food selection and

availability in the habitat leading to diet specialization (Pyke, 1984; Pyke et al., 1977).

However, the partial preference theory postulates that a negative relationship and mixed

diets are to be expected if foraging decisions of herbivores are constrained by food nutrient

content (Pulliam, 1975; Westoby, 1974) or, as the detoxification limitation theory

concludes, if eating a range of foods reduces the negative effects of toxins (Freeland &

Janzen, 1974; Marsh et al., 2006a).

29

Ultimately, the foraging theory provides a unifying framework for all the existing theories

and predicts that wild animals will balance the costs and benefits of different foraging

behaviours and their effects on fitness depending upon the animals’ nutritional needs, the

nutritional value and spatial distribution of forage, as well as competition and the risk of

predation (Stephens & Krebs, 1987). Herbivores perceive the risk of predation and plant

toxins as costs to foraging and are able to quantify, compare, and balance these costs,

therefore, alter their foraging behaviour (Nersesian, Banks, & McArthur, 2011).

Furthermore, the availability and proximity of shelter acting as a refuge from predators may

affect herbivore foraging decisions (Duffy & Hay, 1991; Tuft et al., 2011). Herbivores may

manage a risk of predation by selectively foraging in areas close to safety, such as protective

covers, hollows, or borrows.

Moreover, foraging decisions of herbivores in the wild are not simply a trade-off between

nutrient acquisition and the regulation or avoidance of toxins; animal behaviour can be

affected by other ecological factors, such as competition over limited food resources

(Shipley, Forbey, & Moore, 2009). Competitive interactions between animals can lead to

niche partitioning and resource specialization with some herbivores having a limited

realized diet incorporating just a few plant species considered to be “difficult” or low in

nutrients and high in toxins (Robinson & Wilson, 1998). Generalist herbivores, on the

contrary, consume plant species that are wildly available and avoid those that are difficult to

digest or require specialized behavioural, morphological, or physiological adaptations. They

compete well with other herbivores if less difficult types of food are relatively abundant,

but represent poor competitors if plant availability is restricted to difficult types of food

(Shipley, Forbey, & Moore, 2009).

Studying herbivory in Australian ecosystems, dominated by a single genus Eucalyptus,

provides a unique opportunity to investigate foraging decisions of wild animals in a

“field laboratory” where animals are faced with a limited number of food choices. Namely,

despite being widely available, eucalypts are considered to be a difficult food for leaf-eating

insects and arboreal marsupials since they are rich in toxins and poor in nutrients (Fox &

Morrow, 1983; Noble, 1989; Moore et al., 2004a; Moore et al., 2004b; DeGabriel et al.,

2010). Furthermore, the chemical composition of eucalypt foliage changes depending on

the season, habitat, species, individual tree and leaf age (Moore et al., 2004c; Loney et al.,

2006; Andrew, et al., 2007).

30

The feeding experiments with captive arboreal marsupials have shown that animals are

sensitive to types and concentrations of eucalypt toxins and forage selectively on less

difficult species and individuals of the same species (Lawler et al., 1998; Wallis, Watson, &

Foley, 2000; Wiggins et al., 2006b). Nevertheless, these nutritional studies often fail to

establish food preferences of arboreal marsupials in the wild and to look beyond their

eucalypt-dominated diet.

Circumpassing limitations of the previous studies is especially relevant to this study; as

foraging strategies of arboreal marsupials represent a continuum from obligate specialists,

the koala (Phascolarctos cinereus) and greater glider (Petauroides volans), which feed almost

exclusively on foliage of Eucalyptus spp., to a facultative specialist, the common ringtail

possum (Pseudocheirus peregrinus), which supplements its eucalypt diet with foliage of other

tree species, to a facultative generalist, the common brushtail possum (Trichosurus vulpecula),

which forages on a wide variety of woody and seral species. Nonetheless, both specialist

and generalist arboreal marsupials develop food preferences and select some species,

usually high in nutrients and low in toxins, more often than would be expected from their

availability in the habitat. For instance, although koalas have been found to forage on some

120 eucalypt species, they usually use one or two species at a site (reviewed by Moore &

Foley, 2000).

By contrast, common brushtail possums due to their lower detoxification capacity

compared to specialist herbivores are compelled to feed on a variety of species (Marsh et

al., 2006b). To maximize nutrient intake and minimize the negative effects of toxins,

common brushtail possums are known to diet mix (incorporate different plant species and

individuals into their diet, (Wiggins et al., 2003) and diet switch (modify time intervals

between eating different types of food, (Wiggins et al., 2006a). Common brushtail possums

have been found to forage on Eucalyptus and Acacia trees and shrubs, parasitic plants,

different species of herbs and grasses (Evans, 1992; Fitzgerald, 1984; Freeland & Winter,

1975; How & Hillcox, 2000; Irlbeck & Hume, 2003), as well as exotic, ornamental and

orchard trees (Eymann, Herbert, & Cooper, 2006; Sweetapple, Fraser, & Knightbridge,

2004), flowers, pollen, seeds, and fruit (Cowan, 1990) and even some fungi and animal

material (Cowan, 1989).

A special case of selective foraging by arboreal marsupials is expressed at tree level with

animals choosing to feed on mistletoes, parasitic plants dependent on their tree hosts for

water and nutrients.

31

Acting as mineral sinks, mistletoes have nutrient-rich foliage that are considered to be less

chemically and physically defendant, hence, readily eaten by many species of folivores

(Watson, 2001; Watson, 2011). Anecdotal evidence and cafeteria trails with captive koalas

and common brushtail possums demonstrated that animals prefer feeding on mistletoes

over their eucalypt hosts (reviewed by Reid, 1997; Reid & Yan, 2000). Furthermore,

nutritional studies conducted in New Zealand noted serious declines in populations of

native mistletoes in areas where common brushtail possums were found extensively

foraging (Sweetapple et al., 2002; Sweetapple, 2008).

Apart from food availability and nutritional quality, other factors may affect foraging

decisions of arboreal marsupials, including inter- and intra-specific competition over food

and shelter resources and predator avoidance behaviours. For instance, two closely related

species, the common brushtail possum (Trichosurus vulpecula) and mountain brushtail

possum (T. cunninghami), are known to forage on both Eucalyptus and Acacia foliage; yet, the

former species feeds more often on Eucalyptus while the latter on Acacia foliage (Hume,

1999; Burchfield, Agar, & Hume, 2005). It remains to be tested whether this difference is a

result of competitive exclusion, resource partitioning between two sympatric species, or the

existence of inter-specific differences in morphological and physiological adaptations.

Likewise, arboreal marsupials can compete over trees with hollows used as shelter and

concealment from predators forcing animals to defend their trees from intruders or travel

farther from safety in a search for new foraging and denning sites (Smith & Lindenmayer,

1988).

In the current study, foraging behaviour of a wild population of common brushtail

possums (Trichosurus vulpecula) is investigated in relation to food and shelter availability and

competition with a sympatric species, the mountain brushtail possum (T. cunninghami).

In doing so, this study moves away from explaining foraging decisions of arboreal

marsupials using solely nutritional terms, the foliar concentrations of nutrients and toxins,

and explores complex ecological interactions to highlight the existing foraging patterns.

The questions associated with the scales of food availability, measures of diet breadth and

food preferences, as well as competition over food and shelter and a risk of predation are

given due attention. In subsequent chapters foraging decisions of common brushtail

possums are examined in the context of spatial organization of food and shelter resources

and nutritional value of forage.

32

Aims and hypotheses

1. To document diet breadth and food preferences of common brushtail possums

Hypothesis 1: Common brushtail possums are expected to have a mixed diet,

incorporating in addition to eucalypt foliage other species and their parts.

Hypothesis 2: Common brushtail possums show foraging preferences for certain plant

species irrespective of their availability in the habitat.

2. To investigate factors influencing tree use by common brushtail possums

Hypothesis 1: Common brushtail possums are expected to use certain tree species, as well

as trees parasitized by mistletoes more often than expected from their availability in the

habitat.

Hypothesis 2: Trees with hollows are expected to be used more often by common

brushtail possums than expected from their availability in the habitat.

Hypothesis 3: Tree use of common brushtail possums is expected to be affected by

occurrence patterns of a sympatric species, the mountain brushtail possum (T. cunninghami).

33

Study sites

The current landscape-scale study comprised a 70 km network of linear, roadside

vegetation corridors in the Strathbogie Ranges (36° 48´ S, 145° 45´ E) in north-eastern

Victoria, Australia (Fig. 1 and 2, Table 1). The Strathbogie Ranges plateau is a part of the

Victorian Highlands–Northern Fall Bioregion, which covers an area of 1500 km2 about 150

km north-northeast of Melbourne (Phillips et al., 2002). The plateau is part of a batholith

formed from the Middle to Upper Devonian granitic intrusion (370–390 mya) and consists

of a mildly dissected plateau of rolling to hilly tableland, 320–700 m above sea level

(Fig. 3). The plateau consists of coarse-grained granite (Hergt, Phillips, & Ely, 2002).

Reddish duplex or gradational soils have developed from the weathered granite, with gleyed

sandy gradational soils and gleyed loams along drainage lines. The plateau is a part of the

Broken and Goulburn River water catchments (Rundle & Rowe, 1974).

The closest weather-station to the study area is at the township of Strathbogie (35° 51’ 50”

S, 145° 44’ 51’’ E). Mean maximum annual temperature is 18.5°C (1974–2012), and mean

minimum annual temperature is 6.1°C. Mean annual precipitation is 967.4 mm

(1902–2013). February is the hottest and driest month with a mean maximum temperature

of 27.3°C, mean minimum of 11.7°C, and mean monthly precipitation of 44.8 mm.

Winters are cool and wet; July is the coldest month with a mean maximum temperature of

10.2°C and minimum of 1.6°C; mean maximum precipitation of 122.9 mm. Rain (> 1 mm)

falls for 81.4 days per year on average; and snow falls occasionally (Bureau of Meteorology,

Strathbogie 2013).

The pre-European native vegetation on the Strathbogie plateau consists of Herb-rich

Foothill Forest, with Shrubby Dry Forest on the lower slopes. Mountain Dry Woodland

occupies the upper slopes of the plateau, and grassy dry forest occurs along the major river

valleys. Riparian and swamp woodlands or scrubs (i.e., Perched Boggy Shrubland, Swampy

Woodland, and Swampy Riparian Woodland) occur in seasonally or permanently inundated

areas and along watercourses (Carr et al., 2006). Since the 1850s, the relatively high rainfall

and fertile soils have resulted in extensive clearing of the original vegetation for agriculture

and softwood production (Pinus radiata). This clearing produced a highly heterogeneous

landscape (Fig. 4) consisting of native vegetation (29%), pine plantations (5%), and cleared

pastoral land (66%, Martin & Martin, 2004).

34

The most intact remnants of the original vegetation are preserved in linear roadside

corridors (10–40 m wide) and discrete patches of forest on both private and public land.

Roadside vegetation remnants are believed not to have been logged for the last 100 years or

burnt for at least half a century.

Vegetation species composition and associated arboreal marsupial fauna

The surveyed network of roadside vegetation corridors consisted of a mosaic of Herb-rich

Foothill Forest in drier areas (Fig. 5) and Swampy Riparian Woodland in seasonally or

permanently inundated areas and along watercourses (Fig. 6). Herb-rich Foothill Forest

was dominated by Eucalyptus radiata (narrow-leaved peppermint), E. dives (broad-leaved

peppermint), and E. viminalis (manna gum) with occasional E. globulus subsp. bicostata

(Victorian blue gum) and E. obliqua (messmate). Key mid-stratum and understorey tree

species included Acacia dealbata (silver wattle), A. melanoxylon (blackwood) and Exocarpos

cupressiformis (cherry ballart; a root parasitic shrub or small tree in the Santalaceae family

often associated with Eucalyptus hosts).

Similar mid-stratum and understorey tree species occurred in Swampy Riparian Woodland,

with Eucalyptus camphora (mountain swamp gum) dominating in the upper-storey. In the

study area, two mistletoe species were found parasitizing branches of different Eucalyptus

and Acacia species, the dominant Amyema pendula (drooping mistletoe) and the less

abundant Muellerina eucalyptoides (creeping mistletoe, both Loranthaceae family).

The understorey of Herb-rich Foothill Forest was dominated by shrubby, herbaceous, or

grassy species from the families Fabaceae, Asteraceae, Poaceae, and austral bracken

(Pteridium esculentum) while Swampy Riparian Woodland was dominated by rushes and

sedges. The most common exotic species in the understorey was Rubus fruticosus (common

blackberry).

In the study area, Downes, Handasyde, and Elgar (1997) carried out a fauna composition

survey. Two closely related, generalist herbivores, the common brushtail possum

(Trichosurus vulpecula) and the mountain brushtail possum (T. cunninghami), were found to be

common in the study area. The eucalypt specialists, the common ringtail possum

(Pseudocheirus peregrinus), greater glider (Petauroides volans), and koala (Phascolarctos cinereus),

were widespread in the study area as was the acacia-dependant species, the sugar glider

(Petaurus breviceps). Potential predators in the study area included the European red fox

(Vulpes vulpes), feral cat (Felis catus), and a native raptor, the powerful owl (Ninox strenua).

35

Fig. 1 Location of study sites within the Strathbogie Ranges, north-eastern Victoria, Australia

36

Fig. 2 Landscape view of roadside vegetation corridors across the Strathbogie Ranges plateau (photo by K. Petrović)

Table 1 Length of surveyed roadside vegetation corridors in the study area

Road name

Road type

Total length (km)

Surveyed length (km)

Creek Junction Rd sealed road 13.2 12.0

Bonnie Doon Rd sealed road 9.3 7.0

Boundary Hill Rd sealed road 5.9 5.6

Harrys Creek Rd sealed road 8.8 6.8

Euroa-Strathbogie Rd sealed road 5.3 3.5

Spring Creek Rd sealed road 4.6 4.5

Brookleigh Rd unsealed road 4.5 4.1

McGearys Rd unsealed road 2.9 2.9

Mackrells Rd unsealed road 4.5 4.4

Ankers Rd sealed road 14.7 12.3

Tames Rd

sealed road

0.3

0.3

Total n. a. 74.0 63.4

37

Fig. 3 Digital elevation model of the study sites within the Strathbogie Ranges Fig. 4 Land use across the study sites (Primary classification, BRS 2010)

38

Fig. 5 Herb-rich Foothill Forest in roadside vegetation in the Strathbogie Ranges (photo by K. Petrović)

Fig. 6 Swampy Riparian Woodland in roadside vegetation in the Strathbogie Ranges (photo by K. Petrović)

39

Methods

Description and justification of methods used to establish diet breadth, food preferences, and tree use by

common brushtail possums

To investigate foraging behaviour and resource selection of arboreal marsupials and other

groups of herbivores it is important to employ reliable and replicable research methods.

For the purpose of robust dietary and habitat inferences, Manly and colleagues (2002)

categorized resource units as “available,” “used,” and “unused” by animals with resource

selection measured by comparing any two of the three possible combinations. However,

the current techniques that are used for measuring availability of forage to herbivores

impose some logistical challenges, such as the researcher’s a priori knowledge of the type of

resources that animals actually use. Moreover, identification of the resources used by

herbivores is especially difficult in the case of arboreal marsupials as they are

predominantly nocturnal species foraging high in tree canopies (Wayne et al., 2006).

Direct methods, such as catching animals regularly for assessment of mouth or fistula

content, might be difficult to carry out or stressful to animals. Laboratory analytical

methods for establishing the dietary composition of arboreal marsupials can be considered

unethical or prohibitively destructive and involve the killing of animals to examine their

stomach content (Sweetapple & Nugent, 1998). Contrarily, using non-lethal methods, such

as the faecal content analysis, can introduce a certain level of bias to results by

underestimating or overestimating proportions of different plant species in a diet (Dunnet,

Harvie, & Smit, 1973).

Vegetation survey

In this study, a systematic vegetation survey was carried out to determine the composition,

dominance, and structure of tree species, as well as to estimate the relative availability of

different tree species and mistletoes to common brushtail possums. This approach was

deemed to be more appropriate than random sampling as it allowed a large set of data to be

collected within a limited time frame across a large study area, including 70 km of roadside

vegetation corridors. Vegetation corridors were divided into 200 m long strips, 10–40 m

wide, depending on the width of a surveyed corridor (Fig. 7). These strips were located on

both sides of the road as home ranges of common brushtail possums are known to include

both roads and roadside vegetation corridors (Del Borrello, 2009).

40

A single tree at the start of each 200 m vegetation strip was identified as a “systematic tree”

for which the following information was collected: exact location, tree species, height,

diameter at breast height (DBH), and number of hollows and mistletoes. In total 577 trees

were described across the entire study area.

Fig. 7 Sampling design for establishing tree species availability, possum and mistletoe tree occurrence in the study area

To establish the relative availability of tree species and mistletoes to possums, the canopy

volume of 50 randomly selected trees parasitized by mistletoe was measured, as well as the

volume of 334 mistletoes. The exact location of all studied trees was measured with a hand-

held positioning device Garmin GPS 12 Personal Navigator (Garmin International, Inc.,

Olathe, KS, USA). Tree diameter and height were recorded using a hand-held laser

hypsometer (LaserAce®, MDL Laser Systems, Ltd., York, UK). To collect and store data

in the field, a Trimble® Nomad hand-held computer (Sunnyvale, CA, USA) installed with

ArcPad 8.0 Software integrated with ArcGIS 9.3 Software (2009 ESRI, Redlands, CA) was

used.

41

Techniques used for establishing diet breadth and tree use by common brushtail possums

The selection of resources by common brushtail possums was investigated using two

complementary techniques: the faecal content analysis of remnant cuticle fragments and

nightly spotlighting observations of animal foraging activity. These techniques are standard

methods recommended by Jones and Krockenberger (2007) for studying the foraging of

arboreal, nocturnal, and cryptic herbivores. The authors highlighted advantages and

disadvantages of each method. Namely, direct observations of foraging animals produce

the highest diversity of diet; the faecal content analysis requires the least time in the field;

and tree selection allows intraspecific measures of preference to be determined. On the

other hand, direct observations of foraging animals are time-consuming and difficult within

a dense forest; the faecal content analysis underestimates the importance of species with

fragile cuticles; and tree selection is not directly related to food intake and can be ascribed

to availability of hollows and mistletoes, presence of conspecifics, competitors, and

predators.

Faecal content analysis

The faecal content analysis is a standard technique for quantifying the dietary composition

of herbivores. The analysis relies on the premise that a plant cuticle, a waxy non-cellular

layer protecting a leaf from desiccation, is species specific. Cuticles carry the unique imprint

of leaf surface characteristics, such as a size, shape, and alignment of cell patterns, stomata,

and trachoma. This technique is commonly used in rangeland studies of grazing and

browsing behaviour of large herbivores and is considered to be a relatively reliable

technique for establishing the diet composition of arboreal folivorous marsupials (Ellis,

Carrick, Lundgren, Veary, & Cohen, 1999).

The main disadvantage of cuticular analysis is the differential digestion of various plant

species by herbivores, with species with fragile cuticles being totally or partially broken in

the digestion process; therefore, they are missed or under-represented in analysed samples

(Fitzgerald & Waddington, 1979). Different digestibility of plant cuticles means that

feeding trials with captive animals are required to derive correction factors to adjust the raw

data (Dunnet et al., 1973; Fitzgerald & Waddington, 1979; Vavra & Holechek, 1980).

However, Sweetapple and Nugent (2007) point out that the inevitable random sampling

error in the correction factors reduces statistical precision and suggests the stomach

content analysis as a sound alternative.

42

In the current study, it was technically unfeasible and prohibitively destructive to capture or

kill animals to establish their diet. Hence, proportional representation of some plant groups

(i.e., herbs, grasses and fruits) could have been underestimated while others, such as

sclerophyll foliage Eucalyptus and Acacia species, could have been overestimated in the diet

of possums. This fact was treated with caution when discussing results.

Collection of common brushtail possum faeces

To establish the diet breadth, food selection, and preferences of common brushtail

possums in the study area, animal faeces were collected opportunistically from live traps

located at sites where possums were observed foraging (Fig. 8). During two trapping

periods November–December 2009 and January–February 2010 (see Chapter 3 for details

of the used trapping method), 56 clumps of scats were collected, each of which was

deposited by a different animal. To ascertain independence of samples, cage traps were set

up only once in a given location. Moreover, where possible, each animal was identified

from ear tattoos made by previous researchers working on the same population

(Del Borrello, 2009). Only fresh faeces were collected for the analysis. Faeces were

considered fresh if they were moist, maintained a black coating, and fragments of plants

remained green. They were collected each morning during trap clearing, transported to a

field station, and were permanently frozen to avoid degradation.

43

Fig. 8 Faeces collection sites (photo of possum scats by P. Canty)

44

Microhistological analysis of plant fragments in faeces

To identify plant fragments remaining in possum faeces, a reference slide collection of leaf

cuticles was prepared for all native tree species and parasitic plants, as well as grasses and

herbs that were available to animals in the study area (Fig. 9). Reference slides were

prepared by cutting leaf edges to expose the mesophyll tissue within the cuticular layers.

An area approximately 10 mm x 10 mm was sectioned with a sharp scalpel. The section

was placed in 42 g/L sodium hypochlorite (domestic bleach) until the mesophyll was

sufficiently digested. The cuticles were then peeled apart and washed with distilled water to

remove all the residual sodium hypochlorite. Once rinsed, the cuticles were observed under

a dissecting microscope to check if all the mesophyll was removed. Samples were then

stained with 0.2 g/L gentian violet for 3 minutes. The cuticles were rinsed of excess stain in

distilled water and were mounted on a microscope slide with corn syrup. Prepared slides

were photographed with a Nikon digital camera at x10, x20 and x40 magnifications to

create a cuticle reference library.

To analyse the content of faecal samples, scats were prepared by loosely breaking two or

three randomly chosen scats from each clump of scats. The scats were placed in a small

bottle and covered with 42 g/L sodium hypochlorite (domestic bleach), then left to digest

for 12 to 24 hours until most plant fragments in scats were translucent (modified from Ellis

et al., 1999). The plant fragments were rinsed with distilled water and were then washed

through a 1000 μm sieve, followed by a 500 μm sieve. All particles that were too big for the

1000 μm sieve or too small for the 500 μm sieve were discarded. The remaining plant

fragments were considered large enough to be identified since other similar studies found

that the proportion of fragments compromising each species was consistent between sieve

fractions (Ellis et al., 1999; Tuft, Crowther, & McArthur, 2011).

Next, all remaining plant fragments were placed in a Petri dish containing gentian violet

(0.15 g/L) and stained for 3 minutes. The plant fragments were then tipped onto Whatman

filter paper (11 μm pore diameter) atop a Bruckner Funnel, with slight air suction, and were

rinsed with a stream of distilled water until excess stain had washed away. The fragments

were dried on a filter paper and mounted onto a microscope slide using corn syrup. To

analyse the cuticle content of each scat sample, starting from the edge of a slide, the first

100 clear cellular cuticles were identified under a compound light microscope. Species

accumulation curves indicated that 100 cuticle fragments per scat were adequate to detect

the majority of present species (Fig. 10).

45

Fig. 9 Microhistological leaf tissue structure (x20 magnification) of Acacia dealbata (silver wattle) and A. melonoxylon (blackwood), Eucalyptus radiata (narrow-leaved peppermint) and E. dives (broad-leaved peppermint), E. camphora (swamp gum) and E. viminalis (manna gum), E. globulus (Victorian blue gum) and E. obliqua (messmate), Amyema pendula (drooping mistletoe) and Exocarpos cupressiformis (cherry ballart; photo by H. McGregor)

46

Fig. 10 Example of species accumulation curves representing the number of plant cuticle fragments considered sufficient for detecting majority of plant species present within a faecal pellet of the common brushtail possum

Spotlighting survey

The spotlighting survey or direct observations of nocturnal activity of arboreal herbivores

involve locating animals within tree canopies using a hand-held lamp and making

behavioural observations through binoculars. Canopy spotlighting or scanning searches are

performed at night while walking along established transects. Animals are located through

reflected eye-shine, sudden movement, or vocalization. Known biases of this technique

include low detectability estimates and presence of false absences due to differences in

animal individual behaviour (i.e., light or observer shyness), dark colouration of animals,

the location of animal home ranges in relation to forest edges, and differences in tree

allometry or canopy structure (Bennett et al., 1991). To overcome some of these biases, the

spotlighting survey was conducted twice and a second independent observer was

employed.

To locate common brushtail possums in roadside vegetation, a spotlighting survey was

conducted in the spring and summer of 2009. To avoid possible differences in foraging

behaviour, sections of vegetation corridors connected to intact patches of forest or

neighbouring household gardens were omitted from the sampling. Two observers spent a

total of 140 hours of spotlighting, each using an 80-W spotlight and 8 x 30 binoculars,

starting at least 1 hour after sunset and walking at a steady pace of 1 km/h.

47

All individuals were recorded in each 200 m vegetation strip, with the locations of multiple

individuals cross-checked to avoid multiple records of the same individual. The following

data were collected for trees occupied by common brushtail possums: exact location, tree

species, height, and diameter at breast height (DBH), number of mistletoes and hollows. A

hollow was considered a suitable refuge for common brushtail possums if the entrance was

greater than 20 cm diameter since animals cannot enter smaller openings (Gibbons &

Lindenmayer, 2002). Moreover, to investigate the incidence of resource competition with a

sympatric species, the mountain brushtail possum (Trichosurus cunninghami), tree occupancy

and tree species characteristics were also recorded. The exact tree location was measured

with a hand-held positioning device Garmin GPS 12 Personal Navigator (Garmin

International, Inc., Olathe, KS, USA). Tree diameter and height were recorded using a

hand-held laser hypsometer (LaserAce®, MDL Laser Systems, UK). All sightings of

possums were plotted on a digitized base map of a study area with ArcGIS 9.3 Software

(2009 ESRI, Redlands, CA).

48

Data analysis

Estimation of available biomass to common brushtail possums

To establish dietary preferences of common brushtail possums in the study area it was

important to determine the amount of available leaf biomass in an ecologically meaningful

way. In the current study, it was technically unfeasible and prohibitively destructive to

directly measure the leaf biomass of trees and mistletoes since cutting a number of large

trees with high mistletoe loads and separately drying and weighing leaves of both trees and

mistletoes would be required. Instead, the leaf biomass of trees and mistletoes was

estimated using the calculated total aboveground biomass of trees and measured in the field

volumes of tree canopies and mistletoes. Consequently, the leaf biomass of trees and

mistletoes was expressed as a non-parametric measure of proportional availability of trees

and mistletoes to common brushtail possums in the study area.

For the purpose of this study a number of assumptions have been adopted. First, it was

assumed that the leaf biomass of trees was directly proportional to the total aboveground

biomass (Enquist & Niklas, 2002). Secondly, as it was technically unfeasible to estimate

directly mistletoe leaf biomass, the total volume of mistletoes was used instead and

compared with the total canopy volume of a subset of parasitized trees of the known

aboveground biomass. It was assumed that the total volume of tree canopies is directly

proportional to the total aboveground biomass (Dai et al., 2009).

Moreover, it is important to stress here that different species of trees with different

morphology and canopy structure may affect mistletoe traits, such as growth form and leaf

shape (Barlow & Wiens, 1977). Hence, for the purpose of this study it was adopted that

mistletoe leaf biomass is proportional to tree leaf biomass irrespective of tree species (after

Reid, Yan & Fittler, 1994).

To calculate the total aboveground biomass of trees in the study area, a generalized

allometric equation was employed that used tree diameter (DBH) as a predictor variable of

the total aboveground biomass:

lny = -2.3267 + 2.4855 lnx, where y is the aboveground biomass (kg) and x is tree DBH (cm)

49

This generalized allometric equation was developed by Keith, Barrett, and Keenan (2000)

for a suite of species growing in native sclerophyllous forests. The equation was derived by

combining predicted biomass values from specific allometric equations developed for

individual species or sites. The generalized allometric equation accounted for most of the

variation in biomass predicted for individual species or sites (R2 = 0.96).

To estimate proportional availability of mistletoes and different species of trees in the study

area, coefficient was introduced and expressed as the ratio of availability of mistletoes

( ) to availability of trees ( ):

Assuming that availability is directly proportional to volume, we can further express

coefficient as:

In the above equation is the total volume of mistletoes found parasitizing n

number of trees of the total canopy volume while is the average

volume of mistletoes per tree. is the average canopy volume of n number of trees.

In the present analysis, 531 (N) trees were selected systematically to represent the entire

population of trees in the study area, out of which 66 trees (n1) were found to be

parasitized by mistletoes. However, in this case volumes of tree canopies and mistletoes

were not known. Therefore, an independent sample of 50 trees parasitized by mistletoes

(n2) was selected randomly out of all parasitized trees in the study area, and volumes of tree

canopies and mistletoes were measured.

Next, coefficient was calculated for the 50 trees of the known canopy and mistletoe

volumes:

50

To calculate coefficient with 95% confidence intervals, the bootstrap procedure was

used. Namely, 50 trees corresponding to the number of trees of the known canopy and

mistletoe volumes were randomly drawn from the pool of 50 trees (with replacement).

This procedure was repeated 10,000 times.

To calculate the total volume of mistletoes parasitizing 66 trees (n1) in the entire population

of trees (N = 531), it was assumed that the average volume of mistletoes per tree was equal

in both studied subpopulations ( and ):

Next, coefficient or the ratio of the total volume of mistletoes to the total volume of

tree canopies in the study area was calculated. It was assumed that the total volume of

mistletoes was equal to the total volume of mistletoes parasitizing 66 trees

:

The bootstrap procedure (10,000 repetitions) was used to calculate coefficient with

95% confidence intervals (CI).

Knowing the value of coefficient, it is possible to calculate availability of mistletoes

( ), total availability of trees ( ), and availability of each of the tree species

( ) present in the study area. Assuming that availability ( ) is proportional to biomass

( ), we can express availability as coefficient :

51

Knowing that mistletoe availability is we can express

mistletoe availability using the following formula:

In the above equation, is the total aboveground biomass of all trees present in the

study area that can be expressed using the following formula:

symbolizes the total aboveground biomass of a specific tree species: SW (Acacia

dealbata), BG (Eucalyptus globulus), SG (Eucalyptus camphora), MG (Eucalyptus viminalis), MM

(Eucalyptus obliqua), NLP (Eucalyptus radiata), WC (Exocarpos cupressiformis), and EX (exotic

tree).

Finally, assuming that all availabilities added together equal one, we can derive the

following equations for calculating mistletoe availability ( ) and the availability of

each of the tree species ( ) present in the study area:

Diet and food preferences of common brushtail possums

To verify the assumption that a diet of common brushtail possums consists of at least one

Eucalyptus species and might include other plant groups (i.e., acacias, parasitic plants, herbs,

and grasses), the percent share of different forage types in the overall diet was established

and presented as a mean with standard error (s.e), unless otherwise indicated. The collected

data on the faecal composition of 56 common brushtail possums were pooled together for

further analyses while the individual variation was tested by calculating the coefficient of

variation.

52

To explore food preferences of common brushtail possums in the study area the

standardized selection index (B) was employed (after Manly et al., 2002). The standardized

selection index (B) compares resource use (or consumption of different plant species

reconstructed from the faecal content analysis) with resource availability (or estimated

biomass). In this study, the term “food preferences” refers to the positive selection of

different plant species. Namely, a plant species is preferred when its proportion in the diet

is significantly higher than the proportion of available biomass in the habitat; and,

conversely, a plant species is avoided when its proportion in the diet is significantly lower

than the proportion of available biomass in the habitat. Hence, to be selective refers to

both preference and avoidance of different plant species. When a plant species is

consumed in proportion to its availability, no selection is observed.

Resource selection (B) is defined by the following formula:

Bil = (ui/mi)/∑i (ui/mi), where ui is a number of units in category i in the sample of used

units and mi is the number of available units in category i in the sample (i = 1, 2,…, I)

while l is the number of animals.

The values of this index range from 0 to 1 and are interpreted as the probability that a food

type would be the next selected if all food types were equally available. When food types

were available but not used, the 0 use value was replaced by 0.01 (after Di Stefano &

Newell, 2008).

In this study, resource use (U) was restricted only to species of trees and parasitic plants

that were expressed as the proportion of all species consumed by common brushtail

possums. Tree species, Acacia melanoxylon and Eucalyptus obliqua, were excluded from the

analysis because the sample size of A. melanoxylon was too small to produce meaningful

results while E. obliqua was scarcely distributed within a study area and, hence, available

only to a small number of animals. Also, the understorey species were excluded from the

analysis as it was hard to establish available biomass in a meaningful way. Namely, the

studied common brushtail possums were not restricted to roadside vegetation corridors

and often visited adjacent pastured areas (see Chapter 3). Moreover, as this study focused

on overall selection patterns and not individual variation in consumption, the relative

resource availability (M) was the same for all animals across the landscape.

53

This approach is acceptable only in a highly homogenous habitat or when habitat

heterogeneity is expressed at a micro-scale and associated herbivore assemblages are highly

mobile; the last two premises were fulfilled in this study (Zweifel‐Schielly et al., 2011).

The resource use (U), availability (M), and selection (B) of different plant species were

calculated for each individual in the study area (n = 56) and expressed as medians with 95%

confidence intervals (CI). In this study, median appeared to be a better measure of central

tendency than mean since values were proportions ranging from 0 to 1 (after Zar, 2009).

Confidence intervals and significance of selection (B) were calculated using a bootstrap

procedure (10,000 repetitions) in the statistical program R 2.11.1 (2010 The R Foundation

for Statistical Computing). To determine common brushtail possum preferences for

different genera, the logarithmic differences (median ± 95% confidence intervals) were

calculated and their significance tested. Here, the logarithmic difference between two plant

genera was independent from the other genera for which the selection index was also

calculated. Hence, the effects were expressed as the first named food type minus the

second food type; therefore, positive values meant the first named food type was selected

to a greater degree.

Determinants of tree use by common brushtail possums

Tree species preferences of common brushtail and mountain brushtail possums were

determined using a bootstrap or randomization procedure after Jones, Maclagan, and

Krockenberger (2006). A number of trees corresponding to a number of independent

sightings of possums (38 trees used by common brushtail possums and 49 trees used by

mountain brushtail possums) was randomly selected (with replacement) from the total of

577 trees recorded in the study area, and the tree species was recorded. This procedure was

repeated 10,000 times to produce an expected distribution of randomly selected tree

species. The number of sightings of possums in a specific tree species was compared with

the randomly generated distribution, and the probability of sighting was determined directly

from the randomized distribution. To avoid pseudo replication, all trees that were used by

more than one individual were excluded from the analysis.

54

To test the hypotheses that trees parasitized by mistletoes, trees bearing hollows, as well as

medium and large trees (50–100 cm and 100–200 cm DBH) were used more often by

common brushtail and mountain brushtail possums than expected from their availability in

the habitat, the chi-square goodness of fit test was used. To test the relationship between

tree size and the number of mistletoes and hollows, the Spearman's rank test was used. To

compare differences between all measured variables for trees used by common brushtail

possums and mountain brushtail possums, first the Lilliefors test for normality was

employed and next the Wilcoxon signed ranks test was used to compare medians.

All calculations were performed using the statistical program R 2.11.1 (2010 The R

Foundation for Statistical Computing) with a significance level α = 0.05.

Finally, to explain the relationship between common brushtail possum tree presence or

absence and all measured variables (tree species, height and diameter, biomass, number of

mistletoes and hollows, and the occurrence of the mountain brushtail possum) the

binomial logistic regression, a type of the binomial generalized linear model was used.

The binomial logistic regression analysis examines the influence of various factors on a

dichotomous outcome (1—presence; 0—absence) by estimating the probability of the

event’s occurrence. It examines the relationship between one or more independent

variables (continuous or categorical) and the log odds of the dichotomous outcome by

calculating changes in the log odds of the dependent as opposed to the independent

variable itself.

Next, the step Akaike’s Information Criterion (hereafter AIC) function from the MASS

library was employed to identify the best model with the combination of variables

explaining the possum tree presence (Venables & Ripley, 2002). This function is a stepwise

algorithm that sequentially searches through all possible models for the one that minimizes

the AIC. Namely, it starts from the simplest model (the intercept) and adds one step at a

time to a new variable, considered the best by the AIC, to each of the consecutive models.

On each step, the AIC from the new model is compared to the previous model, and if the

AIC from the new model is smaller, the step is repeated. If the AIC from the new model is

not smaller, the algorithm stops and the best model is considered to be the one from the

previous step. If there are no more variables to be added to the new model, the algorithm

stops and the full model is considered the best. The significance of variables from the final

selected model is estimated using the z-test.

55

The AIC was used in this study because the information-theoretic approach is a

recommended alternative to traditional hypothesis testing since it is more tolerant to

violations of parametric statistics that are commonly encountered in ecological data.

Moreover, the AIC method allows for the comparison of multiple working hypotheses and

does not rely solely on the use of ecologically arbitrary P values for determining

significance (Burnham & Anderson, 2002).

Distribution and density of sympatric species, the common brushtail possum and mountain brushtail

possum

All sightings of common brushtail possums and mountain brushtail possums within

roadside vegetation corridors were plotted on a digitised base map of a study area with

ArcGIS 9.3 Software (2009 ESRI, Redlands, CA). Next, the probability density of these two

sympatric species was established using the ArcGIS Kernel Home Range function that

calculates a fixed kernel home range utilization distribution as a grid coverage using the

least squares cross validation for the smoothing parameter (Worton, 1989).

56

Results

Diet breadth and food preferences of common brushtail possums

Based on a total of 5,617 epidermal fragments from 56 analysed faecal samples, the overall

diet composition of common brushtail possums in the study area was established. In total,

nine upper-storey and mid-strata native tree species were identified from the genera

Eucalyptus and Acacia (5 eucalypt and 4 acacia species), along with root and aerial parasites,

Exocarpos cupressiformis (cherry ballart) and Amyema pendula (drooping mistletoe). Exotic trees

included Pinus radiata (radiata pine), Fraxinus angustifolia (narrow-leaved ash), Quercus robur

(English oak), and Prunus cerasifera (cherry plum); due to a low sample size they were pooled

together for further analyses. Understorey species were ascribed to broad functional groups

of herbs and grasses and were excluded from all statistical analyses (see Methods for

justification).

Trees and parasitic plants contributed to the largest proportion of epidermal fragments

identified in faeces of common brushtail possums (60%, Fig. 11). Eucalyptus species from

Monocalyptus and Symphyomyrtus subgenera were found to represent 21.6% of all species

consumed (dominated by E. radiata 14.1%), followed by exotic trees (17.1%) and Acacia

species (9.7%) with silver wattle (A. dealbata) representing 6.3% of Acacia spp. (Fig. 12).

Parasitic plants Amyema pendula and Exocarpos cupressiformis represented 6.8% and 4.3% of all

consumed species (Fig. 12). Understorey species, including forbs and grasses, represented

15.3% of the possum diet while fruit (i.e., common blackberry, Rubus fruticosus) and pollen

(i.e., Acacia and Eucalyptus) constituted 11% of the total diet (Fig. 12). However, 802

epidermal fragments (14.2%) could not be assigned to any plant species, group, or part

(Fig. 12).

57

Fig. 11 The overall diet composition of common brushtail possums occurring in the roadside vegetation corridors of the Strathbogie Ranges, Victoria

Fig. 12 Mean consumption (%) with standard error (s.e) of different plant species, groups, and parts by common brushtail possums occurring in the roadside vegetation corridors of the Strathbogie Ranges, Victoria

58

On average, six plant species, including from one to three Eucalyptus and Acacia species,

were found in 75% of all faeces analysed, followed by exotic trees that compromised 71%

of all faecal samples. Surprisingly, more than 90% of faeces contained key understorey

groups, forbs and grasses. Finally, more than half of all faeces contained fragments of one

of the two parasitic species, Exocarpos cupressiformis or Amyema pendula, fruit and pollen.

In the present study, differences in use of various plant species/genera by common

brushtail possums and their availability in the habitat produced different selection indices

(B, Table 2). Parasitic species, Exocarpos cupressiformis and Amyema pendula, were consumed

more frequently than expected from their availability in the habitat while Eucalyptus species

were consumed less frequently than expected. (P < 0.05, Table 2, Fig. 13). The most

dominant tree species in the habitat, Eucalyptus radiata and E. viminalis, were selectively

avoided regardless of the magnitude of available biomass in the habitat (P < 0.001,

Table 2, Fig. 13). Moreover, when all genera were compared, parasitic plants, Acacia, and

exotic trees were preferred over genus Eucalyptus (P < 0.001, Fig. 14).

59

Table 2 Relative availability, use, and selection of different tree species and parasitic plants by common brushtail possums occurring in the roadside vegetation corridors of the Strathbogie Ranges, Victoria. Values are presented as medians with 95% confidence intervals (CI). Plus (+) symbolizes positive selection or preference for the species, while minus (–) negative selection or avoidance of the species by common brushtail possums. An asterix indicates significance of selection at P < 0.05

Plant species % Available (A) % Used (U) Selected (B)

Acacia dealbata 0.014 (0.008, 0.022) 0.129 (0.091, 0.171) 0.143 (0.078, 0.229) (+)

Eucalyptus globulus 0.034 (0.007, 0.076) 0.001 (0.000, 0.002) 0.006 (0.002, 0.032) (–)*

Eucalyptus camphora 0.088 (0.050, 0.136) 0.078 (0.037, 0.131) 0.021 (0.008, 0.045) (–)*

Eucalyptus viminalis 0.195 (0.124, 0.278) 0.013 (0.002, 0.031) 0.003 (0.001, 0.008) (–)*

Eucalyptus radiata 0.623 (0.530, 0.710) 0.261 (0.184, 0.344) 0.014 (0.006, 0.027) (–)*

Exotic trees 0.035 (0.005, 0.080) 0.316 (0.229, 0.408) 0.196 (0.097, 0.414) (–)

Exocarpos cupressiformis 0.002 (0.001, 0.003) 0.073 (0.040, 0.113) 0.295 (0.169, 0.443) (+)*

Amyema pendula 0.003 (0.001, 0.006) 0.123 (0.073, 0.183) 0.291 (0.167, 0.448) (+)*

60

Fig. 13 Standardized index of plant species selection (B) by common brushtail possums in the study area. Errors are 95% confidence intervals (CI) of the median. The dotted line corresponds to a critical value of 0.125 (1/n where n is the number of plant species) or no selection (i.e., species selected in proportion to their availability). Any bar above the line signify positive selection or preference for the plant species and any bar below the line signify negative selection or avoidance of the species by common brushtail possums. An asterix indicates significance at P < 0.05

61

Fig. 14 Selection of plant genera by common brushtail possums in the study area Genus abbreviations: SW (Acacia), EU (Eucalyptus), EX (exotic trees), WC (Exocarpos), and MIS (Amyema). Effects are expressed as the first named food type minus the second named food type, so positive values mean the first named type is selected to a greater degree. Errors are 95% confidence intervals (CI) of the difference

62

Determinants of tree use by common brushtail possums

In total, 53 common brushtail possums and 60 mountain brushtail possums were recorded

in the study area. Both species displayed a highly scattered distribution and occupied the

same areas within roadside vegetation corridors (Fig. 15). From ten dominant mid- and

upper-storey species of trees present in the study area, two thirds of sightings of common

brushtail possums were in Eucalyptus camphora (23.7%), E. radiata (21.1%), and E. viminalis

(21.1%). Exotic species of trees were used more often than expected from their availability

in the habitat (15.8%, P < 0.001, Fig. 16). Similarly, mountain brushtail possums were

most often found on E. radiata (34.7%), followed by E. camphora (20.4%), and E. viminalis

(18.4%), which were all used in proportion to their availability (Fig. 16). Surprisingly, Acacia

dealbata (12.2%) was used by mountain brushtail possums less often than expected from its

availability in the habitat (P < 0.05, Fig. 16).

Out of 577 trees recorded in the study area, 11.6% were parasitized by drooping mistletoe

(Amyema pendula) with 4.0 ± 0.5 parasites per tree on average. Mistletoe infection rates were

not uniform across different tree species. Most often mistletoes were found on Eucalyptus

radiata, representing 59.7% of parasitized trees, with on average 3.8 ± 0.7 parasites per tree.

A weak positive correlation was noted between tree diameter and number of mistletoes

(r = 0.24, df = 78, P < 0.05).

Furthermore, almost 12.3% of surveyed trees had hollows large enough to be used as

denning sites by common brushtail possums, with on average 2.4 ± 0.2 hollows per tree.

Tree species with the greatest number of hollows were Eucalyptus radiata (45% of all trees

with hollows, 2.6 ± 0.4 hollows per tree) and E. viminalis (26.7%, 2.5 ± 0.3 hollows per

tree). The number of hollows was positively correlated with tree diameter (r = 0.59, df =

78, P < 0.001).

In the study area, most of the trees were in the small diameter class (5–50 cm; 63.5%),

followed by medium sized trees (50–100 cm, 22.6%) and large trees (> 100 cm, 13.9%).

The species with the largest diameter were E. radiata (69.5 ± 3.8 cm) and E. viminalis (68.3

± 6.2 cm). The medium height of trees in the study area was 11.3 ± 0.2 m and, since height

was highly correlated with tree diameter (r = 0.81, df = 78, P < 0.001), it was omitted from

the further analysis.

63

Fig. 15 Probability density of the common brushtail possums (n = 53) and mountain brushtail possums (n = 60) in the Strathbogie Ranges, Victoria

64

a) b) Fig. 16 The comparison of proportions of above ground biomass (AGB, grey bars) of different tree species used by common brushtail possums (a) and mountain brushtail possums (b), and availability of those species within a landscape (median with 95% confidence intervals, CI). Tree species: BG Eucalyptus globulus (Victorian blue gum), BLP E. dives (broad-leaved peppermint), BW Acacia melanoxylon (blackwood), EX (exotic trees), MG E. viminalis (manna gum), MM E. ovata (messmate), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (mountain swamp gum), SW A. dealbata (silver wattle), WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant difference in use and availability at P < 0.05

65

Despite the expectation, there were no significant differences observed in the use of trees

with or without mistletoes by common brushtail possums (χ2 = 0.528, df = 1, P = 0.608).

A similar pattern was observed for mountain brushtail possums (χ2 = 0.803, df = 1,

P = 0.773). Trees with hollows were used by both common and mountain brushtail

possums more often than expected by chance alone (χ2 = 13.778, df = 1, P < 0.001;

χ2 = 3.741, df = 1, P <0.05 respectively). Common brushtail possums more often used

trees larger in diameter (median 71.5 cm; 95% confidence intervals 64.5–79 cm; χ2 =13.778;

df = 1, P < 0.001) than the median diameter of available trees in the study area (28 cm;

95% confidence intervals 23–30 cm). The similar pattern was observed for mountain

brushtail possums (median 54 cm; 95% confidence intervals 35–77 cm; χ2 = 13.067, df = 1,

P < 0.001) Furthermore, no statistically significant differences were detected between

common and mountain brushtail possums in terms of tree use and specific tree traits:

species (χ2 = 6.140, df = 9, P = 0.828), number of mistletoes (W = 875, P = 0.391),

hollows (W = 1049, P = 0.201), and diameter (W = 1063, P = 0.260).

Finally, when taking into account all measured variables, the tree species, height and

diameter, biomass, number of mistletoes and hollows, and the presence of mountain

brushtail possums (T. cunninghami), the binomial generalized linear model indicated that tree

diameter and presence of T. cunninghami were the best variables for predicting tree

occupancy by common brushtail possums (r2 = 0.22, P < 0.001, Tab. 3). The model

explained 76.3% of common brushtail possum tree presences and 76.3% of tree absences.

Table 3 The binomial generalized linear model predicting tree occupancy by common brushtail possums (AIC = 87.73). An asterix indicates significance at P < 0 ‘***’ and P < 0.001 ‘**’ Coefficients Estimate Std. Error z-value P-value Intercept -1.498e+00 4.855e-01 -3.085 0.002** Diameter 2.283e-02 7.343e-03 3.109 0.002** T. cunninghami 1.868e+01 1.974e+03 0.009 0.992 Binomial generalized linear model

df Dev. Resid. df Resid. Dev. P-value Null 75 105.358 Diameter 1 12.117 74 93.242 0.0005*** T. cunninghami 1 11.512 73 81.730 0.0007***

66

Discussion

Summary of key results

In the present study, the common brushtail possum (Trichosurus vulpecula) demonstrated a

generalist feeding strategy consuming a wide variety of plant species and parts. In addition

to different Eucalyptus species, studied animals consumed substantial amounts of foliage

from Acacia species, exotic species of trees, parasitic plants, understorey species, as well as

some fruit and pollen. It is worth noting here, that the root parasite, Exocarpos cupressiformis,

was recorded for the first time in a diet of common brushtail possums. Also, common

brushtail possums were found consuming fruits of an invasive species, Rubus fruticosus.

Nevertheless, despite their generalist feeding strategy common brushtail possums exhibited

clear foraging preferences at multiple scales of resource heterogeneity ranging from forest

types to tree species and parasitic plants within tree canopies. Parasitic plants, Amyema

pendula and Exocarpos cupressiformis, were consumed more frequently than expected from

their availability in the habitat while the dominant Eucalyptus species were consumed less

frequently. Moreover, Acacia species and exotic trees were consumed more often than

expected, but results were not significant.

Common brushtail possums were most often found in Eucalyptus camphora, E. radiata, and

E. viminalis trees during their nightly foraging activity. They used only exotic species of trees

more often than expected from their availability in the habitat. Contrary to expectation,

trees parasitized by mistletoes were not used selectively by common brushtail possums.

A similar pattern of tree use was observed for a sympatric species, the mountain brushtail

possum (T. cunninghami) with animals found most often in canopies of dominant eucalypt

species and hardly ever foraging in trees with mistletoes. Contrary to expectation, mountain

brushtail possums used Acacia dealbata trees less often than expected from their availability

in the habitat. Both common and mountain brushtail possums preferred trees with hollows

and trees larger in diameter. As a result, no significant differences were observed between

trees used by common brushtail and mountain brushtail possums. Tree diameter and

occurrence of mountain brushtail possums were the best variables to explain the observed

patterns of tree use by common brushtail possums.

67

The diet breadth and food preferences of common brushtail possums

Foraging decisions of arboreal marsupials inhabiting eucalypt forests and woodlands are

complex and depend not only on food availability and nutritional quality, but also on

animal ability to locate different types of food and to learn to eat or avoid difficult forage.

In the current study, common brushtail possums had a broad diet and consumed 18

different types of food, including foliage of woody and seral species, as well as fruit and

pollen. Yet, foliage of habitat dominant Eucalyptus species represented less than one quarter

of all consumed forage. Likewise, previous studies demonstrated a great dietary breadth of

common brushtail possums incorporating different groups of plants and plant parts into

their diet (summarized by Kerle, 2001). For instance, a proportion of Eucalyptus foliage in a

diet of common brushtail possums has been found to vary from 90% in eucalypt

woodlands in dry tropics of North Queensland to being completely absent or comprising

less than 10% in wet sclerophyll forests of Tasmania (Fitzgerald, 1984; DeGabriel et al.,

2009b).

Why common brushtail possums consume different amounts of Eucalyptus foliage in

different parts of their range remains unclear, but some general patterns emerge from

recent studies with captive and wild animals. Eucalyptus foliage is known to be low in

nutrients, especially nitrogen (a proxy of protein content), and has high concentrations of

components that interfere with animal digestion and absorption of protein, such as tannins

and fibre (DeGabriel et al., 2008; DeGabriel et al., 2009a). Common brushtail possums are

less effective in digesting fibre than eucalypt specialists, such as the koala and greater glider,

since they lack a mechanism for separating large food particles that allow for subsequent

microbial fermentation (Sakaguchi & Hume, 1990). Unlike common ringtail possums,

common brushtail possums do not practice caecotrophy to maximize nutrient acquisition

and are more sensitive to tannins and associate their astringent taste with post-ingestive

negative effects (Marsh et al., 2003a). Furthermore, eucalypt leaves contain a range of plant

secondary metabolites that are potentially toxic to common brushtail possums, such as

terpenes, cyanogenic glycosides, and formylated phloroglucinol compounds (FPCs),

powerful antifeedants found only in eucalypts from the subgenus Symphyomyrtus (Brophy &

Southwell, 2002; Eschler et al., 2000; Gleadow et al., 2008; Moore et al., 2004).

The inability of common brushtail possums to satisfy their nutritional requirements from

a single species or for that matter to sustain entire populations on a diet composed

exclusively of Eucalyptus foliage was first recognized by Freeland and Winter (1975).

68

The authors observed that while animals spent 66% of their time foraging on eucalypts,

common brushtail possums usually consumed three different foods per night, two species

of Eucalyptus and one non-eucalypt species, in order to reduce the negative effects of any

single toxin. In the current study, common brushtail possums were found consuming an

average of six plant species, including one to three Eucalyptus species with Eucalyptus radiata

from a subgenus Monocalyptus eaten most often (14.1%) and complemented with negligible

amounts of E. camphora (3.8%) and E. viminalis (0.8%) from a subgenus Symphyomyrtus.

The observed high intake of E. radiata by common brushtail possums was unexpected

since this tree species is noted for its high level of phenolics and terpenes approaching up

to 8% in some chemotypes (Foley, 1992). Furthermore, high concentrations of formylated

phloroglucinol compounds (FPCs) have been found to deter possums from foraging on E.

camphora and E. viminalis (Marsh et al., 2003a; Moore et al., 2004a). In a feeding trial

performed by Dearing and Cork (1999), common brushtail possums lost 6.5 ± 1.1 % of

their body weight when offered solely foliage of E. radiata. Yet, animals simultaneously

offered foliage of E. radiata (Monocalyptus) and E. melliodora (Symphyomyrtus) increased their

food intake and consumed 33% more food than on either diet offered singly. The authors

concluded that a diet breadth of common brushtail possums is a product of their

detoxification abilities as these animals are physiologically constrained by their capacity to

absorb, metabolize, and excrete similar types of toxins found in eucalypt foliage.

Recently, DeGabriel and colleagues (2009b) demonstrated that common brushtail possums

occurring in different habitats have developed local physiological and behavioral

adaptations to deal with difficult forage. Contrary to the authors’ expectations, southern

populations inhabiting forests and woodlands dominated by Eucalyptus species proved to be

more sensitive to concentrations of tannins, terpenes, phenolic glycosides, and formylated

phloroglucinol compounds (FPCs) than northern populations. The authors explained these

findings by noting the higher diversity of plant species in southern forests and woodlands

that allowed possums to have broader diets and switch to less difficult forage when

necessary. A positive relationship between availability of forage and a breadth of diet of

common brushtail possums proved to be the case in the current study, where Herb-rich

Foothill Forest and Swampy Riparian Woodland had a well-developed mid-stratum

dominated by Acacia spp. and species-rich understorey composed of native and exotic

species of herbs and grasses.

69

Parasitic plants, Amyema pendula and Exocarpos cupressiformis, were frequently found

parasitizing stems and roots of Eucalyptus and Acacia trees. As a consequence, the studied

common brushtail possum had a broad diet incorporating foliage of Acacia species (9.7%)

with silver wattle (A. dealbata) representing 6.3%, Amyema pendula 6.8%, and Exocarpos

cupressiformis 4.3% while herbs and grasses represented 15.3% of a possum diet.

It is likely that the studied common brushtail possums learned to avoid Eucalyptus foliage to

minimize the negative effects of toxins and selectively foraged on nitrogen-fixing Acacia

species to maximize their nitrogen intake. Yet, Acacia leaves due to their phyllode structure

tend to be fibrous and have high tannin content, and may contain cyanogenic glycosides

and fluroacetate interfering with possum digestion (Irlbeck & Hume, 2003). Fitzgerald

(1984) studied a population of common brushtail possums inhabiting dry sclerophyll

forests in Tasmania where A. dealbata and A. melanoxylon were the most common

understorey species and recorded Acacia foliage (especially A. dealbata) to be present in the

possum diet throughout the year. Similarly, Evans (1992) studied the diet of common

brushtail possums in central arid Australia and noticed that foliage of A. coriacea was a

preferred dietary constituent along with foliage from other Acacia species.

However, in a feeding trial carried out by McArthur and Turner (1997) when common

brushtail possums were offered foliage of acacia species (A. dealbata and A. melanoxylon) and

eucalypt foliage from Monocalyptus and Symphyomyrtus subgenera, animals preferred feeding

on foliage of Symphyomyrtus species and almost completely avoided A. melanoxylon. The

observed diet mixing behaviour could have been a result of nutrient balancing in foraging

decisions of common brushtail possums governed by the nitrogen content of forage.

Raubenheimer and Simpson (2003) demonstrated that a generalist herbivore, the desert

locust (Schistocerca gregaria), selected a diet with a higher nitrogen content. The authors

concluded that the capability of desert locusts to deal with excess ingested protein might be

related to the greater nitrogen content of selected foods rather than their broad dietary

range.

Furthermore, Wiggins and colleagues (2006a, 2008) argued that frequent switching between

chemically diverse foliage reduces physiological constraints imposed by toxin-rich diet and

enables more efficient nutrient acquisition by common brushtail possums. This reduction

in physiological constraints might explain why in the current study common brushtail

possums preferred feeding on parasitic plants over their Eucalyptus and Acacia tree hosts.

70

Similarly, other experimental and field studies have shown that mistletoes from

Loranthaceae family are preferentially eaten by common brushtail possums, common

ringtail possums, koalas, and greater gliders (summarized by Reid 1997).

Mistletoes are considered to be of greater nutritional quality for herbivores as they are rich

in nutrients, have higher foliar moisture content, and are presumably less toxic than their

tree hosts (Watson, 2001, 2011). Foliage of box mistletoe Amyema miquelii, closely related to

drooping mistletoe A. pendula, has been found to have a higher content of potassium and

sodium that can be limiting to some herbivore populations (March & Watson, 2009). Yet,

plant secondary metabolites of genus Amyema remain largely unknown. Both species of

mistletoes were found to contain herbivory-reducing terpenes (Preston & Watson, 2010)

and flavonoid afzelin (Hensens, Lewis & Mulquiney, 1971).

In the present study, diet switching between parasitic plant and their tree hosts could have

helped common brushtail possums to overcome short-term intake constraints resulting

from toxic effects of each species. The foraging decisions of the studied animals might

have been shaped by small-scale heterogeneity of resources with mistletoes representing

alternative food patches within otherwise uniform tree canopies. Namely, Wiggins et al.

(2008) predicted that heterogeneity of plant patches, in relation to distribution of plant

secondary metabolites, and the scale at which this occurs across the landscape, are critical

factors that influence foraging efficiency and fitness of mammalian herbivores.

Furthermore, preferential foraging of common brushtail possums on a root parasite,

Exocarpos cupressiformis (cherry ballart, Santalaceae), deserves special attention as the current

study is the first one to recognize the importance of this species for arboreal marsupials.

Foliage of E. cupressiformis has been previously detected in a diet of the brushtail-tailed rock-

wallaby (Petrogale penicillata, Van Eeden, Di Stefano, & Coulson 2011). The authors argued

that although E. cupressiformis did not appear to form a large portion of wallabies’ diet it

could represent an important highly nutritious component. Yet, no foliar chemical analysis

has been carried out until the present (see Chapter 4 for details). Furthermore, root

parasites from Santalaceae family have been found to sequester some of the plant

secondary metabolites from their hosts, which in turn can negatively affect the foraging of

herbivores (Woldemichael, & Wink, 2002). In the current study, E. cupressiformis parasitized

both nitrogen-fixing acacias and eucalypts containing different groups of plant secondary

metabolites, yet how these host trees affected palatability of E. cupressiformis remains

unknown.

71

In the present study, it was also impossible to distinguish whether common brushtail

possums were “smoothing out their diet” (eating proportionally less of common and more

of rare species), which increases their dietary breadth, or if the foliar chemistry made

Eucalyptus foliage lower quality food for this generalist arboreal marsupial, or if both

reasons together increased their dietary breadth. The experience from sampling different

types of food, such as parasitic plants and transient food aversion to Eucalyptus foliage,

could not be ruled out as a possible explanation for common brushtail possum food

preferences.

Food aversions often called sensory-specific satiety become pronounced when foods

contain high concentrations of toxins or nutrients or both or are nutritionally imbalanced

or deficient (the satiety hypothesis, Provenza et al., 2003). Namely, when animals or

humans are presented with a variety of foods, they prefer an alternative to the one eaten

most recently because interaction among flavours, nutrients, and toxins can cause transient

aversions. Sampling a variety of foods allows animals to learn about the consequences of

ingesting each food or combination of foods enabling them to activate or maintain

different detoxification pathways (Marsh et al., 2006a, Marsh et al., 2006b).

Consequently, in the wild common brushtail possums consume in addition to tree and

mistletoe foliage many different types of foods, such as grasses and forbs, fruits, seeds, and

pollen. In the present study, grasses and forbs represented 15.3% of the possum diet while

fruit and pollen comprised 11% of the total diet. Grasses and forbs were found to

represent 40–60% of the diet of common brushtail possums in dry sclerophyll forests

adjacent to pastures in Tasmania (Fitzgerald, 1984). On this diet digestibility is high

(i.e., cell walls of grasses are low in fibre and plant secondary metabolites) while forbs are

rich in protein, sugars, and lipids (Provenza et al., 2003). Interestingly, although Downey et

al. (1997) found Rubus fruticosus to be relatively rare in the study area, fruits of this invasive

species were readily eaten by common brushtail possums in the present study.

Likewise, in New Zealand seasonally available fruits and seeds of native and exotic species

were found to compromise 70% of a common brushtail possum diet (Dungan et al., 2002).

Williams (1982) compared nutritional properties of some fruits eaten by common brushtail

possums in New Zealand podocarp-forests and concluded that fruits appeared to be a

source of readily digestible energy (rich in carbohydrates and low in fibre), however, they

did not differ from leaves in levels of crude protein, lipids, and minerals.

72

While no such study exists in Australia, the less folivorous and more frugivorous diet of

common brushtail possums compared to other arboreal marsupials’ diets has been

explained by their less specialized dentition and their small body size.

Furthermore, common brushtail possums extended their diet to incorporate several

introduced species, including plantation, ornamental, and orchard trees (Kerle, 2001).

In the current study, exotic species of trees from Europe (Fraxinus angustifolia, Quercus robur

and Prunus cerasifera) and North America (Pinus radiata) represented 17.1% of a possum diet

despite being a minor component in the habitat. Recently, a comprehensive study of

foraging preferences of generalist herbivores from multiple continents demonstrated their

affinity for exotic plants that are evolutionary naïve or lack specific plant secondary

metabolites (Morrison & Hay, 2011). For example, in a feeding trial South American apple

snails (Pomacea sp.) preferred North American macrophytes while North American crayfish

(Procambarus spiculifer) preferred South American, Asian, and Australian macrophytes that

were less chemically defendant than native species.

In the case of introduced herbivores, such as common brushtail possums in New Zealand,

it has been postulated that the degree of diet generalism and access to novel plant species

with a greater nutritional value and lower plant defences may contribute to the successful

invasion and establishment of common brushtail possums (O’Reilly-Wapstra & Cowan,

2010). Dietary plasticity of common brushtail possums probably allowed this generalist

forager to successfully invade broadleaf-podocarp forests and to a lesser degree southern

beech Nothofagus forests in New Zealand. In these forest communities common brushtail

possums are known to eat approximately 100 plant species, yet show strong preferences for

foliage of relatively rare species, such as red mistletoe Peraxilla tetrapetala and yellow

mistletoe Alepis flavida, contributing to mistletoe declines in some areas (Montague, 2000;

Sweetapple et al., 2002; Sweetapple, 2008).

In Australia and even more in New Zealand, common brushtail possums have been found

to supplement their diet with animal sourced protein from arthropods, lizards, bird eggs,

and chicks, and opportunistically scavenge on animal carcasses (Brown, Innes, & Shorten,

1993; Cowan & Moeed, 1987). In the present study, it was hard to distinguish animal

material in possum faeces since this would require a different methodological approach,

such as establishing differences between nitrogen and carbon isotope ratios among

consumed foods and possum body tissues.

73

Nonetheless, it can be speculated that out of 14% of unidentified forage material some part

was of animal origin while the rest was finely digested plant material with no recognizable

cell structures. Different digestibility of plant material could have underestimated the

proportion of seral species, such as herbs and grasses in a diet of the studied common

brushtail possums. The underestimated proportion of seral species could have been

partially accounted for if smaller sieves than the ones used in this study were utilized to

retain finer leaf fragments for further identification.

Also, the relative importance of fruits, flowers, and pollen was difficult to quantify using

the faecal content analysis method. Difficulties in identifying the cell structure of fruits can

partially explain why mistletoe fruits, favoured by possums in some areas (Reid, 1997), were

not detected in faeces of the studied animals. By contrast, it was relatively easy to identify

the well preserved cuticles of different Eucalyptus species and establish their share in the

overall diet of common brushtail possums. Differential digestion may be accounted for

using correction factors developed from regression equations that compensate over- or

underestimation of cuticle types (Dearden, Pegau, & Hansen, 1975). Correction factors are

produced by feeding captive animals known quantities of forage and then assessing the

relative quantity of each species present in their faeces. This method proved to be

unfeasible in the current study due to ethical and logistic constraints associated with

capture and husbandry of common brushtail possums.

Determinants of tree use by common brushtail possums

For arboreal marsupials spending most of their lives in tree canopies, trees represent food,

shelter, and a stage upon which interactions with conspecifics, competitors, and predators

take place. This recognized polygonal use of trees by arboreal marsupials represents a

challenge for researchers trying to single out reasons for the selection of some trees and the

avoidance of others. In the present study, food choices of common brushtail possum were

expected to be reflected in their tree use measured against tree species availability in the

habitat. From ten dominant mid- and upper-storey tree species, two-thirds of the sightings

of common brushtail possums were in Eucalyptus camphora (23.7%), E. radiata (21.1%), and

E. viminalis (21.1%), yet eucalypt species represented just 21.6% of a possum diet and were

avoided by foraging animals.

74

Only exotic species of trees were used more often by common brushtail possums than

expected from their availability in the habitat (15.8%) and this was also reflected in their

diet with exotic trees representing 17.1% of all consumed species. These findings suggest

that common brushtail possums could have used exotic trees exclusively as foraging sites.

Conversely, Eucalyptus trees could have been used by common brushtail possums for other

reasons, including presence of alternative food patches of mistletoes, occurrence of

hollows used as shelter and concealment from predators, and a tree size or branches big

enough to support animal body weight.

However, against expectations common brushtail possums did not selectively use trees

parasitized by mistletoes. This finding challenges the results from the faecal content

analysis showing that common brushtail possums preferentially foraged on Amyema pendula

over its Eucalyptus and Acacia tree hosts. There are many possible and often competing

explanations for these seemingly contradictory results. First, nocturnal marsupials use

visual cues to locate food, at least from a distance, and have monochromatic vision;

therefore, they are unlikely to be able to detect differences in foliage colour between

mistletoes and their tree hosts (Reid & Yan, 2000). Moreover, some species of mistletoes

tend to resemble their tree hosts in appearance presumably to avoid being eaten by

herbivores (Barlow & Wiens, 1997). It has been argued that mistletoe host-resemblance is a

type of cryptic-mimicry where mistletoes increase their fitness by blending into its

vegetative background under the selective pressure exerted by arboreal marsupials.

The question therefore arises as to how common brushtail possums find mistletoes in tree

canopies. Can they similar to primates remember locations of a wide variety of plant foods

and move directly to such foods when and where available without wasting time and

energy searching randomly? There are no studies known to the author measuring the

cognitive abilities of arboreal marsupials. From the nutritional perspective, it can be argued

that a special “taste” of common brushtail possums for mistletoes is the outcome of their

generalist feeding strategy. Animals might consume mistletoes opportunistically when they

forage on eucalypts. For instance, in the current study Eucalyptus radiata represented a major

component of possum diet. Also, E. radiata was parasitized by A. pendula most often and

had the highest number of mistletoes per tree. High mistletoe availability within E. radiata

canopies could have created a specific preference for A. pendula and promoted diet

switching behaviour in possums.

75

Furthermore, tree selection by common brushtail could have been affected by the

occurrence of a sympatric species, the mountain brushtail possum, as both species did not

vary in their use of trees. Furthermore, these sympatric species were never found sharing

the same tree. These findings were confirmed by another study looking at foraging patterns

of common brushtail possums and mountain brushtail possums in the same study area

(Del Borrello 2009). The author suggested that studying these species both in sympatry and

allopatry could reveal differences in tree use and explore underlying mechanisms. Namely,

if in allopatry common brushtail possums would show preferences for a certain tree, as has

been previously demonstrated, than the presence of a competitor might have altered their

foraging behaviour.

However, a resource overlap does not necessary indicate competition as revealed in a study

by Lindenmayer and Cunningham (1997) that looked at patterns of co-occurrence of

arboreal marsupials in the montane ash forests of Victoria. In these forests mountain

brushtail possums were found to co-occur with greater gliders, another strictly folivorous

species. In the current study, koalas were found often foraging on Eucalyptus viminalis, a

species that represented just 0.8% of the studied common brushtail possums’ diet. While

koalas do not appear to be limited by interspecific competition (Norton & Neave, 1990), it

remains to be tested whether they exclude other smaller species from foraging on E.

viminalis.

The other important tree characteristic that might have affected possum tree selection was

the presence of suitable size hollows used as shelter and concealment from predators.

Namely, trees with hollows were preferred by common brushtail possums, with E. radiata

and E. viminalis having the highest hollow number per tree. Johnson et al. (2001)

determined that at high population densities, hollow availability was a limiting factor for

populations of common brushtail possum in mainland Queensland and New South Wales.

Also, a certain level of competition over hollow-bearing trees could have been exhibited

through the sympatric mountain brushtail possums. However, no pattern was observed

when comparing differences in hollow bearing trees used by two species of possums

indicating low population numbers, increased dispersal rates, or high availability of hollows

in the habitat.

Del Borrello (2009) working in the same study area found no evidence for competition

over hollow bearing trees and exclusion of common brushtail possums from suitable

denning sites by mountain brushtail possums.

76

The studied roadside vegetation corridors have been recognized as sustaining an

exceptionally high number of trees with hollows due to the fact that they have not been

logged for more than 100 years (Martin & Martin 2004; Martin & Martin, 2007).

Differences in body size with mountain brushtail possums being larger than common

brushtail possums, hence requiring larger size hollows, could have alleviated competition

between the sympatric species and facilitated their co-existence.

In the current study, both common brushtail possums and mountain brushtail possums

were more often found on larger trees exceeding 50 cm in diameter. A similar pattern was

observed for the specialist herbivore, the koala (Moore & Foley, 2005). Koalas visited trees

with a wider diameter significantly more often than expected from their availability in the

habitat as these trees provided larger patches of food and accounted for a greater

proportion of available biomass (Moore & Foley, 2005). Larger trees are also known to

sustain higher mistletoe populations (Reid & Lange, 1988) and have a higher number of

hollows than trees smaller in diameter (< 50 cm; Gibbons & Lindenmayer, 2002).

Moreover, larger trees are suitable for bird nesting and common brushtail possums might

opportunistically feed on eggs and chicks to obtain extra protein as shown in New Zealand

studies (Brown, Innes, & Shorten 1993; Brown, Moller, & Innes, 1996). Furthermore, by

foraging in larger trees and moving along branches of neighbouring trees, common

brushtail possums might have reduced the risk of predation by avoiding travelling on the

ground between patchily distributed resources. In the study area, the high activity of

ground predators, such as foxes and feral cats, was frequently observed, which might have

influenced common brushtail possums to seek refuge high in tree canopies.

Finally, it is possible that the presence of observers while spotlighting might have altered

the foraging behaviour of the studied animals driving them off the ground and into trees.

Also, the observed discrepancies in tree species selection and consumption could reflect

limitations of spotlighting as a method for establishing food preferences of arboreal

herbivores. Recorded high possum visitations to E. camphora trees could be an artefact of

their open canopy structure enabling easy detection in comparison with the more dense

canopies of E. radiata and E. viminalis. Some of the observations could also reflect animals

moving between feeding locations.

77

Hence, to disentangle the real reasons for selecting some trees and avoiding others, it is

recommended to carry out focal observations of individuals while foraging and establish

their denning sites using radio-telemetry (see Chapter 3). The present study has

demonstrated that foraging decisions of common brushtail possums are complex and

depend not only on food availability and nutritional quality, but also interplay between

many ecological factors: availability of shelter, competition, and predation. An integrative

approach is needed to better understand why the world is green and full of herbivores and

how animals balance the trade-offs between eating and being eaten.

78

References

Alldredge, M. W., Peek, J. M., & Wall, W. A. (2002). Nutritional quality of forages used by elk in northern Idaho. Journal of Range Management, 55(3), 253–259.

Allen, R. B., Fitzgerald, A. E., & Efford, M. G. (1997). Long-term changes and seasonal patterns in possum (Trichosurus vulpecula) leaf diet, Orongorongo Valley, Wellington, New Zealand. New Zealand Journal of Ecology, 21(2), 181–186.

Andrew, R. L., Peakall, R., Wallis, I. R., & Foley, W. J. Spatial distribution of defense chemicals and markers and the maintenance of chemical variation. Ecology, 88(3), 716–728.

Ball, J. P., Danell, K., & Sunesson, P. (2001). Response of a herbivore community to increased food quality and quantity: An experiment with nitrogen fertilizer in a boreal forest. Journal of Applied Ecology, 37(2), 247–255.

Barlow, B. A., & Wiens, D. (1977). Host-parasite resemblance in Australian mistletoes: The case for cryptic mimicry. Evolution, 31(1), 69–84.

Belovsky, G. E. (1978). Diet optimization in a generalist herbivore: The moose. Theoretical Population Biology, 14(1), 105–134.

Bennett, A. F., Lumsden, L. F., Alexander, J. S. A., Duncan, P. E., Johnson, P. G., Robertson, P., & Silveira, C. E. (1991). Habitat use by arboreal mammals along an environment gradient in north-eastern Victoria. Wildlife Research, 18(2), 125–146.

Bjorndal, K. A. (1989). Flexibility of digestive responses in two generalist herbivores, the tortoises Geochelone carbonaria and Geochelone denticulata. Oecologia, 78(3), 317–321.

Braithwaite, L. W., Binns, D. L., & Nowlan, R. D. (1988). The distribution of arboreal marsupials in relation to eucalypt forest types in the Eden (NSW) woodchip concession area. Australian Wildlife Research, 15(4), 363–373.

Braithwaite, L. W., Dudzinski, M. L., & Turner, J. (1983). Studies on the arboreal marsupial fauna of eucalypt forests being harvested for woodpulp at Eden, N.S.W. II. Relationship between the fauna density, richness and diversity, and measured variables of the habitat. Australian Wildlife Research, 10(2), 231–247.

Braithwaite, L. W., Turner, J., & Kelly, J. (1984). Studies on the Arboreal Marsupial Fauna of Eucalypt Forests Being Harvested for Wood Pulp at Eden, N.S.W. Iii. Relationships Between Faunal Densities, Eucalypt Occurrence and Foliage Nutrients, and Soil Parent Materials. Australian Wildlife Research, 11(1), 41–48.

Brophy, J. J., & Southwell, I. A. (2002). Eucalyptus chemistry. In J. J. W. Coppen (Ed.), Eucalyptus: The Genus Eucalyptus (pp. 102–161). London: Taylor & Francis.

Brown, K., Innes, J., & Shorten, R. (1993). Evidence that possums prey on and scavenge birds’ eggs, birds and mammals. Notornis, 40(3), 169–177.

Brown, K., Moller, H., & Innes, J. (1996). Sign left by brushtail possums after feeding on bird eggs and chicks. New Zealand Journal of Ecology, 20(2), 277–284.

Burchfield, E., Agar, N. S., Hume, I. D. (2005). Effects of terpenes and tannins on some physiological and biochemical parameters in two species of phalangerid possums (Marsupialia: Phalangeriadae). Australian Journal of Zoology, 53(6), 395–402.

Bureau of Meteorology, Strathbogie. (2013). Summary climate statistics. Retrieved 13th February, 2013, from http://www.bom.gov.au/climate/averages/tables/cw_082042.shtml

Burnham, K. P., & Anderson, D. R. (2002). Model Selection and Multimodel Inference: A practical Information-Theoretic Approach (2 ed.). New York: Springer.

Carr, G. W., Moysey, E. D., Mathews, S., Frood, D., White, M., Griffioen, P., Rosengren, N. (2006). The location of Spring Soak and Peatland wetlands in the Goulburn Broken Catchment—Wetland Implementation Plan Stage. A report to the Goulburn Broken Catchment Management Authority.

79

Catchment Scale Land Use Map. (2010). Retrieved from http://adl.brs.gov.au/landuse/index.cfm?fa=main.downloadData

Charnov, E. L. (1976). Optimal foraging, the marginal value theorem. Theoretical Population Biology, 9(2), 129-136.

Cork, S. J. (1990). Digestive physiology of koalas. Proceedings of the Australian Physiological and Pharmacological Society, 21(2), 95–104.

Cork, S. J., & Foley, W. J. (1991). Digestive and metabolic strategies of arboreal mammalian folivores in relation to chemical defenses in temperate and tropical forests. In R. T. Palo & C. T. Robbins (Eds.), Plant defenses against mammalian herbivory (pp. 133–166). Boca Raton: CRC Press.

Cork, S. J., & Foley, W. J. (1997). Digestive and metabolic adaptations of arboreal marsupials for dealing with plant antinutrients and toxins. In N. R. Saunders & L. A. Hinds (Eds.), Marsupial Biology—Recent Research, New Perspectives (pp. 204-226). Sydney: UNSW Press.

Cowan, P. (1990). Fruits, seeds, and flowers in the diet of brushtail possums, Trichosurus vulpecula, in lowland podocarp/mixed hardwood forest, Orongorongo Valley, New Zealand. New Zealand Journal of Zoology, 17(4), 549-566.

Cowan, P. E. (1989). A vesicular-arbuscular fungus in the diet of brushtail possums, Trichosurus vulpecula. New Zealand Journal of Botany, 27(1), 129–131.

Cowan, P. E., & Moeed, A. (1987). Invertebrates in the diet of brushtail possums, Trichosurus vulpecula, in lowland podocarp/broadleaf forest, Orongorongo Valley, Wellington, New Zealand. New Zealand Journal of Zoology, 14(2), 163–177.

Crawley, M. J. (1983). Herbivory: The dynamics of animal—plant interactions. Berkeley: University of California Press.

Dai, X., Jia, X., Zhang, W., Bai, Y., Zhang, J., Wang, Y., & Wang, G. (2009). Plant height–crown radius and canopy coverage–density relationships determine above-ground biomass—density relationship in stressful environments. Biology Letters, 5(4), 571–573.

Dearden, B. L., Pegau, R. E., & Hansen, R. M. (1975). Precision of microhistological estimates of ruminant food habits. The Journal of Wildlife Management, 39: 402–407.

Dearing, M. D. (1996). Disparate determinants of summer and winter diet selection of a generalist herbivore, Ochotona princeps. Oecologia, 108(3), 467–478.

DeGabriel, J. L., Wallis, I. R., Moore, B. D., & Foley, W. J. (2008). A simple, integrative assay to quantify nutritional quality of browses for herbivores. Oecologia, 156(1), 107–116.

DeGabriel, J. L., Moore, B. D., Foley, W. J., & Johnson, C. N. (2009a). The effects of plant defensive chemistry on nutrient availability predict reproductive success in a mammal. Ecology, 90(3), 711–719.

DeGabriel, J. L., Moore, B. D., Shipley, L. A., Krockenberger, A. K., Wallis, I. R., Johnson, C. N., & Foley, W. J. (2009b). Inter-population differences in the tolerance of a marsupial folivore to plant secondary metabolites. Oecologia, 161(3), 539–548.

DeGabriel, J. L., Moore, B. D., Marsh, K. J., & Foley, W. J. (2010). The effect of plant secondary metabolites on the interplay between the internal and external environments of marsupial folivores. Chemoecology, 20(2), 97–108.

DeGabriel, J. L., Moore, B. D., Felton, A. M., Ganzhorn, J. U., Stolter, C., Wallis, I. R., Johnson, C. N., & Foley, W. J (2013). Translating nutritional ecology from the laboratory to the field: Milestones in linking plant chemistry to population regulation in mammalian browsers. Oikos, 123(3), 298–308.

Del Borrello, C. (2009). Home-range and habitat use in sympatric brushtail possums (Trichosurus spp.). (Honours thesis), Melbourne University, Melbourne.

Di Stefano, J., & Newell, G. R. (2008). Diet selection by the swamp wallaby (Wallabia bicolor): Feeding strategies under conditions of changed food availability. Journal of Mammalogy, 89(6), 1540–1549.

80

Digital Elevation Model (DEM-S) Version 1.0. (2002). Retrieved from http://glcf.umd.edu/data/srtm

Downes, S. J., Handasyde, K. A., & Elgar, M. A. (1997). The use of corridors by mammals in fragmented Australian eucalypt forests. Conservation Biology, 11(3), 718–726.

Duffy, J. E., & Hay, M. E. (1991). Food and shelter as determinants of food choice by an herbivorous marine amphipod. Ecology, 72(4), 1286–1298.

Dungan, R. J., O’Cain, M. J., Lopez, M. L., & Norton, D. A. (2002). Contribution by possums to seed rain and subsequent seed germination in successional vegetation, Canterbury, New Zealand. New Zealand Journal of Ecology, 26(2), 121–128.

Dunnet, G. M., Harvie, A. E., & Smit, T. J. (1973). Estimating the proportions of various leaves in the diet of the opossum, Trichosurus vulpecula Kerr, by faecal analysis. Journal of Applied Ecology, 10(3), 737–745.

Ellis, W., Carrick, F., Lundgren, P., Veary, A., & Cohen, B. (1999). The use of faecal cuticle examination to determine the dietary composition of koalas. Australian Zoologist, 31, 127–133.

Enquist, B. J., & Niklas, K. J. (2002). Global allocation rules for patterns of biomass partitioning in seed plants. Science, 295(5559), 1517–1520.

Eschler, B. M., Pass, D. M., Willis, R., & Foley, W. J. (2000). Distribution of foliar formylated phloroglucinol derivatives amongst Eucalyptus species. Biochemical Systematics and Ecology, 28(9), 813–824.

ESRI, ArcGIS Desktop 9.3. (2009). Redlands, CA: Environmental Systems Resource Institute.

ESRI, ArcPad 8.0. (2009). Redlands, CA: Environmental Systems Resource Institute. Evans, M. C. (1992). Diet of the brushtail possum Trichosurus vulpecula (Marsupialia:

Phalangeridae) in central Australia. Australian Mammalogy, 15, 25–30. Eymann, J., Herbert, C., & Cooper, D. (2006). Management issues of urban common

brushtail possums Trichosurus vulpecula: A loved or hated neighbour. Australian Mammalogy, 28(2), 153–171.

Felton, A. M., Felton, A., Raubenheimer, D., Simpson, S. J., Foley, W. J., Wood, J. T., Lindenmayer, D. B. (2009). Protein content of diets dictates the daily energy intake of a free-ranging primate. Behavioral Ecology, 20(4), 685–690.

Fitzgerald, A. E. (1984). Diet of the possum (Trichosurus vulpecula) in three Tasmanian forest types and its relevance to the diet of possums in New Zealand forests. In A. P. Smith & I. D. Hume (Eds.), Possums and gliders (pp. 137–143). Sydney: Surrey Beatty & Sons.

Fitzgerald, A. E., & Waddington, D. C. (1979). Comparison of two methods of fecal analysis of herbivore diet. The Journal of Wildlife Management, 43(2), 468–473.

Fitzgerald, B. M., & Gibb, J. (2001). Introduced mammals in a New Zealand forest: Long-term research in the Orongorongo Valley. Biological Conservation, 99(1), 97–108.

Fox, L. R., & Morrow, P. A. (1983). Estimates of damage by herbivorous insects on Eucalyptus trees. Australian Journal of Ecology, 8(2), 139–147.

Freeland, W. J., & Janzen, D. H. (1974). Strategies in herbivory by mammals: The role of plant secondary compounds. The American Naturalist, 108(961), 269–287.

Freeland, W. J., & Winter, J. W. (1975). Evolutionary consequences of eating:Trichosurus vulpecula (Marsupialia) and the genus Eucalyptus. Journal of Chemical Ecology, 1(4), 439–455.

Gibbons, P., & Lindenmayer, D. (2002). Tree hollows and wildlife conservation in Australia. Collingwood: CSIRO Publishing.

Gleadow, R. M., Haburjak, J., Dunn, J. E., Conn, M. E., & Conn, E. E. (2008). Frequency and distribution of cyanogenic glycosides in Eucalyptus L'Hérit. Phytochemistry, 69(9), 1870–1874.

81

Grant, C. C., & Scholes, M. C. (2006). The importance of nutrient hot-spots in the conservation and management of large wild mammalian herbivores in semi-arid savannas. Biological Conservation, 130(3), 426–437.

Green, W. (1985). A review of ecological studies relevant to management of the common brushtail possum: New Zealand Forest Service.

Green, W. Q., & Coleman, J. D. (1986). Movement of possums(Trichosurus vulpecula) between forest and pasture in Westland, New Zealand: Implications for bovine tuberculosis transmission. New Zealand Journal of Ecology, 9, 57–70.

Hardy, A. C. (1967). Great waters: A voyage of natural history to study whales, plankton and the waters of the Southern Ocean in the old Royal Research Ship "Discovery". London: Collins.

Hergt, J. M., Phillips, G. N., & Ely, K. S. (2002). Strathbogie Igneous Complex, central Victoria. In G. N. Phillips & K. S. Ely (Eds.), Victoria Undercover: Benalla 2002 Conference Proceedings and Field guide: Collaborative Geoscience in Northern Victoria (pp. 43–50). Collingwood: CSIRO Publishing.

Hindell, M. A., Handasyde, K. A., & Lee, A. K. (1985). Tree species selection by free-ranging koala populations in Victoria. Wildlife Research, 12(2), 137–144.

How, R., & Hillcox, S. (2000). Brushtail possum, Trichosurus vulpecula, populations in south-western Australia: Demography, diet and conservation status. Wildlife Research, 27(1), 81–89.

Hume, I. D. (1999). Marsupial nutrition. Cambridge: Cambridge University Press. Hunter, M. D., & Price, P. W. (1992). Playing chutes and ladders: Heterogeneity and the

relative roles of bottom-up and top-down forces in natural communities. Ecology, 73(3), 723–732.

Irlbeck, N. A., & Hume, I. D. (2003). The role of Acacia in the diets of Australian marsupials ? A review. Australian Mammalogy, 25(2), 121–134.

Johnson, A. R., Wiens, J. A., Milne, B. T., & Crist, T. O. (1992). Animal movements and population dynamics in heterogeneous landscapes. Landscape Ecology, 7(1), 63–75.

Johnson, C. N., Clinchy, M., Taylor, A. C., Krebs, C. J., Jarman, P. J., Payne, A., & Ritchie, E. G. (2001). Adjustment of offspring sex ratios in relation to the availability of resources for philopatric offspring in the common brushtail possum. Proceedings of the Royal Society of London. Series B: Biological Sciences, 268(1480), 2001–2005.

Jones, K. M. W., & Krockenberger, A. K. (2007). Determining the diet of cryptic folivores: An assessment of diet analysis techniques using the green ringtail possum (Pseudochirops archeri) as a case study. Wildlife Research, 34(5), 352–358.

Jones, K. M. W., MacLagan, S. J., & Krockenberger, A. K. (2006). Diet selection in the green ringtail possum (Pseudochirops archeri): A specialist folivore in a diverse forest. Australian Ecology, 31(7), 799–807.

Kauffman, M. J., Varley, N., Smith, D. W., Stahler, D. R., MacNulty, D. R., & Boyce, M. S. (2007). Landscape heterogeneity shapes predation in a newly restored predator–prey system. Ecology Letters, 10(8), 690–700.

Keith, H., Barrett, D., & Keenan, R. (2000). Review of allometric relationships for estimating woody biomass for New South Wales, the Australian Capital Territory, Victoria, Tasmania, and South Australia National Carbon Accounting System Technical Report 5B. Canberra: Australian Greenhouse Office.

Kerle, J. A. (2001). Possums: The brushtails, ringtails and greater glider. Sydney: UNSW Press. Kerle, J. A. (1984). Variation in ecology of Trichosurus: Its adaptive significance. In A. P.

Smith & I. D. Hume (Eds.), Possums and gliders (pp. 115–128). Sydney: Surrey Beatty & Sons.

Ketterings, Q. M., Coe, R., van Noordwijk, M., Ambagau, Y., & Palm, C. A. (2001). Reducing uncertainty in the use of allometric biomass equations for predicting above-ground tree biomass in mixed secondary forests. Forest Ecology and Management, 146(1), 199–209.

82

Kirmani, S. N., Banks, P. B., & McArthur, C. (2010). Integrating the costs of plant toxins and predation risk in foraging decisions of a mammalian herbivore. Oecologia, 164(2), 349–356.

Lawler, I. R., Foley, W. J., Eschler, B. M., Pass, D. M., & Handasyde, K. (1998). Intraspecific variation in Eucalyptus secondary metabolites determines food intake by folivorous marsupials. Oecologia, 116(1), 160–169.

Li, H., Madden, J. L., & Potts, B. M. (1995). Variation in volatile leaf oils of the Tasmanian Eucalyptus species—1. Subgenus Monocalyptus. Biochemical Systematics and Ecology, 23(3), 299–318.

Lindenmayer, D. B., & Cunningham, R. B. (1997). Patterns of co-occurrence among arboreal marsupials in the forests of central Victoria, southeastern Australia. Australian Journal of Ecology, 22(3), 340–346.

Loney, P. E., McArthur, C., Potts, B. M., & Jordan, G. J. (2006). How does ontogeny in a Eucalyptus species affect patterns of herbivory by Brushtail Possums? Functional Ecology, 20(6), 982–988.

MacLennan, D. G. (1984). The feeding behaviour and activity patterns of the brushtail possum, Trichosurus vulpecula, in an open eucalypt woodland in Southeast Queensland. In A. P. Smith & I. D. Hume (Eds.), Possums and gliders (pp. 155–161). Sydney: Surrey Beatty & Sons.

Mangione, A. M., Dearing, M. D., & Karasov, W. H. (2000). Interpopulation differences in tolerance to creosote bush resin in desert woodrats (Neotoma lepida). Ecology, 81(8), 2067–2076.

Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L., & Erickson, W. P. (2002). Resource selection by animals: Statistical design and analysis for field studies. London: Kluwer Academic Pub.

March, W. A., & Watson, D. M. (2007). Parasites boost productivity: Effects of mistletoe on litterfall dynamics in a temperate Australian forest. Oecologia, 154(2), 339–347.

March, W. A., & Watson, D. M. (2010). The contribution of mistletoes to nutrient returns: Evidence for a critical role in nutrient cycling. Austral Ecology, 35(7), 713–721.

Maron, J. L., & Crone, E. (2006). Herbivory: Effects on plant abundance, distribution and population growth. Proceedings of the Royal Society B: Biological Sciences, 273(1601), 2575–2584.

Marsh, K. J., Foley, W. J., Cowling, A., & Wallis, I. R. (2003a). Differential susceptibility to Eucalyptus secondary compounds explains feeding by the common ringtail (Pseudocheirus peregrinus) and common brushtail possum (Trichosurus vulpecula). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology, 173(1), 69–78.

Marsh, K. J., Wallis, I. R., & Foley, W. J. (2003b). The effect of inactivating tannins on the intake of Eucalyptus foliage by a specialist Eucalyptus folivore (Pseudocheirus peregrinus) and a generalist herbivore (Trichosurus vulpecula). Australian Journal of Zoology, 51(1), 31–42.

Marsh, K. J., Wallis, I. R., Andrew, R. L., & Foley, W. J. (2006a). The detoxification limitation hypothesis: Where did it come from and where is it going? Journal of Chemical Ecology, 32(6), 1247–1266.

Marsh, K. J., Wallis, I. R., McLean, S., Sorensen, J. S., & Foley, W. J. (2006b). Conflicting demands on detoxification pathways influence how common brushtail possums choose their diets. Ecology, 87(8), 2103–2112.

Martin, J. K., & Martin, A. A. (2004). Wildlife conservation beyond the parks: The Strathbogie Ranges. In D. Lunney (Ed.), Conservation of Australia's Forest Fauna (2nd ed., pp. 508–512). Mosman: Royal Zoological Society of New South Wales.

Martin, J. K., & Martin, A. A. (2007). Resource distribution influences mating system in the bobuck (Trichosurus cunninghami: Marsupialia). Oecologia, 154(1), 227–236.

83

McArthur, C., & Turner, S. (1997). Feeding preferences of captive brushtail possums for eucalypt and acacia foliage. Tasforests Hobart, 9, 155–162.

McArthur, C., Goodwin, A., & Turner, S. (2000). Preferences, selection and damage to seedlings under changing availability by two marsupial herbivores. Forest Ecology and Management, 139(1-3), 157–173.

McArthur, C., Hagerman, A., & Robbins, C. T. (1991). Physiological strategies of mammalian herbivores against plant defenses. In R. T. Palo & C. T. Robbins (Eds.), Plant Defenses against Mammalian Herbivores (pp. 103–114). Boca Raton: CRC Press.

McLean, S., & Foley, W. J. (1997). Metabolism of Eucalyptus terpenes by herbivorous marsupials. Drug Metabolism Reviews, 29(1–2), 213–218.

Montague, T. L. (2000). The brushtail possum: Biology, impact and management of an introduced marsupial. Lincoln: Manaaki Whenua Press.

Moore, B. D., & Foley, W. J. (2000). A review of feeding and diet selection in koalas (Phascolarctos cinereus). Australian Journal of Zoology, 48(3), 317–333.

Moore, B. D., Wallis, I. R., Marsh, K. J., & Foley, W. J. (2004a). The role of nutrition in the conservation of the marsupial folivores of eucalypt forests. In D. Lunney (Ed.), Conservation of Australia’s forest fauna (2 ed., pp. 549–575). Mosman: Royal Zoological Society of New South Wales.

Moore, B. D., Wallis, I. R., Palá-Paúl, J., Brophy, J. J., Willis, R. H., & Foley, W. J. (2004b). Antiherbivore chemistry of Eucalyptus—cues and deterrents for marsupial folivores. Journal of Chemical Ecology, 30(9), 1743–1769.

Moore, B. D., Wallis, I. R., Wood, J. T., & Foley, W. J. (2004c). Foliar nutrition, site quality, and temperature influence foliar chemistry of tallowwood (Eucalyptus microcorys). Ecological Monographs, 74(4), 553–568.

Moore, B. D., & Foley, W. J. (2005). Tree use by koalas in a chemically complex landscape. Nature, 435(7041), 488–490.

Moore, B. D., Lawler, I. R., Wallis, I. R., Beale, C. M., & Foley, W. J. (2010). Palatability mapping: A koala's eye view of spatial variation in habitat quality. Ecology, 91(11), 3165–3176.

Morrison, W. E., & Hay, M. E. (2011). Herbivore preference for native vs. exotic plants: Generalist herbivores from multiple continents prefer exotic plants that are evolutionary naïve. PLoS One, 6(3), e17227.

Nersesian, C. L., Banks, P. B., & McArthur, C. (2011). Titrating the cost of plant toxins against predators: Determining the tipping point for foraging herbivores. Journal of Animal Ecology, 80(4), 753–760.

Nichols, W. F., Killingbeck, K. T., & August, P. V. (1998). The Influence of Geomorphological Heterogeneity on Biodiversity II. A Landscape Perspective. Conservation Biology, 12(2), 371–379.

Noble, I. R. (1989). Ecological traits of the Eucalyptus L’Hérit. Subgenera Monocalyptus and Symphyomyrtus. Australian Journal of Botany, 37(3), 207–224.

Norton, T. W, & Neave, H. M. (1990). Koala habitat: A conceptual functional model. In G. Gordon (Ed.), Koalas: Research for Management (pp. 93-101). Brisbaine: World Koala Research Inc.

O’Reagain, P. J., & Schwartz, J. (1995). Dietary selection and foraging strategies of animals on rangeland. Coping with spatial and temporal variability. Paper presented at the International Symposium on the Nutrition of Herbivores.

O’Reilly-Wapstra, J. M., & Cowan, P. (2010). Native plant/herbivore interactions as determinants of the ecological and evolutionary effects of invasive mammalian herbivores: The case of the common brushtail possum. Biological Invasions, 12(2), 373–387.

Pacala, S. W., & Roughgarden, J. (1982). Spatial heterogeneity and interspecific competition. Theoretical Population Biology, 21(1), 92–113.

84

Pass, D. M., Foley, W. J., & Bowden, B. (1998). Vertebrate herbivory on Eucalyptus—identification of specific feeding deterrents for common ringtail possums (Pseudocheirus peregrinus) by bioassay-guided fractionation of Eucalyptus ovata foliage. Journal of Chemical Ecology, 24(9), 1513–1527.

Pausas, J. G., Braithwaite, L. W., & Austin, M. P. (1995). Modelling habitat quality for arboreal marsupials in the south coastal forests of New South Wales, Australia. Forest Ecology and Management, 78(1-3), 39–49.

Phillips, G. N., Ely, K. S., Hergt, J. M., Cownley, D. G., Paul, B., & Dickins, C. (2002). 3-dimensional geometry of the Strathbogie batholith. In G. N. Phillips & K. S. Ely (Eds.), Victoria Undercover: Benalla 2002 Conference Proceedings and Field guide: Collaborative Geoscience in Northern Victoria (pp. 55–60). Collingwood: CSIRO Publishing.

Provenza, F. D., Villalba, J. J., Dziba, L. E., Atwood, S. B., & Banner, R. E. (2003). Linking herbivore experience, varied diets, and plant biochemical diversity. Small Ruminant Research, 49(3), 257–274.

Provenza, F. D., Villalba, J. J., Wiedmeier, R., Lyman, T., Owens, J., Lisonbee, L.,… & Lee, S (2009). Value of plant diversity for diet mixing and sequencing in herbivores. Rangelands, 31(1), 45–49.

Pulliam, H. R. (1975). Diet optimization with nutrient constraints. The American Naturalist, 109(970), 765–768.

Pyke, G. H. (1984). Optimal foraging theory: A critical review. Annual Review of Ecology and Systematics, 15(1), 523–575.

Pyke, G. H., Pulliam, H. R., & Charnov, E. L. (1977). Optimal foraging: A selective review of theory and tests. Quarterly Review of Biology, 52(2), 137–154.

R: A language and environment for statistical computing. R 2.11.1. (2010). Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.R-project.org

Raubenheimer, D., & Simpson, S. J. (2003). Nutrient balancing in grasshoppers: Behavioural and physiological correlates of dietary breadth. Journal of Experimental Biology, 206(10), 1669–1681.

Reid, N., & Lange, R. T. (1988). Host specificity, dispersion and persistence through drought of two arid zone mistletoes. Australian Journal of Botany, 36(3), 299–313.

Reid, N. (1997). Control of mistletoes by possums and fire: A review of the evidence. The Victorian Naturalist, 114(3), 149–158.

Reid, N., & Yan, Z. (2000). In P. J. Keane, G. A. Kile, F. D. Podger & B. N. Brown (Eds.), Diseases and Pathogens of Eucalypts (pp. 353–375). Collingwood: CSIRO Publishing.

Ritchie, M. E. (2009). Scale, heterogeneity, and the structure and diversity of ecological communities (Vol. 45). Princeton: Princeton University Press.

Robinson, B. W., Wilson, D. S. (1998) Optimal foraging, specialization, and a solution to Liem’s Paradox. The American Naturalist, 151(3), 223–235.

Rundle, A. S., & Rowe, R. K. (1974). A Study of the Land in the Catchment of the Broken River. Kew: Soil Conservation Authority of Victoria.

Sakaguchi, E., & Hume, I. D. (1990). Digesta retention and fibre digestion in brushtail possums, ringtail possums and rabbits. Comparative Biochemistry and Physiology Part A: Physiology, 96(2), 351–354.

Scrivener, N. J., Johnson, C. N., Wallis, I. R., Takasaki, M., Foley, W. J., & Krockenberger, A. K. (2004). Which trees do wild common brushtail possums (Trichosurus vulpecula) prefer? Problems and solutions in scaling laboratory findings to diet selection in the field. Evolutionary Ecology Research, 6(1), 77–87.

Searle, K. R., Hobbs, N. T., & Gordon, I. J. (2007). It's the" foodscape", not the landscape: Using foraging behavior to make functional assessments of landscape condition. Israel Journal of Ecology and Evolution, 53(3), 297–316.

Sedinger, J. S., & Raveling, D. G. (1984). Dietary selectivity in relation to availability and quality of food for goslings of cackling geese. The Auk, 101(2), 295–306.

85

Senft, R. L., Coughenour, M. B., Bailey, D. W., Rittenhouse, L. R., Sala, O. E., & Swift, D. M. (1987). Large herbivore foraging and ecological hierarchies. BioScience, 37(11), 789–799.

Sessions, L. A., & Kelly, D. (2002). Heterogeneity in vertebrate and invertebrate herbivory and its consequences for New Zealand mistletoes. Austral Ecology, 26(6), 571–581.

Shipley, L. A., Forbey, J. S., & Moore, B. D. (2009). Revisiting the dietary niche: When is a mammalian herbivore a specialist? Integrative and Comparative Biology, 49(3), 274–290.

Speiser, B., & Rowell-Rahier, M. (1991). Effects of food availability, nutritional value, and alkaloids on food choice in the generalist herbivore Arianta arbustorum (Gastropoda: Helicidae). Oikos, 62(3), 306–318.

Spotfire S+® Version 8.1 (2010). TIBCO Software Inc., Somerville. Sweetapple, P. J., & Nugent, G. (1998). Comparison of two techniques for assessing

possum (Trichosurus vulpecula) diet from stomach contents. New Zealand Journal of Ecology, 22(2), 181–188.

Sweetapple, P. J., Nugent, G., Whitford, J., & Knightbridge, P. I. (2002). Mistletoe (Tupeia antarctica) recovery and decline following possum control in a New Zealand forest. New Zealand Journal of Ecology, 26(1), 61–72.

Sweetapple, P. J., Fraser, K. W., & Knightbridge, P. I. (2004). Diet and impacts of brushtail possum populations across an invasion front in South Westland, New Zealand. New Zealand Journal of Ecology, 28(1), 19–33.

Sweetapple, P. J. (2008). Spatial variation in impacts of brushtail possums on two Loranthaceous mistletoe species. New Zealand Journal of Zoology, 32(2), 177–185.

Tuft, K. D., Crowther, M. S., Connell, K., Müller, S., & McArthur, C. (2011). Predation risk and competitive interactions affect foraging of an endangered refuge-dependent herbivore. Animal Conservation, 14(4), 447–457.

Tuft, K. D., Crowther, M. S., & McArthur, C. (2011). Multiple scales of diet selection by brush-tailed rock-wallabies (Petrogale penicillata). Australian Mammalogy, 33(2), 169–180.

Turner, M. G. (1989). Landscape ecology: The effect of pattern on process. Annual Review of Ecology and Systematics, 20, 171–197.

Unsicker, S. B., Oswald, A., Köhler, G., & Weisser, W. W. (2008). Complementary effects through dietary mixing enhance the performance of a generalist insect herbivore. Oecologia, 156(2), 313–324.

Van Der Ree, R. (2012). Ecology of arboreal marsupials in a network of remnant linear habitats. (Ph. D. thesis), Deakin University, Geelong.

Van Eeden, L., Di Stefano, J., & Coulson, G. (2011). Diet selection by the brush-tailed rock-wallaby (Petrogale penicillata) in East Gippsland, Victoria. Australian Mammalogy, 32(2), 162–168.

Van Tienderen, P. H. (1991). Evolution of generalists and specialist in spatially heterogeneous environments. Evolution, 45(6), 1317–1331.

Venables, W. N., & Ripley, B. D. (2002). Modern Applied Statistics with S (4 ed.). New York: Springer.

Wallis, I. R., Watson, M. L., & Foley, W. J. (2002). Secondary metabolites in Eucalyptus melliodora: Field distribution and laboratory feeding choices by a generalist herbivore, the common brushtail possum. Australian Journal of Zoology, 50(5), 507–519.

Watson, D. M. (2001). Mistletoe—A Keystone Resource in Forests and Woodlands Worldwide. Annual Review of Ecology and Systematics, 32(1), 219–249.

Watson, D. M. (2011). Mistletoes of southern Australia. Collingwood: CSIRO Publishing. Watts, D. P. (1984). Composition and variability of mountain gorilla diets in the central

Virungas. American Journal of Primatology, 7(4), 323–356.

86

Wayne, A. F., Cowling, A., Rooney, J. F., Ward, C. G., Wheeler, I. B., Lindenmayer, D. B., & Donnelly, C. F. (2006). Factors affecting the detection of possums by spotlighting in Western Australia. Wildlife Research, 32(8), 689–700.

Westoby, M. (1974). An analysis of diet selection by large generalist herbivores. The American Naturalist, 108(961), 290–304.

Westoby, M. (1978). What are the biological bases of varied diets? The American Naturalist, 112(985), 627–631.

Wiggins, N. L., McArthur, C., McLean, S., & Boyle, R. (2003). Effects of two plant secondary metabolites, cineole and gallic acid, on nightly feeding patterns of the common brushtail possum. Journal of Chemical Ecology, 29(6), 1447–1464.

Wiggins, N., McArthur, C., & Davies, N. W. (2006a). Diet switching in a generalist mammalian folivore: Fundamental to maximising intake. Oecologia, 147(4), 650–657.

Wiggins, N., McArthur, C., & Davies, N. W., & McLean, S. (2006b). Behavioral responses of a generalist mammalian folivore to the physiological constraints of a chemically defended diet. Journal of Chemical Ecology, 32(6), 1133–1147.

Wiggins, N. L., McArthur, C., Davies, N. W., & McLean, S. (2008). Spatial scale of the patchiness of plant poisons: A critical influence on foraging efficiency. Ecology, 87(9), 2236–2243.

Williams, C. K. (1982). Nutritional properties of some fruits eaten by the possum Trichosurus vulpecula in a New Zealand broadleaf-podocarp forest. New Zealand Journal of Ecology, 5, 16–20.

Woldemichael, G. M., & Wink, M. (2002). Concomitant occurrence of pyrrolizidine and quinolizidine alkaloids in the hemiparasite Osyris alba L. (Santalacea). Biochemical Systematics and Ecology, 30(2), 139–149.

Worton, B. J. (1989). Kernel Methods for Estimating the Utilization Distribution in Home-Range Studies. Ecology, 7(1), 164–168.

Young, H. E. (1937). The mistletoe problem. Brisbaine: Queensland Forestry Department. Zar, J. H. (2009). Biostatistical Analysis (5 ed.). New Jersey: Pearson.

Zweifel‐Schielly, B., Leuenberger, Y., Kreuzer, M., & Suter, W. (2011). A herbivore's food landscape: Seasonal dynamics and nutritional implications of diet selection by a red deer population in contrasting Alpine habitats. Journal of Zoology, 286(1), 68–80.

87

CH

AP

TE

R

3

Home range attributes of the common brushtail possum (Trichosurus vulpecula):

How does individual variability in resource use, food, and shelter influence spatial patterns?

88

CHAPTER 3

Home range attributes of the common brushtail possum (Trichosurus vulpecula):

How does individual variability in resource use, food, and shelter influence spatial patterns?

Introduction

Understanding the factors that shape home range size, resource use, and the spatial

organization of animals within a landscape is one of the fundamental challenges in ecology.

An animal’s home range is defined as the area traversed by an individual performing daily

life activities, such as foraging, resting, shelter-seeking, mating, and raising young (Burt,

1943). Long ago, Seton (1909, p. 23) observed, “No wild animal roams at random over the

country; each has a home region, even if it has not an actual home.” This area restricted use

of space has profound consequences for many ecological processes including animal

distribution and abundance (Gautestad & Mysterud, 2005), habitat use and resource

selection (Rhodes, McAlpine, Lunney, & Possingham, 2005), community structure (Fagan,

Lutscher, & Schneider, 2007), and predator-prey interactions (Lewis & Murray, 1993).

Home range is a spatial expression of individual behaviours that an animal performs to

survive, reproduce, and avoid predators and competitors; therefore, its movement

pathways are influenced by many different biotic and abiotic factors (summarized by

Börger et al., 2006; DeGabriel, Moore, Marsh, & Foley, 2010, Fig. 1). According to

Moorcroft and Lewis (2006) these macroscopic spatial patterns are determined by a large

number of single movement steps, each of which results from the interactions among

individual animal life traits, physiological state, and the external environment. Furthermore,

herbivores living in nutritionally poor environments modify their movement pathways in

response to the availability and nutritional quality of food and adjust accordingly their

foraging behaviour and physiological state (Dearing, Foley, & McLean, 2005).

89

Fig. 1 Graphical model of animal space-use processes (modified from Börger et al., 2006)

90

However, a home range size is not a universal measure of animal space use; it varies

between species, populations, and individuals and depends on the interactions between

individual energetic requirements and the spatial distribution of limiting resources across a

landscape. This individual variability is associated with differences in body size, sex and age,

reproductive status, habitat type, and availability of food and shelter to individuals

(Cederlund & Sand, 1994; Harestad & Bunnel, 1979; Milton & May, 1976; Perry & Garland

Jr, 2002). While shelter can be a limiting resource, the availability, distribution, and quality

of food usually strongly influence the home range size of individuals (DeGabriel, Moore,

Foley, & Johnson, 2009; Mares et al., 1982; Vedder, 2005). Individuals in food-rich

environments are expected to meet their energy requirements over a reduced area,

compared to animals living in areas with sparse or patchily distributed resources (McIntyre

& Wiens, 1999; Said et al., 2009). Social structure and mating patterns can also affect the

home range size of different members of a population with males usually having larger

home ranges than females to maximize mating opportunities and reproductive success

(Martin & Martin, 2007; Sandell, 1989).

Despite this recognition, variability in resource selection and space-use among individuals

within the same population has long been dismissed as ecological noise around the mean

(Ringler, 1983). As summarized by Bourke, Magnan, and Rodriguez (2005), although

individual variability in foraging patterns of fishes was well described by Ivlev (1961), its

ecological significance for the home ranging behaviour and social structure of different

animal species was only recognized much later (Lomnicki, 1988). Roughgarden (1972,

1979) identified two components of individual variability in resource use: the within-

individual component represents the variation in resource use by a single individual

through time, and the between-individual component represents resource diversity used by

different individuals in a population. Between-individual variability may develop from

physiological, behavioural, and morphological differences among animals while within-

individual variability may be caused by specific physiological states, such as hunger and

fear, and behavioural contexts, such as animal experience, food type or distribution, and

motivation. Individual variability in turn can create two types of spatial organisation of

animals: The first type affects an individual’s home range size, while the second type affects

the spacing of animals within a landscape.

91

Arboreal marsupials, with their versatility in diet, dependence on large hollows for shelter

and reproduction, and broad habitat tolerances, are an ideal group to clarify determinants

of individual variability in space use and resource selection. In particular, the study of

arboreal marsupial home ranging behaviour can answer questions regarding whether

individual variability in a home range size is the outcome of differences in energy needs

(i.e., females rearing young, males competing over females or dispersing subadult males),

and availability and nutritional quality of forage.

Previous nutritional studies have shown that variability in the chemical composition of

foliage of closely related Eucalyptus species and individuals of the same species can explain

some of the observed spatial patterns at scales ranging from a couple of meters to

thousands of kilometres (Andrew, Peakall, Wallis, & Foley, 2007; DeGabriel, Moore,

Shipley, et al., 2009; Lawler, Foley, & Eschler, 2000; Moore, Lawler, Wallis, Beale, & Foley,

2010). In addition, the patchy distribution of hollow-bearing trees is considered a limiting

resource in Australia’s fragmented and intensively managed forests and woodlands

affecting the spatial organization of entire populations of arboreal marsupials (Gibbons &

Lindenmayer, 2002).

The present study examines the effects of within- and between-individual variability in

resource use, food and shelter, on the home ranging behaviour of a model generalist

herbivore, the common brushtail possum (Trichosurus vulpecula: Marsupialia, Phalangeridae).

Following the findings presented in Chapter 2 that highlighted the importance of parasitic

plants and different species of trees in a possum diet, this chapter will explore their

combined effects on space use and resource selection by individual animals. This study also

provides baseline data on tree use and the foraging of possums on specific species of trees

and individual trees which will be used in Chapter 4 to measure the nutritional quality of

preferred and avoided trees. Studying individual responses to variation in food resources

and other potentially limiting factors will allow for a better understanding of the common

brushtail possum ecology, its behavioural adaptations to nutritionally poor environments,

spatial organization, and evolutionary success.

92

Aims and hypotheses

The main aim of this study was to determine how availability of resources (i.e., food and shelter) shapes

spatial patterns, foraging, denning behaviour, and tree use of the common brushtail possum (Trichosurus

vulpecula).

In particular, the following hypotheses were tested:

Hypothesis 1: Total home range size and core areas differ among common brushtail

possum individuals and sexes and are influenced by food and shelter availability.

Hypothesis 2: Common brushtail possums show preferences for specific trees and

differentiate trees used as denning sites from those used as a foraging sites.

Hypothesis 3: Trees parasitized by drooping mistletoe (Amyema pendula) are used more

often by common brushtail possums while foraging than would be expected from their

availability within home ranges.

Hypothesis 4: Trees with suitably sized hollows are used as denning sites by common

brushtail possums more often than would be expected from their availability within home

ranges.

93

Study sites

This study was carried out in the Strathbogie Ranges (36° 48´ S, 145° 45´ E) in north-

eastern Victoria, Australia (see Chapter 2 for details; Fig. 2). It compromised a network of

vegetation corridors along Bonnie Doon Rd., Boundary Hill Rd., Creek Junction Rd., and

Ankers Rd (Fig. 3). The surveyed roadsides supported a fine-scale mosaic of Herb-rich

Foothill Forest in drier areas and Swampy Riparian Woodland in seasonally or permanently

inundated areas and along watercourses. The dominant tree species in Herb-rich Foothill

Forest were Eucalyptus radiata (narrow-leaved peppermint), E. viminalis (manna gum), and a

root parasitic shrub/small tree, Exocarpos cupressiformis (cherry ballart). In Swampy Riparian

Woodland, the dominant tree species were Eucalyptus camphora (mountain swamp gum) and

Acacia dealbata (silver wattle). Two mistletoe species were found parasitizing Eucalyptus and

Acacia trees, the abundant Amyema pendula (drooping mistletoe) and the relatively scarce

Muellerina eucalyptoides (creeping mistletoe).

Fig. 2 Herb-rich Foothill Forest and Swampy Riparian Woodland within roadside vegetation corridors in the Strathbogie Ranges (photo by K. Petrović)

94

Fig. 3 Surveyed roadside vegetation corridors in the Strathbogie Ranges, north-eastern Victoria

95

Methods

Description and justification of methods used for establishing a home range size, foraging, and denning

patterns of common brushtail possums

To determine a home range size, foraging, and denning patterns of individual common

brushtail possums, very high frequency (VHF) radio-telemetry was used to track individual

animals from a distance. The presence of a common brushtail possum in a tree at night was

considered a useful proxy of its foraging activity. This method has been widely used in

studies of foraging behaviour of nocturnal arboreal herbivores (Jones & Krockenberger,

2007; Martin & Handasyde, 1999; Moore & Foley, 2005; Scrivener et al., 2004) which are

cryptic or easily disturbed and are, therefore, rarely observed foraging. However, this

method can produce biased results as animals use trees for different reasons. For instance,

animals can use trees as movement routes between foraging patches, as den sites and

escape from predators, or to facilitate social interactions. This method allowed for trees

used as den sites to be identified with full certainty because animals were found resting in a

single tree during the day.

Animal capture

Initial information on the occurrence of common brushtail possums in the study area was

obtained from the spotlighting survey (Chapter 2). From this information, a trapping

procedure was designed that reduced capture of non-target species and minimized stress to

animals. All trapping, handling, and radio-tracking of possums was conducted under the

auspices of the CSU Animal Care and Ethics Committee, approval No 08/090 and

Department of Sustainability and Environment Flora and Fauna permit, no 10004604.

Animal trapping was conducted over 60 nights spread over two trapping periods that

varied in length and location of the capture sites:

1) 15 November–15 December 2008

2) 3 January–2 February 2009

At each site 10–20 large wire traps (Professional Trapping Supplies 600 mm x 260 mm x

240 mm) were set one hour before dusk for three consecutive nights until suitable animals

had been captured (Fig. 4). Traps were baited with apple and peanut butter and placed on

the ground at the base of a tree with hollows or where animals were previously seen

foraging.

96

Traps were partly covered with a durable plastic cover and leaf litter to protect captured

animals from inclement weather. Traps were checked before dawn, and animals not needed

for the study were immediately released at the site of capture. Suitable animals were

transferred from a trap to a hessian bag and transported to the field station for further

processing.

Animal handling

All captured common brushtail possums were weighed to the nearest 0.01 kg and sedated

with an intramuscular injection of tiletamine/zolazepam (Zoletil©, Virbac Australia Pty.

Ltd, Peakhurst, NSW, Australia). A single dose of Zoletil© 10 mg/kg animal body weight

was administrated to facilitate handling and minimise distress for the animals (after Jackson

2003). Ear pinnae of all individuals were tattooed using a tattoo punch and given a unique

number for subsequent identification and future studies. Animal sex was recorded and

standard morphometric measurements were taken to the nearest 0.1 mm (i.e., head length,

head width, hind foot or pes length, and ear length). The pouch condition and reproductive

status of females were recorded; if a pouch young was present, its sex was checked and

head length measured (an index of age). Male scrotal width and testis size were measured

(i.e., length, width, and depth). In addition, a tooth wear class, an index of age based on

wear patterns of the first molar in the upper jaw (1–7; Winter, 1980), and a rump wear

class, an index of body condition (0–3), were established. Based on the criteria set up by

Martin and colleagues (2007) that included body weight, female pouch condition, male

testis size, and tooth wear class, each individual was classified as juvenile, sub-adult, or

adult.

Next, nine randomly selected individuals (3 males and 6 females) out of 16 trapped were

fitted with 30 g VHF radiocollars (Sirtrack Pty. Ltd., Havelock North, New Zealand); each

collar was marked with a reflective tape to facilitate identification at night (Fig. 4). Each

animal was given an apple and left to recover from sedation and handling in a large

wooden box with fresh air for at least 6 hours before being released. In all instances,

possums recovered well without showing any signs of distress and consumed the food that

was provided. Animals were released one hour after dusk at the point of capture or at the

base of their den tree if known. At the end of the study period all animals wearing radio-

collars were retrapped to retrieve the collars and were released at the point of capture. No

animal health or welfare issues arose during the capture, handling, and radio-tracking of

animals.

97

Radio-telemetry survey

To determine tree use (i.e., foraging and denning behaviour) and establish home range size,

animals were fitted with radio-collars operating at discrete frequencies (151–152 MHz);

their movements were radio-tracked using a R100 telemetry receiver (Communications

Specialist, Inc., Orange, CA, USA) and a hand-held three-element yagi antenna (A. F.

Antronics, Inc., Urbana IL, USA). Animals were radio-tracked for 30 days and 30 nights

over four months (November 2008 to February 2009) to collect 60 fixes for each animal.

The previous radio-telemetry studies have found that 60 fixes per animal is the

recommended minimum number of fixes to correctly establish the home range size of

sedentary animals (after Kenward, 2001; Millspaugh & Marzluff, 2001). To minimize the

confounding effects of mating and territorial behaviour and to avoid any inadvertent

impacts on animal welfare, animals were not radio-tracked during a breeding season in

March and April.

98

Fig. 4 Common brushtail possum captured in a wire trap, fitted with a VHF radio-collar and released to its den tree (photo by K. Petrović)

99

Radio-tracking was carried out during the spring and summer, which was considered the

best time to explore how animals responded to the chemical variability of trees containing

the highest concentrations of plant secondary metabolites (see Chapter 4). Moreover,

spreading radio-tracking over a longer period minimized the bias introduced by spatio-

temporal autocorrelation in the animal movement analysis (Hansteen, Andreassen, & Ims,

1997). The same principle was applied when radio-tracking animals over a single night.

Namely, the time at which tracking commenced and the order in which individuals were

tracked was altered each time. All individuals were located between 1 hour after dusk and

approximately 2 hours before dawn or after they emerged from their dens and started

foraging. Radio-tracking was continued during their activity through the night and before

they returned to their dens (Fig. 5). To eliminate the possibility of spatio-temporal

autocorrelation between night fixes, the location of each individual was recorded only once

per night. In all cases, animals were identified with the use of an 80-W spotlight with a red

filter and 8 x 30 binoculars before recording their exact location using a hand-held

positioning device, a Garmin GPS 12 Personal Navigator (Garmin International, Inc.,

Olathe, KS, USA).

Common brushtail possums rested in one location during the day; hence, only one fix per

day was required to determine the location of den trees (Fig. 5). Day radio-fixes or passive

fixes represented animal den sites while night radio-fixes or active fixes represented

foraging activity. Both day and night fixes were considered in the home range analysis as

den sites have been found to influence ranging behaviour of brushtail possums in roadside

vegetation corridors (Del Borrello, 2009; Martin, Handasyde, & Taylor, 2007). Finally,

to estimate a home range size, all day and night fixes were plotted on a digitised base map

with ArcGIS 10.0 Software (ESRI, Redlands, CA, USA) using the animal movement

program, an ESRI ArcView extension (Hooge, Eichenlaub, & Solomon, 1999).

Vegetation survey

Within the established home range boundaries of the nine studied common brushtail

possums, the following information was collected regarding used trees: exact location, tree

species, height and diameter at breast height (DBH), and the number of hollows and

mistletoes. A hollow was considered suitable for possums if its entrance was greater than

10 cm in diameter (Gibbons et al., 2002). Trees used as foraging and denning sites were

described, and each tree was marked with a numbered tag for subsequent identification and

collection of leaf samples for the chemical analyses as described in Chapter 4.

100

To determine the availability of different tree species and the availability of trees with

hollows and mistletoes within each possum’s home range, additional information about

tree species composition and structure was collected. Trees were surveyed systematically at

20 m intervals along transect that was established across the entire home range area on

both sides of the road (Fig. 6). The exact location and identity of all studied trees was

recorded with a hand-held positioning device (Garmin GPS 12 Personal Navigator,

Garmin International, Inc., Olathe, KS, USA). Tree diameter and height were measured

using a hand-held laser hypsometer (LaserAce®, MDL Laser Systems, Ltd., York, UK).

Fig. 5 Night and day radio-tracking of common brushtail possums within roadside vegetation corridors in the Strathbogie Ranges (photo by A. Eberhard)

101

Fig. 6 Schematic diagram depicting vegetation sampling design used within the estimated home range of each common brushtail possum

102

Data analysis

Home range attributes

A home range size was estimated for each of nine studied common brushtail possums

using two different methods: the minimum convex polygon (MCP; Mohr, 1947) and fixed

kernel (FK; Hooge et al., 1999; Worton, 1989). The minimum convex polygon method is

useful due to its simplicity and comparability between studies; it calculates the area

bounded by the outermost fixes. However, there are several problems with this method,

particularly the strong influence of outliers which can inflate estimates of home range size

(Harris et al., 1990; Powell, 2000). The fixed kernel method is, therefore, used in many

studies. This non-parametric statistical method calculates a kernel home range estimate

using Hooge’s algorithm (Hooge et al., 1999). It is considered less sensitive to outliers

because it identifies the areas of greatest use for each individual (Pope, Lindenmayer, &

Cunningham, 2004; Powell, 2000).

In this study, the fixed kernel method was used with a smoothing factor determined by the

least-square cross validation. This method calculated the core home range area (50%), the

area with the highest animal activity, and the total home range area (90%) using both day

(i.e., representing den trees) and night fixes (i.e., representing food trees). The same home

range areas were calculated by the minimum convex polygon method. Following the

findings of Martin and colleagues (2007), the entire width of the habitat, including the road,

was considered when calculating the total home range size as common brushtail possums

were observed crossing the road on numerous occasions. To identify whether the total

home range estimates were calculated correctly, minimum convex polygon areas were

plotted against the number of position fixes for each individual. Home range asymptotes

approached a plateau in all nine cases (Fig. 7).

103

Fig. 7 The minimum convex polygon area (MCP) accumulation curves for the nine radio-tracked common brushtail possums (3 males and 6 females)

Using two methods, the differences between sexes of the calculated total and core home-

range sizes were studied. First, the Shapiro-Wilk test was completed to test for the

normality of data. Next, the F-test was performed to test if variances of the two

populations were equal. If the obtained P-value was > 0.05, then the t-test was used to

compare two populations under the assumption of equal variances. However, if the P-value

of the F-test was < 0.05 then the non-parametric, the Wilcoxon signed ranks test, was used.

To determine the entire area of common brushtail possum arboreal activity, including all

trees that were used as den or foraging sites by individual animals, 100% minimum convex

polygon estimates, as well as foraging and denning areas, were calculated (after

Lindenmayer, Welsh, & Donnelly, 1997; Martin, 2006). In this study, repeated use of the

same den or food tree by common brushtail possums was treated as a single data point to

avoid biasing range estimates by multiple records. The observed differences in denning and

foraging ranges between sexes were compared using the t-test and the Wilcoxon signed

ranks test. Finally, the Pearson product-moment correlation was used to investigate

whether foraging and denning areas influenced the total home range size.

104

Determinants of tree use by common brushtail possums

To identify individual variability in tree use, each animal was treated as a separate case and

results were pooled only when it was possible to make general conclusions about the

studied animals. All calculations were performed in R version 2.11.1 (Copyright (C) 2010

The R Foundation for Statistical Computing) with a significance level α = 0.05 unless

otherwise indicated.

Foraging preferences of each common brushtail possum were determined using a

bootstrap or randomization procedure (after Jones, Maclagan and Krockenberger 2006).

First, the above ground biomass of each tree species in each home range was calculated

using a general allometric equation (after Keith et al., 2000, see Chapter 2 for details):

lny = -2.3267 + 2.4855 lnx, where y is the aboveground biomass (kg) and x is tree DBH (cm)

Next, 30 records (corresponding with the number of nightly tree sightings for each

common brushtail possum) were randomly selected (with replacement) from the total

number of trees sampled within each home range, and the tree species was recorded. This

procedure was repeated 10,000 times to produce an expected distribution of randomly

selected tree species. The number of sightings of common brushtail possums in a specific

tree species was compared with the randomly generated distribution, and the probability of

sighting was determined directly from the randomized distribution.

To establish whether trees with hollows of a suitable size for common brushtail possums

or trees parasitized by mistletoes were used more often than would be expected from their

availability within animal home ranges, contingency tables were produced. The significance

of the difference between the two proportions was assessed using the Pearson's chi-square

test with Yates' continuity correction to avoid overestimation of P values. When

proportions were close or equal to zero, the bootstrap procedure (100,000 repetitions) was

used instead. The Wilcoxon rank-sum test was used to identify possible differences in the

distance of single and multiple used trees from the closest den tree. The same analysis was

performed to ascertain differences in the distance between trees parasitized by drooping

mistletoe or free from mistletoe and the closest den tree.

105

To investigate the relationship between used and unused trees by common brushtail

possums and predictor variables (i.e., tree species, height and diameter, biomass, and

number of mistletoes and hollows) the binomial generalized linear mixed model (GLMM)

was fitted with an animal considered a random factor. The same model was used to

investigate the relationship between food or den trees and predictor variables. Firstly, any

highly intercorrelated explanatory variables (Spearman’s Rho > 0.7) were excluded from

the analysis to avoid multicollinearity in the modelling process (Tabachnick & Fidell, 1996).

Next, GLMM was fitted with Laplace approximation used to generate parameter estimates.

Finally, the step Akaike’s Information Criterion (hereafter AIC) function from the MASS

library was employed to identify the best model with a combination of variables explaining

the possum tree presence (Venables & Ripley, 2002). This function is a stepwise algorithm

that sequentially searches through all possible models for the one that minimizes the AIC.

Namely, it starts from the simplest model (the intercept) and adds one step at a time to a

new variable, considered the best by the AIC, to each of the consecutive models. On each

step, the AIC from the new model is compared to the previous model, and if the AIC from

the new model is smaller, the step is repeated. If the AIC from the new model is not

smaller, the algorithm stops and the best model is considered to be the one from the

previous step. If there are no more variables to be added to the new model, the algorithm

stops and the full model is considered the best. The significance of variables from the final

selected model is estimated using the z-test.

It is worth noting, that the AIC (Akaike’s Information Criterion) was used because the

information-theoretic approach is a recommended alternative to traditional hypothesis

testing since it is more tolerant to violations of parametric statistics that are commonly

encountered in ecological data. Moreover, the AIC method allows for the comparison of

multiple working hypotheses and does not rely solely on the use of ecologically arbitrary

P values for determining significance (Burnham & Anderson, 2002). Analysis was carried

out using the statistical program Spotfire S+® Version 8.1 (2010, TIBCO Software Inc.

Somerville, MA).

106

Results

Home range attributes

In this study, the nine common brushtail possums (3 males and 6 females, Table 1) were

radio-tracked over a four-month period (Table 2) to establish their total home range size

(90%), core area (50%), as well as their denning and foraging ranges (100%). The location

of home ranges of the nine radio-tracked individuals within a landscape of the Strathbogie

Ranges is shown in Fig. 8. The average total home range size (90%), calculated using the

minimum convex polygon method for both sexes combined, was 2.2 ± 0.6 ha (mean ± s.e.)

while the average core area (50%) was 0.4 ± 0.2 ha. The average total home range of males

was 4.1 ± 0.6 ha and of females 1.2 ± 0.4 ha. The total home range of males was

significantly larger than those of females (t = - 3.98, df = 7, P < 0.05; Table 3).

The average core area of males was more than twice the size of females, but the result was

not significant (0.8 ± 0.4 ha males; 0.3 ± 0.2 ha; W = 3, P = 0.167; Table 3). The total and

core home range areas of the nine common brushtail possums (3 males and 6 females)

calculated using the minimum convex polygon method are shown in Fig. 9.

Similarly, the average total home range (90%) of the nine common brushtail possums

calculated using the fixed kernel method was 2.4 ± 1.0 ha, and a core area 0.5 ± 0.2 ha,

with males having larger total and core home ranges than females (t = -2.25, df = 7, P =

0.06; t = - 2.52, df = 7, P < 0.05 respectively; Table 3). The average total home range of

males was 4.9 ± 2.2 ha and of females 1.2 ± 0.5 ha while the average core area of males

was 1.0 ± 0.4 ha and of females 0.3 ± 0.1 ha (Table 3). The total and core home range

areas of the nine common brushtail possums (3 males and 6 females), which were

calculated using the fixed kernel method, are shown in Fig. 10.

The total home range (100%) and foraging and denning ranges of the nine common

brushtail possums (3 males and 6 females) were established using the minimum convex

polygon method. The total home range and foraging ranges of the nine radio-tracked

common brushtail possums were strongly correlated (r = 0.91, P < 0.001) whereas the total

home range was less strongly correlated with denning ranges of all studied animals (r =

0.64, P = 0.07). The percent of overlap between the foraging area and total-home range

was 72.3% for males and 98% for females. The average foraging area of males was 4.8 ±

0.9 ha and of females’ 2.0 ± 0.8 ha while the average denning area of males was 1.5 ± 0.8

ha and of females’ 0.2 ± 0.1 ha (Table 4).

107

Males had larger foraging and denning ranges than females, but the result was not

significant (t = -2.18, df = 7, P = 0.07; W = 2, P = 0.10 respectively). Both sexes had

significantly larger foraging than denning home ranges (W = 0, P < 0.05). The total home

range (100%) and denning and foraging ranges of the nine common brushtail possums

(3 males and 6 females), which were calculated using the minimum convex polygon

method, are shown in Fig. 11.

Determinants of tree use by common brushtail possums

Approximately two-thirds of nightly sightings of the nine common brushtail possums

(3 males and 6 females) were in exotic species of trees (30.3%) and Eucalyptus radiata

(28.7%), followed by less frequently used E. camphora (13.4%) and E. viminalis (12.2%), and

occasionally used Exocarpos cupressiformis (3.9%) and Acacia dealbata (3.5%). In 5% of cases,

common brushtail possums were recorded foraging on the ground. Of the six tree species

most commonly used by common brushtail possums while foraging, Eucalyptus species

were used in accordance with their availability expressed as above ground biomass. Exotic

species of trees and a root parasite Exocarpos cupressiformis were used proportionally to their

availability and only Acacia dealbata was used significantly less often than expected (P <

0.001, Fig. 12).

Out of 520 recorded trees available to common brushtail possums within their home

ranges, 9.6% were parasitized by drooping mistletoe (Amyema pendula) with on average 3.7

± 0.7 mistletoe clumps per tree. Mistletoe infection rates were not uniform across tree

species. Most often mistletoes were found on E. radiata, representing almost half of all

parasitized trees (46%) with on average 5.0 ± 1.2 mistletoe clumps per tree, followed by E.

viminalis (24%) with 1.6 ± 0.2 mistletoe clumps per tree. By contrast, out of 210 trees used

by common brushtail possums while foraging, 20.5% were parasitized by drooping

mistletoe with on average 4.7 ± 0.9 mistletoe clumps per tree. Eucalyptus radiata represented

67.4% parasitized trees with on average of 5.6 ± 1.3 mistletoe clumps per tree. The chi-

square test showed that common brushtail possums used mistletoe parasitized trees more

often while foraging than expected by chance alone (χ2 = 14.911, df = 1, P < 0.001).

108

Table 1 Morphometric features and a life stage of the nine radio-tracked common brushtail possums (3 males and 6 females)

Animal Weight Head Ear Pes Tooth Rump Life ID

(kg)

Length (mm)

Width (mm)

Length (mm)

Length (mm)

Wear class

Wear class

Stage

Male 1 3.0 91.4 49.1 59.3 53.3 4 0 Adult

Male 2 3.3 94.6 51.1 57.5 43.4 5 2 Adult

Male 3 2.4 98.3 57.7 73.9 60.1 2 2 Subadult

Female 1 1.6 83.6 44.9 61.6 59.2 2 1 Subadult

Female 2 2.9 93.7 50.9 79.6 61.9 5 0 Lactating

Female 3 2.0 92.2 52.1 67.7 63.4 3 0 Adult

Female 4 3.0 100.2 60.4 63.4 57.8 4 0 Adult

Female 5 2.7 100.8 56.7 77.0 60.2 6 0 Lactating

Female 6 2.6 93.6 52.2 56.3 63.3 2 3 Adult

109

Table 2 Details of a radio-telemetry survey of the nine common brushtail possums (3 males and 6 females)

Animal ID

Tracking period

Days tracked

Total No. passive fixes (Den trees)

Total No. active fixes (Food trees)

Male 1

16.11.2008.–07.04.2009.

32

32 (15)

30 (27)

Male 2 25.11.2008.–12.04.2009. 30 30 (4) 29 (27)

Male 3 28.1120.08.–19.04.2009. 32 29 (12) 32 (28)

Female 1 25.11.2008.–18.04.2009. 30 30(13) 30 (25)

Female 2 12.01.2009.–19.04.2009. 30 30 (2) 30 (28)

Female 3 12.01.2009.–23.04.2009. 30 30 (3) 29 (25)

Female 4 11.01.2009.–21.04.2009. 30 30 (5) 30 (27)

Female 5 03.12.2008.–08.04.2009. 31 31 (3) 30 (25)

Female 6 14.12.2008.–18.04.2009. 30 30 (1) 30 (26)

110

Fig. 8 Location of home ranges of the nine common brushtail possums (3 males and 6 females) within a landscape of the Strathbogie Ranges, Victoria

111

Table 3 The total home range (90%) and core area (50%) of the nine common brushtail possums (3 males and 6 females) calculated using the minimum convex polygon (MCP) and fixed kernel (FK) method. A zero value represents repeated occupancy of the same tree by the animal

Table 4 The total home range (100%), foraging and denning ranges of the nine common brushtail possums (3 males and 6 females) calculated using the minimum convex polygon (MCP) method. A zero value represents repeated occupancy of the same tree by the animal

Animal ID MCP 90% (ha) MCP 50% (ha) FK 90% (ha) FK 50% (ha)

male01 3.07 1.29 5.05 1.26 male02 4.99 0.02 0.93 0.24 male03 4.39 1.02 8.71 1.56 female01 1.66 1.00 3.54 0.67 female02 0.30 0.09 0.70 0.19 female03 2.02 0.43 1.45 0.37 female04 2.68 0.00 1.13 0.33 female05 0.17 0.01 0.05 0.01 female06 0.23 0.00 0.09 0.03

Animal ID

Total home-range (ha)

Foraging range(ha)

Denning range (ha)

male01 6.36 3.47 1.55 male02 6.59 6.59 0.06 male03 7.14 4.46 2.87 female01 3.96 3.89 0.58 female02 0.34 0.30 0.01 female03 2.04 1.94 0.50 female04 4.81 4.79 0.01 female05 0.26 0.26 0.00 female06

0.61

0.61

0.00

112

Fig. 9 The total home range (90%) and core area (50%) of the nine common brushtail possums (3 males and 6 females) calculated using the minimum convex polygon method. For the size of individual home range refer to the scale

113

Fig. 10 The total home range (90%) and core area (50%) of the nine common brushtail possums (3 males and 6 females) calculated using the fixed kernel method. For the size of individual home range refer to the scale

114

Fig. 11 The total home range (100%) and foraging and denning ranges of the nine common brushtail possums (3 males and 6 females) calculated using the minimum convex polygon method. For the size of individual home range refer to the scale

115

Fig. 12 The comparison of proportions of above ground biomass (AGB) of tree species most commonly used by common brushtail possums while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

Within the studied home ranges there were 72 trees with hollows large enough for

common brushtail possums, and these trees comprised 12.9% of the total number of trees

(n = 560). Standing dead trees (stumps) represented 22.2% of trees found to have hollows,

with on average 2.8 ± 0.7 hollows per tree. Of living trees, Eucalyptus radiata had the most

hollows (31.9%, 2.9 ± 0.4 hollows/tree). Common brushtail possums were found resting

during the day in only 10.7% of cases in limbs of Eucalyptus globulus and exotic species of

trees and spent the majority of their time denning in different Eucalyptus trees. In most

cases, common brushtail possums were found in Eucalyptus radiata (65%) with on average

6.2% ± 0.2 hollows per tree and in 17.7% of cases in stumps (5.2% ± 0.8 hollows/tree).

The chi-square test showed that common brushtail possums used trees with hollows more

often than would be expected from their availability (χ2 = 274.084, df = 1, P < 0.001), and

they used these trees exclusively as their den sites (χ2 = 124.534; df = 1, P < 0.001).

116

Common brushtail possums used on average 6.7 ± 1.9 different trees as den sites, with

males having a significantly higher number of repeatedly used den trees than females (11.3

± 3.7; 4.3 ± 1.8 respectively, P < 0.001).

Moreover, common brushtail possums used medium-sized trees (50–100 cm in diameter

and 10–20 m high) and large-sized trees (100–150 cm and > 150 cm in diameter and > 20

m high) significantly more often than small-sized trees (5–50 cm in diameter, <10 m high;

χ2 = 352.668, df = 3, P < 0.001; χ2 = 207.956, df = 2, P < 0.001 respectively). The number

of mistletoes was weakly correlated with tree diameter and height (r = 0.18, r = 0.26

respectively, P < 0.001) while the number of hollows was strongly correlated with tree

diameter (r = 0.76, P < 0.001) and less strongly with tree height (r = 0.48, P < 0.001).

The binomial generalized linear mixed model showed that the key parameters that

separated trees used by common brushtail possums from unused ones were tree species,

diameter, and the number of hollows (r2 = 0.37, P < 0.001). Trees with a larger diameter

and a higher number of hollows were used more often by animals (Table 5).

The model explained 75.7% of common brushtail possum tree presences and 86.1% of

absences. Furthermore, tree species, diameter, number of hollows, and mistletoes were the

most important factors separating common brushtail possum food trees from den trees (r2

= 0.60, P < 0.001). Eucalyptus camphora and exotic trees were used as food trees while

stumps and E. radiata were used as den trees; E. viminalis was used both as a food and den

tree by common brushtail possums. Food trees had a smaller diameter, less hollows, and

more mistletoes than den trees (Table 6). The model explained 90.0% of common

brushtail possum foraging records and 88.2% of denning records.

117

Table 5 The binomial generalized linear mixed model predicting tree occupancy by common brushtail possums (AIC = 921.27, BIC = 990.55). Tree species: BLP Eucalyptus dives (broad-leaved peppermint), BW Acacia melanoxylon (blackwood), EX (exotic trees), MG E. viminalis (manna gum), MM E. obliqua (messmate), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), ST (stump), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significance at P < 0 ‘***’, P < 0.001 ‘**’, P < 0.01 ‘*’

Random effects:

Var. Std. Dev. Animal (Intercept) 0.257 0.507 Fixed effects:

Estimate Std. Error z-value P-value Intercept -0.964 0.449 -2.148 0.032* Species BLP -4.306 1.026 -4.196 2.72e-05*** Species BW -0.285 0.852 -0.335 0.738 Species EX -0.409 0.442 -0.927 0.354 Species MG -1.521 0.471 -3.226 0.001** Species MM -0.626 0.768 -0.816 0.414 Species NLP -1.504 0.411 -3.661 0.0003*** Species SG -1.545 0.477 -3.242 0.001** Species ST -1.855 0.526 -3.527 0.0004*** Species SW -2.170 0.528 -4.106 4.03e-05*** Species WC -0.805 0.582 -1.384 0.166346 Diameter 0.031 0.004 8.441 < 2e-16*** Hollow 0.354 0.065 5.437 5.41e-08*** Binomial generalized linear mixed model: Null fit: Visitation ~ 1 + Animal Fit: Visitation ~ Species + Diameter + Hollow + Animal df AIC BIC logLik Dev. Chi sq. Chi df P-value Null fit 2 1432.59 1442.48 -714.29 1428.59 Fit 14 921.27 990.55 -446.64 893.27 535.32 12 < 2.2e-16***

118

Table 6 The binomial generalized linear mixed model predicting tree use (food and den) by common brushtail possums (AIC = 295.88, BIC = 358.55). Tree species: BLP Eucalyptus dives (broad-leaved peppermint), BW Acacia melanoxylon (blackwood), EX (exotic trees), MG E. viminalis (manna gum), MM E. obliqua (messmate), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), ST (stump), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significance at P < 0 ‘***’, P < 0.001 ‘**’, P < 0.01 ‘*’

Random effects:

Var. Std. Dev. Animal (Intercept) 0.843 0.918 Fixed effects:

Estimate Std. Error z-value P-value Intercept 1.253e+00 7.208e-01 1.738 0.082 Species BLP 2.183e+01 5.619e+02 0.039 0.969 Species BW 1.809e+01 2.906e+03 0.006 0.995 Species Stump -1.677e+01 4.261e+03 -0.004 0.997 Species EX 2.040e+00 7.034e-01 2.900 0.004** Species MG 2.469e+00 7.250e-01 3.406 0.001*** Species MM 2.351e+01 6.050e+04 0.000 0.999 Species NLP 2.969e+00 6.019e-01 4.932 8.12e-07*** Species SG 4.672e+00 9.880e-01 4.729 2.25e-06*** Species SW 1.436e+01 7.583e+02 0.019 0.985 Species WC 1.850e+01 6.084e+03 0.003 0.998 Diameter -2.959e-02 6.886e-03 -4.297 1.73e-05*** Mistletoe 2.128e-01 9.910e-02 2.147 0.032* Hollow -6.276e-01 1.067e-01 -5.882 4.06e-09*** Binomial generalized linear mixed model: Null fit: Visitation ~ 1 + Animal Fit: Visitation ~ Species + Diameter + Mistletoe + Hollow + Animal

Df AIC BIC logLik Dev. Chi sq. Chi df P-value Null fit 2 664.18 672.54 330.09 660.18 Fit 15 295.88 358.55 132.94 265.88 394.31 13 < 2.2e-16***

119

Individual variability in tree use

Tree use varied among the studied common brushtail possum individuals (3 males and 6

females) with animals found foraging in 6.7% to 86.7% of cases in exotic species of trees.

In addition, tree use of Eucalyptus radiata was highly uneven ranging from 6.7% to 60.7%,

and a similar pattern was observed for E. viminalis and E. camphora trees (Table 7). When

the biomass of trees used as foraging sites by common brushtail possums was compared to

available tree biomass that was measured within each individual animal home range, strong

individual differences were observed among the studied animals. For instance, male

possum 1 and females 1 and 5 used E. radiata proportionally to its availability within the

home ranges while male 2 and female 3 used it less often than expected from its availability

(P < 0.05, Fig. 13). Male 2 and female 3 used E. viminalis more often than expected by

chance alone (P < 0.001) while female 5 avoided this species (P < 0.05). E. camphora was

used less often by females 1, 3, 4 and 6 (P < 0.05). Similarly, in all home ranges where

Acacia dealbata was found, common brushtail possums avoided using this species during

their nightly foraging activity (P < 0.05). However, when foraged trees where plotted

against the number of night radio fixes taken for each animal, the accumulation curves did

not reach a plateau in all cases (Fig. 14). This suggests that further radio-tracking could

produce different visitation rates to analysed trees and could change possum species

preferences.

Common brushtail possums were much less versatile in their use of den trees; most

animals (2 males and 4 females) were found significantly more often denning in Eucalyptus

radiata trees (P < 0.001), with only female 3 using E. camphora (P = 0.19) and female 6 E.

viminalis (P < 0.001) as their primary den sites. Here, accumulation curves plateaued for six

of the studied animals (1 male and 6 females) and increased for male 1 and 3 and female 1

(Fig. 15). Almost 90% of the nightly sightings of common brushtail possums were in trees

that were used only once during the sampling period (Fig. 16). There was a significant

difference in distances traversed by common brushtail possums from den trees to single

and multiple used trees (P < 0.001). Multiple used trees were on average 56.5 ± 17.5 m

from den trees while single used trees were on average 130.9 ± 8.8 m from den trees. No

significant difference was observed in distances travelled by common brushtail possums

between their den trees and trees parasitized by mistletoes (109.3 ± 21.5 m) or free from

mistletoes (123.0 ± 8.9 m, P = 0.788).

120

Table 7 Percentage of available and used trees by the nine common brushtail possums (3 males and 6 females) during their nightly foraging activity within each home range. Maximum and minimum visits (%) to each of the tree species by different individuals are marked in red

Animal ID Male 1 Male 2 Male 3 Female 1 Female 2 Female 3 Female 4 Female 5 Female 6

Species Avail Use Avail Use Avail Use Avail Use Avail Use Avail Use Avail Use Avail Use Avail Use Acacia dealbata 27.6 3.6 50.7 17.9 16.7 0.0 25.0 0.0 11.6 0.0 4.0 3.6 4.0 0.0 33.3 6.7 12.2 0.0

Acacia melanoxylon 0.0 3.6 0.0 0.0 0.0 0.0 5.0 10.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.1 0.0

Eucalyptus camphora 1.7 0.0 1.3 3.6 0.0 0.0 35.0 56.7 0.0 0.0 46.0 7.1 46.0 50.0 2.2 0.0 12.2 0.0

Eucalyptus globulus 10.3 3.6 0.0 0.0 4.8 4.0 0.0 0.0 1.4 0.0 0.0 0.0 0.0 0.0 6.7 23.3 0.0 0.0

Eucalyptus viminalis 0.0 0.0 5.3 0.0 6.0 8.0 0.0 3.3 20.3 6.7 6.0 10.7 6.0 0.0 17.8 20.0 32.7 65.4

Eucalyptus dives 3.4 0.0 0.0 0.0 26.2 8.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Eucalyptus radiata 48.3 60.7 29.3 35.7 28.6 48.0 35.0 30.0 23.2 6.7 24.0 32.1 24.0 21.4 13.3 26.7 2.0 0.0

Eucalyptus obliqua 0.0 3.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 13.3 13.3 0.0 0.0

Exocarpos cupressiformis 0.0 0.0 0.0 0.0 7.1 4.0 0.0 0.0 1.4 0.0 4.0 0.0 4.0 3.6 8.9 3.3 36.7 26.9

Exotic trees 8.6 25.0 13.3 42.9 10.7 28.0 0.0 0.0 42.0 86.7 16.0 46.4 16.0 25.0 4.4 6.7 0.0 7.7

121

Fig. 13 The comparison of proportions of above ground biomass (AGB) of tree species used by individual common brushtail possums (3 males and 6 females) while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

122

Fig. 13 cont. The comparison of proportions of above ground biomass (AGB) of tree species used by individual common brushtail possums (3 males and 6 females) while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

123

Fig. 13 cont. The comparison of proportions of above ground biomass (AGB) of tree species used by individual common brushtail possums (3 males and 6 females) while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

124

Fig. 13 cont. The comparison of proportions of above ground biomass (AGB) of tree species used by individual common brushtail possums (3 males and 6 females) while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

125

Fig. 13 cont. The comparison of proportions of above ground biomass (AGB) of tree species used by individual common brushtail possums (3 males and 6 females) while foraging (grey bars) and availability of those species within animal home ranges (median with 95% confidence intervals, CI). Tree species: EX (exotic trees), MG Eucalyptus viminalis (manna gum), NLP E. radiata (narrow-leaved peppermint), SG E. camphora (swamp gum), SW Acacia dealbata (silver wattle), and WC Exocarpos cupressiformis (cherry ballart). An asterix indicates significant differences in use and availability at P < 0.001

126

Fig. 14 Accumulation curves of the number of food trees recorded during night radio-tracking of the nine common brushtail possums (3 males and 6 females)

Fig. 15 Accumulation curves of the number of den trees recorded during day radio-tracking of the nine common brushtail possums (3 males and 6 females)

127

Fig. 16 Percentage (%) of trees used by common brushtail possums once or more during their nightly foraging activity

128

Discussion

Summary of key results

In the present study, male common brushtail possums had larger home ranges than

females. The total home range size of both males and females was strongly correlated with

their foraging ranges and only weakly correlated with their denning ranges, suggesting that

food availability was a more limiting factor for common brushtail possums than the

availability of hollow-bearing trees. Moreover, the studied common brushtail possums

preferred different species of trees when foraging and repeatedly used individuals of the

same species, as well as trees parasitized by mistletoes. Possums differentiated between

trees used as foraging and denning sites based on species, diameter, and number of hollows

and mistletoes, with trees having a higher number of hollows used more often as denning

sites.

Home range attributes

In general, a home range size of arboreal marsupials inhabiting eucalypt forests and

woodlands is considered to be the outcome of animal foraging, denning, and mating

behaviours. In this study, areas traversed by common brushtail possums during their

nightly foraging activity, along with areas occupied during the day by animals resting in a

single hollow-bearing tree, produced different estimates of the total home range size. Male

common brushtail possums had significantly larger home ranges (4.1 ± 0.6 ha) than

females (1.2 ± 0.4 ha). This finding is consistent with other studies looking at differences in

home range size between sexes, with male common brushtail possums having home ranges

from 1 to 11 ha and females from 0.7 to 7.4 ha in size (summarized by Kerle, 2001).

Sex differences in the home range size of males and females (5.1 ± 0.8 ha; 2.1 ± 0.3 ha

respectively) have been found previously for a closely related species, the mountain

brushtail possum (Trichosurus cunninghami), occurring in the same study area (Martin et al.,

2007). The authors suggested that males of this polygynous species had to range over larger

areas than females to access sufficient food resources and maximize reproductive success

by overlapping their home ranges with home ranges of multiple females.

129

Similarly, common brushtail possums have a predominantly polygynous mating system

(Isaac & Johnson, 2003), and males often have overlapping home ranges and compete for

access to multiple estrous females (Winter, 1976). Hence, larger male home ranges could be

attributed to different reproductive imperatives of males and females.

Moreover, the sexual dimorphism observed in the studied common brushtail possums,

with males being larger than females, could have affected the differences in home range

size between the sexes. McNab (1963) suggested that the home range size of herbivores is

related to animal body weight and energetic requirements. Namely, herbivores’ energetic

requirements increase with body weight forcing animals to forage over larger areas,

especially in areas with patchily distributed resources. Yet, in habitats of greater

environmental productivity and with aggregated resources, herbivores are expected to

range over smaller areas according to the habitat productivity hypothesis (Harestad &

Bunnel, 1979). This could partially explain the relatively small home ranges of both male

and female common brushtail possums occurring in the study area. The roadside

vegetation corridors in the Strathbogie Ranges are considered highly productive habitats

since they have not been logged for the last 100 years or burnt for the last 50 years (Martin

& Martin, 2007). As a result, these habitats are characterised by a high density of large trees

used as foraging and denning sites by arboreal marsupials.

Furthermore, other studies carried out in New Zealand have demonstrated that patches of

high quality habitat can maintain higher densities of common brushtail possums, and in

consequence, individuals can have relatively small home ranges (summarized by Cowan,

2001). For instance, where common brushtail possums live on farmland with scattered

patches of remnant forest or scrub, they show two types of ranging behaviour (Cowan &

Clout, 2000): Some have small home ranges centred on preferred habitats, such as stream-

side willows or swamps and never venture far out onto farmland; others range up to 1600

m over open pasture and have annual home ranges of up to 60 ha (Brockie et al., 1997).

Hence, the mobility of a species can affect the individual home range size, especially in the

habitats where resources are rare and patchily distributed. In these habitats, more mobile

animals have larger home ranges in order to include a sufficient number of suitable food

trees (DeGabriel et al., 2010).

130

Finally, abundance and quality of food have been previously recognized as the major

factors shaping home range size of herbivores (Säid et al., 2009). DeGabriel and colleagues

(2009a) found that the size of common brushtail possum home ranges was affected by the

nutritional quality of trees, with larger home ranges occurring in areas of lower nutritional

quality. The significantly larger foraging ranges than denning home ranges observed in the

studied common brushtail possums could have been the outcome of their generalist

feeding strategy that forced animals to range over larger areas in search of different types of

food. In the current study, the total home range size of both males and females was

strongly correlated with foraging area and only weakly correlated with denning area,

suggesting that food availability might have been a more limiting factor for common

brushtail possums than availability of hollow-bearing trees.

In the current study, male common brushtail possums had a significantly higher number of

dens than females (11.3 ± 3.7, 4.3 ± 1.8 respectively). By contrast, How (1981) found that

in north-eastern New South Wales male common brushtail possums used on average 1.5

dens and females 3.5 dens. In Queensland Winter (1976) found that only one to three dens

were used on a regular basis with females often using the same den for months. The small

number of dens recorded in earlier studies could indicate that the number of trees with

suitable size hollows for possums is limited in many parts of Australia subjected to heavy

deforestation and fragmentation. However, in the Strathbogie Ranges the roadside

vegetation corridors have been identified as sustaining a large number of trees bearing

hollows allowing mountain brushtail possums to occupy multiple den trees (males

16.5 ± 1.5 and females 7.4 ± 0.6; Martin et al., 2007).

In summary, resources available to animals within their home ranges do not affect animals

separately. Recently Di Stefano and colleagues (2011) proposed a conceptual framework

for quantifying the influence of multiple resources on the home range size of herbivores.

According to the authors, separate resources (i.e., food and shelter) interact to form a

continuous, multi-dimensional surface exhibiting an emergent environmental pattern with a

home range size more closely correlated with resource heterogeneity than with simple

additive or interactive effects of separate resources. They demonstrated that the availability

of different types of food (i.e., forbs, shrubs, ferns, monocots, and trees) and shelter

(i.e., lateral cover) affected the home range size of a terrestrial herbivore, the swamp

wallaby (Wallabia bicolor).

131

Individual variability in tree use

It has been recognized previously that individuals of the same species can exploit resources

differently to fulfil their requirements for growth, survival, and reproduction, and maximise

in this way their fitness contribution to future generations (Dall, McNamara, & Houston,

2004). For herbivores, several regulating mechanisms have been identified, including

variation in structure, abundance, and spatial distribution of forage within their home

ranges (Spalinger & Hobbs, 1992; Hobbs et al., 2003). In the current study, individual

possums used different tree species during their nightly foraging activity, and individual

home ranges differed in terms of tree species composition and abundance. For instance,

some home ranges spanned across both Herb-rich Foothill Forest and Swampy Riparian

Woodland (Female 1, 3 and 4) while others were dominated by a single species, Eucalyptus

radiata (Male 1 and 3). Interestingly, one of the older females carrying a second young in

season (Female 2) had a home range dominated by exotic species of trees. Hence, selection

against particular tree species by one common brushtail possum could have been due to its

high abundance within animal home range while selection for the same species by a

different possum could have been due to its rarity within its home range.

Alternatively, resource partitioning among the nine studied common brushtail possums

could have occurred with individuals using different species of trees as their foraging sites

within otherwise relatively uniform habitats. Individual differences in dietary preferences

have been identified as “individual specialization” or the expression of animal personalities

in a situation where generalist populations are composed of specialized foragers with

dietary preferences representing subsets of the broader species-specific generalist diet

(Bolnick et al., 2003). For example, the gracile mouse opossum (Gracilinanus microtarsus), an

omnivorous arboreal marsupial that inhabits the Atlantic rainforest and cerrado in Brazil,

has been found to exhibit significant individual specialization with diets of specialists being

nested within the diets of generalists (Araújo et al., 2009). Alternatively, the diet differences

may be exaggerated by direct interference between individuals, with dominant individuals

securing preferred foods by forcing subordinate individuals to feed on suboptimal

resources.

Apart from observed differences in tree species preferences, the studied common brushtail

possums repeatedly used individual trees of the same species that were usually closer to

their den trees.

132

Similar patterns were observed for opossum (Didelphis virginiana), with its den sites located

within close proximity to food (Allen, Marchinton, & Lentz, 1985). Opposite to

expectation, common brushtail possums were more likely to visit trees closer to dens

irrespective of mistletoe presence (i.e., they were not prepared to travel further to visit trees

with mistletoe although they preferred them). The observed inconsistencies in the findings

of this study could have been caused by a low sample size and if further radio-tracking was

possible it would allow for the separation of different factors.

Likewise, common brushtail possums do not always choose trees based solely on their

availability and proximity to shelter; other factors, such as competition with co-occurring

arboreal marsupials, can shape their tree choices. In the current study, possums showed a

negative selection for Acacia dealbata which is the preferred food of mountain brushtail

possums known to exhibit a certain level of territoriality in protecting their key resources

from competitors (Martin & Martin, 2007). Del Borrello (2009) looked at differences in

tree use by common and mountain brushtail possums in the same study area and suggested

a certain level of dietary partitioning between the sympatric species. Furthermore, two

species of possums are known to have different tolerances to plant secondary metabolites,

such as tannins and terpenes, with the latter being found only in Eucalyptus species eaten by

common brushtail possums (Burchfield, Agar, & Hume, 2006). Similarly, in tropical

rainforests in north-east Queensland, sympatric folivores (i.e., Lumholtz’s tree kangaroo,

Dendrolagus lumholtzi; lemuroid ringtail possum, Hemibelideus lemuroides; Herbert River ringtail

possum, Pseudochirulus herbertensis; and coppery brushtail possum, Trichosurus johnstonii)

partitioned resources based on their tolerance of different plant secondary metabolites

(Kanowski, Irvine, & Winter, 2003).

The current study has demonstrated that the distances travelled between foraging and

denning sites and the intensity of tree use by common brushtail possums shape their home

range size. Animals that exhibit the observed behaviour are altering the rate and bias of

their movement in response to heterogeneous features of the landscape (i.e. the availability

of different tree species, mistletoes and hollows, and the chemical composition of forage).

Moreover, intra-specific differences were observed with individual animals differing in their

responsiveness to environmental variability across space and time. Further investigation of

the mechanisms underlying individual variability in foraging and denning behaviour and

space use patterns can yield interesting insights into individual animal personalities, fitness

consequences, and the evolutionary success of this generalist herbivore.

133

References

Allen, C. H., Marchinton, R. L., & Lentz, W. M. (1985). Movement, habitat use and denning of opossums in the Georgia Piedmont, 113(2). American Midland Naturalist, 408–412.

Andrew, R. L., Peakall, R., Wallis, I. R., & Foley, W. J. (2007). Spatial distribution of defense chemicals and markers and the maintenance of chemical variation. Ecology, 88(3), 716–728.

Araújo, M. S., Martins, E. G., Cruz, L. D., Fernandes, F. R., Linhares, A. X., Dos Reis, S.

F., & Guimarães, P. R. (2009). Nested diets: A novel pattern of individual‐level resource use. Oikos, 119(1), 81–88.

Bolnick, D. I., Svanbäck, R., Fordyce, J. A., Yang, L. H., Davis, J. M., Hulsey, C. D., & Forister, M. L. (2003). The ecology of individuals: Incidence and implications of individual specialization. The American Naturalist, 161(1), 1–28.

Börger, L., Franconi, N., Ferretti, F., Meschi, F., De Michele, G., Gantz, A., & Coulson, T. (2006). An Integrated Approach to Identify Spatiotemporal and Individual-Level Determinants of Animal Home Range Size. The American Naturalist, 168(4), 471–485.

Bourke, P., Magnan, P., & Rodriguez, M. (2005). Individual variations in habitat use and morphology in brook charr. Journal of Fish Biology, 51(4), 783–794.

Brockie, R. E., Ward, G. D., & Cowan, P. E. (1997). Possums (Trichosurus vulpecula) on Hawke's Bay farmland: Spatial distribution and population structure before and after a control operation. Journal of the Royal Society of New Zealand, 27(2), 181–191.

Burchfield, E., Agar, N., & Hume, I. (2006). Effects of terpenes and tannins on some physiological and biochemical parameters in two species of phalangerid possums (Marsupialia: Phalangeridae). Australian Journal of Zoology, 53(6), 395–402.

Burdette, W. J., & Gehan, E. A. (1970). Planning and analysis of clinical studies. Springfield: Thomas.

Burt, W. H. (1943). Territoriality and home range concepts as applied to mammals. Journal of Mammalogy, 24(3), 346–352.

Cederlund, G., & Sand, H. (1994). Home-range size in relation to age and sex in moose. Journal of Mammalogy, 75(4), 1005–1012.

Chilcott, M., & Hume, I. (1985). Coprophagy and selective retention of fluid digesta: Their role in the nutrition of the common ringtail possum, Pseudocheirus peregrinus. Australian Journal of Zoology, 33(1), 1–15.

Comport, S., Ward, S., & Foley, W. (1996). Home ranges, time budgets and food-tree use in a high-density tropical population of greater gliders, Petauroides volans minor (Pseudocheiridae : Marsupialia). Wildlife Research, 23(4), 401–419.

Cowan, P. E. (2000). Factors Affecting Possum Re-infestation—Implications for Management: Department of Conservation Wellington.

Cowan, P. E. (2001). Advances in New Zealand mammalogy 1990–2000: Brushtail possum. Journal of the Royal Society of New Zealand, 31(1), 15–29.

Cowan, P. E., & Clout, M. (2000). Possums on the move: Activity patterns, home ranges, and dispersal. In T. Montague, T. (Ed.) The brushtail possum: Biology, impact and management of an introduced marsupial (pp. 24–34). Lincoln: Manaaki Whenua Press.

Dall, S. R., Houston, A. I., & McNamara, J. M. (2004). The behavioural ecology of personality: Consistent individual differences from an adaptive perspective. Ecology Letters, 7(8), 734–739.

Dearing, M. D., Foley, W. J., & McLean, S. (2005). The influence of plant secondary metabolites on the nutritional ecology of herbivorous terrestrial vertebrates. Annual Review of Ecology, Evolution, and Systematics, 36, 169–189.

134

DeGabriel, J. L. (2008). Demographic constraints imposed by plant nutrients and secondary chemistry on the common brushtail possum (Trichosurus vulpecula Kerr.). (Ph. D. Thesis), the Australian National University, Canberra.

DeGabriel, J. L., Moore, B. D., Foley, W. J., & Johnson, C. N. (2009a). The effects of plant defensive chemistry on nutrient availability predict reproductive success in a mammal. Ecology, 90(3), 711–719.

DeGabriel, J., Moore, B. D., Shipley, L., Krockenberger, A. K., Wallis, I. R., Johnson, C. N., & Foley, W. J. (2009b). Inter-population differences in the tolerance of a marsupial folivore to plant secondary metabolites. Oecologia, 161(3), 539–548.

DeGabriel, J., Moore, B. D., Marsh, K. J., & Foley, W. J. (2010). The effect of plant secondary metabolites on the interplay between the internal and external environments of marsupial folivores. Chemoecology, 20(2), 97–108.

Del Borrello, C. (2009). Home-range and habitat use in sympatric brushtail possums (Trichosurus spp.). (Honours thesis), the University of Melbourne, Melbourne.

Di Stefano, J., Coulson, G., Greenfield, A., & Swan, M. (2011). Resource heterogeneity influences home range area in the swamp wallaby Wallabia bicolor. Ecography, 34(3), 469–479.

Dussault, C., Ouellet, J.-P., Courtois, R., Huot, J., Breton, L., & Jolicoeur, H. (2005). Linking moose habitat selection to limiting factors. Ecography, 28(5), 619–628.

Efford, M., Warburton, B., & Spencer, N. (2000). Home-range changes by brushtail possums in response to control. Wildlife Research, 27(2), 117–127.

Ellis, W., Melzer, A., Carrick, F., & Hasegawa, M. (2002). Tree use, diet and home range of the koala (Phascolarctos cinereus) at Blair Athol, central Queensland. Wildlife Research, 29(3), 303–311.

Enoksson, B., & Nilsson, S. G. (1983). Territory size and population density in relation to food supply in the nuthatch Sitta europaea (Aves). The Journal of Animal Ecology, 52(3), 927–935.

ESRI, ArcGIS Desktop 10.0. (2012). Redlands, CA: Environmental Systems Resource Institute.

Fagan, W. F., Lutscher, F., & Schneider, K. (2007). Population and Community Consequences of Spatial Subsidies Derived from Central Place Foraging. The American Naturalist, 170(6), 902–915.

Fisher, D. O., & Lara, M. C. (1999). Effects of body size and home range on access to mates and paternity in male bridled nailtail wallabies. Animal Behaviour, 58(1), 121–130.

Fisher, D. O., & Owens, I. P. (2000). Female home range size and the evolution of social organization in macropod marsupials. Journal of Animal Ecology, 69(6), 1083–1098.

Ganzhorn, J. U. (1988). Food partitioning among Malagasy primates. Oecologia, 75(3), 436–450.

Gautestad, A. O., & Mysterud, I. (2005). Intrinsic scaling complexity in animal dispersion and abundance. The American Naturalist, 165(1), 44–55.

Gibbons, P., & Lindenmayer, D. (2002). Tree hollows and wildlife conservation in Australia. Collingwood: CSIRO.

Hansteen, T. L., Andreassen, H. P., & Ims, R. A. (1997). Effects of spatiotemporal scale on autocorrelation and home range estimators. The Journal of Wildlife Management, 61(2), 280–290.

Harestad, A. S., & Bunnel, F. L. (1979). Home range and body weight—a reevaluation. Ecology, 60(2), 389–402.

Harris, S., Cresswell, W. J., Forde, P. G., Trewhella, W. J., Woollard, T., & Wray, S. (1990). Home-range analysis using radio-tracking data—a review of problems and techniques particularly as applied to the study of mammals. Mammal Review, 20(2–3), 97–123.

135

Harvey, P. H., & Clutton-Brock, T. H. (1981). Primate home-range size and metabolic needs. Behavioral Ecology and Sociobiology, 8(2), 151–155.

Hews, D. K. (1993). Food resources affect female distribution and male mating opportunities in the iguanian lizard Uta palmeri. Animal Behaviour, 46(2), 279–291.

Hobbs, T. N., Gross, J. E., Shipley, L. A., Spalinger, D. E., & Wunder, B. A. (2003). Herbivore functional response in heterogeneous environments: A contest among models. Ecology, 84(3), 666–681.

Hooge, P. N., Eichenlaub, W., & Solomon, E. (1999). The animal movement program. USGS, Alaska Biological Science Center.

How, R. A. (1972). The ecology and management of Trichosurus species (Marsupialia) in N.S.W. (Ph.D. thesis), the University of New England, Armidale.

How, R. A. (1981). Population Parameters of Two Congeneric Possums, Trichosurus spp., in North Eastern New South Wales. Australian Journal of Zoology, 29(2), 205–215.

Isaac, J. L., & Johnson, C. N. (2003). Sexual dimorphism and synchrony of breeding: Variation in polygyny potential among populations in the common brushtail possum, Trichosurus vulpecula. Behavioral Ecology, 14(6), 818–822.

Ivlev, V. S. (1961). Experimental ecology of the feeding of fishes. New Haven: Yale University Press.

Johnson, D. H. (1980). The comparison of usage and availability measurements for evaluating resource preference. Ecology, 61(1), 65–71.

Johnson, C. N., Clinchy, M., Taylor, A. C., Krebs, C. J., Jarman, P. J., Payne, A., & Ritchie, E. G. (2001). Adjustment of offspring sex ratios in relation to the availability of resources for philopatric offspring in the common brushtail possum. Proceedings of the Royal Society of London. Series B: Biological Sciences, 268(1480), 2001–2005.

Jones, K. M. W., & Krockenberger, A. K. (2007). Determining the diet of cryptic folivores: An assessment of diet analysis techniques using the green ringtail possum (Pseudochirops archeri) as a case study. Wildlife Research, 34(5), 352–358.

Kanowski, J., Irvine, A. K., & Winter, J. W. (2003). The relationship between the floristic composition of rain forests and the abundance of folivorous marsupials in north-east Queensland. Journal of Animal Ecology, 72(4), 627–632.

Kenward, R. (2001). A manual for wildlife radio tagging. London: Academic Press. Kerle, A. (2001). Possums: The brushtails, ringtails and greater glider. Sydney: UNSW Press. Lawler, I. R., Foley, W. J., & Eschler, B. M. (2000). Foliar concentration of a single toxin

creates habitat patchiness for a marsupial folivore. Ecology, 81(5), 1327–1338. Lewis, M. A., & Murray, J. D. (1993). Modelling territoriality and wolf-deer interactions.

Nature, 366(6457), 738–740. Lindenmayer, D. B. (1997). Differences in the biology and ecology of arboreal marsupials

in forests of southeastern Australia. Journal of Mammalogy, 1117–1127. Lindenmayer, D. B., Cunningham, R. B., Tanton, M. T., Smith, A. P., & Nix, H. A. (1991).

Characteristics of hollow-bearing trees occupied by arboreal marsupials in the montane ash forests of the Central Highlands of Victoria, south-east Australia. Forest Ecology and Management, 40(3–4), 289–308.

Lomnicki, A. (1988). Population Ecology of Individuals. Princeton: Princeton University Press. Main, M. B. (2008). Reconciling competing ecological explanations for sexual segregation

in ungulates. Ecology, 89(3), 693–704. Mares, M. A., Lacher Jr, T. E., Willig, M. R., Bitar, N. A., Adams, R., Klinger, A., & Tazik,

D. (1982). An experimental analysis of social spacing in Tamias striatus. Ecology, 63(2), 267–273.

Martin, J. K. (2005). Behavioural ecology of bobuck (Trichosurus cunninghami). (Ph. D. thesis), the University of Melbourne, Melbourne.

Martin, J. K., Handasyde, K. A., & Taylor, A. C. (2007). Linear roadside remnants: Their

influence on den‐use, home range and mating system in bobucks (Trichosurus cunninghami). Austral Ecology, 32(6), 686–696.

136

Martin, J. K., Handasyde, K. A., Taylor, A. C., & Coulson, G. (2007). Long-term pair-bonds without mating fidelity in a mammal. Behaviour, 144(11), 1419–1445.

Martin, J. K., & Martin, A. A. (2007). Resource distribution influences mating system in the bobuck (Trichosurus cunninghami: Marsupialia). Oecologia, 154(1), 227–236.

Martin, R. W., & Handasyde, K. A. (1999). The Koala: Natural history, Conservation and Management. Sydney: University of New South Wales.

McGlynn, T. P., Shotell, M. D., & Kelly, M. S. (2003). Responding to a variable environment: home range, foraging behavior, and nest relocation in the Costa Rican rainforest ant Aphaenogaster araneoides. Journal of Insect Behavior, 16(5), 687–701.

McIntyre, N. E., & Wiens, J. A. (1999). Interactions between landscape structure and animal behavior: The roles of heterogeneously distributed resources and food deprivation on movement patterns. Landscape Ecology, 14(5), 437–447.

McNab, B. K. (1963). Bioenergetics and the determination of home range size. American Naturalist, 97(894), 133–140.

Millspaugh, J. J., & Marzluff, J. M. (2001). Radio tracking and animal populations. San Diego: Academic Press.

Milton, K., & May, M. L. (1976). Body weight, diet and home range area in primates. Nature, 259(5543), 459–462.

Mohr, C. O. (1947). Table of equivalent populations of North American small mammals. American Midland Naturalist, 37(1), 223–249.

Montague, T. L. (2000). The brushtail possum: Biology, impact and management of an introduced marsupial. Lincoln: Manaaki Whenua Press.

Moorcroft, P., & Lewis, M. A. (2006). Mechanistic home range analysis. Princeton: Princeton University Press.

Moore, B. D., & Foley, W. J. (2005). Tree use by koalas in a chemically complex landscape. Nature, 435(7041), 488–490.

Moore, B. D., Lawler, I. R., Wallis, I. R., Beale, C. M., & Foley, W. J. (2010). Palatability mapping: A koala's eye view of spatial variation in habitat quality. Ecology, 91(11), 3165–3176.

Nersesian, C., Banks, P., & McArthur, C. (2012). Influences of plant toxins and their spatial distribution on foraging by the common brushtail possum, a generalist mammalian herbivore. Journal of Chemical Ecology, 38(12), 1544–1551.

Ostfeld, R. S. (1985). Limiting resources and territoriality in microtine rodents. American Naturalist, 126(1), 1–15.

Paterson, B., Morris, R., Weston, J., & Cowan, P. (1995). Foraging and denning patterns of brushtail possums, and their possible relationship to contact with cattle and the transmission of bovine tuberculosis. New Zealand Veterinary Journal, 43(7), 281–288.

Perry, G., & Garland Jr, T. (2002). Lizard home ranges revisited: Effects of sex, body size, diet, habitat, and phylogeny. Ecology, 83(7), 1870–1885.

Pope, M. L., Lindenmayer, D. B., & Cunningham, R. B. (2004). Patch use by the greater glider (Petauroides volans) in a fragmented forest ecosystem. I. Home range size and movements. Wildlife Research, 31(6), 559–568.

Powell, R. A. (2000). Animal home ranges and territories and home range estimators. In L. Boitani & T. K. Fuller (Eds.), Research techniques in animal ecology: Controversies and consequences (pp. 65–110). New York: Columbia University Press.

Rhodes, J. R., McAlpine, C. A., Lunney, D., & Possingham, H. P. (2005). A spatially explicit habitat selection model incorporating home range behavior. Ecology, 86(5), 1199–1205.

Ringler, N. H. (1983). Variation in foraging tactics of fishes. In D. L. G. Noakes, D. G. Lindquist, G. S. Helfman & J. A. Ward (Eds.), Predation and prey in fishes (pp. 159–171). Hague: Junk.

Roughgarden, J. (1972). Evolution of niche width. The American Naturalist, 106(952), 683–718.

137

Roughgarden, J. (1979). Theory of population genetics and evolutionary ecology. New York: Macmillan.

Säid, S., Gaillard, J., Widmer, O., Débias, F., Bourgoin, G., Delorme, D., & Roux, C.

(2009). What shapes intra‐specific variation in home range size? A case study of female roe deer. Oikos, 118(9), 1299–1306.

Sandell, M. (1989). The mating tactics and spacing patterns of solitary carnivores. In J. L. Gittleman (Ed.), Carnivore behavior, ecology, and evolution (pp. 164–182). New York: Springer.

Scrivener, N. J., Johnson, C. N., Wallis, I. R., Takasaki, M., Foley, W. J., & Krockenberger, A. K. (2004). Which trees do wild common brushtail possums (Trichosurus vulpecula) prefer? Problems and solutions in scaling laboratory findings to diet selection in the field. Evolutionary Ecology Research, 6(1), 77–87.

Seton, E. T. (1909). Life-histories of northern animals: An account of the mammals of Manitoba (Vol. 2). New York: Charles Scribner's Sons.

Smith, A. C., & Schaefer, J. A. (2002). Home-range size and habitat selection by American marten (Martes americana) in Labrador. Canadian Journal of Zoology, 80(9), 1602–1609.

Smith, G. C., Mathieson, M., & Hogan, L. (2007). Home range and habitat use of a low-density population of greater gliders, Petauroides volans (Pseudocheiridae: Marsupialia), in a hollow-limiting environment. Wildlife Research, 34(6), 472–483.

Spalinger, D. E., & Hobbs, N. T. (1992). Mechanisms of foraging in mammalian herbivores: New models of functional response. American Naturalist, 140(2), 325–348.

Sullivan, T. P., Sullivan, D. S., & Krebs, C. J. (1983). Demographic responses of a chipmunk (Eutamias townsendii) population with supplemental food. The Journal of Animal Ecology, 52(3), 743–755.

Tufto, J., Andersen, R., & Linnell, J. (1996). Habitat use and ecological correlates of home range size in a small cervid: The roe deer. Journal of Animal Ecology, 65(6), 715–724.

Van Beest, F. M., Rivrud, I. M., Loe, L. E., Milner, J. M., & Mysterud, A. (2011). What determines variation in home range size across spatiotemporal scales in a large browsing herbivore? Journal of Animal Ecology, 80(4), 771–785.

Vedder, A. L. (2005). Movement patterns of a group of free‐ranging mountain gorillas (Gorilla gorilla beringei) and their relation to food availability. American Journal of Primatology, 7(2), 73–88.

Winter, J . W. (1976). The behaviour and social organisation of the brush-tail possum (Trichosurus vulpecula Kerr). (Ph.D. thesis), University of Queensland, St Lucia.

Winter, J. W. (1980). Tooth wear as an age index in a population of the brush-tailed possum, Trichosurus vulpecula (Kerr). Australian Wildlife Research, 7(3), 359–363.

Worton, B. J. (1989). Kernel Methods for Estimating the Utilization Distribution in Home-Range Studies. Ecology, 7(1), 164–168.

Youngentob, K., Wallis, I. R., Lindenmayer, D. B., Wood, J., Pope, M., & Foley, W. J. (2011). Foliage chemistry influences tree choice and landscape use of a gliding marsupial folivore. Journal of Chemical Ecology, 37(1), 71–84.

Youngentob, K. N., Yoon, H.-J., Coggan, N., & Lindenmayer, D. B. (2012). Edge effects influence competition dynamics: A case study of four sympatric arboreal marsupials. Biological Conservation, 155, 68–76.

138

CH

AP

TE

R

4

Dietary choices of the common brushtail possum (Trichosurus vulpecula) in a chemically

complex environment: When not to take a green smoothie?

139

CHAPTER 4

Dietary choices of the common brushtail possum (Trichosurus vulpecula) in a chemically complex

environment: When not to take a green smoothie?

Introduction

In 1977 Janzen wrote,

Plants are not just food for animals, and animals are not just decorations on the

vegetation. The world is not green. It is coloured lectin, tannin, cyanide, caffeine,

aflatoxin, and canavanine. And there is a lot of cellulose thrown in to make the mix

even more inedible. Animals are not ambulatory bomb calorimeters. They starve,

they ache, they abort, they vomit, they remember, they die, and they evolve (Janzen,

2007: 706).

Throughout the last 50 years of herbivore nutritional ecology, numerous studies have

focused on the influence of various plant chemical and physical properties on animal

dietary choices (reviewed by Agrawal & Fishbein, 2006; Carmona, Lajeunesse, & Johnson,

2011). This work has primarily focused on three dominant plant properties: nutritional

quality, including minerals, nitrogen and water content (Scriber, 1977; Mattson, 1980;

McNaughton, 1988); anti-herbivore activity of plant secondary metabolites (Bennett &

Wallsgrove, 1994; Dearing, Foley, & McLean, 2005); and mechanical properties of fibre

and leaf toughness (Lucas, Turner, Dominy, & Yamashita, 2000).

In the past, researchers have acknowledged the ecological importance of both nutrients and

plant secondary metabolites for herbivores (Coley & Barone, 1996; Molyneux & Ralphs,

1992). Yet, they have rarely managed to reconcile the complex interactions between these

opposite, but mutually complementary, aspects of plant chemistry. Many researchers

considered the availability of plant protein, expressed as foliar total nitrogen content, as a

single limiting factor for wild populations of herbivores, directly affecting their

maintenance, growth and reproduction (Cockfield, 1988; Joern & Behmer, 1997; White,

1984). Other studies analysed the toxic effects of different groups of plant secondary

metabolites (i.e. phenolics, terpenes, cyanogenic glycosides) on the metabolism and

foraging behaviour of herbivores (Rosenthal & Berenbaum, 1992).

140

Conversely, one of the most ubiquitous groups of plant secondary metabolites, tannins,

do not exhibit toxic effects on herbivores and instead act as forage digestibility reducers

and herbivory deterrents (Mole & Waterman, 1987; Schultz, 1989).Tannins are known to

represent 5–10% of plant biomass and are water-soluble phenolic compounds that bind

and precipitate proteins and other macronutrients within aqueous solutions, severely

decreasing the nutritive value of browse by reducing its digestibility (Barbehenn & Peter

Constabel, 2011; Mueller-Harvey, 2006). Herbivores not only consume tannins that render

nitrogen unavailable, but also tannins that may cause oxidative stress on an animal’s

digestive tract via auto-oxidation at high pH or enzymatic oxidation at lower pH

(Barbehenn & Peter Constabel, 2011; Salminen & Karonen, 2011). Nutritional

requirements of herbivores are further constrained as tannins decrease fibre digestibility by

binding bacterial enzymes or forming indigestible complexes with cell wall constituents or

both (Reed, 1995).

Differential activity of tannins has encouraged the evolution of various strategies by

herbivores, such as avoiding tannins or tolerating the negative effects of tannins (Iason &

Villalba, 2006). Some animals remain highly sensitive to even the smallest amounts of

tannins and cease feeding after experiencing the distinctively astringent taste, e.g., rodents

(Shimada & Saitoh, 2003), lagomorphs (Clausen, Provenza, Burritt, Reichardt, & Bryant,

1990) and primates (Takemoto, 2003). Other groups, like ruminants, have developed post-

ingestive physiological adaptations to counteract tannins, including the secretion of tannin-

binding proteins in the saliva, the increased gastrointestinal mucus production, the

degradation of tannins by gut microorganisms, and the activation of detoxifying enzymes

(reviewed by McArthur, Hagerman, & Robbins, 1991). Insects for instance, maintain highly

alkaline gut conditions to dissolve tannin-protein complexes (i.e., lepidopteran larvae,

Berenbaum, 1980) or use enzymes β-glucosidase to hydrolyse tannic acid or adsorb it on

the peritrophic membrane (i.e., locust, Bernays, 1981).

On the other side, arboreal marsupials inhabiting Australian forests and woodlands

dominated by a single genus Eucalyptus subsist dominantly on nitrogen-poor, but tannin-

rich and highly fibrous foliage. Tannins can represent up to 27% and nitrogen 0.8–2% of

leaf dry matter, and arboreal marsupials have developed a suite of morphological,

physiological, and behavioural adaptations to digest low quality forage (Macauley & Fox,

1980; Moore et al., 2004a). For instance, a eucalypt specialist, the koala (Phascolarctos

cinereus), possesses gut microflora actively metabolizing tannin-protein complexes.

141

Namely, Streptococcus bovis biotope I and tannin-protein complex-enterobacterium, occurring

frequently in the digestive tract of koala, have a unique property of degrading proteins

complexed with hydrolysable tannins (Osawa, 1991).

Another eucalypt specialist, the common ringtail possum (Pseudocheirus peregrinus), is less

sensitive to tannins and has a well-developed caecum that selectively retains bacteria and

other nitrogen-rich small particles and a well-developed dentition that reduces highly

fibrous foliage to small particles (Hume, 1999). This species also practices caecotrophy to

return nitrogen-rich caecal contents to the digestive tract (Marsh et al., 2003; McArthur &

Sanson 1991, 1993). Conversely, a dietary generalist, the common brushtail possum

(Trichosurus vulpecula), is more sensitive to tannins because of the immediate oral sensation

of the astringent taste of foliage. In addition, the common brushtail possum has less

specialized dentition and a small caecum relative to its body size with no evidence of a

separation mechanism (Hume, 1999).

High sensitivity of common brushtail possums to tannins has been demonstrated by Foley

and Hume (1987), who added polyethylene glycol 4000 (PEG), a tannin-binding agent, to

the drinking water of captive animals maintained on the sole diet of Eucalyptus melliodora

foliage. E. melliodora foliage contains 20–29% total phenolics and supplementation with

PEG increased the common brushtail possums dry matter intake by 32%, their digestibility

of neutral-detergent fibre by 77%, and consequently increased their intake of metabolisable

energy by 67%. The effects were attributed to the reversal of the tannin-microbial enzyme

complexes in the hindgut. Furthermore, common brushtail possums are known to possess

tannin-binding proteins in saliva, but their efficacy against detrimental effects of tannins is

limited (McArthur et al., 1995).

Apart from tannins Eucalyptus foliage typically contains a plethora of toxins, including

terpenes, cyanogenic glycosides and potent antifeedants, which are formylated

phloroglucinol compounds (FPCs) found only in eucalypts from subgenus Symphyomyrtus

(Eschler et al., 2000; Moore et al., 2004b; Gleadow et al., 2008). Nevertheless, common

brushtail possums can overcome the negative effects of toxins by having a broad diet and

activating different detoxification pathways (Marsh et al., 2006b; Wiggins, McArthur, &

Davies, 2006; Nersesian et al., 2012).

142

For instance, in contrast to more specialized mountain brushtail possums, common

brushtail possums are tolerant of sesquiterpenes, one of the most widespread terpene

groups among eucalypt species, due to their high glucuronidation capacity involved in the

detoxification of terpenes (Trichosurus cunninghami; Burchfield, Agar & Hume 2005).

Furthermore, common brushtail possums can recognise cyanogenic glycosides, such as

cyanide, and avoid these highly toxic compounds while foraging (Clapperton et al., 1996).

Also, formylated phloroglucinol compounds known to deter koalas and common ringtail

possums from foraging on many eucalypt species and individuals of the same species have

been found to affect common brushtail possums to a lesser extent (Marsh et al., 2003).

Consequently, this recognized tolerance of plant secondary metabolites by common

brushtail possums has shifted once again nutritional ecologists’ attention from plant

toxicity to plant nutritional quality.

Recently, a new foraging theory, the nitrogen availability hypothesis, has been proposed by

DeGabriel et al. (2009a, 2010) to explain herbivore foraging behaviour, habitat suitability,

and life strategies. According to this hypothesis, foliar concentration of available nitrogen,

rather than total nitrogen, tannins, and fibre individually, is a more ecologically meaningful

measure of plant nutritional quality for herbivores. A method developed by DeGabriel and

colleagues (2008) uses a simple in vitro assay to integrate foliar total nitrogen, fibre, tannins,

and dry matter digestibility into a single measure of available nitrogen to herbivores (see

Methods for detailed description).

The nitrogen availability hypothesis has been tested empirically on arboreal marsupials

(DeGabriel et al., 2009a; Stalenberg, 2010; Youngentob et al., 2011), moose (McArt et al.,

2009), and spider monkeys (Felton et al., 2009b). These nutritional studies ranged from

laboratory based feeding trials to foraging observations of wild populations and uniformly

found that both specialist and generalist herbivores preferred leaves from species and

individuals with lower levels of tannins and higher available nitrogen. Furthermore,

DeGabriel et al. (2009) demonstrated that the natural variation of available nitrogen rather

than total nitrogen among eucalypts growing in otherwise homogenous habitats has

affected the reproductive success of individual common brushtail possums. Hence, tannins

were identified as the main dietary constituent that limits the availability of nitrogen and

best explains reproductive output, explicitly linking herbivore nutrition and population

dynamics.

143

Wallis et al. (2012) argued that the main advantage of measuring foliar concentrations of

available nitrogen is in its broad, simple approach that allows the comparison of diverse

vegetation groups using the same currency. This new approach is especially relevant in the

Australian context, where the previous studies rarely acknowledged the dietary breadth of

arboreal marsupials and focused only on measuring the nutritional quality of eucalypts. A

great deal of work has been dedicated to explain how the Eucalyptus chemical composition

and variability affect arboreal marsupial foraging behaviour; however, herbivory on native

Acacia species has received much less attention (Dynes & Schlink, 2002; Irlbeck & Hume,

2003).

Additionally, hardly any data exist on the chemical composition of plants parasitising roots

and branches of different Eucalyptus and Acacia species and their overall effects on

herbivores (Close, Davidson, & Davies, 2006; Hensens, Lewis, & Mulquiney, 1971;

Preston, An, & Watson, 2010). Hence, the present study assessed for the first time the

nutritional quality of co-occurring Eucalyptus and Acacia species and parasitic plants,

Amyema pendula and Exocarpos cupressiformis, for a generalist arboreal marsupial, the common

brushtail possum (Trichosurus vulpecula).

144

Aims and hypotheses

The specific aims of this work focused on two research areas with underlying hypotheses

drawn from the existing research:

1. To investigate the chemical variability of trees and parasitic plants available to common brushtail

possums within their home ranges.

Hypothesis 1: Nitrogen-fixing Acacia species will have the highest foliar total nitrogen

concentrations, followed by nitrogen-accumulating parasitic plants and nitrogen-poor

Eucalyptus species (Irlbeck & Hume, 2003; Watson, 2009).

Hypothesis 2: Species from the same genera will have similar concentrations of foliar total

nitrogen, but will vary in concentrations of available nitrogen, tannins, and fibre

(DeGabriel, 2008; Wallis, Nicolle, & Foley, 2010).

Hypothesis 3: Parasitic plants should have the highest dry matter digestibility and lowest

dry matter content, followed by sclerophyllous Eucalyptus species and Acacia species

(McLeod, 1973; Umucalilar, Gulsen, Coskun, Hayirli, & Dural, 2007).

Hypothesis 4: Intra-specific chemical variability will be observed among different tree

species and parasitic plants, with the latter being more variable depending on the species of

the tree host (Lawler, Foley, Eschler, Pass, & Handasyde, 1998; Snyder, Fineschi, Linhart,

& Smith, 1996).

2. To establish the influence of chemical variability on the common brushtail possum’s dietary choices

and their home range size.

Hypothesis 1: Common brushtail possums are expected to prefer genera, species, and

individuals with higher foliar concentrations of available nitrogen, dry matter digestibility,

and lower dry matter content (after DeGabriel et al., 2009a).

Hypothesis 2: The size of common brushtail possum home ranges will be affected by the

nutritional quality of browse, with larger home ranges occurring in areas of lower

nutritional quality (after DeGabriel et al., 2009a).

145

Methods

Description and justification of methods used for measuring the nutritional quality of forage

The key factors explaining herbivore dietary choices are foliar concentrations of total

nitrogen, a proxy of protein content; tannins, one of the most ubiquitous plant secondary

metabolites; and fibre as a cell-wall constituent. In the past, many studies of the nutritional

quality of forage for herbivores overlooked the complex interactions between these

mutually complementary factors. Only recently, DeGabriel and colleagues (2008) proposed

a new integrative approach for measuring the additive and interactive effects of tannins and

fibre on nitrogen availability to herbivores. This method begins with two-stage in vitro

enzymatic digestion to establish the amount of foliar nitrogen and fibre digested by

herbivores; it then uses polyethylene glycol (PEG), a tannin binding agent, to determine the

amount of nitrogen inhibited by tannins. Application of PEG provides a direct method for

measuring the biological effects of tannins on herbivores without a need to classify them

into hydrolysable or condensed tannins (Silanikove et al., 1996). Additionally, this method

provides a uniform comparative approach for studying cross-taxa chemical variability of

plants, overcoming the problem of otherwise incomparable genera and species (W. J. Foley,

pers. comm., February 2009). Hence, this simple integrative approach was considered to be

the most suitable method for quantifying and comparing the nutritional quality of different

trees and parasitic plants available to common brushtail possums within their home ranges.

Another problem encountered in this study was finding a robust method for chemically

analysing a large number of leaf samples from the common brushtail possums’ relatively

large home ranges (see Chapter 3). Near Infrared Spectroscopy (NIRS) proved to be a

simple, time- and cost-effective alternative to conventional analytical methods (Foley et al.,

1998). This method employs multivariate statistical models to relate spectral characteristics

of a plant material to its chemical properties. In particular, when samples are irradiated with

near-infrared light their reflectance spectra represent a mixture of chemical bonds (C—H,

N—H, and O—H) characteristic for a specific chemical compound. These spectra are then

statistically calibrated against the reference values for a specific chemical compound that is

analysed using traditional laboratory methods. Next, multivariate regression equations are

developed using the partial least square regression method to estimate a chemical

composition of multiple samples. Thus, NIRS represents a practical and robust alternative

to the complicated, slow, and expensive chemical methods used for estimating foliar

chemical composition of multiple species over large geographic areas (Ebbers et al., 2002).

146

Sampling design and collection of leaf material

To describe the foraging environment of common brushtail possums and quantify and

compare nutritional quality of different trees and parasitic plants, leaf material was collected

from six randomly chosen home ranges (2 male and 4 female, as described in Chapter 3).

The sampling design included trees used (n1 = 98) and unused by common brushtail

possums (n2 = 98) while foraging at night, as well as trees parasitized by drooping mistletoe

(Amyema pendula; n3 = 75) and unparasitized trees (n4 = 75; Fig. 1). Additionally, 75 samples

of mistletoe foliage were collected from both Eucalyptus and Acacia tree hosts. Used trees

were considered to represent an accurate proxy of possum foraging; however, the

possibility of false absences could not be ruled out (Martin & Handasyde, 1999).

Figure 2 illustrates locations of all sampled trees, including possum used and mistletoe

parasitized trees. Tree diameter and height were recorded using a hand-held laser

hypsometer (LaserAce®, MDL Laser Systems, UK). Since diameter and height were highly

correlated (Spearman’s Rho = 0.79, P < 0.001), only tree diameter was reported as a proxy

of tree age in this study. Moreover, biomass of all sampled trees was calculated using a

general allometric equation developed by Keith, Barrett, and Keenan (2000; see Chapter 2

for details).

The collection of leaf material was carried out during a two week sampling period in April

2009 following the end of a radio-tracking survey of common brushtail possum performed

to establish space use patterns and resource selection (see Chapter 3). A single sampling

effort was considered to be an accurate representation of the nutritional quality of browse

available to possums at a given point in time; therefore, overcoming the problem of

seasonal fluctuations in foliar nitrogen and tannin levels (Cooper, Owen-Smith, & Bryant,

1988; Fox & Macauley, 1977). In total, 418 leaf samples were collected, with on average

69.7 ± 4.2 samples per home range. These samples included nine species from Eucalyptus,

two from Acacia, and the parasitic species, Amyema pendula and Exocarpos cupressiformis

(Table 1).

For the purpose of this study, the genus Eucalyptus was divided into subgenera Monocalyptus

and Symphyomyrtus, since the two subgenera are known to differ in plant secondary

metabolites and dependant groups of herbivores (Eschler et al., 2000; Gleadow et al., 2008;

Noble, 1989). Despite a small sample size (n = 5), Eucalyptus obliqua from the subgenus

Monocalyptus was included in the analysis since its foliar chemical composition has not been

previously studied.

147

To establish chemical differences between parasitic plants and their tree hosts, only

samples of Amyema pendula were analysed since it was impossible to identify tree hosts of

the root parasite, Exocarpos cupressiformis, without extensive root excavation (Table 2). A

single terminal branch containing both mature and juvenile leaves was collected from the

centre of a tree canopy using handheld secateurs mounted on a telescopic aluminium pole

or with the help of a professional arborist (Tree Tactics, Pty Ltd). Since common brushtail

possums are known to forage on both mature and juvenile leaves, they (50–100 g) were

analysed together to establish average browse nutritional quality (Loney, McArthur, Potts,

& Jordan, 2006). Collected samples were placed on ice in the field and later transferred to a

-20°C freezer to prevent chemical degradation of leaf material. Leaves were then freeze

dried (CSIRO, Canberra, ACT) and passed through a 1 mm sieve in a Cyclotec 1093 mill

(Tecator, Sweden).

148

Fig. 1 Schematic diagram of the sampling design employed to establish the nutritional quality of forage available to common brushtail possums. Depicted here is the within home range comparison of possum used (A), unused/unparasitized (B) and mistletoe parasitized trees (C) and between home range comparison of trees used by common brushtail possums(D)

149

Fig. 2 Location of the sampled trees within home ranges of the six common brushtail possums (CBP; 2 male and 4 female) in the Strathbogie Ranges, Victoria

150

Table 1 Summary of foliage samples collected within the six common brushtail possum home ranges

Table 2 Summary of Amyema pendula foliage samples collected from different species of tree hosts within the six common brushtail possum home ranges

Genus

Species

No samples

Acacia

Acacia dealbata 41 Acacia melanoxylon

21

Amyema

Amyema pendula

75

Exocarpos

Exocarpos cupressiformis

22

Monocalyptus

Eucalyptus radiata 95 Eucalyptus dives Eucalyptus obliqua

17 5

Symphyomyrtus

Eucalyptus camphora 75 Eucalyptus viminalis 39 Eucalyptus globulus

21

Genus

Species

No mistletoe samples

Acacia

Acacia dealbata

12

Monocalyptus

Eucalyptus radiata 31 Eucalyptus dives

7

Symphyomyrtus

Eucalyptus camphora 12 Eucalyptus viminalis 8 Eucalyptus globulus

5

151

Near-Infrared Spectroscopy (NIRS)

Collection of near-infrared spectra of leaf material

To minimize residual moisture and particle size affecting readings of near-infrared spectra,

all freeze dried and ground leaf samples (uniform sample particle size 1 mm) were placed in

an oven at 40⁰C for at least 1 h to remove residual moisture. Samples were then spread

evenly in a scanning quarter cup and were scanned with a near-infrared reflectance

spectrometer with a spinning cup attachment (Model 6500, Foss NIRSystems, Silver

Spring, MD) at the Australian National University, Canberra, ACT (after Foley et al., 1998).

Reflectance spectra of all samples were collected between 408–1093 nm and 1108–2493

nm. Each sample was scanned twice or until the root mean square of the two scans (scored

as log (1/reflectance)) was less than 3.0 x 10-4. The spectra were then averaged.

For an independent validation set, 20 samples including all studied species were selected.

Calibration set and NIRS equations development

The purpose of the calibration process in this study was to develop Near-Infrared

Spectroscopy equations used to predict the concentrations of chemical compounds for all

samples based on their near-infrared spectra (n = 418; see Laboratory Analyses for

detailed description of selected chemical compounds). To identify samples most suitable

for NIRS equations development, the algorithm SELECT was used in the NIRS 3, version

4.00 Software (WinISI Infrasoft International, Port Matilda, PA). The algorithm SELECT

employs principal component analysis (PCA) and Mahalanobis distances (H statistics) to

identify a subset of samples with spectra representing the full spectral variability for all

collected samples; these samples formed the calibration set (Shenk & Westerhaus, 1991).

The calibration set was then subjected to chemical analyses to obtain reference values for

developing the best NIRS equations. In this study, 158 samples were selected as the most

representative for a calibration set and their chemical composition was analysed in the

laboratory.

152

The next step in developing NIRS equations was to choose by a trial-and-error approach

the equations with lowest standard error of calibration (SEC) and highest coefficient of

determination (R2) for each of the studied chemical compounds using different regression

models and mathematical treatments to reduce background effects (Foley et al., 1998;

Woolnough & Foley, 2002). After trying a number of regression models, the modified

partial-least squares regression (MPLS) proved to be the most suitable for developing

NIRS equations apart from one case where partial-least squares regression (PLS) was used

(Table 3). The MPLS and PLS techniques combine the principal component analysis

(PCA) and multiple linear regression analyses. PCA is used to reduce the number of

measured near infrared spectra to just a few combinations suitable for describing most

spectral variability while multiple linear regression relates spectra to the reference values of

the calibration set. In addition to MPLS and PLS, a variety of mathematical treatments

were used to reduce the potential variation in spectra due to differences in sample particle

size and residual moisture, as well as to reduce multicollinearity caused by developing

equations based on more than one wavelength (Batten, 1998; Table 3). For example, a

mathematical treatment of “2, 4, 4, 1” refers to using the second derivative, leaving the gap

of four wavebands between calculated values, performing a first smoothing over four

wavebands and then a second smoothing over one waveband.

153

Table 3 Calibration statistics and mathematical treatments of the Near-Infrared Spectroscopy equations†.† Abbreviations: constituent = foliar chemical

attribute measured in the presence and absence of polyethylene glycol (PEG); n = number of samples used to develop the NIRS equations; = mean value of samples; SD = standard deviation from the mean; SECV = standard error of cross validation; R² and 1-VR = degree of correlation between predicted and actual measures; Scatter correction = standard normal variate (SNV) and detrend of mathematical transformations; Math treatment = Savitzy-Golay spectral-smoothing function; Regression type = modified partial least square (MPLS) and partial least square regression (PLS)

Constituent

N

SD

R²

SECV

1-VR

Wavelength used

Scatter correction

Math treatment

Regression type

Total N 154 1.59 0.48 0.99 0.07 0.98 All SNV Detrend 2, 4, 4, 1 MPLS

DMD + PEG 148 46.76 7.76 0.85 3.42 0.81 All SNV 1, 8, 8, 1 MPLS

DMD 153 48.29 8.38 0.86 4.09 0.76 All SNV 2, 8, 8, 1 MPLS

N digestibility + PEG 151 0.70 0.07 0.55 0.05 0.45 1108–2492.8 None 2, 4, 4, 1 PLS

N digestibility 150 0.44 0.16 0.86 0.08 0.79 All SNV Detrend 2, 6, 4, 1 MPLS

Available N + PEG 153 1.11 0.37 0.94 0.11 0.91 1108–2492.8 SNV 2, 8, 8, 1 MPLS

Available N

150

0.74

0.41

0.93

0.12

0.91

All

SNV Detrend

2, 4, 4, 1

MPLS

154

Validation procedures

The validity of the NIRS equations was tested twice. The first test was used when

performing the cross-validation procedure during the development of equations for all

analysed samples (Shenk & Westerhaus, 1991). This cross-validation procedure involves

separating all samples into groups and performing calibrations on all but one of the groups,

which is then used as an independent validation set. The procedure is repeated multiple

times until all groups of samples have been cross-validated. In this way, SECV (standard

error of cross-validation) and the coefficient of the determination of cross-validation

(1-VR) are generated, which explained the error associated with predictions (Table 3).

In the second test, the validity of the NIRS equations was evaluated only for a calibration

set (n = 158) by fitting simple linear regression models of predicted concentrations of

selected chemical compounds against concentrations of the same compounds measured in

the laboratory. Predicted concentrations of the selected foliar compounds strongly

correlated with the concentrations measured in the laboratory (P < 0.001, Fig. 3).

Moreover, to further investigate the accuracy and robustness of NIRS equations, an

independent set of 20 samples was analysed using conventional laboratory techniques. The

validity of the NIRS equations was again assessed using simple linear regression. The

predicted concentrations of the selected foliar compounds fitted well with the measured

ones (P < 0.001, Fig. 4).

155

a)

b)

c) Fig. 3 Linear regression of Near-Infrared Spectroscopy predicted and measured in the laboratory concentrations of the selected foliar compounds for a calibration set (n = 158). Figures represent a) total nitrogen (% DM); b) available nitrogen (% DM); c) dry matter digestibility (% DMD). Dotted lines at y = x are for reference only

156

a)

b)

c) Fig. 4 Linear regression of Near-Infrared Spectroscopy predicted and measured in the laboratory concentrations of the selected foliar compounds for a validation set (n = 20). Figures represent a) total nitrogen (%DM); b) available nitrogen (%DM); c) dry matter digestibility (%DMD). Dotted lines at y = x are for reference only

157

Laboratory chemical analyses

In vitro determinations of foliar digestibility and nutrient availability

Considering that only limited amounts of foliar dry matter and total nitrogen are actually

digestible by herbivores, a multiple-step chemical analysis was performed to establish the

amount of nitrogen available to herbivores (Fig. 5). First, to quantify foliar dry matter

digestibility and nitrogen digestibility, the in vitro two-stage enzymatic digestion was

performed using pepsin/cellulase following the method of DeGabriel et al. (2008).

The in vitro pepsin/cellulase digestion emulated digestion of foliar material in hindgut

fermenters, with pepsin simulating nitrogen digestion and cellulase simulating fibre

digestion. In detail, samples from a calibration set (n = 158) and an independent validation

set (n = 20) were weighed into pre-weighed filter bags (ANKOM F57, ANKOM

Technology, Macedon, NY) in four 0.80 ± 0.01 g replicates. Next, duplicate samples were

treated with polyethylene glycol (PEG 4000 analytical grade, Sigma, St. Louis, MO) in

0.05 M Tris-BASE buffer. The remaining control samples were treated only with buffer.

The samples were placed in an incubator at 33°C and gently stirred for 24 h. All samples

were washed, dried, and weighed to a constant mass. Samples were then incubated in

pepsin (2.00 g 1:10,000 pepsin in 1 L 0.1 N HCl) for 24 h, washed, and placed in cellulase

(6.25 g cellulase in 1 L of 100 mmol acetic acid buffer) for 48 h and stirred gently at 33°C.

Following enzymatic digestion and washing, samples were oven-dried at 60°C to a constant

mass and remaining indigested residue weighed to calculate dry matter digestibility (%).

To estimate nitrogen digestibility (Equation 1, Fig. 5), the concentrations of total nitrogen

(% DM) and nitrogen left after digestion (% DM) were measured using the Dumas

technique with a LECO TrueSpec combustion N analyser calibrated against EDTA

(LECO Corporation, St. Joseph, MI). Next, available nitrogen (% DM) concentration was

obtained by multiplying nitrogen digestibility with total nitrogen (% DM) in the presence

and absence of PEG (Equation 2). Finally, by calculating the amount of available nitrogen

(% DM) bound with PEG, it was possible to establish amounts of tannins and fibre

bounded with nitrogen (Equations 3 and 4). The concentration of available nitrogen,

tannin- and fibre-bound nitrogen summed up to total N (% DM; Equation 5, Fig. 5).

158

Equations: 1. N digestibility (% Total N) = (Total N (mg) – N remaining in the indigested residue (mg))/ Total N (mg) 2. Available N (% DM) = Digestible N (% Total N) x Total N (% DM) 3. Tannin-bound N (% DM) = Available N with PEG (% DM) – Available N (% DM) 4. Fibre-bound N (% DM) = Total N (% DM) – Available N with PEG (% DM) 5. Total N (% DM) = Available N (% DM) + Tannin-bound N (% DM) + Fibre-bound N (% DM)

Fig. 5 Example of the five step calculation process for determining different nitrogen fractions in foliage

N digestibility is the amount of Total N lost during in vitro digestion

Available N is the amount of Total N available to herbivores

Tannin-bound N is the amount of Total N precipitated by tannins

Fibre-bound N is the amount of Total N immobilised by fibre

159

Data analysis

To investigate relationships between the studied chemical compounds, the Spearman’s

rank-order correlation and the simple linear regression were performed. To compare inter-

and intra-specific chemical variability of trees and parasitic plants available to common

brushtail possums within their home ranges, the one-way analysis of variance (ANOVA)

analysis with the post-hoc Tukey’s multiple comparisons test was carried out. Data were

tested for normality using the Shapiro-Wilk test and were log transformed if they departed

from normality. To investigate differences in proportions of available nitrogen, tannin-

bound nitrogen, and fibre-bound nitrogen across different species, the test for equality of

proportions was used, followed by a pair-wise comparison of species with the Bonferroni

correction for alpha.

To establish differences between trees used and unused by common brushtail possums, as

well as trees parasitized and unparasitized by mistletoes, the Welch’s t-test was used. The

Kruskal-Wallis test was used to compare species differences across common brushtail

possum home ranges. To identify which chemical compounds best explained possum tree

use and mistletoe tree parasitism, the binomial generalized linear model was fitted to all

species and next to each species individually (Nelder & Wedderburn, 1972). Firstly, any

highly inter-correlated explanatory variables (Spearman’s Rho > 0.7) were excluded from

the analysis to avoid multicollinearity in the modelling process (Tabachnick & Fidell, 1996).

Secondly, the stepAIC function from the MASS library was used to find the best

combination of explanatory variables. The stepAIC function used AIC as the step criterion

in both directions. All results were reported as observed means ± s.e. and statistical

significance followed the convention of P < 0.05. All statistical analyses were performed in

Spotfire S+® Version 8.1 (2010, TIBCO Software Inc. Somerville, MA).

160

Results

Foliar chemical attributes

Foliar concentrations of total and available nitrogen (% DM) were strongly positively

correlated in all studied plant genera (Spearman’s Rho = 0.86, P < 0.001, Fig. 6) with

R2 ranging from 0.86 for Acacia to 0.96 for Amyema (Fig. 8). An exception was the

subgenus Monocalyptus, in which foliar concentrations of available nitrogen were more

variable than those observed for total nitrogen, leading to a much lower coefficient of

correlation (R2 = 0.37) compared to subgenus Symphyomyrtus (R2 = 0.88). Further

examination of variation in the subgenus Monocalyptus showed strong differences between

species with R2 values ranging from 0.70 for Eucalyptus radiata to 0.30 for E. obliqua. Both

total and available nitrogen were negatively correlated with tree diameter across all genera

(Spearman’s Rho = - 0.53, - 0.58 respectively, P < 0.001).

No correlation was observed between foliar concentrations of total nitrogen and tannin-

bound nitrogen (% DM) when considering all species (Spearman’s Rho = 0.06, P < 0.001,

Fig. 6). However, when analysed separately, tannin-bound nitrogen of Acacia foliage

proved to be moderately correlated with total nitrogen (% DM; R2 = 0.39, Fig. 9).

Conversely, tannin-bound nitrogen was negatively correlated with available nitrogen (%

DM) while both total and available nitrogen were positively correlated with fibre-bound

nitrogen (% DM; Fig. 6). Foliar concentrations of fibre-bound nitrogen (% DM) were

highly variable for genera Amyema and Symphyomyrtus producing lower R2 values compared

to the other genera (Fig. 10). Tree diameter was negatively correlated with fibre-bound

nitrogen (% DM; Spearman’s Rho = - 0.46, P < 0.001) and positively correlated with

tannin-bound nitrogen (% DM; Spearman’s Rho = 0.16, P < 0.001).

Nitrogen digestibility (% total N) and dry matter digestibility (% DMD) were positively

correlated (Spearman’s Rho = 0.68, P < 0.001, Fig. 7). Furthermore, both nitrogen

digestibility and dry matter digestibility were positively correlated with available nitrogen

(% DM; Spearman’s Rho = 0.78, 0.36 respectively), negatively correlated with tannin-

bound nitrogen (% DM; Spearman’s Rho = - 0.63, - 0.73 respectively), and negatively

correlated with foliar dry matter content (Spearman’s Rho = - 0.36, - 0.30 respectively, P <

0.001, Fig. 7).

161

Fig. 6 Spearman’s Rho correlations of selected chemical variables for all species (n = 413). An asterix (*) indicates a significant correlation at P < 0.001. Arrows show direction and strength of correlation

Figure 7 Spearman’s Rho correlations of foliar nitrogen and dry matter digestibility for all species (n = 413). An asterix (*) indicates a significant correlation at P < 0.001. Arrows show direction and strength of correlation

162

Fig. 8 Linear regression of total nitrogen (% DM) against available nitrogen (% DM) of different genera and Eucalyptus subgenera. The equation and R2 value refer to linear regression performed on all samples (n = 413)

Fig. 9 Linear regression of total nitrogen (% DM) against tannin-bound nitrogen (% DM) of different genera and Eucalyptus subgenera. The equation and R2 value refer to linear regression performed on all samples (n = 413)

163

Fig. 10 Linear regression of total nitrogen (% DM) against fibre-bound nitrogen (% DM) of different genera and Eucalyptus subgenera. The equation and R2 value refer to linear regression performed on all samples (n = 413)

164

Foliar nitrogen composition of trees and parasitic plants

The foliar total nitrogen concentration was highest in nitrogen-fixing Acacia, followed by

the root parasite, Exocarpos, and eucalypts from subgenus Symphyomyrtus (% DM; P < 0.001,

Fig. 11). Against expectations, an aerial parasite Amyema and eucalypts from the subgenus

Monocalyptus had the lowest concentrations of total nitrogen (% DM; P < 0.001, Fig. 11).

Plant genera varied in foliar concentrations of available nitrogen, tannin- and fibre-bound

nitrogen (% DM; P < 0.001, Fig. 12). Genus Monocalyptus had the lowest concentration of

available nitrogen and was followed by genera Amyema and Symphyomyrtus, which had similar

concentrations of available nitrogen. The genus with the highest concentration of available

nitrogen was Acacia, which had also the highest concentration of fibre-bound nitrogen

while Monocalyptus had the highest concentration of tannin-bound nitrogen (% DM;

P < 0.001, Fig. 12).

165

Fig. 11 Mean foliar concentration of total nitrogen (% DM) with standard error (s.e.) of different plant genera and Eucalyptus subgenera. Letters above the bars represent significant differences between the groups at P < 0.001

Fig. 12 Mean foliar concentrations of available nitrogen (black bar), tannin-bound nitrogen (grey), and fibre-bound nitrogen (white) with standard error (s.e.) of different plant genera and Eucalyptus subgenera. The foliar concentrations of available nitrogen, tannin-bound nitrogen and fibre-bound nitrogen sum to total nitrogen (% DM) concentration

166

From all species Acacia dealbata had the highest concentration of foliar total nitrogen

(%DM) while Eucalyptus camphora had the highest total nitrogen (% DM) of all Eucalyptus

species (P < 0.001, Fig. 13, Table 4). Of the two parasitic species, Exocarpos cupressiformis

had higher concentration of foliar total nitrogen (% DM) than Amyema pendula (P < 0.001,

Fig. 13, Table 4). Acacia dealbata had the highest available nitrogen (% DM), followed by

Exocarpos cupressiformis and Eucalyptus camphora while species from Monocalyptus subgenera

had the lowest (P < 0.001, Fig. 13, Table 4). Moreover, Monocalyptus species had the

significantly highest foliar concentrations of tannin-bound nitrogen (% DM), especially

Eucalyptus radiata (P < 0.001, Fig. 13, Table 4). Foliar concentrations of tannin-bound

nitrogen (% DM) were the lowest in E. viminalis and Amyema pendula while fibre-bound

nitrogen (% DM) was the highest in the two Acacia species, A. dealbata and A. melanoxylon

(P < 0.001, Fig. 13, Table 4).

All studied species had similar proportions of fibre-bound nitrogen ranging from 26 to

33% (χ2 = 1.65, df = 9, P = 0.996 Fig. 14). Also, most of the species had similar

proportions of tannin-bound nitrogen (13–17%), apart from the species from the subgenus

Monocalyptus (χ2 = 34.49, df = 9, P < 0.001) with Eucalyptus radiata having 44 % of total

nitrogen bound with tannins. Consequently, due to extremely high proportions of tannins

in Monocalyptus species, only small amounts of total nitrogen (5–27 %) were actually

available to herbivores in comparison to other species where the proportions of available

nitrogen ranged from 41 % for E. globulus to 61 % for Exocarpos cupressiformis (χ2 = 24.19,

df = 9, P < 0.05).

167

Fig. 13 Mean foliar concentrations of available nitrogen (black bar), tannin-bound nitrogen (grey), and fibre-bound nitrogen (white) with standard error (s.e.) of different plant genera, Eucalyptus subgenera and species. The foliar concentrations of available nitrogen, tannin-bound nitrogen, and fibre-bound nitrogen sum to total nitrogen (% DM) concentration

168

Fig. 14 Percentage (%) of available nitrogen (black bar), tannin-bound nitrogen (grey), and fibre-bound nitrogen (white) in foliage of different plant species. The foliar concentrations of available nitrogen, tannin-bound nitrogen, and fibre-bound nitrogen sum to total nitrogen (% DM) concentration

169

Dry matter digestibility and dry matter content of trees and parasitic plants

The species with the highest dry matter digestibility (% DMD) was drooping mistletoe

Amyema pendula (P < 0.001, Fig. 15, Table 4), followed by Exocarpos cupressiformis and

Eucalyptus viminalis. The species with the lowest dry matter digestibility (% DMD) were

Eucalyptus dives and E. radiata, both from genus Monocalyptus (P < 0.001, Fig. 15, Table 4).

The genera with the significantly lowest dry matter content (%) were parasitic Amyema and

Exocarpos while Monocalyptus had the highest (P < 0.001, Fig. 16).

Fig. 15 Mean foliar dry matter digestibility (% DMD) with standard error (s.e.) of different plant genera and Eucalyptus subgenera. Letters above the bars represent significant differences among the groups at P < 0.001

Fig. 16 Mean foliar dry matter content (%) with standard error (s.e.) of different plant genera and Eucalyptus subgenera. Letters above the bars represent significant differences among the groups at P < 0.001

170

Table 4 Summary statistics of foliar chemical attributes of different tree species and parasitic plants available to common brushtail possums within their home ranges. Letters represent significant differences among the groups at P < 0.001

Species Statistics Acacia dealbata

Acacia melanoxylon

Amyema pendula

Exocarpos cupressiformis

Eucalyptus dives

Eucalyptus obliqua

Eucalyptus radiata

Eucalyptus camphora

Eucalyptus globulus

Eucalyptus viminalis

No. samples

41 21 75 22 17 5 95 75 21 41

Total N Mean 2.60e 2.27d 1.44b 1.89c 1.20a 1.08a 1.43b 1.76c 1.11a 1.49b

(% DM) s.e. 0.04 0.04 0.03 0.05 0.02 0.03 0.01 0.02 0.02 0.02

CV (%) 9.39 8.46 20.21 11.29 8.41 5.86 8.40 10.88 8.39 8.30

Digestible N Mean 0.54c 0.52c 0.55c 0.59d 0.28a 0.31a 0.26a 0.52c 0.45b 0.56c

(% DM) s.e. 0.01 0.01 0.01 0.01 0.01 0.02 0.01 0.00 0.01 0.01

CV (%) 7.19 7.54 9.37 8.65 19.57 17.98 22.90 7.39 8.14 8.06

Available N Mean 1.35d 1.24d 0.80b 1.15d 0.43a 0.31a 0.39a 0.95c 0.45a 0.82b

(% DM) s.e. 0.03 0.03 0.03 0.04 0.03 0.03 0.01 0.02 0.02 0.02

CV (%) 13.99 11.89 28.15 15.53 24.39 23.56 28.16 14.54 15.99 13.55

Tannin-bound N Mean 0.44b 0.31a 0.22a 0.24a 0.45b 0.46b 0.63c 0.33a 0.22a 0.18a

(% DM) s.e. 0.01 0.01 0.01 0.02 0.01 0.02 0.01 0.01 0.02 0.02

CV (%) 16.19 20.30 36.68 42.05 13.26 9.20 9.46 33.90 34.32 55.51

Fibre-bound N Mean 0.82d 0.73c 0.41b 0.50b 0.33a 0.31a 0.42b 0.48b 0.44b 0.49b

(% DM) s.e. 0.01 0.02 0.01 0.02 0.01 0.02 0.01 0.01 0.02 0.02

CV (%) 7.98 13.72 17.35 14.66 9.55 16.75 13.95 21.43 17.00 19.72

DM Mean 50.10b 48.22a 44.59a 45.64a 53.24b 47.48a 54.55b 48.65a 55.13b 56.24b

(%) s.e. 1.12 1.34 0.60 1.10 0.54 1.32 0.62 0.60 1.39 0.88

CV (%) 14.34 12.77 11.61 11.28 4.21 6.23 11.06 10.76 11.59 10.03

DMD Mean 47.68b 43.21a 59.81e 45.87b 37.46a 40.30a 39.71a 51.46c 50.12c 53.20d

(%) s.e. 0.28 0.41 0.24 0.35 0.51 1.05 0.29 0.22 0.54 0.24

CV (%) 3.71 4.38 3.48 3.61 5.62 5.82 7.04 3.70 4.93 2.89

DBH Mean 17.22a 24.05a n. a. 16.09a 74.47d 62.60c 79.36 44.20b 41.57b 65.41c

(%) s.e. 1.52 3.40 n. a. 1.66 5.83 13.65 3.30 2.20 5.98 3.59

CV (%) 56.67 64.79 n. a. 48.29 32.29 48.75 40.47 44.15 65.87 35.18

171

Intra-specific variability in the foliar attributes

Chemical variability within species was expressed as the coefficient of variation, which

describes the variation as a proportion of the mean (Table 4). There was little within-

species variation observed for all studied tree species in terms of total nitrogen. Tree

species with the highest coefficient of variation of available nitrogen were from the genus

Monocalyptus exceeding 20% while Symphyomyrtus species were the most variable in terms of

tannins and fibre. From all species, Eucalyptus viminalis was the most variable in foliar tannin

levels (> 50%), followed by Exocarpos cupressiformis (42%) while Eucalyptus camphora was

most variable in terms of fibre (Table 4).

The most variable in terms of foliar concentrations of total and available nitrogen was

drooping mistletoe (Amyema pendula, Table 4). Drooping mistletoe’s total nitrogen (% DM)

was lower than the total nitrogen of its tree hosts, Acacia dealbata (t = 9.19, df = 21.79,

P < 0.001, Fig. 17) and Eucalyptus camphora (t = 5.13, df = 21.83, P < 0.001, Fig. 17), but

higher than E. dives (t = -2.36, df = 4.81, P = 0.0667). No significant differences in total

nitrogen (% DM) were found between mistletoes and their tree hosts, E. radiata or E.

viminalis (t = 1.66, df = 47.61, P = 0.1026; t = 2.067, df = 12.78, P = 0.06 respectively,

Fig. 17).

Foliar concentrations of total and available nitrogen (% DM) of Amyema pendula depended

on the species of host and were highest in mistletoes parasitizing Acacia dealbata (P < 0.001,

Fig. 18). No significant differences in concentrations of total and available nitrogen

(% DM) were detected in mistletoes growing on eucalypt species from the Monocalyptus and

Symphyomyrtus subgenera (% DM, Fig. 18). Mistletoe tannin-bound nitrogen was not

significantly different among different species of hosts while fibre-bound nitrogen was

highest in mistletoes parasitizing Amyema pendula. No significant relationship was found

between foliar concentrations of total nitrogen (% DM) between mistletoes and their tree

hosts (P < 0.05, Fig. 19).

Mistletoes parasitizing different species of tree hosts had significantly lower dry matter

content (% DM) and higher dry matter digestibility (% DMD) than their hosts (P < 0.001).

Mistletoe foliar dry matter content (% DM) varied depending on the species of tree host

and was highest in mistletoes parasitizing E. radiata (46.11 ± 0.95) and lowest in mistletoes

parasitizing E. dives (39.67 ± 1.65; P < 0.001).

172

Dry matter digestibility (% DMD) was highest in mistletoes parasitizing E. camphora (60.98

± 0.32) while lowest in mistletoes parasitizing E. dives (58.20 ± 0.50; P < 0.001).

The Welch’s t-test revealed no significant differences in foliar concentrations of total

nitrogen in among trees parasitized and free from Amyema pendula. Only the parasitized

trees of E. dives and Acacia dealbata had lower foliar concentrations of total nitrogen

(% DM) than unparasitized ones (t = - 2.49, df = 6.42, P < 0.05; t = - 2.62, df = 20.64,

P < 0.05 respectively). Also, parasitized trees Eucalyptus dives and Acacia dealbata had lower

foliar concentrations of available nitrogen (% DM) than unparasitized ones (t = - 3.34,

df = 11.99, P < 0.05; t = - 2.35, df = 20.38, P < 0.05 respectively). No significant

differences were observed among parasitized and unparasitized trees in terms of foliar dry

matter content (% DM) and dry matter digestibility (% DMD). Mistletoe parasitized E.

radiata and E. camphora had a significantly higher diameter than trees free from Amyema

pendula (t = 2.11, df = 59.14, P < 0.05; t = 3.09, df = 18.23, P < 0.001 respectively).

Fig. 17 Mean foliar total nitrogen content (% DM) with standard error (s.e.) of different tree hosts and drooping mistletoe (Amyema pendula). An asterix (*) indicates significant differences between the groups at P < 0.001

*

173

Fig. 18 Mean foliar concentrations of available nitrogen (black bar), tannin-bound nitrogen (grey), and fibre-bound nitrogen (white) with standard error (s.e.) of dropping mistletoe (Amyema pendula) parasitizing different species of tree hosts. The foliar concentrations of available nitrogen, tannin-bound nitrogen, and fibre-bound nitrogen sum to total nitrogen (% DM) concentration

Figure 19 Linear regression of total nitrogen (% DM) of different tree hosts against total nitrogen (% DM) of drooping mistletoe (Amyema pendula)

174

The binomial generalized linear model showed that a key parameter separating trees

parasitized by Amyema pendula from those free from mistletoes was tree diameter (r2 = 0.03,

P < 0.05). Trees larger in diameter were more often parasitized by Amyema pendula

(Table 5). The model explained 56% of mistletoe tree presences and 64% of absences.

Table 5 The binomial generalized linear model predicting tree occupancy by drooping mistletoe Amyema pendula (AIC = 205.45). An asterix indicates significance at P < 0 ‘***’, P < 0.001 ‘**’ and P < 0.05 ‘*’

Coefficients Estimate Std. Error z-value P-value Intercept 0.762 0.350 2.178 0.029 * Diameter -0.012 0.005 -2.452 0.014 * Binomial generalized linear model

df Dev. Resid. df Resid. Dev. P-value

Null 149 207.94 Diameter 1 6.4938 148 201.45 0.011*

Inter-specific differences of trees used by common brushtail possums

Among all trees used by common brushtail possums, rarely-used Acacia dealbata (see

Chapter 2 and 3) had the highest foliar concentrations of total and available nitrogen (%

DM) and the highest concentration of fibre-bound nitrogen (P < 0.001, Fig. 20). From

five Eucalyptus species present in the study area, E. camphora, the second most used tree

species by common brushtail possums, had the highest concentrations of total and

available nitrogen (% DM, P < 0.001, Fig. 20). Conversely, the most often used tree

species, Eucalyptus radiata, had the lowest foliar concentration of available nitrogen (% DM)

and the highest tannin-bound nitrogen concentration of all tree species (P < 0.001, Fig.

20). The rarely used species, Eucalyptus obliqua and Eucalyptus globulus, had the lowest

concentration of total nitrogen (% DM; P < 0.001, Fig. 20). Furthermore, Eucalyptus

viminalis had the highest concentration of fibre-bound nitrogen of all Eucalyptus species (%

DM; P < 0.001, Fig. 20) and the highest dry matter content (58. 36 ± 1.21%), but also the

highest dry matter digestibility of all studied species (52.84 ± 0.44%). Tree species with the

lowest dry matter content was E. camphora (47.60 ± 1.21%) while E. radiata had the lowest

dry matter digestibility of all studied species (40.27 ± 0.41%).

175

Fig. 20 Mean foliar concentrations of available nitrogen (black bar), tannin-bound nitrogen (grey), and fibre-bound nitrogen (white) with standard error (s.e.) of different tree species used by common brushtail possums. The foliar concentrations of available nitrogen, tannin-bound nitrogen, and fibre-bound nitrogen sum to total nitrogen (% DM) concentration

176

Intra-specific differences of trees used by common brushtail possums

Hardly any significant differences were observed among trees used by common brushtail

possums and the closest unused trees of the same species in terms of all measured chemical

attributes (P > 0.05, Fig. 21 and 22, Table 6). A significant difference in diameter was

revealed for Eucalyptus radiata trees (P < 0.05). However, across common brushtail possum

home ranges, few significant differences between trees of the same species were discerned;

for example, Eucalyptus viminalis and E. camphora varied in foliar concentrations of total and

available nitrogen, dry matter content, and dry matter digestibility (P < 0.05).

a)

b)

Fig. 21 Intra-specific differences in foliar chemistry of used and unused trees by common brushtail possums within their home ranges. Figures show (a) total nitrogen (% DM) and (b) available nitrogen (% DM) with standard error (s.e.) of different tree species

177

a)

b) Fig. 22 Intra-specific differences in foliar chemistry of used and unused trees by common brushtail possums within their home ranges. Figures show (a) dry matter digestibility (% DMD) and (b) dry matter content (%) with standard error (s.e.) of different tree species

178

Table 6 Summary statistics of foliar chemical attributes of different tree species used and unused by common brushtail possums within their home ranges. An asterix (*) indicates significant differences between the groups at P < 0.05

Species Stats Eucalyptus camphora

Eucalyptus viminalis

Eucalyptus globulus

Eucalyptus radiata

Eucalyptus obliqua

Acacia dealbata

Acacia melanoxylon

Exocarpos cupressiformis

No. samples 42 26 12 80 6 6 8 16

Total N (% DM) Mean ± s.e.

Used 1.76 ± 0.04 1.55 ± 0.04 1.14 ± 0.03 1.44 ± 0.02 1.05 ± 0.03 2.67 ± 0.14 2.19 ± 0.07 1.83 ± 0.07

Unused 1.83 ± 0.04 1.46 ± 0.04 1.06 ± 0.06 1.44 ± 0.02 1.10 ± 0.04 2.37 ± 0.02 2.33 ± 0.02 1.80 ± 0.08

Digest. N (% total N) Mean ± s.e.

Used 0.52 ± 0.01 0.57 ± 0.01 0.44 ± 0.01 0.27 ± 0.01 0.35 ± 0.01* 0.55 ± 0.03 0.51 ± 0.02 0.58 ± 0.02

Unused 0.53 ± 0.01 0.57 ± 0.01 0.44 ± 0.02 0.26 ± 0.01 0.26 ± 0.02 0.52 ± 0.01 0.53 ± 0.01 0.56 ± 0.02

Avail. N (% DM) Mean ± s.e.

Used 0.95 ± 0.03 0.87 ± 0.03 0.45 ± 0.02 0.41 ± 0.01 0.33 ± 0.05 1.39 ± 0.14 1.17 ± 0.07 1.10 ± 0.06

Unused 1.01 ± 0.03 0.80 ± 0.03 0.44 ± 0.04 0.39 ± 0.02 0.27 ± 0.03 1.20 ± 0.02 1.28 ± 0.01 1.09 ± 0.06

Tannins (% DM) Mean ± s.e.

Used 0.33 ± 0.03 0.17 ± 0.02 0.25 ± 0.03 0.62 ± 0.01 0.43 ± 0.02* 0.41 ± 0.01 0.32 ± 0.02 0.19 ± 0.01

Unused 0.35 ± 0.02 0.14 ± 0.02 0.17 ± 0.04 0.64 ± 0.01 0.50 ± 0.01 0.32 ± 0.04 0.32 ± 0.02 0.15 ± 0.01

Fibre (% DM) Mean ± s.e.

Used 0.48 ± 0.03 0.50 ± 0.03 0.43 ± 0.02 0.41 ± 0.01 0.29 ± 0.04 0.88 ± 0.02 0.70 ± 0.05 0.54 ± 0.02

Unused 0.46 ± 0.02 0.52 ± 0.02 0.46 ± 0.02 0.41 ± 0.01 0.33 ± 0.01 0.85 ± 0.03 0.72 ± 0.03 0.57 ± 0.02

DM (%) Mean ± s.e.

Used 47.60 ± 1.21 58.36 ± 1.21 54.73 ± 2.36 56.00 ± 1.09 48.38 ± 1.28 56.27 ± 6.26 51.04 ± 1.94 48.46 ± 2.73

Unused 46.66 ± 0.75 55.66 ± 1.52 58.05 ± 3.48 54.85 ± 0.91 45.11 ± 2.07 47.33 ± 4.43 52.90 ± 4.86 44.50 ± 0.49

DMD (%) Mean ± s.e.

Used 51.11 ± 0.36 52.84 ± 0.44 50.65 ± 1.18 40.27 ± 0.41* 41.50 ± 1.21 47.09 ± 0.94 43.29 ± 0.70 45.85 ± 0.77

Unused 50.71 ±0.30 53.18 ± 0.48 49.07 ± 1.19 39.84 ± 0.45 38.13 ± 0.74 47.02 ± 0.54 44.28 ± 0.70 45.66 ± 0.18

DBH (cm) Mean ± s.e.

Used 46.76 ±3.67 68.69 ± 6.05 60.00 ± 15.55 91.30 ± 6.13 58.67 ± 21.15 18.33 ± 5.84 29.00 ± 4.93 15.75 ± 3.76

Unused 45.05 ± 4.34 64.31 ± 5.57 38.50 ± 7.23 67.41 ± 3.46 75.67 ± 14.33 16.00 ± 1.00 26.75 ± 4.26 17.25 ± 1.89

179

Determinants of tree use by common brushtail possums

To discriminate which of the measured tree attributes best explained overall tree use by

common brushtail possums the binomial generalized linear model was fitted to all tree

species and each species individually. The final model indicated that tree diameter (P <

0.05) and foliar concentration of available nitrogen (P = 0.09) were the best variables for

predicting the presence of common brushtail possums across all analysed trees (Table 7).

The model explained 60.2% of common brushtail possum tree presences and 61.2% of

absences. For the most visited species, Eucalyptus radiata, the model predicted that diameter

(P < 0.001) was the best variable for predicting 60% of common brushtail possum

presences and 72.5% of absences while for E. camphora, foliar concentration of available

nitrogen (P = 0.14) explained 66.7% of common brushtail possum tree presences and

71.4% of absences.

Table 7 The binomial generalized linear model predicting tree use by common brushtail possums (AIC = 269.08). An asterix indicates significance at P < 0 ‘***’, P < 0.001 ‘**’ and P < 0.05 ’*’ Coefficients Estimate Std. Error z-value P-value Intercept -1.501 0.613 -2.448 0.014* Diameter 0.015 0.006 2.791 0.005* Available N 0.852 0.510 1.668 0.095 Binomial generalized linear model

df Dev. Resid. df Resid. Dev. P-value

Null 195 271.71 Diameter 1 5.798 194 265.92 0.016 * Available N 1 2.837 193 263.08 0.092

Relationships between tree attributes and common brushtail possum home range size

Finally, when comparing the nutritional quality (mean foliar concentration of available

nitrogen) of individual possum home ranges, significant differences (P < 0.001) were

observed between individual animals (female 1 and 5, female 4 and female 5, female 4 and

male 3, and female 1 and male 3). However, the Spearman’s rank-order correlation showed

that none of the measured foliar attributes influenced possum home range size (P > 0.05).

180

Discussion

Summary of key results

In the present study, the phytochemical environment of common brushtail possums

included species of different nutritional quality. The nitrogen-fixing Acacia dealbata had the

highest concentrations of foliar total and available nitrogen, as well as the highest

concentration of fibre-bound nitrogen, of all species. The co-occurring root parasite,

Exocarpos cupressiformis, and eucalypts from subgenus Symphyomyrtus had relatively high

concentrations of available nitrogen with Eucalyptus viminalis found to have the lowest

concentration of tannin-bound nitrogen of all species. The aerial parasite, Amyema pendula,

had moderate concentrations of available nitrogen, which varied among individuals

depending on a species of host, with the most foliar nitrogen found in mistletoes

parasitizing Acacia dealbata trees. Mistletoes also had the lowest dry matter content and

highest dry matter digestibility. Eucalypts from the subgenus Monocalyptus were found to be

of the poorest nutritional quality for common brushtail possums, due to their low

concentrations of available nitrogen, high concentrations of tannin-bound nitrogen, and

low dry matter digestibility.

Contrary to expectation, common brushtail possums were found most often foraging in

Eucalyptus radiata (Monocalyptus), which contains the lowest foliar concentration of available

nitrogen and highest concentration of tannin-bound nitrogen. The common brushtail

possums also foraged in E. camphora and E. viminalis (Symphyomyrtus), which contain the

highest available nitrogen concentrations of all Eucalyptus species. However, possums were

found only occasionally on Acacia dealbata and A. melanoxylon that had the highest available

nitrogen concentrations of all studied species. Despite the fact that parasitic plants were of

high nutritional quality for common brushtail possums, the results of this study were

inconclusive. These inconclusive results could be attributed to a small sample size of

Exocarpos cupressiformis and a lack of certainty about tree use in the case of trees parasitized

by mistletoes. Surprisingly, there were few significant intra-specific differences stated

between trees used by common brushtail possums and the closest unused ones. In this

study, none of the measured foliar chemical attributes influenced a home range size while

tree diameter and available nitrogen explained 60.2% of common brushtail possum tree

presences and 61.2% of absences.

181

The nutritional value of different tree species and parasitic plants available to common brushtail possums

To understand the nature of chemically complex “foodscapes” and their ecological effects

on common brushtail possums, it is important to ascertain the nutritional quality of forage

available to herbivores within their home ranges and to describe any between- and within-

species variability affecting animal foraging choices. In this study, the chemical analysis of

the nutritional quality of forage for common brushtail possums revealed that, despite being

subjected to similar environmental conditions, co-occurring genera and species exhibited

broad differences in foliar concentrations of total and available nitrogen, reflecting their

diverse life-history strategies (modes of nitrogen acquisition and partitioning), plant

genotype differences, and environmental variation affecting chemotypes.

For instance, acacias are known to reach the highest concentrations of foliar total nitrogen

through the symbiotic relations with root prokaryotic bacteria (Burdon et al., 1999). Yet,

acacias are considered to be of low nutritional quality for herbivores due to presumed high

concentrations of tannins (Dynes & Schlink, 2002). Nevertheless, in the present study

Acacia species had similar concentrations of tannin-bound nitrogen (c.a. 20%) as parasitic

plants and eucalypts from the subgenus Symphyomyrtus. It is important to stress here that

crude measurements of polyethylene glycol (PEG) tannin binding capacity in this study

ignore the structural complexity of different groups of tannins and their diverse effects on

herbivores.

In a more detailed study, Cork and Pahl (1984) investigated nutritional quality, including

total nitrogen, total tannins, and condensed tannins, of Acacia melanoxylon and different

Monocalyptus and Symphyomyrtus species for common ringtail possums. This study showed

that A. melanoxylon had the highest total nitrogen concentration, but similar concentration

of total tannins as Monocalyptus species, and a lower concentration of condensed tannins

than both Monocalyptus and Symphyomyrtus species. On the other side, the rarity of Australian

herbivores found feeding on Acacia foliage might be attributed to the presence of plant

secondary metabolites, such as alkaloids, cyanogenic glycosides, flavonoids and small

amounts of terpenoids (Dynes & Schlink, 2002). By contrast, Eucalyptus species that

represent a large portion of the diet of arboreal marsupials are heavily defended by several

groups of plant secondary metabolites, including terpenes, cyanogenic glycosides, and

formylated phloroglucinol compounds (FPCs; Moore et al., 2004b).

182

Between species differences in foliar concentrations of plant secondary metabolites further

complicate the situation and require a holistic metabolomic approach to identify chemical

traits that consistently vary between Acacia and Eucalyptus subgenera as suggested by Tucker

and colleagues (2010).

In the present study, foliar total nitrogen concentrations varied among Eucalyptus subgenera

from moderate to low, with Symphyomyrtus species having slightly higher concentrations of

total nitrogen than Monocalyptus species, with a single Symphyomyrtus species Eucalyptus

camphora having the highest concentration of total nitrogen of all eucalypts. Conversely,

distinct differences were observed in the foliar concentrations of available nitrogen with

Symphyomyrtus species containing more than half the available nitrogen of that of subgenus

Monocalyptus. Wallis, Nicolle and Foley (2010) found that 31 species of Monocalyptus

contained slightly less total nitrogen than 82 Symphyomyrtus species (1.00 vs. 1.12% DM) and

less than half of the available nitrogen (0.27 vs. 0.59% DM).

Ecological differences between subgenera Monocalyptus and Symphyomyrtus have been known

since the 1980s, with the tannin-rich Monocalyptus subgenera recognised to carry a lower

diversity of herbivores and suffer less leaf loss and damage than Symphyomyrtus (Noble,

1989). For instance, Youngentob et al. (2010) demonstrated that for a eucalypt specialist,

the greater glider (Petauroides volans), lower concentrations of available nitrogen in E. radiata

compared to E. viminalis explained the tree use of Monocalyptus species. By contrast, the

concentration of formylated phloroglucinol compounds, syderoxylonals, was more

important than available nitrogen in the case Symphyomyrtus species.

In the present study, the parasitic plants, Amyema pendula and Exocarpos cupressiformis, were

expected to have moderate to slightly higher foliar concentrations of total nitrogen than

their tree hosts due to mineral absorption and accumulation, hence they would experience

higher levels of herbivory (Bannister, 1989; Canyon & Hill, 1997). However, this study,

contrary to other studies (Bannister, Strong, & Andrew, 2002; March & Watson, 2010),

showed that A. pendula contained lower concentrations of total nitrogen than most of its

tree hosts. Additionally, mistletoe foliar concentrations of total nitrogen varied across

different species of hosts, being highest in mistletoes parasitizing acacias with hardly any

differences observed in those parasitizing eucalypt species. By contrast, the root parasite

Exocarpos cupressiformis had significantly higher concentration of total nitrogen than A.

pendula possibly due to its tendency to selectively parasitize nitrogen-fixing hosts or acquire

nitrogen independently from the soil (Sinclair, 2006).

183

Furthermore, this study has shown that foliar concentrations of tannin-bound nitrogen of

parasitic plants were similar to those observed for both Acacia and Eucalyptus host species,

suggesting that some other factors may affect the dietary choices of herbivores. Mistletoes

are known to accumulate large amounts of potassium and phosphorus, macronutrients

important in herbivore diets, but at the same time sequester plant secondary metabolites

from their tree hosts, such as alkaloids and cyanogenic glycosides (Martin Cordero et al.,

1993; Schulze, Turner, & Glatzel, 1984). Conversely, the secondary chemistry of the root

parasite, Exocarpos cupressiformis, is poorly researched. Only recently, the presence of

exocarpic acid was discovered in a tropical rainforest species, Exocarpos latifolius. This acid is

known to exhibit antibacterial properties but has unknown effects on herbivores (Koch et

al., 2009).

Differences in dry matter digestibility (% DMD) revealed one more level of chemical

variability between studied species and genera, with Amyema pendula being the most

digestible, and fibrous species from Monocalyptus being the least. Higher dry matter

digestibility observed in Symphyomyrtus than Monocalyptus species was also confirmed by

other studies (Wallis et al., 2010; Youngentob et al., 2011). In general, high fibre content

was suggested as the main cause of the low digestibility of Eucalyptus species as microscopic

observations of plant fragments in the caecum and faeces of common brushtail possums

revealed few bacteria attached to fibrous tissues (Foley & Hume, 1986). Furthermore,

Acacia species were found to be moderately digestible with phyllodinous Acacia melanoxylon

being higher in fibre and less digestible than the bipinnate leaves of Acacia dealbata.

Common brushtail possums are thought to select Acacia leaves based on their nutrient and

fibre content since they have a relatively small gut capacity (Crowe & Hume, 1997). Thus,

enzymatic digestion in the small intestine is relatively more important, and microbial

fermentation in the hindgut is less so, hence the possum’s strong selection against dietary

fibre.

In conclusion, there is no currently accepted uniform cross-taxa definition of the

nutritional quality of forage for herbivores, making it hard to generalise which genera and

species are more suitable and readily consumed by herbivores. The measured effects of

tannins and fibre on nitrogen availability can explain nutritional differences between

Eucalyptus genera. However, total nitrogen is a better nutritional measure for Acacia and

parasitic plant species.

184

Similarly, if we identify genera and species with high foliar concentrations of total and

available nitrogen that herbivores avoid eating, we might wish to explore other reasons for

this behaviour, such as the presence and high concentrations of different plant secondary

metabolites. Hence, the next section will focus on animal foraging choices in a chemically

complex environment, hopefully overcoming our inherently biased perception of what

represents “good food” for herbivores.

Common brushtail possum foraging choices in a chemically complex environment

Unlike the findings of other similar studies, the present study demonstrated that common

brushtail possums used trees from both Eucalyptus subgenera, Monocalyptus and

Symphyomyrtus (see Chapters 2 and 3 for details). A currently accepted view is that common

brushtail possums rarely feed on Monocalyptus species in order to avoid the high costs of

tannins on their nitrogen metabolism and instead prefer Symphyomyrtus species in mixed

eucalypt stands (Marsh et al., 2003; Moore et al., 2004). However, in this study the most

often used eucalypts, Eucalyptus radiata (Monocalyptus) and E. camphora (Symphyomyrtus), were

diametrically different in their nutritional quality. Eucalyptus radiata had the highest foliar

concentration of tannin-bound nitrogen and consequently the lowest available nitrogen

from all used species while Eucalyptus camphora had the highest total and available nitrogen

from all eucalypts.

The fact that common brushtail possums were most often found foraging on tannin-rich

Eucalyptus radiata foliage was unexpected as possums are believed to be sensitive to tannins

due to their distinctively astringent taste and nauseating effect that cause animals to cease

feeding (Marsh et al., 2003). To counteract the negative effects of tannins, animals secrete

proline-rich salivary proteins that bind tannins; yet, McArthur and colleagues (1995) noted

low rates of protein secretion in common brushtail possums. Other detoxification

strategies can be employed by animals, such as associations with gut microbes that

dissociate tannin-protein complex or secrete tannase (DeGabriel et al., 2009a, 2009b).

However, in this study the Symphyomyrtus species, despite having low tannin concentrations

were found to be eaten less often by possums, likely due to presence of other potent

herbivore deterrents, formylated phloroglucinol compounds (FPCs, Eschler et al., 2000).

185

Formylated phloroglucinol compounds are known to damage enterochromaffin cells in the

gut of herbivores, releasing serotonin that is subsequently detected by the emetic system

producing the sensation of nausea (Lawler et al., 1998). Research with captive koalas,

common brushtail possums, and ringtail possums has shown that herbivores regulated their

intake of formylated phloroglucinol compounds when they were isolated from eucalypts

and added to artificial diets (Lawler et al., 1998; Wallis, Watson, & Foley, 2002). Scrivener

and colleagues (2004) demonstrated that wild common brushtail possums largely avoided

trees with high concentrations of formylated phloroglucinol compounds while their use of

trees with low concentrations of these compounds was highly variable suggesting the

importance of tannins and other plant secondary metabolites. Recently a new group of

chemicals has been discovered in Monocalyptus, free flavanones with no substitution in Ring

B; yet, their effects on herbivores remain to be verified (Tucker et al., 2010).

Regardless of the nature of the chemical differences between Eucalyptus subgenera, this

study has demonstrated that common brushtail possums have a much broader diet than

previously assumed and consume both Symphyomyrtus and Monocalyptus species depending

on their availability (see Chapters 2 and 3 for details). Similar patterns were observed in

the study of foraging choices of co-occurring greater gliders in a habitat dominated by

Eucalyptus radiata (Monocalyptus) and E. viminalis (Symphyomyrtus) by Youngentob et al. (2011).

The authors explained this phenomenon as the result of animals using different metabolic

detoxification pathways for different groups of plant secondary metabolites. For instance,

gliders will eat foliage from Symphyomyrtus species until the physiological effects of

formylated phloroglucinol compounds force them to stop; then they switch to Monocalyptus

while detoxification processes remove formylated phloroglucinol compounds. Next, they

would revert to eating Symphyomyrtus foliage before ingesting debilitating amounts of

tannins from Monocalyptus species.

Furthermore, intra-specific differences in the concentrations of nutrients and chemical

defences have been recognised for a long time as drivers of dietary choices of arboreal

marsupials (Lawler, Foley, & Eschler, 2000; Lawler et al., 1998). Genetically and

environmentally driven intra-specific chemical differences in foliar concentrations of

available nitrogen, tannins, and formylated phloroglucinol compounds among Eucalyptus

species have been postulated to influence tree selectivity and foliage intake of common

brushtail possums and koalas (Andrew et al., 2010; O'Reilly-Wapstra, McArthur, & Potts,

2002; O'Reilly-Wapstra et al., 2005).

186

However, in this study no significant differences were observed between used trees and the

closest unused trees of the same species in terms of all measured tree variables, apart from

a few exceptions. The lack of difference between the trees may be explained by physical

barriers in pollen dispersal or by habitat uniformity or autocorrelation processes causing

genetically similar trees to occur in close proximity to each other (Andrew et al., 2007). It is

important to stress here that a low sample size could have affected the results, creating a

large number of false negatives or trees used by possums that were not recorded in this

study. Nevertheless, across possum home ranges, few significant intra-specific differences

were discerned. This suggests that intra-specific differences might occur on a broader scale

across landscapes and regions, exerting adverse effects on different herbivore populations.

Recently, the Geometric Framework for nutrition has been proposed by Raubenheimer,

Simpson and Mayntz (2009) that enables a better understanding of animal nutritional

strategies and interpretation of food choices that are otherwise difficult to explain. The

focus of this framework is on observing free food choices of individual animals when faced

with diverse dietary options and relating these choices to several nutritionally relevant

measures within simple geometrical models. The nutritional measures can include the

following: the optimal balance and amount of nutrients required to be ingested and

allocated to growth over a given time period (the intake and growth targets), the animal’s

current state in relation to these requirements, available foods and the consequences for the

animal upon ingesting them, and the amount of nutrients and plant secondary metabolites

that are retained and detoxified by the animal. Future studies on nutritional requirements of

arboreal marsupials should consider incorporating the Geometric Framework into their

research design.

187

References

Agrawal, A. A. (2001). Phenotypic plasticity in the interactions and evolution of species. Science, 294(5541), 321–326.

Agrawal, A. A. (2011). Current trends in the evolutionary ecology of plant defence. Functional Ecology, 25(2), 420–432.

Agrawal, A. A., & Fishbein, M. (2006). Plant defense syndromes. Ecology, 87(7), 132–149. Andrew, R. L., Peakall, R., Wallis, I. R., & Foley, W. J. (2007). Spatial distribution of

defense chemicals and markers and the maintenance of chemical variation. Ecology, 88(3), 716–728.

Andrew, R. L., Wallis, I. R., Harwood, C. E., & Foley, W. J. (2010). Genetic and environmental contributions to variation and population divergence in a broad-spectrum foliar defence of Eucalyptus tricarpa. Annals of Botany, 105(5), 707–717.

Bannister, P. (1989). Nitrogen concentration and mimicry in some New Zealand mistletoes. Oecologia, 79(1), 128–132.

Bannister, P., Strong, G. L., & Andrew, I. (2002). Differential accumulation of nutrient elements in some New Zealand mistletoes and their hosts. Functional Plant Biology, 29(11), 1309–1318.

Barbehenn, R. V., & Peter Constabel, C. (2011). Tannins in plant-herbivore interactions. Phytochemistry, 72(13), 1551–1565.

Batten, G. D. (1998). Plant analysis using near infrared reflectance spectroscopy: The potential and the limitations. Australian Journal of Experimental Agriculture, 38(7), 697–706.

Bell, T. L., & Adams, M. A. (2011). Attack on all fronts: Functional relationships between aerial and root parasitic plants and their woody hosts and consequences for ecosystems. Tree Physiology, 31(1), 3–15.

Bennett, R. N., & Wallsgrove, R. M. (1994). Tansley Review No. 72. Secondary metabolites in plant defence mechanisms. New Phytologist, 127(4), 617–633.

Berenbaum, M. (1980). Adaptive significance of midgut pH in larval Lepidoptera. The American Naturalist, 115(1), 138–146.

Bernays, E. A. (1981). Plant tannins and insect herbivores: An appraisal. Ecological Entomology, 6(4), 353–360.

Boland, D. J., Brophy, J. J., & House, A. P. N. (1991). Eucalyptus leaf oils: Use, chemistry, distillation and marketing. Melbourne: Inkata Press.

Braithwaite, L. W., Turner, J., & Kelly, J. (1984). Studies on the Arboreal Marsupial Fauna of Eucalypt Forests Being Harvested for Wood Pulp at Eden, N.s.w. Iii. Relationships Between Faunal Densities, Eucalypt Occurrence and Foliage Nutrients, and Soil Parent Materials. Australian Wildlife Research, 11(1), 41–48.

Burchfield, E., Agar, N. S., & Hume, I. D. (2006). Effects of terpenes and tannins on some physiological and biochemical parameters in two species of phalangerid possums (Marsupialia: Phalangeridae). Australian Journal of Zoology, 53(6), 395–402.

Burdon, J. J., Gibons, A. H., Searle, S. D., Woods, M. J., & Brockwell, J. (1999). Variation in the effectiveness of symbiotic associations between native rhizobia and temperate Australian Acacia: Within-species interactions. Journal of Applied Ecology, 36(3), 398–408.

Canyon, D. V., & Hill, C. J. (1997). Mistletoe host-resemblance: A study of herbivory, nitrogen and moisture in two Australian mistletoes and their host trees. Australian Journal of Ecology, 22(4), 395–403.

Carmona, D., Lajeunesse, M. J., & Johnson, M. T. J. (2011). Plant traits that predict resistance to herbivores. Functional Ecology, 25(2), 358–367.

188

Choo, G. M., Waterman, P. G., McKey, D. B., & Gartlan, J. S. (1981). A simple enzyme assay for dry matter digestibility and its value in studying food selection by generalist herbivores. Oecologia, 49(2), 170–178.

Clapperton, B. K., Matthews, L. R., Fawkes, M. S., & Pearson, A. J. (1996). Lithium and cyanide-induced conditioned food aversions in brushtail possums. The Journal of Wildlife Management, 60(1), 195–201.

Clausen, T. P., Provenza, F. D., Burritt, E. A., Reichardt, P. B., & Bryant, J. P. (1990). Ecological implications of condensed tannin structure: A case study. Journal of Chemical Ecology, 16(8), 2381–2392.

Close, D. C., Davidson, N. J., & Davies, N. W. (2006). Seasonal fluctuations in pigment chemistry of co-occurring plant hemi-parasites of distinct form and function. Environmental and Experimental Botany, 58(1-3), 41–46.

Cockfield, S. D. (1988). Relative availability of nitrogen in host plants of invertebrate herbivores: Three possible nutritional and physiological definitions. Oecologia, 77(1), 91–94.

Coley, P. D., & Barone, J. A. (1996). Herbivory and plant defenses in tropical forests. Annual Review of Ecology and Systematics, 27, 305–335.

Cooper, S. M., Owen-Smith, N., & Bryant, J. P. (1988). Foliage acceptability to browsing ruminants in relation to seasonal changes in the leaf chemistry of woody plants in a South African savanna. Oecologia, 75(3), 336–342.

Crowe, O., & Hume, I. D. (1997). Morphology and function of the gastrointestinal tract of Australian folivorous possums. Australian Journal of Zoology, 45(4), 357–368.

Cyr, H., & Face, M. L. (1993). Magnitude and patterns of herbivory in aquatic and terrestrial ecosystems. Nature, 361(6408), 148–150.

Dearing, M. D., Foley, W. J., & McLean, S. (2005). The influence of plant secondary metabolites on the nutritional ecology of herbivorous terrestrial vertebrates. Annual Review of Ecology, Evolution, and Systematics, 36, 169–189.

DeGabriel, J. L. (2008). Demographic constraints imposed by plant nutrients and secondary chemistry on the common brushtail possum (Trichosurus vulpecula Kerr.). (Ph. D. thesis), The Australian National University, Canberra.

DeGabriel, J., Wallis, I., Moore, B., & Foley, W. (2008). A simple, integrative assay to quantify nutritional quality of browses for herbivores. Oecologia, 156(1), 107–116.

DeGabriel, J. L., Moore, B. D., Foley, W. J., & Johnson, C. N. (2009a). The effects of plant defensive chemistry on nutrient availability predict reproductive success in a mammal. Ecology, 90(3), 711–719.

DeGabriel, J., Moore, B. D., Shipley, L., Krockenberger, A. K., Wallis, I. R., Johnson, C. N., & Foley, W. J. (2009b). Inter-population differences in the tolerance of a marsupial folivore to plant secondary metabolites. Oecologia, 161(3), 539–548.

DeGabriel, J., Moore, B. D., Marsh, K. J., & Foley, W. J. (2010). The effect of plant secondary metabolites on the interplay between the internal and external environments of marsupial folivores. Chemoecology, 20(2), 97–108.

Di Stefano, J., & Newell, G. R. (2008). Diet selection by the swamp wallaby (Wallabia bicolor): Feeding strategies under conditions of changed food availability. Journal of Mammalogy, 89(6), 1540–1549.

Dynes, R. A., & Schlink, A. C. (2002). Livestock potential of Australian species of Acacia. Conservation Science Western Australia, 4(3), 117–124.

Ebbers, M. J. H., Wallis, I. R., Dury, S., Floyd, R., & Foley, W. J. (2002). Spectrometric prediction of secondary metabolites and nitrogen in fresh Eucalyptus foliage: Towards remote sensing of the nutritional quality of foliage for leaf-eating marsupials. Australian Journal of Botany, 50(6), 761–768.

Eschler, B. M., Pass, D. M., Willis, R., & Foley, W. J. (2000). Distribution of foliar formylated phloroglucinol derivatives amongst Eucalyptus species. Biochemical Systematics and Ecology, 28(9), 813–824.

189

Felton, A. M., Felton, A., Lindenmayer, D. B., & Foley, W. J. (2009a). Nutritional goals of wild primates. Functional Ecology, 23(1), 70–78.

Felton, A. M., Felton, A., Raubenheimer, D., Simpson, S. J., Foley, W. J., Wood, J. T., ... & Lindenmayer, D. B. (2009b). Protein content of diets dictates the daily energy intake of a free-ranging primate. Behavioral Ecology, 20(4), 685–690.

Foley, W. J., & Hume, I. D. (1987). Digestion and metabolism of high-tannin Eucalyptus foliage by the brushtail possum (Trichosurus vulpecula)(Marsupialia: Phalangeridae). Journal of Comparative Physiology B, 157(1), 67–76.

Foley, W. J., McIlwee, A., Lawler, I. R., Aragones, L., Woolnough, A. P., & Berding, N. (1998). Ecological applications of near infrared reflectance spectroscopy—a tool for rapid, cost-effective prediction of the composition of plant and animal tissues and aspects of animal performance. Oecologia, 116(3), 293–305.

Forbey, J. S., Dearing, M. D., Gross, E. M., Orians, C. M., Sotka, E. E., & Foley, W. J. (2013). A pharm-ecological perspective of terrestrial and aquatic plant–herbivore interactions. Journal of Chemical Ecology, 39(4), 465–480.

Fox, L. R., & Macauley, B. J. (1977). Insect grazing on Eucalyptus in response to variation in leaf tannins and nitrogen. Oecologia, 29(2), 145–162.

Freeland, W. J., & Janzen, D. H. (1974). Strategies in herbivory by mammals: The role of plant secondary compounds. The American Naturalist, 108(961), 269–287.

Glatzel, G., & Geils, B. W. (2009). Mistletoe ecophysiology: Host–parasite interactions. Botany, 87(1), 10–15.

Gleadow, R. M., Haburjak, J., Dunn, J. E., Conn, M. E., & Conn, E. E. (2008). Frequency and distribution of cyanogenic glycosides in Eucalyptus L'Hérit. Phytochemistry, 69(9), 1870–1874.

Hansen, A. P., Pate, J. S., & Atkins, C. A. (1987). Relationships between acetylene reduction activity, hydrogen evolution and nitrogen fixation in nodules of Acacia spp.: Experimental background to assaying fixation by acetylene reduction under field conditions. Journal of Experimental Botany, 38(1), 1–12.

Hensens, O. D., Lewis, K. G., & Mulquiney, C. E. (1971). Some extractives from mistletoe. Australian Journal of Chemistry, 24(2), 431–434.

Iason, G., & Villalba, J. (2006). Behavioral strategies of mammal herbivores against plant secondary metabolites: The avoidance-tolerance continuum. Journal of Chemical Ecology, 32(6), 1115–1132.

Irlbeck, N. A., & Hume, I. D. (2003). The role of Acacia in the diets of Australian marsupials ? A review. Australian Mammalogy, 25(2), 121–134.

Janzen, D. H., (1977). Promising Directions of Study in Tropical Animal–Plant Interactions. Annals of the Missouri Botanical Garden, 64(4), 706-736.

Joern, A., & Behmer, S. T. (1997). Importance of dietary nitrogen and carbohydrates to survival, growth, and reproduction in adults of the grasshopper Ageneotettix deorum (Orthoptera: Acrididae). Oecologia, 112(2), 201–208.

Kavanagh, R. P., & Lambert, M. J. (1990). Food selection by the greater glider, Petauroides volans: Is foliar nitrogen a determinant of habitat quality? Wildlife Research, 17(3), 285–299.

Keith, H., Barrett, D., & Keenan, R. (2000). Review of allometric relationships for estimating woody biomass for New South Wales, the Australian Capital Territory, Victoria, Tasmania, and South Australia National Carbon Accounting System Technical Report 5B. Canberra: Australian Greenhouse Office.

Koch, M., Bugni, T. S., Pond, C. D., Sondossi, M., Dindi, M., Piskaut, P., Barrows, L. R. (2009). Antimycobacterial activity of Exocarpos latifolius is due to exocarpic acid. Planta Med, 75(12), 1326–1330.

Lawler, I. R., Foley, W. J., & Eschler, B. M. (2000). Foliar concentration of a single toxin creates habitat patchiness for a marsupial folivore. Ecology, 81(5), 1327-1338.

190

Lawler, I. R., Foley, W. J., Eschler, B. M., Pass, D. M., & Handasyde, K. (1998). Intraspecific variation in Eucalyptus secondary metabolites determines food intake by folivorous marsupials. Oecologia, 116(1), 160–169.

Lawler, I. R., Foley, W. J., Pass, G. J., & Eschler, B. M. (1998). Administration of a 5HT3 receptor antagonist increases the intake of diets containing Eucalyptus secondary metabolites by marsupials. Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology, 168(8), 611–618.

Lawler, I. R., Stapley, J., Foley, W. J., & Eschler, B. M. (1999). Ecological example of conditioned flavor aversion in plant-herbivore interactions: Effect of terpenes of Eucalyptus leaves on feeding by common ringtail and brushtail possums. Journal of Chemical Ecology, 25(2), 401–415.

Loney, P. E., McArthur, C., Potts, B. M., & Jordan, G. J. (2006). How does ontogeny in a Eucalyptus species affect patterns of herbivory by Brushtail Possums? Functional Ecology, 20(6), 982–988.

Lucas, P. W., Turner, I. M., Dominy, N. J., & Yamashita, N. (2000). Mechanical defences to herbivory. Annals of Botany, 86(5), 913–920.

March, W. A., & Watson, D. M. (2010). The contribution of mistletoes to nutrient returns: Evidence for a critical role in nutrient cycling. Austral Ecology, 35(7), 713–721.

Marsh, K. J., Foley, W. J., Cowling, A., & Wallis, I. R. (2003). Differential susceptibility to Eucalyptus secondary compounds explains feeding by the common ringtail (Pseudocheirus peregrinus) and common brushtail possum (Trichosurus vulpecula). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology, 173(1), 69–78.

Marsh, K. J., Wallis, I. R., & Foley, W. J. (2003). The effect of inactivating tannins on the intake of Eucalyptus foliage by a specialist Eucalyptus folivore (Pseudocheirus peregrinus) and a generalist herbivore (Trichosurus vulpecula). Australian Journal of Zoology, 51(1), 31–42.

Marsh, K. J., Wallis, I. R., Andrew, R. L., & Foley, W. J. (2006a). The detoxification limitation hypothesis: Where did it come from and where is it going? Journal of Chemical Ecology, 32(6), 1247–1266.

Marsh, K. J., Wallis, I. R., McLean, S., Sorensen, J. S., & Foley, W. J. (2006b). Conflicting demands on detoxification pathways influence how common brushtail possums choose their diets. Ecology, 87(8), 2103–2112.

Martin Cordero, C., Gil Serrano, A. M., & Ayuso Gonzalez, M. J. (1993). Transfer of bipiperidyl and quinolizidine alkaloids to Viscum cruciatum Sieber (Loranthaceae) hemiparasitic on Retama sphaerocarpa boissier (Leguminosae). Journal of Chemical Ecology, 19(10), 2389–2393.

Martin, R. W., & Handasyde, K. A. (1999). The Koala: Natural history, Conservation and Management. Sydney: University of New South Wales.

Mattson, W. J. (1980). Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics, 11(1), 119–161.

McArt, S. H., Spalinger, D. E., Collins, W. B., Schoen, E. R., Stevenson, T., & Bucho, M. (2009). Summer dietary nitrogen availability as a potential bottom-up constraint on moose in south-central Alaska. Ecology, 90(5), 1400–1411.

McArthur, C., Hagerman, A., & Robbins, C. T. (1991). Physiological strategies of mammalian herbivores against plant defenses. In R. T. Palo & C. T. Robbins (Eds.), Plant Defenses against Mammalian Herbivores (pp. 103–114). Boca Raton: CRC Press.

McArthur, C., Sanson, G. D., & Beal, A. M. (1995). Salivary proline-rich proteins in mammals: Roles in oral homeostasis and counteracting dietary tannin. Journal of Chemical Ecology, 21(6), 663–691.

McLean, S., & Foley, W. J. (1997). Metabolism of Eucalyptus terpenes by herbivorous marsupials. Drug Metabolism Reviews, 29(1–2), 213–218.

191

McLean, S., Brandon, S., Boyle, R. R., & Wiggins, N. L. (2008). Development of tolerance to the dietary plant secondary metabolite 1, 8-cineole by the brushtail possum (Trichosurus vulpecula). Journal of Chemical Ecology, 34(5), 672–680.

McLeod, M. N. (1973). The digestibility and the nitrogen, phosphorus and ash contents of the leaves of some Australian trees and shrubs. Australian Journal of Experimental Agriculture, 13(62), 245–250.

McNaughton, S. J. (1988). Mineral nutrition and spatial concentrations of African ungulates. Nature, 334, 343–345.

Mole, S., & Waterman, P. G. (1987). Tannins as Antifeedants to Mammalian Herbivores—Still an Open Question? In G. R. Waller (Ed.), Allelochemicals: Role in Agriculture and Forestry (Vol. 330, pp. 572–587): American Chemical Society.

Molyneux, R. J., & Ralphs, M. H. (1992). Plant toxins and palatability to herbivores. Journal of Range Management, 45(1), 13–18.

Moore, B. D., Wallis, I. R., Marsh, K. J., & Foley, W. J. (2004a). The role of nutrition in the conservation of the marsupial folivores of eucalypt forests. In D. Lunney (Ed.), Conservation of Australia’s Forest Fauna (2 ed., pp. 549–575). Mosman: Royal Zoological Society of New South Wales.

Moore, B. D., Wallis, I. R., Palá-Paúl, J., Brophy, J. J., Willis, R. H., & Foley, W. J. (2004b). Antiherbivore chemistry of Eucalyptus—cues and deterrents for marsupial folivores. Journal of Chemical Ecology, 30(9), 1743–1769.

Moore, B. D., & Foley, W. J. (2005). Tree use by koalas in a chemically complex landscape. Nature, 435(7041), 488–490.

Mueller-Harvey, I. (2006). Unravelling the conundrum of tannins in animal nutrition and health. Journal of the Science of Food and Agriculture, 86(13), 2010–2037.

Nelder, J. A., & Wedderburn, R. W. M. (1972). Generalized linear models. Journal of the Royal Statistical Society. Series A (General), 135(3), 370–384.

Nersesian, C. L., Banks, P. B., Simpson, S. J., & McArthur, C. (2012). Mixing nutrients mitigates the intake constraints of a plant toxin in a generalist herbivore. Behavioral Ecology, 23(4), 879–888.

Noble, I. R. (1989). Ecological traits of the Eucalyptus subgenera Monocalyptus and Symphyomyrtus. Australian Journal of Botany, 37(3), 207–224.

O'Reilly-Wapstra, J., McArthur, C., & Potts, B. M. (2002). Genetic variation in resistance of Eucalyptus globulus to marsupial browsers. Oecologia, 130(2), 289–296.

O'Reilly-Wapstra, J. M., Potts, B. M., McArthur, C., & Davies, N. W. (2005). Effects of nutrient variability on the genetic-based resistance of Eucalyptus globulus to a mammalian herbivore and on plant defensive chemistry. Oecologia, 142(4), 597–605.

Press, M. C., & Phoenix, G. K. (2005). Impacts of parasitic plants on natural communities. New Phytologist, 166(3), 737–751.

Preston, A. L., An, M., & Watson, D. M. (2010). Chemical profile differences in endemic parasitic weeds: A study of host-parasite chemical profiles in select mistletoe and eucalypt species. Paper presented at the Seventeenth Australasian Weeds Conference, Christchurch, New Zealand.

Raubenheimer, D., Simpson, S. J., & Mayntz, D. (2009). Nutrition, ecology and nutritional ecology: Toward an integrated framework. Functional Ecology, 23(1), 4–16.

Reed, J. D. (1995). Nutritional toxicology of tannins and related polyphenols in forage legumes. Journal of Animal Science, 73(5), 1516–1528.

Rosenthal, G. A., & Berenbaum, M. R. (1992). Herbivores: Their interactions with secondary plant metabolites: Ecological and Evolutionary Processes (Vol. 2). San Diego: Academic Press.

Salminen, J. P., & Karonen, M. (2011). Chemical ecology of tannins and other phenolics: We need a change in approach. Functional Ecology, 25(2), 325–338.

Schultz, J. C. (1989). Tannin—insect interactions. In R. W. Hemingway & J. J. Karchesy (Eds.), Chemistry and Significance of Condensed Tannins (pp. 417–433). New York: Plenum Press.

192

Schulze, E. D., Turner, N. C., & Glatzel, G. (1984). Carbon, water and nutrient relations of two mistletoes and their hosts: A hypothesis. Plant, Cell & Environment, 7(5), 293–299.

Scriber, J. M. (1977). Limiting effects of low leaf-water content on the nitrogen utilization, energy budget, and larval growth of Hyalophora cecropia (Lepidoptera: Saturniidae). Oecologia, 28(3), 269–287.

Scrivener, N. J., Johnson, C. N., Wallis, I. R., Takasaki, M., Foley, W. J., & Krockenberger, A. K. (2004). Which trees do wild common brushtail possums (Trichosurus vulpecula) prefer? Problems and solutions in scaling laboratory findings to diet selection in the field. Evolutionary Ecology Research, 6(1), 77–87.

Shenk, J. S., & Westerhaus, M. O. (1991). Population definition, sample selection, and calibration procedures for near infrared reflectance spectroscopy. Crop Science, 31(2), 469–474.

Shimada, T., & Saitoh, T. (2003). Negative effects of acorns on the wood mouse Apodemus speciosus. Population Ecology, 45(1), 7–17.

Silanikove, N., Gilboa, N., Nir, I., Perevolotsky, A., & Nitsan, Z. (1996). Effect of a daily supplementation of polyethylene glycol on intake and digestion of tannin-containing leaves (Quercus calliprinos, Pistacia lentiscus, and Ceratonia siliqua) by goats. Journal of Agricultural and Food Chemistry, 44(1), 199–205.

Simpson, S. J., & Raubenheimer, D. (2012). The nature of nutrition: A unifying framework. Australian Journal of Zoology, 59(6), 350–368.

Sinclair, S. J. (2006). The influence of dwarf cherry (Exocarpos strictus) on the health of river red gum (Eucalyptus camaldulensis). Australian Forestry, 69(2), 137–141.

Snyder, M. A., Fineschi, B., Linhart, Y. B., & Smith, R. H. (1996). Multivariate discrimination of host use by dwarf mistletoe Arceuthobium vaginatum subsp. cryptopodum: Inter- and intra-specific comparisons. Journal of Chemical Ecology, 22(2), 295–305.

Stalenberg, E. (2010). Spatial variation in habitat quality effects fine-scale resource use by a low density koala population. (Honours thesis), the Australian National University, Canberra.

Spotfire S+® Version 8.1. (2010). TIBCO Software Inc., Somerville. Tabachnick, B. G., & Fidell, L. S. (1996). Multivariate statistics. New York: Harper Collins

College Publishing. Takemoto, H. (2003). Phytochemical determination for leaf food choice by wild

chimpanzees in Guinea, Bossou. Journal of Chemical Ecology, 29(11), 2551–2573. Tixier, H., Duncan, P., Scehovic, J., Yant, A., Gleizes, M., & Lila, M. (1997). Food selection

by European roe deer (Capreolus capreolus): Effects of plant chemistry, and consequences for the nutritional value of their diets. Journal of Zoology, 242(2), 229–245.

Tucker, D., Wallis, I. R., Bolton, J., Marsh, K. J., Rosser, A., Brereton, I.,Foley, W. J. (2010). A metabolomic approach to identifying chemical mediators of mammal–plant interactions. Journal of Chemical Ecology, 36(7), 727–735.

Umucalilar, H., Gulsen, N., Coskun, B., Hayirli, A., & Dural, H. (2007). Nutrient composition of mistletoe (Viscum album) and its nutritive value for ruminant animals. Agroforestry Systems, 71(2), 77–87.

Venables, W. N., & Ripley, B. D. (2002). Modern Applied Statistics with S (4 ed.). New York: Springer.

Wallis, I. R., Watson, M. L., & Foley, W. J. (2002). Secondary metabolites in Eucalyptus melliodora: Field distribution and laboratory feeding choices by a generalist herbivore, the common brushtail possum. Australian Journal of Zoology, 50(5), 507–519.

Wallis, I. R., Nicolle, D., & Foley, W. J. (2010). Available and not total nitrogen in leaves explains key chemical differences between the eucalypt subgenera. Forest Ecology and Management, 260(5), 814–821.

193

Wallis, I. R., Edwards, M. J., Windley, H., Krockenberger, A. K., Felton, A., Quenzer, M., ... & Foley, W. J. (2012). Food for folivores: Nutritional explanations linking diets to population density. Oecologia, 169(2), 281–291.

Watson, D. M. (2001). Mistletoe—a keystone resource in forests and woodlands worldwide. Annual Review of Ecology and Systematics, 32(1), 219–249.

Watson, D. M. (2009). Determinants of parasitic plant distribution: The role of host quality. Botany, 87(1), 16–21.

White, T. C. R. (1984). The abundance of invertebrate herbivores in relation to the availability of nitrogen in stressed food plants. Oecologia, 63(1), 90–105.

Wiggins, N., McArthur, C., & Davies, N. W. (2006). Diet switching in a generalist mammalian folivore: Fundamental to maximising intake. Oecologia, 147(4), 650–657.

Woolnough, A. P., & Foley, W. J. (2002). Rapid evaluation of pasture quality for a critically endangered mammal, the northern hairy-nosed wombat (Lasiorhinus krefftii). Wildlife Research, 29(1), 91–100.

Youngentob, K., Wallis, I. R., Lindenmayer, D. B., Wood, J., Pope, M., & Foley, W. J. (2011). Foliage chemistry influences tree choice and landscape use of a gliding marsupial folivore. Journal of Chemical Ecology, 37(1), 71–84.

Zweifel‐Schielly, B., Leuenberger, Y., Kreuzer, M., & Suter, W. (2011). A herbivore's food landscape: Seasonal dynamics and nutritional implications of diet selection by a red deer population in contrasting Alpine habitats. Journal of Zoology, 286(1), 68–80.

194

CH

AP

TE

R

5

Herbivory in the heterogeneous world

195

CHAPTER 5

Herbivory in the heterogeneous world

Synthesis of results and conclusions

Herbivores are known to make decisions and alter their movements and foraging

behaviour at different scales of resource heterogeneity, ranging from geographic regions

and landscapes to plant communities, species, individuals of the same species, and even

plant parts. To understand the ecology of herbivores in heterogeneous environments it is

important to ascertain the critical scales at which the processes involved in resource

selection occur. Johnson (1980) identified four orders of resource selection by animals:

the geographic range of a species, the home range of an individual, the various habitats

used by an individual within its home range, and the individual resources selected within

each habitat. Hence, resource selection of herbivores is influenced by broad scale

processes, such as regional climate and soil types; landscape-scale processes, such as the

amount and configuration of habitat; and fine-scale processes, such as the chemistry of

individual plants or plant parts (Moore & DeGabriel, 2012; Senft et al., 1987).

The present work has focused on exploring underlying patterns of resource selection

(i.e., food and shelter) and space use-patterns of the common brushtail possum (Trichosurus

vulpecula, Marsupialia: Phalangeridae). This generalist arboreal herbivore has been observed

to respond to multiple scales of resource heterogeneity (i.e., regional, landscape, and

habitat), and plant internal heterogeneity or chemical variability of different tree species,

individuals, and tree parts (DeGabriel, Moore, Marsh, & Foley, 2010). Hence, this system

provides a unique opportunity to study complex plant–herbivore interactions in a species

rich and chemically heterogeneous environment and to relate these interactions to other

spatial ecological processes, such as competition over resources (i.e., food and shelter) and

a potential risk of predation. In the present study, additive and interacting effects of

environmental heterogeneity on resource selection by common brushtail possums were

studied at three spatial scales: the landscape, home range, and tree scale.

196

In Chapter 2, resource selection by the entire population of common brushtail

possums (Trichosurus vulpecula) and indirectly measured competition with a sympatric

species, the mountain brushtail possum (T. cunninghami), were studied at a

landscape-scale.

Next, in Chapter 3, individual variability in resource use (i.e., food and shelter) and

space-use patterns of selected common brushtail possums were studied at a home

range scale.

Finally, in Chapter 4, food selection by common brushtail possums in response to

the nutritional quality of forage was explored at a tree-scale.

By studying resource selection and space-use patterns of common brushtail possums at

different scales, landscape, home range, and individual tree, the opportunity arose to

compare and contrast findings across different ecological metrics and answer more general

questions about herbivory: What is the realized diet breadth of generalist herbivores and

are they facultative specialists in a heterogeneous world of many choices? Do individual

differences in resource use, such as selective foraging on different species of plants and

individuals of the same species, lead to the expression of different animal personalities?

Is using a single currency to measure the nutritional value of forage for herbivores

obscuring the complex effects of multiple nutrients and toxins?

In Chapter 2 resource selectivity by common brushtail possums was examined in the

context of food availability (species composition of trees and parasitic plants), hollow

availability (used as shelter), as well as potential competition with a sympatric species, the

mountain brushtail possum. The main aim of this study was to document the diet breadth

and food preferences of common brushtail possums and to investigate underlying factors

influencing their tree use at a broad landscape-scale. As demonstrated in other similar

studies (Freeland & Winter, 1975; Marsh, Wallis, McLean, Sorensen, & Foley, 2006;

Nersesian, Banks, Simpson, & McArthur, 2012), common brushtail possums behaved as

true generalist herbivores consuming a wide variety of foods apart from Eucalyptus leaves,

including leaves of different species of Acacia; exotic trees; parasitic plants, Amyema pendula

and Exocarpos cupressiformis; as well as understorey species and fruits of Rubus fruticosus.

197

However, despite their generalist feeding strategy, common brushtail possums exhibited

clear foraging preferences, ranging across forest types and species of trees to discrete

patches of mistletoes within tree canopies. Acacia dealbata, exotic trees, and parasitic plants

were eaten more frequently than expected from their availability in the habitat while habitat

dominant Eucalyptus species were consumed less frequently than expected. Additional

factors such as the availability of hollows and competition with mountain brushtail

possums might have shaped tree use by common brushtail possums.

In Chapter 3, space-use patterns and resource selection (i.e., food and shelter) of common

brushtail possums were explored within their home ranges. As expected from previous

studies (DeGabriel, Moore, Foley, & Johnson, 2009; Martin & Martin, 2007), males had

larger home ranges than females. The total home range size of both males and females was

strongly correlated with foraging area and only weakly correlated with denning area,

suggesting that food availability might have been a more limiting factor for common

brushtail possums than availability of hollow-bearing trees. Moreover, individual animals

preferred different species of trees for foraging and denning and repeatedly used individual

trees of the same species, as well as trees parasitized by mistletoes. Common brushtail

possums differentiated between trees used as foraging and denning sites based on the

species and the number of hollows and used trees with a greater number of hollows more

frequently as denning sites. Individual animals showed preferences for different species of

trees used as foraging sites, reflecting different tree species composition or the structural

heterogeneity of home ranges leading to dietary specialization.

In Chapter 4, a single measure of forage nutritional quality, available nitrogen, was used

for cross-taxa comparisons of forage available to common brushtail possums within their

home ranges. This measure combined the foliar concentrations of total nitrogen and

digestibility-reducing tannins and fibre on the amount of nitrogen that was readily available

to herbivores. The chemical analysis of the nutritional quality of forage revealed that,

despite being subjected to similar environmental conditions, co-occurring tree genera and

species reflected their diverse life strategies by exhibiting broad differences in foliar

concentrations of available nitrogen, tannins, and fibre. Unlike previous studies (Marsh,

Foley, Cowling, & Wallis, 2003; Moore et al., 2004), the present research demonstrated that

common brushtail possums used both Monocalyptus and Symphyomyrtus species while

foraging at night.

198

The most often used species, Eucalyptus radiata (Monocalyptus) and E. camphora

(Symphyomyrtus), differed in their nutritional quality. The former species had the lowest

available nitrogen and highest tannin-bound nitrogen concentrations of all used species

while the latter had the highest available nitrogen of all analysed eucalypts. Occasionally

consumed Acacia dealbata had the highest concentration of available nitrogen while parasitic

plants which were preferred by common brushtail possums, Exocarpos cupressiformis and

Amyema pendula, had moderate to high concentrations of available nitrogen, with foliar

nitrogen concentrations of mistletoe varying depending on the species of tree host.

Contrary to expectations, hardly any intraspecific differences were observed among trees

repeatedly used by common brushtail possums, suggesting that some other unmeasured

plant chemical and physical factors combined with a potential risk of predation might have

been involved in animal foraging decisions.

In summary, the findings of this study have highlighted the previous studies’ over-

simplification of the complex food-web interaction in Australian forests and woodlands.

The previous studies focused exclusively on the availability and nutritional quality for

arboreal marsupials of one genus, Eucalyptus, and at the same time excluded other tree

genera and parasitic plants. This study highlighted the need to move on from a traditional

division of arboreal marsupials into either specialist or generalist herbivores, by showing

that generalist species, such as the common brushtail possum, can be selective at multiple

scales of resource heterogeneity, ranging across landscapes, home ranges, and individual

trees to discrete patches of mistletoes within tree canopies (Fig. 1).

199

Fig. 1 Drivers of resource selectivity by common brushtail possums at different spatial scales

200

Furthermore, based on the recognized dietary preferences of entire populations of

generalist herbivores and the specific “food tastes” of individual members of a population,

we can conceptualise generalist herbivores as a specialized group of consumers when

offered a smörgåsbord. Moreover, this study demonstrated that applying a single currency,

available nitrogen, to measure the nutritional quality of different plant species and taxa for

generalist herbivores simplifies complex interactions between multiple nutrients and toxins

that shape animal dietary choices. Only by acknowledging the complexities of generalist

herbivores’ diet balancing processes and by taking into account other confounding

processes, such as competition and predation, it is possible to understand how animals

select their diets in heterogeneous environments.

Future studies need to reconcile the five nutritional goals of herbivores: nitrogen (protein)

maximization (Mattson, 1980), energy maximization (Belovsky, 1986), avoidance and

regulation of plant secondary metabolites (Freeland & Janzen, 1974), limitation of fibre

(Demment & Soest, 1985), and nutrient balancing (Westoby, 1978). To achieve this

synthesis, I suggest structuring research around six levels of inquiry: Level (1) What food is

available within the animal’s home range? Level (2) What is the animal dietary breadth?

Level (3) Which properties of food determine inclusion in the overall diet? Level (4) Which

properties determine the preference ranking of foods that are a part of the overall diet?

Level (5) What are the nutritional goals that underlie the choice and ranking of foods?

Level (6) What are other confounding factors affecting herbivore nutrition?

The current study provides a baseline for future nutritional studies by highlighting the

specialized nature of generalized diets of herbivores expressed over multiple scales of

resource heterogeneity and encourages a more holistic approach in studies of nutritional

quality of forage for herbivores. Recently, DeGabriel and colleagues (2013) identified four

steps that need to be achieved in order to understand nutritional complexity of herbivores:

(1) knowing what foods and how much of these foods wild browsers eat, as well as what

they avoid eating; (2) knowing the relevant aspects of plant nutritional and defensive

chemistry to measure in a given system and how to measure them; (3) understanding the

spatial distribution of nutrients and plant secondary metabolites in plant communities, the

costs they impose on foraging, and the effects on animals’ distributions; and (4) having

appropriate statistical tools to analyze the data.

201

Yet, as the current study has demonstrated, identifying and quantifying what herbivores eat

and even more what they avoid eating is difficult and subjected to many biases and

technical difficulties. Namely, direct foraging observations are difficult in the case of

arboreal marsupials as discussed in Chapter 2 and 3; yet, some technological advances,

such as audio-telemetry and video-tracking are promising (Logan & Sanson, 2002).

Similarly, the fecal content analysis that underestimates participation of species with highly

digestible cuticles can be replaced by methods that use plant cuticular markers (i.e., n-

alkanes, alcohols and long-chain fatty acids coupled with carbon isotope ratios or genetic

markers to determine herbivore diets from faeces (Dove & Mayes, 2006; Bezabih et al.,

2011; Valentini et al., 2009).

Also, measuring relevant nutrients and toxins for herbivore dietary choices represents a

challenge for nutritional ecologists as discussed in Chapter 4. So far, we know more about

type and concentrations of toxins that make herbivores avoid certain foods than

macronutrients that make them consume others. The Geometric Framework (Simpson &

Raubenheimer, 2001) provides a unifying approach to explore effects of both toxins and

nutrients on animal food intake and digestion. Moreover, types and concentrations of

nutrients and toxins vary within and between species, regions, and seasons, and in order to

measure environmental chemical variability, rapid and cheap methods for analysing a large

number of samples are needed. One well-established solution is near infrared reflectance

spectroscopy (NIRS; see Chapter 4 for details), and recent advances include the

development of portable devices to record spectra in the field and the use of airborne

hyper-spectral sensing techniques to measure the spectra from whole tracts of forests

(HYMAP; Dury, Turner, & Foley, 2001). Furthermore, new statistical tools are needed to

deal with the complexity of nutritional data and the non-linear relationship between

nutrient and toxin concentrations and the food intake of herbivores as discussed by

DeGabriel and colleagues (2013). Finally, the effects of increased atmospheric CO2 on

foliar concentrations of carbon-based secondary or structural compounds may have far-

reaching consequences for herbivory and should be incorporated in future studies

exploring the foraging patterns of herbivores.

202

References

Belovsky, G. E. (1986). Optimal foraging and community structure: Implications for a guild of generalist grassland herbivores. Oecologia, 70(1), 35–52.

Bezabih, M., Pellikaan, W. F., Tolera, A., & Hendriks, W. H. (2011). Evaluation of n-alkanes and their carbon isotope enrichments (δ13C) as diet composition markers. Animal, 5(01), 57–66.

DeGabriel, J., Moore, B. D., Marsh, K. J., & Foley, W. J. (2010). The effect of plant secondary metabolites on the interplay between the internal and external environments of marsupial folivores. Chemoecology, 20(2), 97–108.

DeGabriel, J. L., Moore, B. D., Foley, W. J., & Johnson, C. N. (2009). The effects of plant defensive chemistry on nutrient availability predict reproductive success in a mammal. Ecology, 90(3), 711–719.

DeGabriel, J. L., Moore, B. D., Felton, A. M., Ganzhorn, J. U., Stolter, C., Wallis, I. R., ... & Foley, W. J. (2013). Translating nutritional ecology from the laboratory to the field: Milestones in linking plant chemistry to population regulation in mammalian browsers. Oikos, 123(3), 298–308.

Demment, M. W., & Soest, P. J. V. (1985). A Nutritional Explanation for Body-Size Patterns of Ruminant and Nonruminant Herbivores. The American Naturalist, 125(5), 641–672.

Dove, H., & Mayes, R. W. (2006). Protocol for the analysis of n-alkanes and other plant-wax compounds and for their use as markers for quantifying the nutrient supply of large mammalian herbivores. Nature Protocols, 1(4), 1680–1697.

Dury, S. J., Turner, B. J., & Foley, W. J. (2001). Can hyperspectral data be used to map koala and possum habitat?. In Geoscience and Remote Sensing Symposium, 2001. IGARSS'01. IEEE 2001 International (Vol. 4, pp. 1648–1650). IEEE.

Freeland, W. J., & Janzen, D. H. (1974). Strategies in herbivory by mammals: The role of plant secondary compounds. The American Naturalist, 108(961), 269–287.

Freeland, W. J., & Winter, J. W. (1975). Evolutionary consequences of eating:Trichosurus vulpecula (Marsupialia) and the genus Eucalyptus. Journal of Chemical Ecology, 1(4), 439–455.

Johnson, D. H. (1980). The comparison of usage and availability measurements for evaluating resource preference. Ecology, 61(1), 65–71.

Logan, M., & Sanson, G. D. (2002). The effects of tooth wear on the activity patterns of free-ranging koalas (Phascolarctos cinereus Goldfuss). Australian Journal of Zoology, 50(3), 281–292.

Marsh, K. J., Foley, W. J., Cowling, A., & Wallis, I. R. (2003). Differential susceptibility to Eucalyptus secondary compounds explains feeding by the common ringtail (Pseudocheirus peregrinus) and common brushtail possum (Trichosurus vulpecula). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology, 173(1), 69–78.

Marsh, K. J., Wallis, I. R., McLean, S., Sorensen, J. S., & Foley, W. J. (2006). Conflicting demands on detoxification pathways influence how common brushtail possums choose their diets. Ecology, 87(8), 2103–2112.

Martin, J. K., & Martin, A. A. (2007). Resource distribution influences mating system in the bobuck (Trichosurus cunninghami: Marsupialia). Oecologia, 154(1), 227–236.

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

Moore, B. D., & DeGabriel, J. L. (2012). Integrating the effects of PSMs on vertebrate herbivores across spatial and temporal scales. In G. R. Iason, M. Dicke & S. E. Hartley (Eds.), The Ecology of Plant Secondary Metabolites: From Genes to Global Processes (pp. 226–246). Cambridge: Cambridge University Press.

203

Moore, B. D., Wallis, I. R., Marsh, K. J., & Foley, W. J. (2004). The role of nutrition in the conservation of the marsupial folivores of eucalypt forests. In D. Lunney (Ed.), Conservation of Australia’s forest fauna (2 ed., pp. 549–575). Mosman: Royal Zoological Society of New South Wales.

Nersesian, C. L., Banks, P. B., Simpson, S. J., & McArthur, C. (2012). Mixing nutrients mitigates the intake constraints of a plant toxin in a generalist herbivore. Behavioral Ecology, 23(4), 879–888.

Senft, R. L., Coughenour, M. B., Bailey, D. W., Rittenhouse, L. R., Sala, O. E., & Swift, D. M. (1987). Large herbivore foraging and ecological hierarchies. BioScience, 37(11), 789–799.

Simpson, S. J., & Raubenheimer, D. (2001). The geometric analysis of nutrient—allelochemical interactions: A case study using locusts. Ecology, 82(2), 422–439.

Valentini, A., Miquel, C., Nawaz, M. A., Bellemain, E. V. A., Coissac, E., Pompanon, F., ... & Taberlet, P. (2009). New perspectives in diet analysis based on DNA barcoding and parallel pyrosequencing: the trnL approach. Molecular Ecology Resources, 9(1), 51–60.

Westoby, M. (1978). What are the biological bases of varied diets? The American Naturalist, 112(985), 627–631.