How are the dynamics of woodland dominants influenced by ...

How are the dynamics of woodland dominants

influenced by climate and disturbances in south

eastern Australia?

Janet S. Cohn

Bachelor of Natural Resources (Honours)

Thesis submitted to Charles Sturt University for the degree of Doctor of

Philosophy

August 2010

Faculty of Science

School of Environmental Sciences

i

Table of Contents

Acknowledgements iii

Substantial Contributions to the thesis iv

Publications v

Abstract vi

Chapter 1. General introduction and thesis overview 1

Chapter 2. Demographic patterns of Callitris glaucophylla are associated

with rainfall and disturbances along rainfall gradients 21

Abstract and Introduction 22

Methods 25

Results 29

Discussion 37

Chapter 3. Competitive feedback between Eucalyptus species and

Callitris glaucophylla trees and C. glaucophylla regeneration influences

stand dynamics along a rainfall gradient 43


Methods 47

Results 52

Discussion 62

Chapter 4. Feedback between patches of Callitris glaucophylla and fire

intensity influences Callitris survival in flammable Eucalyptus woodlands

in different fire-weather conditions 67


Methods 71

Results 76

Discussion 85

Chapter 5. Synthesis 89

Chapters 2-4 90

The dynamics of woodland dominants in geographic zones 96

Predicted dynamics of woodland dominants in geographic zones under

climate change 98

Further work 100

References 102

ii

Certificate of Authorship

iii

Acknowledgements

Firstly, I would like to thank my supervisors Ian Lunt, Ross Bradstock and Karen Ross.

They were all of immense help during the production of this thesis. Their different

backgrounds and interests complimented one another very well. I am particularly

grateful to Ian, who was always generous with his time, enthusiasm and extraordinary

sense of logic. We had many creative and exciting discussions which helped to shape

the research and I look forward to many more.

Jack Baker and Klaus Koop agreed to my leave from DECCW NSW to pursue this

research. John Whittall, also from DECCW NSW smoothed the way for the

flammability work. Thanks to Stephen Campbell and Alan Smith from State Forests

NSW for their assistance and permission to take tree samples. Statisticians Simon

McDonald (CSU) and Terry Koen (DECCW NSW) were wonderfully generous with

their time and patience.

Finally, I would like to thank my dear family and friends. Andrew, Katrina and Bernie

Cohn, Grace Bradstock, Livia Tigwell, John Warren and Kate Hammill helped me in

the field and in many other ways, for which I am grateful.

iv

Substantial Contributions to the thesis

The research was funded by the Australian Research Council (ARC) Linkage Grant

between Charles Sturt University (CSU) and NSW Department of Environment and

Climate Change (NSW DECC; Project number LP066829) and Australian Institute for

Nuclear Science and Energy (AINSE; AINGRA09075).

Discussions with supervisors Ian Lunt, Ross Bradstock and Karen Ross variously

assisted with the formulation of objectives, methods and data interpretation. They also

edited drafts of the thesis. I carried out the fieldwork with the help of assistants and I

wrote the thesis. Statisticians Simon McDonald and Terry Koen helped to advise me on

the appropriate analyses for difficult datasets and they assisted me in carrying out some

analyses.

v

Publications

Cohn, J. S., Lunt I.D., Ross K. A. & Bradstock R. A. (in review) How do slow-growing,

fire-sensitive conifers survive in flammable eucalypt woodlands? Journal of Vegetation

Science.

The relative contributions of the authors to this manuscript are similar to those outlined

in Substantial Contributions to the Thesis (page iv).

vi

Abstract

Woodlands in south eastern Australia are dominated by trees of contrasting functional

types. Callitris glaucophylla is a slow growing, obligate-seeding conifer and Eucalyptus

species are fast growing, resprouting angiosperms. The success of these functional types

is likely to vary with climate and different levels and rates of disturbances along

climatic gradients. I examined how climate and disturbances influenced the dynamics of

these woodland dominants, by using surveys of varying scales.

Regeneration failure of Callitris has led to ageing populations (mid - late 1800s) below

405 mm annual rainfall in the winter rainfall zone. Failure was associated with farms,

which graze livestock and where rabbits are common. In contrast, Callitris populations

are younger (since 1950s) and expanding on tenures other than farms (roadside,

travelling stock routes, State Forests), where livestock grazing levels are low-moderate,

and rabbits and fires are less frequent above 405 mm mean annual rainfall.

Competition from canopy trees had little effect on the success of Callitris regeneration.

Although regeneration under canopy trees had lower densities, were smaller and less

likely to be reproducing than those in gaps, they have persisted 50 years after their

establishment. In contrast, competition from regeneration had a significant effect on

canopy tree resilience during drought. Callitris canopy trees surrounded by regeneration

had higher mortality and Eucalyptus canopy trees had increased stress levels at lower

rainfall.

In large woodland remnants, where unplanned fires occur (e.g. National Parks), dense

Callitris regeneration can reduce incoming fire intensity from the flammable Eucalyptus

matrix in low-moderate fire weather conditions. This increases the survival probability

of larger Callitris, allowing coexistence with Eucalyptus.

These results highlight how the dynamics of woodland dominants are associated with

disturbance regimes and climate along climatic gradients. Disturbance regimes were

associated with rainfall through different land use types. As climate changes, rainfall,

land use and disturbance regimes will shift along climatic gradients, influencing the

distributions of these woodland dominants.

Chapter 1. General introduction and thesis overview

1

Chapter 1

General introduction and thesis overview

Callitris glaucophylla and Eucalyptus species woodland, Merrimerriwa Range, central western New South Wales (Source: K. Ross)


2

Chapter 1

General introduction and thesis overview

Introduction

In recent centuries woodlands worldwide have undergone dramatic changes (Belsky &

Blumenthal 1997; Wright et al. 2004; Noss et al. 2006) and those in south eastern

Australia are no exception (Clarke 2000; Lunt et al. 2006). Their occurrence in

agricultural zones has led to landscape-scale changes in disturbance regimes (Sivertsen

& Clarke 2000). Introduced grazers, tree removal, and changes in fire regimes and soils

have resulted in a mosaic of woodland compositional and structural states (Austin &

Williams 1988; Westoby et al. 1989; Clarke 2000; Lunt et al. 2006). Interactions among

these processes are often complex and poorly understood (Clarke 2000; Burrows 2002;

Graz 2008).

Woodlands are widespread and extend from the temperate to the arid zone in south

eastern Australia (Fig. 1.1; Specht & Specht 1999; Keith 2004). Within this climatic

range, different processes are likely to control the dynamics of woody plants. Climate

can directly affect species distributions through physiological limitations, imposed by

rainfall and temperature, or indirectly by influencing disturbance regimes (Gworek et al.

2007). For example, in a study along an elevational gradient, the dynamics of Pinus

jeffreyi were variously associated with interactions between rainfall (drought),

parasitism by mistletoe and seed predation by insects (Gworek et al. 2007). Disturbance

regimes may also be influenced by predominant human land uses that vary with climate.

For example, as rainfall increases from the arid to the temperate zone, a shift from

grazing to a range of land uses becomes possible at higher rainfall (e.g. forestry and

cropping; Dale 1997; Bureau of Rural Sciences 2009). Changes in regional land use can

further alter disturbances (e.g. fires and grazing) within patches of native vegetation.

Trees of contrasting functional types dominate woodlands in south eastern Australia.

Callitris glaucophylla, a slow growing, fire-sensitive obligate-seeder often coexists with

a range of Eucalyptus species, which are fast growing, fire-tolerant resprouters (Lacey

1972, 1973; Clayton-Greene 1981; FCNSW 1988; Florence 1996; Specht & Specht

1999). Differences in growth rates and methods of regeneration between these

evergreen trees are likely to result in different responses to climatic conditions and


3

disturbances, leading to variations in age-structure and abundances across their ranges.

At extremes, Callitris can occur as dense regeneration or as widely spaced mature trees

(Lacey 1972, 1973; Noble 1997). The regeneration success of Eucalyptus is also

spatially variable (Reid & Landsberg 2000). Despite their prominence, there are gaps in

our understanding of how the comparative dynamics of Callitris and Eucalyptus vary in

relation to disturbances and climate across their ranges (Clayton-Greene 1981; Bowman

et al. 1988; Bowman & Wilson 1988; Clayton-Greene & Ashton 1990).

Figure 1.1 The distribution of woodland types in Australia. The study area is circled in

blue (Source: Lindenmayer 2005).

Interactions among functional types, productivity and rainfall are hypothesised as

determinants of a species’ success (Bellingham & Sparrow 2000; Huston 2003). Few

studies have examined these interactions using woody plants (Loehle 2000; Bond et al.

2001a; Zavala & Zea 2004; Sankaran et al. 2005; Pausas & Bradstock 2007). Of those

that have, most have concentrated on relationships between shrubs or between grasses

and trees (Scholes & Archer 1997; Bond et al. 2001; Sankaran et al. 2005), rather than

between tree species (Loehle 2000; Zavala & Zea 2004). A useful first step in predicting

patterns of Callitris and Eucalyptus success in inland south eastern Australia is to use

theoretical models to examine the relative success of plant functional types along

productivity and disturbance gradients (Bellingham & Sparrow 2000; Huston 2003).


4

The aim of this thesis is to examine how the dynamics of woodland dominants of

contrasting functional types, Callitris glaucophylla and Eucalyptus species, are

influenced by climate and disturbance regimes in south eastern Australia.

In this first chapter, I give an account of the current knowledge of the dynamics of

woodland dominants, Callitris glaucophylla and Eucalyptus species. The chapter begins

with the climatic setting of study region, a description of disturbance regimes,

physiological and life history characteristics of functional types (Callitris and

Eucalyptus) and our knowledge of how functional types respond to varying rainfall and

disturbances. Then I predict functional type success (Callitris and Eucalyptus) along

productivity and disturbance gradients, by examining a range of theoretical models.

Finally, I give an outline of the thesis.

Callitris and Eucalyptus woodlands

Climate and habitat

The study area lies within woodlands in south eastern Australia (Fig. 1.1). It extends

from the temperate zone in the east to the arid zone in the west (Specht & Specht 1999),

a distance of approximately 600 kms. Mean annual rainfall ranges from 600 to 200 mm.

From north to south, there is a change in seasonal rainfall from summer to winter

dominance (Colls & Whittaker 2001; BOM 2010a). Under climate change, much of the

study area is predicted to have increases in drought severity and frequency and a switch

to summer dominant rainfall (Nicholls 2004; CSIRO & BOM 2007; Hennessey et al.

2008; DECCW NSW 2009). Only in the north east is the water balance expected to

remain relatively unchanged (DECCW NSW 2009).

Woodlands decline in height and become more open from the temperate to the arid zone

(Specht & Specht 1999). They are dominated by Callitris glaucophylla and range of

Eucalyptus species (Fig. 1.2). Whilst both genera can coexist at higher rainfall (Lindsay

1967; Keith 2004), in the drier parts of their range, Eucalyptus become confined to

intermittent water courses or rock outcrops (Brooker & Kleinig 1990). Callitris

glaucophylla most commonly occurs on sandy to loamy textured soils, whilst

Eucalyptus species can occur on a wider range of soils (Lacey 1973; Wardell-Johnson et

al. 1997).


5

Figure 1.2 A mixed stand of Callitris glaucophylla and Eucalyptus species (left

foreground) in central New South Wales. Callitris which established in the 1950s is on

the right beside a 2m tall standard and on the left is a Callitris which established in the

late 1800s (Source: J. Cohn).

Disturbance regimes

Woodlands in south eastern Australia have undergone major changes in disturbance

regimes since European management in the 1800s (Austin & Williams 1988; Westoby

et al. 1989; Clarke 2000; Lunt et al. 2006). Landscape-scale manipulation of tree

densities, changes in fire regimes and introduced herbivores have led to a mosaic of

disturbance regimes and resultant stand types (Noble 1997; Lunney 2001; Lunt &

Spooner 2005).

Extensive areas of woodlands have been cleared for grazing, cropping and building

materials (Sivertsen & Clarke 2000; Cox et al. 2001). Selective ringbarking of

Eucalyptus to improve growth rates of the commercially valuable Callitris (Allen 1995),

has resulted in a switch from Eucalyptus to Callitris dominance in State Forests (Lunt et

al. 2006).


6

Little information is available on past and present fire regimes across the study area.

However, it is generally agreed that fire extent and frequencies have declined since

European management, largely as a result of habitat fragmentation, a reduction in

biomass caused by livestock grazing, as well as active fire suppression and exclusion

(Wilson 1990; Pulsford et al. 1992; Noble 1997, 2001). Unplanned fires still occur

regularly in large woodland remnants, such as conservation reserves (Brookhouse

1999).

Sheep and cattle grazing were introduced during the 1800s (Noble 1997). As a result,

farms have historically had higher grazing levels than more fragmented tenures like

roadside verges, travelling stock routes, State Forests and National Parks (Lunt 1995;

Fensham 1998a). Rabbits were introduced to the region in the late 1800s and spread

rapidly reaching ‘plague proportions’ over much of south eastern Australia until the

1950s (Austin & Williams 1988; Myers et al. 1994), when their numbers declined

dramatically after the release of myxomatosis, a mosquito borne virus (Noble 1997).

Since then, rabbit populations have been maintained at lower levels (Norris et al. 1991;

Tiver & Andrew 1997) with the help of rabbit haemorrhagic disease virus since the

1990s (Scanlan et al. 2006).

Callitris and Eucalyptus functional types and responses to disturbances

Physiological and life history attributes

Callitris glaucophylla and Eucalyptus species are long-lived evergreen trees, capable of

reaching a maximum height of approximately 30m (Cunningham et al. 1992; Fig. 1.2).

The two genera represent contrasting functional types (Table 1.1). Callitris

glaucophylla is a slow-growing, shade tolerant, obligate-seeding conifer, whilst

Eucalyptus species are faster-growing, shade-intolerant, resprouting angiosperms

(Lacey 1972, 1973; Clayton-Greene 1981; FCNSW 1988; Florence 1996). These life-

history differences are likely to influence their respective responses to productivity and

disturbance regimes (Bellingham & Sparrow 2000).

A major difference between Callitris and Eucalyptus is in their responses to fire.

Following 100% canopy scorch, Callitris die, whilst coexisting Eucalyptus resprout

(Gill & Bradstock 1992). Following death, persistence of Callitris is dependent on

seeds. Given these responses, Callitris is termed an obligate-seeder and Eucalyptus a

resprouter (Gill & Bradstock 1992).


7

Table 1.1 Characteristics of Eucalyptus species and Callitris glaucophylla (Lacey 1972,

1973; Hodgkinson et al. 1980; Specht 1981; Boland et al. 1984; FCNSW 1988;

Bowman & Harris 1995; Read 1995; Noble & Gitay 1996; Benson & McDougall 1998;

Zeppel & Eamus 2008).

Callitris Eucalyptus

Life history

Life-form tree tree

Maximum height (m) 30 30

Longevity (years) >200 >200

Leaf-form scale-like broad

Leaf mass per area (g m-2

) 2201 182

1

Leaf longevity (months) not known 18

Regeneration seed seed, resprout

Seed size larger (winged) very small

No. seed g-1

50-150 114-2560

Fruiting frequency annual annual

Physiology

Maximum photosynthetic rate slow fast

Shade tolerance high low

Response to drought reduce canopy leaf loss

conductivity

1 Leaf mass per area = 1/specific leaf area

2Callitris intratropica and Eucalyptus tetrodonta (Prior et al. 2003)

The success of annual seed production of Eucalyptus and Callitris is variable and may

depend on a range of factors including rainfall, stand density, pollination and damage by

birds and insects (Lacey 1973; Burrows & Burrows 1992; Benson & McDougall 1998).

Seeds of both genera develop in woody cones or capsules, and are released on maturity

(Harden 2002). There is no evidence of seed storage, otherwise known as serotiny, in

Callitris glaucophylla (Lacey 1973). Seeds of Callitris are bigger, fewer and winged

compared with the small, unwinged Eucalyptus seeds (Mortlock & Lloyd 2001; Harden

2002). Most seeds of both genera fall within one to three tree heights of the adult,

although wind, water and invertebrates may distribute them further (Lacey 1972;

FCNSW 1988; Bowman & Harris 1995). Successful establishment of Callitris seedlings

requires above average rainfall, especially during summer when drought can lead to


8

desiccation (FCNSW 1988). Eucalyptus regeneration occurs more commonly following

fires and above average rainfall (FCNSW 1988; Bell & Williams 1997; Gill 1997). Fire

contributes to Eucalyptus establishment by stimulating mass seed release, reducing

competition and favourably altering the microenvironment (Bell & Williams 1997).

Callitris grows slowly compared with coexisting Eucalyptus, especially in the earlier

stages of development (Clayton-Greene 1981). Self-thinning of dense Callitris saplings

occurs at slow rates, as they reduce growth rates to compensate for intraspecific

competition (Lacey 1972, 1973; FCNSW 1988). In comparison Eucalyptus self-thin at

faster rates (Clayton-Greene 1981; FCNSW 1988). Slower growth rates of Callitris

increase their time to reproductive maturity, thereby increasing the period of

vulnerability to biotic and abiotic factors (Bond 1989). Without intense competition

individual Callitris may reach reproductive maturity in approximately six years (Lacey

1972). Following germination Eucalyptus quickly develop a lignotuber, from which

they can resprout (Bell & Williams 1997). Callitris and Eucalyptus are similarly long-

lived, reaching ages of at least 200 years (Bowman & Harris 1995; Williams & Brooker

1997).

Callitris have small scale-like leaves, which form a denser canopy than the pendulous,

broader leaves of Eucalyptus. Both leaf types are considered sclerophyllous (Berry &

Roderick 2002), which is interpreted as an adaptation to seasonal drought in

Mediterranean type systems (Cunningham et al. 1999; Ninemets 2001; Wright et al.

2001; Lamont et al. 2002). Studies during droughts have found that Callitris and

Eucalyptus similarly reduce water usage, but by different means. Callitris reduce

canopy conductance or gas exchange and Eucalyptus shed leaves (Attiwill & Clayton-

Greene 1984; Zeppel & Eamus 2008). The recovery of Eucalyptus is dependent on their

ability to resprout (Pook et al. 1966; Attiwill & Clayton-Greene 1984), which may be

tested during periods of prolonged drought.

Differences in leaf size influence the flammability of leaf litter through differential

spatial packing (Schwilk & Ackerly 2001; Scarff & Westoby 2006). Larger leaves like

those of Eucalyptus provide better ventilated ground fuels which increases fire intensity

compared with the smaller leaves of Callitris which pack more tightly (Scarff &

Westoby 2006). The formation of tightly packed litter advantages slow growing,


9

obligate-seeders like Callitris, by increasing their chances of surviving fire and

increasing their probability of reaching reproductive maturity (Scarff & Westoby 2006).

Rainfall

The success of Callitris regeneration is strongly influenced by rainfall. Widespread

establishment of Callitris was observed in the late 1800s and the 1950s following above

average rainfall during La Nina events (Austin & Williams 1988; Noble 1997).

Regeneration during these periods contributed to the formation of dense thickets of

Callitris regeneration, giving it ‘woody weed’ status in parts of the semi-arid zone

(Noble 1997; Fig. 1.3). Conversely, during periods of drought, desiccation of Callitris

seedlings has contributed to Callitris mortality in arid areas (Johnson 1969; Read 1995).

Regeneration of Eucalyptus similarly coincides with above average rainfall (Fensham et

al. 2009).

Survival of Callitris and Eucalyptus trees is negatively affected by drought (Lacey

1972, 1973; Allen 1998; Fensham 1998b; Fensham & Holman 1999; Fensham et al.

2009). During drought higher mortality of Eucalyptus and Callitris trees has been

observed or recorded at higher tree densities (Lacey 1972, 1973; Fensham & Holman

1999). There is anecdotal information that Eucalyptus and Callitris trees surrounded by

dense Callitris sapling regeneration are selectively killed (Lacey 1972, 1973; Allen

1998). Foresters have also observed higher mortality of Callitris trees in lower rainfall

regions (Lacey 1972).

Fire regimes

Given its sensitivity to fire, fire regimes are likely to have a major influence on the

population dynamics of Callitris. A regime of frequent high intensity fire is believed

responsible for contracting populations of Callitris intratropica in northern Australia

(Bowman et al. 2001). Callitris intratropica is a slow-growing obligate-seeder like

Callitris glaucophylla (Bowman & Harris 1995). Patchy burning has been observed in

monospecific stands of Callitris compared with continuous burning in adjacent

Eucalyptus stands (Clayton-Greene 1981; Bradstock & Cohn 2002). Minimal damage to

dense patches of Callitris regeneration has been observed in ‘severe fires’ (Lacey 1972).

Low biomass of grasses, shrubs and lower leaf litter flammability are likely to

contribute to lower flammability associated with dense stands of Callitris (Clayton-

Greene 1981; Bowman & Wilson 1988; Scarff & Westoby 2006).


10

Figure 1.3 Dense regeneration of Callitris glaucophylla in central New South Wales.

Callitris saplings that established in the 1950s surround an adult Callitris that

established in the late 1800s. The standard is 4m tall (Source: J. Cohn).

Interactions between rainfall, grazing and fire regimes are hypothesised as determinants

of woody species encroachment in semi-arid woodlands (Westoby et al. 1989; Walker

1993; Noble 1997; Burrows 2002). Regular fires following above average rainfall and

herbaceous fuel production are believed to have reduced opportunities for woody plant

establishment prior to European management. However, under high levels of introduced

grazers, reductions of herbaceous fuels have contributed to declines in fire frequency,

allowing the establishment of woody species, like Callitris (Noble 1997).

Grazing regimes

Grazing by sheep and cattle has reduced or prevented the regeneration success of both

Callitris (FCNSW 1988; Lacey 1973; Allcock & Hik 2004) and Eucalyptus (Parker &

Lunt 2000; Allcock & Hik 2004). Widespread grazing by large numbers of rabbits

prevented the regeneration of Callitris from 1890 to 1950 (Austin & Williams 1988;

Noble 1997), and is still affecting regeneration in some locations (Allcock & Hik 2004).


11

Summary

Whilst this literature review provides an indication of how rainfall, and disturbance

regimes (logging, fire, grazing) influence Callitris and Eucalyptus dynamics, it also

highlights the relative incompleteness of this knowledge. It reveals that: above average

rainfall stimulates establishment and drought results in mortality of Callitris and

Eucalyptus; high frequencies and intensities of fires do not favour the obligate-seeding

Callitris; regeneration of Callitris and Eucalyptus are negatively affected by grazing.

However, we know little about Callitris survival during low-moderate intensity fires

and how this may vary with site productivity. We do not know how different levels of

grazing may affect Callitris and Eucalyptus success and how this may vary with site

productivity. We also know little about the comparative effects of drought on Callitris

and Eucalyptus, which is especially important as droughts increase in frequency and

severity under global warming (Nicholls 2004; CSIRO & BOM 2007; Hennessey et al.

2008; DECCW NSW 2009). By examining theoretical models of functional type

responses, we can predict the relative success of Callitris and Eucalyptus at varying

levels of productivity and disturbances.

Models of functional type success along productivity and disturbance

gradients

Gymnosperms dominated global vegetation up until the late Cretaceous (Krassilov

1978). The fossil record in Australia provides evidence of a dramatic reduction in

conifer diversity and distribution during the Tertiary (Wells & Hill 1989; Hill 1995).

Possible explanations include competition with angiosperms, which were spreading and

diversifying at that time (Bond 1989), increasing aridity (Brodribb & Hill 1998) and an

increase in disturbance frequency (Hill & Brodribb 1999).

Competition, drought and disturbance regimes are worthy of exploration in explaining

contemporary dominance shifts in Eucalyptus species and C. glaucophylla across their

ranges. By examining a range of theories that use physiological and life history traits to

predict dominance along gradients of productivity and disturbance (Specht 1981;

Keeley 1986; Bond 1989; Westoby 1999; Bellingham & Sparrow 2000; Huston 2003), I

aim to generate hypotheses on the likely mechanisms for the relative success of

Eucalyptus and Callitris in south eastern Australia. Models are ordered from those

which predict patterns along (a) productivity gradients, (b) disturbance gradients, and

(c) productivity and disturbance combinations.


12

Physiological traits and productivity

It is widely accepted that competition for resources, especially light determines

dominance at higher productivity sites (Gurevitch et al. 2006). However, there is debate

on whether competition for water and nutrients (Tilman 1988) is more important than

stress tolerance (Grime 1979) in determining dominance at lower productivity sites. In

the ‘slow seedling’ hypothesis, Bond (1989) proposed that the competitive ability of

conifers in the regeneration phase is limited by physiological characteristics. Because of

intense competition from faster growing angiosperms on higher productivity sites,

conifers are left more vulnerable to disturbances, restricting them to lower productivity

sites.

Leaf traits reflect resource uptake and use-efficiency strategies in plants, so are good

indicators of whole-plant function (Reich et al. 1999). Comparative studies suggest

significant differences in leaf traits between conifers and angiosperms (Ackerly & Reich

1999; Lusk et al. 2003; Prior et al. 2003). Conifers have smaller leaves with longer life

span, lower leaf specific area, lower photosynthetic capacity, stomatal conductance, and

higher area based nitrogen concentration. The implications of these traits are discussed.

The higher photosynthetic capacity per unit leaf mass of evergreen angiosperms is likely

to give angiosperms a competitive advantage over conifers in productive habitats (Lusk

& Matus 2000; Lusk et al. 2003). The difference in photosynthetic performance

between evergreen conifers and angiosperms is partly attributable to variation in

stomatal conductance, which is higher on an area basis in angiosperms (Lusk et al.

2003). Lower stomatal conductivity in conifers is consistent with the theory that

conifers may still be disadvantaged by stomatal traits evolved under higher atmospheric

levels of CO2 (Beerling & Woodward 1996).

Conifers may be advantaged at lower productivity sites by their lower photosynthetic

capacity placing fewer demands on resources (water, nutrients; Grime 1979). Wright et

al. (2001) argued that leaves with higher area based nitrogen concentration, as were

found on conifers (Ackerly & Reich 1999), make greater investments in photosynthetic

‘machinery’ allowing a given photosynthetic rate to be achieved at lower stomatal

conductivity. This encourages water-use efficiency (Schulze et al. 1998; Lamont et al.

2002; Llorens et al. 2004). The longer-lived leaves of conifers allows nutrients to be

accrued over a longer residence time (Roderick et al. 2000; Lusk et al. 2003; Wright et


13

al. 2004), which helps persistence in lower productivity sites. These characteristics

suggest that conifers may be best described as stress-tolerators according to Grime’s

(1979) Competitor-Stress-tolerator-Ruderal scheme. Keeley & Zedler (1998) believe

that this scheme is useful in distinguishing the qualities that separate Pinus species from

woody angiosperms in their ability to tolerate abiotically stressful environments.

Leaf specific area is the ratio of projected leaf area to leaf dry mass (Roderick et al.

2000). Many studies have shown that leaf specific area increases with soil water

availability along aridity gradients (e.g. Pierce et al. 1994; Schulze et al. 1998; Reich et

al. 1999). This adds further weight to the prediction that conifers and angiosperms may

dominate at the lower and higher ends of productivity gradients, respectively.

The gap availability hypothesis relates recruitment opportunities of seeders (non-

resprouters) to the availability of gaps along resource gradients (Specht 1981; Bond &

van Wilgen 1996). In humid environments, regeneration of resprouters predominates

because competition from dense stands of resprouters reduces the successful

establishment of seeders (Specht 1981). In contrast, seeder establishment is more

successful in drier environments, because lower densities of resprouters result in more

open vegetation. Thus the proportional representation of resprouters and seeders shifts

along gradients of water availability. In large-scale gradient studies comparable with

Specht’s (1981), resprouter success increased with increasing moisture availability

(Pausas & Bradstock 2007) and increasing summer rainfall dominance (Ojeda 1998).

However, this trend may be confounded by fire frequency, which Pausas & Bradstock

(2007) found to increase with increasing rainfall. Disentangling the effects of water

availability from fire regimes is important if we are to understand the drivers of the

relative success of resprouters and seeders (Pausas & Bradstock 2007).

At a local scale, seeders are often more abundant on drier sites than resprouters in a

range of Mediterranean ecosystems (Keeley 1986; Keith 1991; Pausas et al. 1999;

Meentemeyer et al. 2001; Clarke & Knox 2002). Since resprouters tend to be deeper

rooted than seeders (Keeley 1986; Bell 2001) and water potential for resprouters is

usually higher during summer (Davis et al. 1999; Ackerly 2004), the above ground

organs of seeders must have physiological, chemical and/or structural compensations

(Paula & Pausas 2006). Comparative studies on physiological tolerance of seeders and

resprouters to drought vary. Some studies have found that seeders are more drought

tolerant than resprouters because they have greater resistance to embolisms caused by


14

water stress and later stomatal closure than resprouters (Davis et al. 1999), or higher

potential for structural resistance to drought and higher water use efficiency (Ackerly

2003, 2004; Paula and Pausas 2006). In other studies resprouters have shown higher

carbon assimilation at low water potentials and later stomatal closure indicating greater

drought tolerance than seeders (Calamassi et al. 2001; Vilagrosa et al. 2003). Although

no pattern in the relative drought tolerance of seeders and resprouters has also been

identified (Smith et al. 1992), Knox & Clarke (2005) found that they respond differently

to the availability of water and nutrients. Seedlings of resprouters may be advantaged at

more fertile sites because of allocation to the formation of roots and starch reserves,

giving them a competitive advantage over seeders after fires (Knox & Clarke 2005).

The studies on leaf traits of conifers and angiosperms support the slow ‘seedling

hypothesis’ (Bond 1989). Faster growth rates may contribute to competitive advantage

of angiosperms over slower growing conifers at higher productivity sites. This restricts

conifers to lower productivity sites where slower growth rates and longer-lived leaves

contribute to their tolerance of stress. There is some support for the gap availability

hypothesis (Specht 1981), which predicts that seeders and resprouters will be more

successful at lower and higher productivity sites, respectively. Seeders may be more

drought tolerant than resprouters allowing better survival in lower productivity sites and

resprouters may have a competitive advantage over seeders in higher productivity sites.

Life history and disturbances

Bellingham & Sparrow (2000) predict that woody seeders can only survive where

disturbances are less frequent, mild and predictable enough for maturity to be reached

before they are killed. Conversely, woody resprouters should increase with increasing

frequency of disturbance (Bellingham & Sparrow 2000; Bond & Midgley 2001), and

decline at the highest frequencies where it is very difficult to grow and maintain storage

levels. Here, herbaceous seeders are more likely to be successful (Bellingham &

Sparrow 2000; Sparrow & Bellingham 2001). In fire-prone Mediterranean systems the

success of resprouters and seeders in relation to fire frequency is variable. In some

ecosystems non-respouters are more successful at low to intermediate frequencies of

fire (Keeley & Zedler 1978; Hilbert 1987; Burgman & Lamont 1992; Enright et al.

1998; Clarke 2002; Clarke & Knox 2002). Conversely, resprouters are more successful

at the lowest and higher frequencies, where seeders senesce and do not reach

reproductive maturity, respectively (Keeley & Zedler 1978; Kruger & Bigalke 1984;


15

Keeley 1986; Enright et al. 1998; Pausas 1999). In other ecosystems, such as Australian

heathlands, the relative success of seeders and resprouters with fire frequency can not be

generalised (Pausas et al. 2004). Traits such as time to maturity, shade tolerance,

serotiny and plant longevity may also need consideration when predicting species’

responses to fire regimes (Pausas 2001).

The fire frequency hypothesis assumes that space limits the regeneration of seeders

(Keeley & Zedler 1978; Specht 1981; Keeley 1986). As fire interval lengthens, more

gaps should occur as mortality rates of resprouters increase due to senescence and death

from higher intensity fires (Keeley & Zedler 1978; Keeley 1986). Consequently, there

will be more opportunities for seeders to establish, while sprouters would be more

common at shorter fire intervals. Whilst a number of field studies support this

hypothesis (Keeley & Zedler 1978; Kruger & Bigalke 1984; Traubaud 1991), Kruger

(1983) found that resprouters were prominent in fynbos shrublands, irrespective of fire

regime.

Westoby (1999) predicts that certain plant traits are more likely to be favoured under

different grazing regimes. Moderate intensity grazing is likely to select for species with

higher specific leaf area (SLA; e.g. Eucalyptus), which are considered more palatable,

because of lower structural strength than leaves with lower specific leaf area (e.g.

Callitris; Reich et al. 1997). Under this grazing intensity, species with lower specific

leaf area are likely to dominate. Under higher grazing intensity where selectivity is low,

species with higher specific leaf area may be more successful, as they can grow more

quickly, especially during early life (Reich et al. 1997; Ninemets 2001; Wright et al.

2001). Although this theory is sound, the comparatively small differences between

specific leaf areas of Callitris and Eucalyptus may not be sufficient to distinguish

different responses to grazing regimes.

Empirical studies indicate that the ability to predict grazing responses using plants traits

may vary with location (Vesk et al. 2004). In the productive grasslands of Argentina

and Israel, grazing resistance of herbaceous species was associated with a range of traits

like higher specific leaf area, short stature and smaller leaves (Diaz et al. 2001), but no

such associations were found in shrubs of the less productive arid and semi-arid

woodlands of Australia (Vesk et al. 2004). Vesk et al. (2004) believe that the different

outcomes of these studies may be related to structural differences in vegetation. In the


16

tropical grasslands where swards are simple and continuous, taller plants are selected, as

there is a ‘top down’ approach to grazing. However in drier woodlands, where the

vegetation is more open, allowing easier movement of grazers and there is a higher

diversity of growth forms (Vesk et al. 2004), simple traits are unlikely to be good

predictors of grazing response (Landsberg et al. 1999). These trends highlight the need

for further empirical studies on plant traits and grazing responses, especially in relation

to site productivity (Vesk et al. 2004).

Gaps created by grazing are most likely to favour plants which produce the highest

quantity of seed per unit biomass (Westoby 1999). Since seed mass is inversely related

to seed output per m2 of canopy (Henery & Westoby 2001), smaller seeded species (e.g.

Eucalyptus) are likely to be more successful in colonising than larger seeded species

(e.g. Callitris) under grazing (Westoby 1999). The applicability of this prediction may

be restricted to higher productivity sites, where the vegetation is denser, rather than

more open lower productivity sites, such as semi-arid woodlands where successful

establishment is dependent on above average rainfall, rather than gap availability (Vesk

et al. 2004).

The theories on the relative success of seeders and resprouters along disturbance

gradients are relatively consistent in their predictions. Seeders (e.g. Callitris) are likely

to succeed at lower intensities and frequencies of disturbances because, firstly they are

usually killed by higher intensity disturbances, and secondly they need sufficient time

between disturbances to reach reproductive maturity. In contrast, resprouters (e.g.

Eucalyptus) are more likely to succeed at higher frequencies and intensities, because of

faster growth rates and their ability to resprout, respectively.

Life history, productivity and disturbances

Relationships between site productivity and disturbances are likely to influence the

relative success of plants with different life histories (Bellingham & Sparrow 2000;

Huston 2003). Site productivity determines growth rates, which determine the time to

reproductive maturity. Since the time required to reach reproductive maturity is longer

at lower productivity sites, resprouters are likely to be more successful than seeders at

lower than higher productivity sites, given a similar disturbance regime (Bellingham &

Sparrow 2000).


17

Productivity and disturbances may also influence life history success through

establishment opportunities. The storage effect (Warner and Chesson 1985) has been

suggested as an important selective pressure for resprouting in drought affected fire-

prone systems. Here seedlings are likely to face a high probability of drought death, so

resprouters are more likely to be successful (Enright & Lamont 1992). Resprouters have

an advantage over seeders in being able to ‘store’ their reproductive potential via their

resprouting ability (Warner and Chesson 1985).

Feedback between vegetation and disturbance regimes can influence the success of life

history types. Vegetation can influence disturbance regimes, since higher biomass

production has the potential to support higher frequencies and intensities of disturbances

(Huston 2003). For example, in fire-prone systems of south eastern Australia, fires are

driven by herbaceous fuels in drier regions and leaf litter in wetter regions (Bradstock

2010). As the climate becomes drier (Nicholls 2004; CSIRO & BOM 2007; Hennessey

et al. 2008; DECCW NSW 2009), fire potential is predicted to decline in drier regions,

as herbaceous biomass declines (Bradstock 2010). In contrast, in wetter regions where

moist fuels may dry, fire potential may increase (Bradstock 2010). Differences in fire

regimes are likely to contribute to varying proportions of life history types along

productivity gradients.

Studies on the success of life history types in relation to productivity and disturbances

are rare. However in one comprehensive study, the evolution of life history types in the

genus Pinus was determined by the interaction of site productivity and fire frequency

(Keeley & Zedler 1998). Resprouting was a common trait where fires were predictable

and stand replacing. In contrast, bark thickness and self-pruning were common traits

where predictable ground fires predominated. Bark thickness and self-pruning were also

common traits at productive sites where less predictable stand replacing fires occurred.

In latter circumstances, the fire return interval may exceed the life span of some Pinus

species, so the ability to resprout was not an advantage.

The theories predict that resprouters are likely to be more successful than seeders at

lower productivity sites when disturbances are present, because of slower growth rates

and limited establishment opportunities for seeders.


18

Summary of predictions

The models are relatively consistent in their predictions of Eucalyptus and Callitris

success with varying productivities and disturbances. Eucalyptus is more likely to

succeed in sites of higher productivity because faster growth rates give Eucalyptus a

competitive advantage over Callitris. Callitris can dominate at lower productivity sites

because of higher water use efficiency. Eucalyptus are predicted to succeed at higher

frequencies and intensities of disturbances, irrespective of productivity, because faster

growth rates allow quicker recovery than Callitris, which is restricted to sites with lower

frequencies and intensities of disturbances.


19

Thesis aims

The aim of this thesis is to examine how the dynamics of woodland dominants of

contrasting functional types, Callitris glaucophylla and Eucalyptus species, are

influenced by climate and disturbance regimes in south eastern Australia.

Specific aims are addressed in chapters 2-4, and the final chapter summarises and

integrates the findings.

Chapter 2. Examines how demographic patterns of Callitris glaucophylla are

associated with rainfall and disturbances along rainfall gradients. This study extended

across the entire study area (Fig. 1.4).

Chapter 3. Examines how competitive feedback between Eucalyptus species and

Callitris glaucophylla trees and C. glaucophylla regeneration influences stand dynamics

along a rainfall gradient. This study occurred along a rainfall gradient in the south east

of the study area.

Chapter 4. Examines how feedback between patches of Callitris glaucophylla and fire

intensity influences Callitris survival in flammable Eucalyptus woodlands in different

fire-weather conditions. This study took place in a large fire-prone reserve in the north

east of the study area.

Chapter 5. Summarises and then integrates the findings using a spatial approach to

illustrate and explain, firstly the current dynamics and distributions of Callitris and

Eucalyptus, and secondly future dynamics under predicted climate changes.

2

3

4

Figure 1.4 A satellite image of Australia showing the scale of each study, numbered by

chapter (Source: Geosciences, Australia)


20

Chapter 2. Demographic patterns of Callitris

21

Chapter 2

Demographic patterns of Callitris glaucophylla are associated with rainfall

and disturbances along rainfall gradients.

Callitris glaucophylla of varying ages in north eastern New South Wales (Source: J. Cohn)


22

Chapter 2

Demographic patterns of Callitris glaucophylla are associated with rainfall

and disturbances along rainfall gradients.

Abstract

Predicting species distributions with changing climate has often relied on climatic

variables, but increasingly there is recognition that disturbance regimes should also be

included in distribution models. I examined how changes in rainfall and disturbances

along climatic gradients determined demographic patterns in a widespread and long-

lived tree species, Callitris glaucophylla in SE Australia. I examined recruitment since

1950 in relation to annual (200-600 mm) and seasonal (summer, uniform, winter)

rainfall gradients, edaphic factors (topography) and disturbance regimes (vertebrate

grazing (tenure and species), fire). A switch from recruitment success to failure

occurred at 405 mm mean annual rainfall, coincident with a change in grazing regime.

Recruitment was lowest on farms with rabbits below 405 mm rainfall (mean = 0-0.89

cohorts) and highest on less disturbed tenures with no rabbits, above 405 mm rainfall

(mean = 3.25 cohorts). Moderate levels of recruitment occurred where farms had no

rabbits or less disturbed tenures had rabbits, above and below 405 mm rainfall (mean =

1.77-1.71 cohorts). These results show that low annual rainfall and high levels of

introduced grazing has led to ageing, contracting populations, whilst higher annual

rainfall with low levels of grazing has led to younger, expanding populations. This

study demonstrates how demographic patterns vary with rainfall and spatial variations

in disturbances, which are linked in complex ways to climatic gradients. Predicting

changes in tree distribution with climate change requires knowledge of how rainfall and

key disturbances (tenure, vertebrate grazing) will shift along climatic gradients.


23

Introduction

A current challenge in ecology is to determine the processes influencing species

distributions, so shifts in distributions under climate change can be predicted. Many

models identify empirical relationships between current distributions of species and

climatic variables that are used to estimate future distributions of species under climate

scenarios (Thomas et al. 2004; Araujo & New 2007). Since they do not consider

population dynamics, biotic factors and disturbances, it is uncertain whether species will

track predicted shifts in suitable climate space (Pearson & Dawson 2003; Guisan &

Thuiller 2005; Keith et al. 2008). In the few studies where these factors have been

measured, biotic interactions and disturbance regimes were just as important as climate

in determining distribution and thus potential range shifts (Miller & Halpern 1998;

Garcia et al. 1999; Gworek et al. 2007; Boulant et al. 2009). This highlights the need to

include biotic factors and disturbances along with climate in studies identifying drivers

of species distributions.

For widespread species, different processes are likely to control demographic patterns in

different parts of their distributions (Gworek et al. 2007). Climate can directly affect

species distributions by limiting rainfall and temperature (Gworek et al. 2007). For

example, establishment of woody species in low rainfall systems is more dependent on

the infrequent periods of above average rainfall (Denham & Auld 2004; Castro et al.

2005; Holmgren et al. 2006a; Squeo et al. 2007), resulting in fewer opportunities than

in high rainfall systems. Climate can indirectly affect species distributions through

disturbance regimes (Gworek et al. 2007), which may be determined by land use, the

distribution of which is determined by climate (Dale 1997; Luck 2008). For example,

high levels of livestock grazing characteristic of arid systems (Walker 1993; Bureau of

Rural Sciences 2009) have contributed to ageing populations of woody species (Garcia

et al. 1999; Meyer & Pendleton 2005; Auld & Keith 2009). By comparison, in higher

rainfall systems, where a range of land uses are possible (Dale 1997), disturbance

regimes may similarly vary. Livestock grazing levels are likely to be higher on farms

than on other tenures (e.g. roadsides; Lunt 1995; Fensham 1998a), which may act as

refuges, especially for palatable woody species. Whilst fire frequencies have declined in

some agricultural systems, leading to an expansion of fire sensitive woody species

(Belsky & Blumenthal 1997; Noble 1997; Noss et al. 2006), large patches of woodland

may maintain a regime of regular fires (Brookhouse et al. 1999). Identifying where and


24

how processes are associated with demographic patterns along climatic gradients, will

allow more precise predictions of distributions under climate change.

Large-scale surveys of the demographic patterns of widespread and long-lived species

can be used to identify the climatic factors and disturbance regimes associated with their

distribution (Peters et al. 2007). Species with widespread distributions offer

comparatively more information on the factors influencing distributions than those with

restricted distributions, since their ranges will encompass greater variability in abiotic

factors (e.g. rainfall), environmental factors (e.g. topography, soil type) and disturbance

regimes (e.g. land use, grazing, fire; Pearson & Dawson 2003). The distribution of trees

is especially valuable because their longevity tells us about past influences (Dale et al.

2001). Seedlings of trees often act as demographic ‘bottle necks’ (Grubb 1977;

Richardson & Bond 1991), because their recruitment is more sensitive to changes in

climate and disturbances than the mortality of adults (Lloyd 1997; Hanson & Weltzin

2000). Changes in the distribution of climatic and disturbance regimes associated with

the demographic patterns of a species will influence its future distribution (Thuiller et

al. 2008).

My aim was to use a long-lived and widespread tree species to examine how rainfall

and disturbances are associated with demographic patterns along climatic gradients.

Callitris glaucophylla provides an ideal model to address this issue. It is a widespread,

long-lived canopy dominant of woodlands ranging from arid to temperate regions of

Australia (Bowman & Harris 1995). It is a slow growing obligate-seeding conifer,

which is sensitive to disturbances, since it does not resprout (Bowman & Harris 1995).

A number of small scale studies have identified varying demographic patterns, which

have been variously linked to climatic and disturbance variables. Recruitment success

has been associated with above average rainfall and recruitment failure with introduced

grazing and high intensity fires (Austin & Williams 1988; Bowman & Latz 1993; Noble

1997). However, it is unclear how these factors vary across climatic gradients and our

ability to predict future changes in distribution is dependent on how these factors

interact in space and time. I predict that: (1) disturbance regimes will vary along rainfall

gradients; and in the absence of inter- and intra-specific competition (2) fewer cohorts

are more likely where rainfall is low, especially where disturbances are frequent; and (3)


25

more cohorts are expected where productivity is higher, especially where disturbances

are less frequent.

Methods

Stands of Callitris glaucophylla Thompson & Johnson were sampled at 56 sites along

three transects of declining mean annual rainfall and varying seasonal rainfall

dominance in New South Wales, Australia (Colls & Whitaker 2001; BOM 2010a; Fig.

2.1). Transects extended from the arid zone in the west (200 mm annual rainfall) to the

temperate zone in the east (600 mm annual rainfall; Specht & Specht 1999; Bureau of

Meteorology 2006). Rainfall varies seasonally from north to south with summer

dominant rainfall in the north (60% excess in summer cf winter, Lacey 1973), uniform

rainfall in the middle and winter dominant rainfall in the south (40% excess rainfall in

winter cf summer). Hereafter, transects will be referred to as the summer (149.76oS, -

29.36oE to 141.44

oS, -29.03

oE

), uniform (148.17

oS, -33.15

oE to 141.63

oS, -32.77

oE)

and winter rainfall zones (147.20oS, -34.80

oE to 141.05

oS, -34.01

oE). Each transect was

approximately 600 kms in length. Fieldwork was undertaken from May to August 2008.

S1

S2

S3

S4

S5

S6

0 200

Kilometres

Figure 2.1 The location of transects used to survey Callitris recruitment along gradients

with mean annual rainfall (200 - 600 mm) from east to west in seasonal rainfall zones

from summer in the north, uniform in the middle to winter in the south in New South

Wales. Sites where wood samples were taken for radiocarbon dating are marked from

S1 to S6.


26

Sites were chosen to sample the maximum number of cohorts at intervals along each

transect. In the higher rainfall area to the east, sites were on average 30 km apart,

however in the lower rainfall areas in the west, sites were less frequent, and previous

vegetation surveys were used to identify sites within 50 kms north or south of each

transect (Fox 1991; Scott 1992; Porteners 1993, Porteners et al. 1997; Keith 2004,

Benson et al. 2006). A total of 56 sites were sampled including 17 in the winter, 23 in

the uniform and 16 in the summer rainfall zone.

Prerequisites for site inclusion were the presence of mature C. glaucophylla trees (circa

1890 recruitment; Whipp 2009) as a seed source, gaps between canopies and at least

10% bare ground to provide space for potential recruitment, flat to undulating

topography, no evidence of earthworks, logging or fires since the 1950s and minimal

stock grazing. Where possible, fenced roadsides and state forests were sampled, since

these were generally less intensely grazed than travelling stock routes and paddocks

(Fensham 1998a; Lunt 2005). However, the availability of less disturbed sites declined

in the lower rainfall areas in the west, where roadsides were often unfenced from

adjacent farms, and state forests were less common.

At each site a range of variables were measured within a 0.1 ha quadrat. The number of

cohorts which had established since the 1950s was counted. Cohorts were distinguished

by height differences, canopy shape, presence of bark lichens and lower stems and local

knowledge of land managers (Read 1995; Whipp 2009). Cohorts were allocated to a

relative age category from A to D, with A being the oldest and D the youngest at each

site. Where there was uncertainty in differentiating between two cohorts I lumped them

rather than split them, resulting in a higher probability of underestimating than

overestimating cohort numbers. The height and diameter of the two tallest individuals

from each cohort were measured.

An attempt to age Callitris samples using ring counts was not pursued, because of its

imprecision. Instead radiocarbon dating, based on the signature of the bomb-spike was

used (Hua 2009). Radiocarbon dating (14

C) of representative basal disks of Callitris

from two distant sites along each transect (where possible) was used to confirm

recruitment dates (Hua 2009; Fig. 2.1). This involved ageing one replicate from each

cohort at four sites (winter and uniform transects) and two replicates from each cohort

to examine within site reliability of cohort assessment at two sites (summer transect).


27

Disks from the younger trees were removed near ground level, but disks from the oldest

trees were removed 20 cm from the ground, reducing the precision of dating the year of

recruitment. Radiocarbon dating used the three innermost rings of each sample and was

confirmed by counting the total number of rings (Hua 2009). Since the precision of

radiocarbon dating is low for the most recent recruits (> 2003), they were aged by ring

counting only.

Topographic position and topsoil texture were recorded in each quadrat (McDonald et

al. 1990). Recent grazing was assessed by recording the presence (presence-absence

data) of native macropods and introduced vertebrates including cattle (Bos taurus

Linnaeus, 1758), sheep (Ovis aries Linnaeus, 1758), European rabbits (Oryctolagus

cuniculus Linnaeus, 1758), and goats (Capra hircus Linnaeus, 1758). Herbivores were

either sighted directly or indirectly by scats, warrens or diggings. Evidence of fire since

the 1950s relied on charred stems or stags of trees that established after 1950. Land

tenure was used as a surrogate for longer term grazing intensity, measured as low to

moderate and high: low to moderate = non-farms, including roadsides fenced from

adjacent farms, travelling stock routes, state forests and national parks; high = farms and

adjacent unfenced roadsides.

At a number of sites in the winter rainfall zone, C. glaucophylla and C. gracilis subsp.

murrayensis J. Garden coexisted. Despite their taxonomic distinction (Hill 1998; Pye et

al. 2003), it was apparent from diagnostic characteristics (cones and pedicels) that they

were producing hybrids (M. Pye pers. comm.). These sites were excluded from

sampling.

Statistical analyses

I used regression tree analyses to examine the suite of variables associated with the

number of cohorts at each site (Table 2.1). Regression trees are one of the few methods

that can model interactions without a priori choice or using stepwise procedures as with

generalised linear modelling (GLM, Breiman et al. 1993). They also accept mixes of

categorical and continuous variables, which is not possible with generalised additive

models (GAM, Thuiller et al. 2003) and they do not assume relations are linear across

the dataset (Breiman et al. 1993). Overfitting was avoided by pruning the tree to

minimise the cross-validated error (Venables & Ripley 2002). Categorical variables

which occurred in fewer than 30% of sites were excluded from the analyses.


28

Spearman’s rank coefficient was used to examine collinearity between remaining pairs.

In the first regression tree analysis all explanatory variables with r2 < 0.5 were included

(Tabachnick & Fidell 2007), i.e. annual rainfall (mm), latitude, topographic position

(flat, dune), land use (non-farm, farm), and the presence of rabbits, sheep, cattle, goats

and kangaroos. In the second regression tree analysis, annual rainfall was removed, as a

number of explanatory variables were significantly correlated with annual rainfall

including land use (r2=-0.461, P < 0.001), rabbits (r

2=-0.48, P < 0.0001), goats (r

2=-

0.34, P < 0.011) and topographic position (r2=-0.34, P < 0.011). No other explanatory

variables were significantly correlated. All statistical analyses were undertaken in

SPLUS software version 8 for windows (Insightful Corp. 2007).

Based on the results from the regression tree analysis, I examined trends in the number

of cohorts with annual rainfall in each seasonal rainfall zone, by fitting linear, loess and

sigmoidal relationships. The curve of best fit was chosen using the residual sum of

squares (RSS) and Akaike Information Criterion (AIC). I used RSS and AIC to

discriminate between the models, with lower values of either indicating a better fit

(Quinn & Keough 2002).

I used an analysis of covariance (ANCOVA) to examine the effects of mean annual

rainfall (covariate) and seasonal rainfall (summer, uniform, winter) on an index of

potential opportunities for Callitris regeneration during summer, since 1950. This index

was derived at each site by assuming that Callitris regeneration is dependent on high

rainfall in summer, a well documented phenomenon (FCNSW 1988). Calculation of the

index involved a number of steps. Firstly, within the estimated time of establishment for

each radiocarbon dated sample (Fig. 2.2), I chose the highest summer rainfall

(December-February). Secondly, I derived the minimum summer rainfall required for

regeneration by selecting the lowest rainfall figure from this dataset of high summer

rainfall. Thirdly, I counted the frequency of this minimum value for each site from 1950

to 2008, and from this was able to construct an index of potential opportunities for

Callitris regeneration along each seasonal rainfall transect. Lastly an analysis of

variance (ANOVA) was used to examine the effects of seasonal rainfall (summer,

winter, uniform) and annual average rainfall (< 405 mm, >405 mm) on the index of

potential opportunities for Callitris regeneration. In both ANCOVA and ANOVA, data


29

were square root transformed to satisfy homogeneity of variances. Post hoc pairwise

comparisons used Boniferoni correction.

I used non-parametric proportion tests to examine the proportional occurrence of

disturbances, namely farms, rabbits, sheep, cattle, goats, and kangaroos with annual

rainfall (< 405 mm, > 405 mm) and seasonal rainfall zone (winter, uniform, summer).

Fires were too infrequent to analyse. Post hoc tests used the Bonferroni correction.

Table 2.1 A description of the variables used in the regression tree analyses

Variable Values Assessed

Annual rainfall (mm yr-1

) 200-600 BOM (2006)

Seasonal rainfall winter, uniform, summer BOM (2010a)

Landuse fenced roadside, travelling tenure maps and

stock route, state forest observation

national park, farm and

unfenced roadside

Herbivores (rabbits, sheep, present / absent direct (sighting)

cattle, goats) indirect (scats, warrens,

diggings)

Fire since 1950 present / absent charred stems or stags of

Callitris that established

after 1950

Results

Radiocarbon dating and ring counting were used to estimate periods of establishment

for representative Callitris samples. The results indicate that whilst field recognition of

discrete cohorts within sites was reliable, recognition of the same age class between

sites was not always reliable. Cohort A was consistently dated from 1959-1964,

although it is likely that these established in the 1950s, since the samples were cut 20

cm above the ground (Fig. 2.2). Cohort B was considered older than C in the field,

however dating showed overlap between sites. Cohort B dated from 1972 to 1996 and

cohort C from 1988 to 2002. Ring counts supported these trends and indicated that

cohort D, which established from 2003 to 2006 was the youngest. In five of the six

cases where replicate samples were dated within a site, both replicates were from the


30

same recruitment year. On the one occasion where this did not occur, there was only

one year difference between the two replicates.

The first regression tree analysis indicated that cohort number was associated with

annual rainfall, seasonal rainfall (i.e. latitude) and land tenure (adjusted R2

= 0.65, Fig.

2.3a). The root or primary division in the regression tree occurred at 405 mm annual

rainfall. There were fewer cohorts below 405 mm annual rainfall, especially in the

winter rainfall zone compared with the summer rainfall zone (0.15 cf 1.78). Above 405

mm annual rainfall, farms had fewer cohorts than less disturbed tenures, such as

roadsides, travelling stock routes, state forests and national parks (2.13 cf 3.28).

The second regression tree analysis, which excluded mean annual rainfall, indicated that

cohort number was associated with combinations of rabbit occurrence and land tenure

(adjusted R2=0.65, Fig. 2.3b). The number of cohorts was lowest on farms with rabbits

in the winter and uniform rainfall zones (0.00) and marginally higher in the summer and

uniform rainfall zones (0.89). Most of these sites occurred below 405 mm annual

rainfall. The highest number of cohorts occurred where there were no rabbits on less

disturbed tenures, namely roadsides, travelling stock routes, state forests and national

parks (3.25). Most of these sites occurred above 405 mm annual rainfall. Moderate

numbers of cohorts were on farms without rabbits (1.77) and less disturbed tenures with

rabbits (1.71). These sites were distributed equally above and below 405 mm annual

rainfall.

Curve fitting identified linear and non-linear trends in the number of cohorts with

annual rainfall, depending on the seasonal rainfall zone (Fig. 2.4). Abrupt declines in

the number of cohorts at approximately 400 mm annual rainfall in the winter and

uniform rainfall zones were described by a sigmoid and loess curve, respectively, based

on low values for the residual sum of squares and AIC (Table 2.2). In the summer

rainfall zone, either a linear or a loess curve best describes the relatively unchanging

relationship between cohort number and annual rainfall, depending on whether AIC or

residual sum of squares is used (Table 2.2).


31

1940

1950

1960

1970

1980

1990

2000

2010

200 300 400 500 600

winter

uniform

summer R1

summer R2

La Nina

Mean annual rainfall (mm)

Recru

itm

ent

perio

d (year)

A

B

C

D

Figure 2.2 Results of radiocarbon dating of selected Callitris samples, showing recruitment

periods with mean annual rainfall in each seasonal rainfall zone (winter, uniform, summer

R1=replicate 1, R2=replicate 2). Cohorts identified in the field are indicated by A, B, C or D.

Cohort A is likely to be older than the dating indicates, as disks were taken at least 20 cm above

the ground. La Nina events are shown as grey shading.

|

Annual rainfall (mm)

>405

FarmsNon-

farms

1.78

9

3.28

18

0.15

21

<405

TenureLatitude

2.13

8

>33o <33o

Win

ter

/unif

orm

Sum

mer/

unif

orm

|

FarmsNon-

farms

0.89

9

7

2

3.25

16

2

14

0.00

11

11

0

Rabbits No Rabbits

FarmsNon-

farms

1.77

13

7

6

1.71

7

3

4

TenureTenure

Winter

/uniform

>33o

Summer/

uniform

<33o

(a) Rainfall included (b) Rainfall excluded

No. cohorts (mean)

No. sites

No. sites <405 mm

No. sites >405 mm

Figure 2.3 The results from regression tree analyses showing the factors associated with the

average number of cohorts since 1950, when mean annual rainfall was included (a) and

excluded (b). In (b) the number of sites < and > 405 mm mean annual rainfall is given below

each branch.


32


Num

ber

of

co

ho

rts

(a) Summer rainfall

(b) Uniform rainfall

(c) Winter rainfall

Figure 2.4 Curves of best fit for the number of cohorts since 1950 with mean average rainfall in

each seasonal rainfall zone.

An index of potential opportunities for Callitris regeneration (square root transformed)

during summer, since 1950, was significantly and positively associated with mean

annual rainfall (Table 2.3, Fig. 2.5) and significantly associated with seasonal rainfall

(summer, uniform, winter). Post-hoc pairwise tests indicated that the index was

significantly greater in summer than either the uniform or winter rainfall zones (mean =

5.18, sd = 1.19, mean = 3.53, sd = 0.99, mean=3.50, sd = 0.83, respectively), but not

significantly different between the winter and uniform rainfall zones.


33

Table 2.2 Results of fitting curves to the number of cohorts with mean annual rainfall in

each seasonal rainfall zone. Models are ordered from best to worst according to AIC.

Rainfall Curve Residual AIC

zone sum squares

Summer linear 32.96 15.56

loess 26.96 20.40

Uniform loess 5.42 -21.43

linear 29.53 9.75

sigmoid 7.40 -18.08

Winter sigmoid 1.83 -29.86

loess 2.20 -22.87

linear 9.24 -6.36

Table 2.3 Results of ANCOVA used to examine the effects of mean annual rainfall

(covariate) and seasonal rainfall (summer, uniform, winter) on an index of

potential opportunities for Callitris regeneration during summer, since 1950. The index

was square root transformed. Adjusted R2 = 0.925.

Source SS df MS F P

Corrected model 79.33 3 26.44 227.96 0.000

Intercept 0.17 1 0.17 1.42 0.238

Annual rainfall (mm yr-1

) 47.88 1 47.88 412.74 0.000

Seasonal rainfall (s,u,w) 12.79 2 6.39 55.11 0.000

Error 6.03 52 0.116

Total 978.00 56

Corrected total 85.37 55


34

0

2

4

6

8

200 250 300 350 400 450 500 550 600 650

winter

uniform

summer

Mean annual rainfall (mm yr-1)

Index o

f re

genera

tion p

ote

ntia

l (s

quare

root tr

ansf

orm

ed)

Figure 2.5 Linear trends in an index of potential opportunities for Callitris regeneration during

summer, since 1950 in different seasonal rainfall zones (summer, uniform, winter) with mean

annual rainfall.

An index of potential opportunities for Callitris regeneration in summer since 1950

(square root transformed), was significantly associated with mean annual rainfall (< 405

mm, > 405 mm) and seasonal rainfall (summer, uniform, winter), but not their

interaction (Table 2.4). There were fewer opportunities for regeneration below 405 mm

than above 405 mm mean annual rainfall (mean = 3.26, se = 0.34 vs mean = 4.43, se =

0.06 ). Opportunities were greater in the summer rainfall zone than in the uniform or

winter (mean = 4.93, se = 0.156; mean = 3.70, se = 0.13; mean = 3.62, se =0.15,

respectively). There was no significant difference in opportunities between winter and

uniform rainfall zones.

Proportion tests were used to examine the occurrence of disturbances with annual

rainfall and with seasonal rainfall (Table 2.5, Fig. 2.6). A significantly greater

proportion of sample sites contained farms, rabbits and sheep below 405 mm annual

rainfall than above 405 mm annual rainfall. The proportion of sites with farms, sheep,

cattle and kangaroos were not significantly different between the seasonal rainfall

zones. The proportion of sites with rabbits was significantly greater in the winter than

the summer rainfall zone and for goats was significantly greater in the uniform than the

winter rainfall zone.


35

Table 2.4 Results of ANOVA which examined the effects of seasonal timing of rainfall

(summer, uniform, winter) and annual average rainfall (< 405 mm, >405 mm) on the

index of potential opportunities for Callitris regeneration, since 1950. The index was

square root transformed.

Source SS df MS F P

Corrected model 67.38 5 13.48 37.47 0.000

Intercept 867.43 1 867.43 2411.90 0.000

Seasonal rainfall (s,u,w) 17.23 2 8.61 23.95 0.000

Annual rainfall (< or > 405) 35.04 1 35.04 97.42 0.000

Seasonal*Annual rainfall 0.78 2 0.39 1.08 0.348

Error 17.98 50 0.36

Total 978.00 56

Corrected total 85.37 55

Table 2.5 Results of proportions tests examining the occurrence of disturbances with

mean annual rainfall (below 405 mm, above 405 mm) and seasonal rainfall (winter W,

uniform U, summer S).

Disturbance Mean annual rainfall (mm) Seasonal rainfall zone

X2 df P X

2 df P

Farms 13.80 1 0.002 below>above 3.37 2 0.190

Rabbits 10.20 1 0.001 below>above 6.86 2 0.030

W>S

Sheep 3.91 1 0.048 below>above 0.41 2 0.810

Cattle 1.31 1 0.250 1.49 2 0.480

Goats 1.31 1 0.290 9.61 2 0.008 U>W

Kangaroos 0.04 1 0.840 0.11 2 0.950


36

Pro

po

rtio

n o

f sites


0.0

0.2

0.4

0.6

0.8

1.0Farms

0.0

0.2

0.4

0.6

0.8

1.0Rabbits

0.0

0.2

0.4

0.6

0.8

1.0

<405 >405

Sheep

Cattle

winter

uniform

summer

Kangaroos

0.0

0.2

0.4

0.6

0.8

1.0 Dunes

Goats

<405 >405

Fire

Figure 2.6 Proportion of sample sites containing farms, grazing species and dunes in

annual (< and > 405 mm) and seasonal rainfall zones (winter, uniform, summer). Means

and standard errors are given.


37

Discussion

These results indicate that demographic patterns of this widespread and long-lived tree

species varied greatly along rainfall and disturbance gradients. Ageing Callitris

populations were associated with low annual rainfall and high grazing levels, whilst

younger, expanding populations were associated with higher annual rainfall and low to

moderate grazing levels. Although these processes cannot be easily disentangled at the

survey scale, the results indicate the inextricable linking of climate and disturbance

regimes.

Annual rainfall

Demographic patterns varied widely across the species distribution. Recruitment failure

has led to ageing Callitris populations below 405 mm annual rainfall in the uniform and

winter rainfall zones. Today, these populations are largely represented by trees which

established in the 1850s or 1890s (Austin & Williams 1990). This is despite a number

of opportunities for recruitment since 1950, provided by favourable high rainfall during

summer, necessary for seedling survival (Fig. 2.5). Studies in this region (Auld & Keith

2009) and in other low rainfall systems worldwide have similarly reported ageing

populations of long-lived species (Garcia et al. 1999; Meyer & Pendleton 2005).

Elsewhere in the region where annual rainfall was higher, Callitris populations are

younger, similar to other semi-arid systems (Kraaij & Ward 2006; Graz 2008; Sankaran

et al. 2008; Boulant et al. 2009). These populations are consistently represented by trees

which established in the 1950s and then at various times through to 2007. A maximum

of four cohorts was recorded at any one site, but this is likely to be an underestimate,

especially where structural complexity was high. These trends are consistent with the

predictions that cohort numbers were more likely to increase with rainfall or

productivity.

Seasonal rainfall

Recruitment success was greater in the summer than either the winter or uniform

rainfall zones. Summer drought limits recruitment, especially in low rainfall systems, so

significant rainfall during this period can increase the recruitment success, by increasing

growth rates and reducing the probability of desiccation and subsequent damage by

grazers (Denham & Auld 2004; Castro et al. 2005; Holmgren et al. 2006b; Squeo et al.

2007). However, an index of opportunities for seedling recruitment during summer was


38

comparable between all seasonal rainfall zones below 405 mm mean annual rainfall.

This suggests that disturbances were responsible for the absence of Callitris recruitment

in the winter and uniform rainfall zones compared with some success in the summer

rainfall zone, below this threshold in annual rainfall. This is consistent with the

hypothesis that fewer cohorts were more likely at lower rainfall, especially where

disturbances were more frequent.Since long-lived species need few recruitment events

to maintain their populations (Watson et al. 1997), one recruitment event in 120 - 160

years may be the difference between local extinction or survival, given that the

longevity of Callitris is approximately 250 years (Bowman & Harris 1995).

Herbivory and fire

As hypothesised the distribution of disturbance regimes varied inextricably with rainfall

gradients. Below 405 mm annual rainfall, most of the sites were on farms which grazed

livestock, a common form of agriculture in lower rainfall systems (Walker 1993; Kraaij

& Ward 2006; Bureau of Rural Sciences 2009). Above 405 mm annual rainfall, a

broader range of land uses reflected greater agricultural flexibility with higher rainfall.

Since the study’s intention was to examine a site’s maximum potential for recruitment, I

was able to select tenures with lower levels of disturbance above 405 mm annual

rainfall. These land uses often included fenced roadsides, travelling stock routes, state

forests and national parks, all of which have historically lower levels of grazing than

farms, thus acting as refuges for native plants (Lunt 1995; Fensham 1998a). The

difference in grazing levels either side of 405 mm annual rainfall was reflected in the

higher occurrence of sheep and rabbits below 405 mm. Fires were almost absent above

405 mm annual rainfall, occurring in only 1 of 30 sites compared with 7 of 26 sites

below 405 mm annual rainfall. Noble (1997) believes there has been a decline in fire

frequency in this fragmented agricultural landscape since European management in the

1850s, suggesting that pre European fire regimes may have made establishment more

difficult for Callitris. These overall trends suggest higher levels of disturbances and or

fewer refuges from disturbances below 405 mm annual rainfall, since European

management.

The distribution of disturbances also varied with seasonal rainfall. Smaller herbivores,

like rabbits and sheep, prefer C3 grasses (Andrzejewska & Gyllenberg 1980; Van Dyne

et al. 1980; Heckathorn et al. 1999), which are more abundant in the winter rainfall


39

zone of southern Australia (Hattersley 1983; Murphy & Bowman 2007), whereas larger

herbivores, like cattle, prefer C4 grasses, which are more abundant in the summer

rainfall zone of northern Australia. Although I did not detect a pattern in the distribution

of cattle and sheep with seasonal rainfall, rabbits, whose successful breeding from

autumn to early summer is dependent on forage produced from winter rainfall, were

more common in the south than the north (Cooke 1977; Parer 1987). Goats were most

common in the uniform rainfall zone. Their generalist browsing diet (Landsberg & Stol

1996), may give them a greater tolerance to drought than sheep. Kangaroos were

ubiquitous. These trends suggest that at least rabbits are distributed in relation to food

availability, which is largely determined by seasonal rainfall.

Relationships between annual rainfall and grazing regimes were closely associated with

trends in recruitment success. In particular, the distributions of land tenure and rabbits

were associated with recruitment success across the species range. Callitris recruitment

was lowest below 405 mm annual rainfall, where rabbits occurred on farms which

grazed mainly sheep, but also goats and cattle. In other studies of low rainfall systems,

recruitment failure of woody species is related to a combination of drought and high

levels of grazing (Holmgren et al. 2006b; Gutierrez et al. 2007; Auld & Keith 2009).

While I cannot easily disentangle grazing from climatic effects in the survey, numerous

studies have illustrated the negative effects of grazing on recruitment in low rainfall

systems (see Holmgren et al. 2006b; Auld & Keith 2009). In particular, introduced

rabbits and livestock have reduced or prevented the regeneration of a range of palatable

woody species, in the arid zones of Australia (see Auld & Keith 2009) and elsewhere in

the world (Holmgren et al. 2006b; Gutierrez et al. 2007). From 1890 to 1950, Callitris

recruitment was prevented by high numbers of rabbits in SE Australia (Austin &

Williams 1988; Noble 1997). Widespread recruitment resumed in the 1950s following

widespread above average rainfall (La Nina) and the release of a virus which reduced

rabbit numbers (Austin & Williams 1988; Noble 1997). Trends in recruitment in my

study agree with this. However, the infrequency of recruitment below 405 mm annual

rainfall in the winter and uniform rainfall zones, suggests, that at least in some

locations, rabbit numbers did not decline to levels allowing recruitment. Indeed, the

lowest recruitment occurred in the winter rainfall zone where rabbits and Callitris often

coexisted on dunes. Warrens, which are necessary for the reproductive success of


40

rabbits in arid zones, are easily constructed in the sandy soils of dunes (Parer 1987).

Higher recruitment in the uniform and summer rainfall zones where rabbits were absent,

suggests patchily distributed livestock grazing (Holmgren et al. 2006a, 2006b; Gutierrez

et al. 2007), either from chance, or management or a lag in return to pre-drought levels,

since wet conditions usually follow drought (Meserve et al. 2003; Letnic et al. 2005).

These trends suggest that a combination of high levels of grazing, typical of rangelands

and rainfall outside the growing season has led to recruitment failure since the mid to

late 1800s. Exclusion of grazers is likely to increase the probability of recruitment

success (Holmgren et al. 2006b; Auld & Keith 2009).

Expanding populations were associated with lower levels of grazing. The highest

recruitment success was in less disturbed tenures with no rabbits. These tenures

included roadsides, travelling stock routes, state forests and national parks. Callitris is

sensitive to grazing, especially by rabbits and livestock (Lacey 1973; Allcock & Hik

2004). The recruitment successes of other woody species have been similarly associated

with lower rather than higher levels of grazing (Lunt 2005; Boulant et al. 2009), but not

in all cases (Richardson & Bond 1991; Brown & Archer 1999). Callitris is also fire

sensitive, since it is often killed by fire and must regenerate from seed. A decline in fire

frequency since European management (Noble 1997) has no doubt contributed to

recruitment success of Callitris, as it has with fire sensitive species in other systems

(Cooper 1960; Belsky & Blumenthal 1997). This contrasts markedly with ageing and

contracting populations of the fire sensitive Callitris intratropica in northern Australia,

where fires have increased in frequency and intensity since European management

(Bowman & Panton 1993). The relative absence of disturbances in certain tenures above

405 mm annual rainfall, is undoubtedly advantageous to the recruitment of the slow

growing and shade tolerant Callitris.

Climate change

Climate change is likely to have both direct and indirect effects on recruitment and thus

the future distributions of Callitris. The predicted increases in frequency, areal extent

and severity of droughts in SE Australia (Nicholls 2004; CSIRO & BOM 2007;

Hennessy et al. 2008), is likely to most severely affect arid and semi-arid regions

(Holmgren et al. 2006a). These conditions may further reduce opportunities for

recruitment below or immediately above 405 mm annual rainfall, if there is an easterly


41

shift in this isohyet, by increasing the chance of desiccation, identified by other studies

as contributing to recruitment failure in Callitris (Johnson 1969; Read 1995). It may

also accelerate senescence of ageing Callitris, removing the only remaining seed

sources in the south west (Foden et al. 2007; Jump et al. 2009). Indirectly, changes or

shifts in key disturbances, such as land tenure and grazing species, might also influence

the distribution of Callitris, as in other woody species (Loehle & Le Blanc 1996;

Turner, et al. 1998; Dale et al. 2001). If farms which graze livestock expanded east to

displace unprofitable crop lands (Dale 1997; Garnaut 2010), recruitment failure may

shift in the same direction, especially if refuges like travelling stock routes disappear.

Predicted declines in the populations of rabbits and livestock as pasture biomass

declines with higher temperatures (Scanlan et al. 2006; Gunasekera et al. 2008) may not

be enough to compensate the direct effects of drought on Callitris seedling recruitment

in an expanding low rainfall zone.

Summary

Demographic patterns in Callitris, a widespread and long-lived tree species were closely

associated with the spatial arrangement of disturbance regimes and their relationship

with rainfall along rainfall gradients. Disturbance regimes were associated with rainfall

through different land use types. As climate changes, rainfall, land use and disturbance

regimes will shift along climatic gradients, influencing species distributions. To predict

changes in the distribution of Callitris and other species, we need to predict shifts in

both disturbance regimes and climate, and disentangle their relative importance along

climatic gradients.

Acknowledgments

Australian Research Council (ARC) Linkage Grant between Charles Sturt University

(CSU) and NSW Department of Environment and Climate Change (DECC) provided

the funding for this research (Project number LP066829). A grant from Australian

Institute for Nuclear Science and Energy (AINSE) provided funding for the radiocarbon

dating of Callitris wood samples (AINGRA09075). I would like to thank Stephen

Campbell and Alan Smith from State Forests NSW for their assistance and permission

to take samples. I would also like to thank John Warren and Kate Hammill for

assistance in the field and Simon McDonald for statistical advice.


42

Chapter 3. Competitive feedback between canopy trees and regeneration

43

Chapter 3

Competitive feedback between Eucalyptus species and Callitris

glaucophylla trees and C. glaucophylla regeneration influences stand

dynamics along a rainfall gradient

Eucalyptus species canopy tree (left background) surrounded by dense Callitris glaucophylla regeneration in central western New South Wales (Source: K. Ross ).


44

Chapter 3

Competitive feedback between Eucalyptus species and Callitris

glaucophylla trees and C. glaucophylla regeneration influences stand

dynamics along a rainfall gradient

Abstract

Little attention has been given to the influence of canopy tree competition on the

encroachment of woody species in semi-arid woodlands of SE Australia. I examined

how canopies of Eucalyptus species and Callitris glaucophylla influenced the survival,

growth and reproduction of C. glaucophylla saplings and how these in turn influenced

the survival of the canopy trees along a rainfall gradient during severe drought

conditions. I sampled in six forests along a rainfall gradient (400 - 600 mm). In each

forest I sampled saplings under canopies of Eucalyptus and Callitris and in adjacent

gaps (4-6 replicates per canopy species). Canopy tree survival was assessed along two

transects in each forest, one with and one without regeneration. Saplings under canopies

had lower densities, were smaller and were less likely to be reproducing than those in

gaps. Sapling densities under canopies decreased from 1.5 to 1.3 m-2

as annual rainfall

increased from 400 to 600 mm. Mortality was low for canopy trees not surrounded by

Callitris regeneration (0 – 10%), but Callitris and Eucalyptus canopy trees surrounded

by Callitris regeneration had higher mortality and stress levels, respectively. Callitris

canopy tree mortality was relatively uniform across the rainfall gradient (0 – 25%) and

the proportion of Eucalyptus canopy trees with higher stress levels varied from 0.95 to

0.30 as annual rainfall increased from 400 to 600 mm. These results suggest that

canopies do not exclude but rather moderate Callitris sapling densities, which in turn

increase the mortality and stress levels of Callitris and Eucalyptus canopy trees during

drought. As drought severity increases under climate change, and ageing Eucalyptus

canopy trees die and are not replaced, the system may shift further from Eucalyptus and

Callitris co-dominance, typical of woodlands prior to European management to a

simplified Callitris dominated system.


45

Introduction

Woody plant encroachment is a major issue in semi-arid regions throughout the world

(Scholes & Archer 1997; Asner et al. 2004). Mechanisms responsible for its occurrence

are still poorly understood, but have variously been attributed to increased CO2, and

changes in climate, soils, grazing and fire regimes (Bond & Midgley 2000; Sankaran et

al. 2005; Fensham et al. 2005). The most commonly cited mechanism is a decline in fire

frequency and intensity, resulting from reductions in grass fuels by introduced grazers

(Noble 1997; Scholes & Archer 1997; Asner et al. 2004). In particular, this has

contributed to the encroachment of fire-sensitive woody plants (Belsky & Blumenthal

1997; Burrows 2005).

The effects of existing tree canopies on woody plant encroachment are less easily

predicted from the ecological literature. Patterns vary from establishment beneath

canopies or in gaps or continuous filling from gap to canopy (Harrington et al. 1981;

Haase et al. 1996; Belsky & Blumenthal 1997; Barnes & Archer 1999). The success of

woody species establishment under canopies is partly dependent on their requirements

for water, nutrients and light and how these are modified by canopy tree species along

rainfall gradients (Veblen 1992; Bertness & Callaway 1994). Most evidence for positive

effects of canopies on woody establishment is at lower rainfall, where canopies improve

micro-climatic conditions (e.g. temperature and water), thereby facilitating

establishment (Holmgren et al. 1997).

The influence of woody encroachment on canopy plants is generally negative (Callaway

& Walker 1997; Belsky & Blumenthal 1997; Barnes & Archer 1999). The increased

competition can accelerate mortality of canopy plants by reducing their vigour and

increasing their vulnerability to insect attack and pathogens, especially during periods of

water deficit (Belsky & Blumenthal 1997; Barnes & Archer 1999). Increased

competition can also reduce opportunities for establishment of canopy plants by

reducing their seed production or making conditions unsuitable for their establishment

by pre-empting resources (McPherson et al. 1988; Barnes & Archer 1999) or changing

fire regimes (Covington & Moore 1994). So, strong inter-specific competition has the

potential to lead to major changes in species composition (Belsky & Blumenthal 1997,

Barnes & Archer 1999).


46

Climate, especially rainfall influences the demography of woody plants in semi-arid

systems (Schwinning et al. 2004; Fensham et al. 2005; Holmgren et al. 2006a;

Shinneman & Baker 2009). Woody plants establish during periods of above average

rainfall (Schwinning & Sala 2004; Fensham et al. 2005; Kraaij & Ward 2006) and have

increased rates of mortality during droughts (Holmgren et al. 2006a; Fensham et al.

2009; Shinneman & Baker 2009). As droughts are predicted to increase in severity in

many semi-arid systems (IPCC 2007), there may be fewer opportunities for woody

species establishment (Garcia et al. 1999; Castro et al. 2005) and accelerated rates of

thinning (Fensham et al. 2009). Given that droughts preferentially thin at higher levels

of competition (Belsky & Blumenthal 1997; Guarin & Taylor 2005; Fensham et al.

2009), dense woody regeneration and ageing canopy trees that overtop them are likely

to suffer higher levels of mortality than those at lower levels of competition (Belsky &

Blumenthal 1997).

Understanding how climate influences competitive interactions between canopy trees

and regeneration are vital for predicting future changes in semi-arid systems, especially

under climate change. Within this system, competitive interactions may vary along a

rainfall gradient. For example, canopies more commonly facilitate seedling

establishment at lower than higher rainfall (Holmgren et al. 1997). Canopy tree

densities are likely to decline with declining rainfall (Specht & Specht 1999) or may be

influenced by disturbances, like logging (Lunt et al. 2006). Drought may also have a

greater negative effect on plant survival at lower than higher rainfall (Lacey 1972). So,

the combined effects of competition and drought may variably drive plant dynamics

along a rainfall gradient.

Semi-arid SE Australia underwent dense encroachment by a range of woody species,

including Callitris glaucophylla, after European settlement in the 1800s (Hodgkinson &

Harrington 1985; Noble 1997). This encroachment coincided with widespread changes

in fire and grazing regimes and canopy tree removal (Noble 1997). In many places

woodland dominants, C. glaucophylla and a range of Eucalyptus species were both

removed or ringbarked at different rates, shifting dominance from Eucalyptus species to

C. glaucophylla (Lunt et al. 2006). Although the dynamics of monospecific stands of C.

glaucophylla have been well researched at the stand scale (Lacey 1972, 1973; FCNSW

1988), there is comparatively little research on dynamics within mixed stands of


47

Eucalyptus species and C. glaucophylla (Clayton-Greene 1981; Clayton-Greene &

Ashton 1990).

Eucalyptus species are fast growing, resprouting angiosperms with large open canopies,

whilst C. glaucophylla is a slow growing, obligate-seeding gymnosperm with a

relatively small dense canopy (Lacey 1973; FCNSW 1988; Bowman & Harris 1995;

Williams & Woinarski 1997). Both genera are considered drought tolerant, and achieve

low water usage by different means (Zeppel & Eamus 2008). Callitris glaucophylla

reduce canopy conductance, whereas Eucalyptus species reduce leaf area by shedding

leaves before recovery by resprouting (Pook et al. 1966; Attiwill & Clayton-Greene

1984; Clayton-Greene & Ashton 1990). During drought, mortality of a range of

Eucalyptus species and C. glaucophylla has been greatest at higher levels of

competition and in drier areas (Lacey 1972, 1973; Allen 1998; Fensham & Holman

1999; Fensham et al. 2009). These patterns have only rarely been quantified (e.g.

Fensham & Holman 1999; Fensham et al. 2005; Fensham et al. 2009). Quantification

will allow better predictions of woody dynamics as climate changes in these semi-arid

systems.

My aim was to examine how canopy trees of Eucalyptus species and C. glaucophylla

influenced the survival, growth and reproduction of C. glaucophylla saplings and how

these in turn influenced the survival of the canopy trees along a rainfall gradient during

drought conditions. The following hypotheses were proposed: (1) densities of saplings

will be greater in gaps than under canopies, and as rainfall increases, densities will

increase in gaps and decrease under canopies; (2) growth rates and the probability of

reproducing will be greater for saplings without canopy or sapling competition and will

increase in areas of higher rainfall; (3) sapling densities, growth rates and the

probability of reproducing will be greater under Eucalyptus than Callitris canopies; (4)

denser saplings will experience greater mortality during a drought; (5) recent drought

mortality of Eucalyptus and Callitris canopy trees will be greater at lower rainfall when

dense sapling regeneration is present.

Methods

The study was restricted to the zone of coexistence of Callitris glaucophylla Thompson

& Johnson and a range of Eucalyptus species in semi-arid central New South Wales,

south-eastern Australia. The study area encompassed a 250 km long gradient of mean


48

annual rainfall, from 600 mm in the east (148.21oS, -33.34

oE) to 400 mm in the west

(145.57oS, -32.74

oE). Sampling occurred within six State Forests distributed along this

rainfall gradient. Fieldwork was undertaken from October to December 2008. A

drought which started in 2001 resulted in extremely dry years in 2002 and 2006 (CSIRO

2008; Fig. 3.1). Mean annual rainfall at each site from 2001 to 2008 was lower than the

long-term mean annual rainfall at each site: from west to east sites experienced 0.88,

0.76, 0.78, 0.88, 0.72 and 0.79 of their long-term mean annual rainfall (Fig. 3.1; Bureau

of Meteorology 2010).

Callitris saplings

Within each of the six sites, I randomly selected canopy trees of Callitris and

Eucalyptus that were surrounded by dense Callitris saplings (Fig. 3.2a). I defined two

zones, the canopy zone under the canopy and the gap zone which extended beyond the

edge of the canopy and was free from any overhead foliage. At each selected canopy

tree, I sampled Callitris saplings within a quadrat under the canopy and another in the

adjacent gap. Quadrats were 8 m2 and there were 4-6 replicates of each canopy tree

species at each site. I sampled the dominant cohorts, namely Callitris and Eucalyptus

canopy trees which established in the 1890s and Callitris saplings from the 1950s

(Lacey 1972; Allen 1998; Whipp 2009). Comparative sampling of Eucalyptus saplings

was not undertaken because of their infrequency in the region.

The density of live and recently dead Callitris saplings was recorded in each quadrat.

Recently dead Callitris saplings were identified by the presence of brown leaves or fine

twigs. To compare the intensity of competition between canopies and gaps, the height,

diameter at breast height (DBH) and the presence of fruiting cones on a range of live

Callitris sapling types were measured: closest to the trunk of canopy trees; random in

gaps (randomly throwing a stone); dominant (tallest) in gaps; and isolated from other

saplings or tree canopies by at least 1m in gaps. Isolated saplings were measured outside

the gap quadrat. Saplings closest to the trunk of the canopy tree approximated maximum

canopy competition and minimal sapling competition and are hereafter referred to as

saplings under canopies. I measured one of each sapling type at each canopy tree and

also recorded the height and canopy area of the latter.


49

0

200

400

600

800

1000

1200 (a) Yathong NR

0

200

400

600

800

1000

1200 (b) Yelkin SF

0

200

400

600

800

1000

1200 (c) Weelah SF

0

200

400

600

800

1000

1200 (d) East Cookeys Plains SF

0

200

400

600

800

1000

1200 (e) Weddin SF

0

200

400

600

800

1000

1200

1950

1953

1956

1959

1962

1965

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

2001

2004

2007

(f ) Backyamma SF

Year

Rain

fall

(mm

yr-1

)

Figure 3.1 Annual rainfall at the six sites from west to east (top to bottom) along the

rainfall gradient from 1950 to 2008 (BOM 2010). The period of recent drought from

2001 to 2008 is shaded grey. Long-term mean annual rainfall is represented by the

horizontal lines.


50

Dense

Callitris

saplings

Callitris

canopy tree

Eucalyptus

canopy tree

(a) Callitris sapling sampling in gaps

and under canopies

gap

gap canopy

canopy

Dense

Callitris

saplings

Callitris

canopy

tree

Eucalyptus

canopy tree

(b) Canopy tree sampling along transects without (lef t) and with dense

Callitris sapling regeneration (right)

Figure 3.2 The design of (a) Callitris sapling sampling showing quadrats under

canopies and in gaps and (b) Callitris and Eucalyptus canopy tree sampling with and

without dense Callitris sapling regeneration along transects.

Canopy trees

Recent mortality of Eucalyptus and Callitris canopy trees was recorded along two

transects at each site, one with and one without dense Callitris regeneration (Fig. 3.2b).

Recently dead canopy trees were identified by the presence of brown leaves or fine

twigs and the absence of resprouting. Twenty Callitris and 20 Eucalyptus canopy trees

were sampled along each transect, the dimensions of which varied with tree density.

Leaf loss on Eucalyptus canopy trees can indicate water stress, so I measured the degree

of leaf loss on each Eucalyptus tree using four categories: 0-5, 6-20, 21-50, >50%. A


51

full canopy was considered to be one where twigs and smaller branches had the

potential to have leaves, so gaps in this potential were measured as a percentage loss.

Stem damage progresses from the apex to the base, so lower resprouting positions

indicate higher stress levels (Vesk & Westoby 2003). I recorded the lowest reprouting

position on each Eucalyptus canopy tree using four categories: base, trunk, branch,

none.


I examined how a range of variables influenced sapling density, height and DBH in

three separate analyses. Since measurements of saplings under each canopy and in the

adjacent gap were not independent, they required an analysis that could accommodate a

split design and also continuous (rainfall) and categorical variables. So, I used nested

analyses of variance and I partitioned the five degrees of freedom for site into linear,

quadratic and cubic effects to examine regressions between rainfall and the response

variables (Steel & Torrie 1960, Quinn & Keough 2002). Testing for linear, quadratic

and cubic effects allowed me to examine the shape of the relationship between rainfall

and the response variables. More specifically, I examined the effects of site (1-6), mean

annual rainfall, canopy species (Eucalyptus, Callitris) and canopy position (gap,

canopy) on the density of live Callitris saplings. I also examined the effects of site,

rainfall, canopy species and sapling type (under canopy, random in gap, dominant in

gap, isolated in gap) on the height and DBH of live Callitris saplings. Post hoc Fishers

LSD tests were used for paired comparisons. Diagnostic checks of these and subsequent

models included examination of the residuals for outliers and goodness of fit. Analyses

were undertaken in GenStat (GenStat 2009).

Linear regressions were used to examine the recent mortality of Callitris saplings in

gaps and under canopies. In gaps, I examined the effects of the total density of saplings

(live and recently dead) and mean annual rainfall on the density of recently dead

saplings. Under canopies, I examined the effects of the same variables and canopy tree

species (Callitris, Eucalyptus) on the density of recently dead Callitris saplings. The

density of recently dead saplings was log transformed to satisfy Levene’s test of

homogeneity and to improve normality. These tests and the following were undertaken

in SPLUS software v8 for Windows (Insightful Corp. 2007).


52

Wilcoxon signed-rank tests were used to examine differences in canopy area, height and

DBH between Callitris and Eucalyptus canopy trees. Data within sites were paired, to

take into account between site differences. Non-parametric testing was used because

data on canopy area and DBH could not be transformed to satisfy the assumption of

normal distribution, required for parametric testing (Quinn& Keough 2002).

Wilcoxon signed-rank tests (paired) were used to examine whether the proportion of (a)

Callitris that died, (b) Eucalyptus with leaf stress (>5% leaf loss), and (c) Eucalyptus

resprouting from the trunk, were influenced by the presence of dense Callitris sapling

regeneration. I then used Spearman’s rank correlation tests to examine relationships

between the proportion of (a-c) where Callitris sapling regeneration was present, and a

range of factors. These factors included mean annual rainfall, density of Callitris or

Eucalyptus canopy trees, and the mean density of Callitris saplings (live and recently

dead) under the respective canopies of Callitris and Eucalyptus at each site. There were

no analyses of Eucalyptus mortality, because only two deaths were recorded.

Results

Canopy effects on Callitris sapling density along the rainfall gradient

There was a significant linear relationship between Callitris sapling density and canopy

position. As predicted sapling densities were greater in gaps than under canopies and as

mean annual rainfall increased, the density of Callitris saplings increased in gaps and

decreased under canopies (Table 3.1, Fig. 3.3). A significant cubic relationship between

the Callitris sapling density and mean annual rainfall indicated site effects, when

saplings in gaps and under canopies were not differentiated. These regression results

agreed with the significant effects of site, canopy position and their interactions on

Callitris sapling density in the ANOVAs (Table 3.1). Sapling densities varied with site,

with canopy position and their interaction. Contrary to predictions, densities under

Callitris canopies were not significantly different from those under Eucalyptus

canopies.

Canopy effects on Callitris sapling growth and reproduction along the rainfall

gradient

The height and DBH of saplings was significantly associated with site, canopy species,

sapling type and their interactions, but not mean annual rainfall (Table 3.1). As

predicted, Callitris saplings in gaps had higher growth rates than those under canopies,


53

but contrary to predictions growth rates were not greater at higher rainfall. Callitris

saplings under canopies (Eucalyptus and Callitris combined) were similar in height to

random saplings in gaps, both being shorter than dominant saplings in gaps, which were

shorter than isolated saplings in gaps (Fig. 3.4a). Similarly, Callitris saplings under

canopies (Eucalyptus and Callitris combined) had bigger DBHs than random saplings in

gaps but smaller than dominant saplings in gaps which were smaller than isolated in

gaps (Fig. 3.4b). As predicted, Callitris saplings under Eucalyptus canopies were taller

and had larger DBHs than those under Callitris canopies (Figs 3.4a, 3.4b). Significant

cubic and quadratic relationships between mean annual rainfall and Callitris sapling

height and DBH indicated site effects, as did the significant cubic relationship between

height or DBH of sapling types and rainfall.

The proportion of Callitris saplings with fruiting cones varied with sapling type along

the rainfall gradient (Fig. 3.4c). As predicted Callitris saplings in gaps were more likely

to reproduce than those under canopies, but contrary to predictions this probability did

not appear to increase with rainfall. Whilst Callitris saplings under canopies and

random saplings in gaps rarely produced cones, dominant saplings in gaps were more

likely to produce cones at low rainfall. Most isolated Callitris saplings in gaps produced

cones except at two sites around 500 mm mean annual rainfall.

Callitris sapling mortality

As predicted the density of recently dead Callitris saplings was significantly associated

with the combined densities of live and recently dead Callitris saplings in both gaps and

under canopies (Table 3.2). The density of recently dead saplings increased as the

combined density increased (Figs 3.5a, 3.5b). The density of recently dead Callitris

saplings in gaps and under canopies was not significantly associated with annual

rainfall. This implies that the combined densities of saplings were not linearly related to

annual rainfall, but are thinning to densities which are (Fig. 3.3). The density of recently

dead Callitris saplings in gaps and under canopies was not significantly associated with

canopy species.


54

Mean annual rainfall (mm yr-1)

y = 0.0028x + 0.6475R² = 0.1362

y = -0.0018x + 2.2541R² = 0.0974

0

1

2

3

4

400 450 500 550 600

gap

canopy

Density (no

. m

-2)

Figure 3.3 Linear regressions of live Callitris sapling densities in gaps (solid lines) and

under canopies (dotted lines) with mean annual rainfall (95% CI). Each data point

shows mean sapling density at each site.


55

Table 3.1 Results of the partly nested analyses of variance with orthogonal

polynomials, used to examine the effects of a range of variables on (a) the density of

saplings, (b) sapling height and (c) sapling diameter at breast height (sapling location: c

closest to trunk under canopy; r random in gap; d dominant in gap; i isolated in gap).

Source of variation d.f. SS MS F P

(a) Sapling density

Tree stratum

Site 5 23.50 4.70 3.67 0.006

Linear 1 0.22 0.22 0.18 0.680

Quadratic 1 3.40 3.40 2.65 0.110

Cubic 1 6.48 6.48 5.07 0.030

Deviations 2 13.40 6.70 5.24 0.008

Species (Eucalypt, Callitris) 1 2.03 2.03 1.59 0.210

Site.Species 5 12.85 2.57 2.01 0.090

Residual 52 66.50 1.28 2.50

Tree.Position stratum

Position (gap, canopy) 1 19.16 19.16 37.48 <0.001

Position.Site 5 6.90 1.38 2.70 0.030

Position.Linear 1 3.45 3.45 6.74 0.012

Position.Quadratic 1 0.09 0.09 0.18 0.676

Position.Cubic 1 0.11 0.11 0.21 0.649

Deviations 2 3.26 1.63 3.19 0.050

Position.Species 1 0.05 0.05 0.10 0.756

Position.Site.Species 5 0.92 0.18 0.36 0.872

Residual 52 26.58 0.51

Total 127 158.50

(b) Sapling height

Tree stratum

Site 5 29.72 5.95 4.62 0.001

Linear 1 2.53 2.53 1.96 0.167

Quadratic 1 8.39 8.39 6.52 0.014

Cubic 1 7.36 7.36 5.72 0.020

Deviations 2 11.45 5.73 4.45 0.016


Site.Species 5 6.29 1.26 0.98 0.440

Residual 52 66.94 1.28 1.57


56

Table 3.1 cont.

Source of variation d.f. SS MS F P

(b) Sapling height cont.


Sapling location (c,r,d,i) 3 314.32 104.77 127.94 <0.001

Sapling location.Site 15 45.41 3.03 3.70 <0.001

Sapling location.Linear 3 2.00 0.67 0.81 0.490

Sapling location.Quad ratic 3 1.89 0.63 0.77 0.510

Sapling location.Cubic 3 24.14 8.05 9.83 <0.001

Deviations 6 17.37 2.89 3.54 0.003

Sapling location.Species 3 7.47 2.49 3.04 0.031

Sapling location.Site.Species 15 21.99 1.47 1.79 0.041

Residual 149 122.02 0.82

Total 248 604.31

(c) Sapling diameter at breast height

Tree stratum

Site 5 5.81 1.16 6.40 <0.001

Linear 1 0.12 0.12 0.67 0.416

Quadratic 1 2.46 2.46 13.58 <0.001

Cubic 1 2.63 2.63 14.51 <0.001

Deviations 2 0.59 0.30 1.63 0.206


Site.Species 5 1.02 0.20 1.12 0.361

Residual 52 9.43 0.18 1.83


Sapling location (c,r,d,i) 3 60.67 20.22 203.99 <0.001

Sapling location.Site 15 5.37 0.36 3.610 <0.001

Sapling location.Linear 3 0.65 0.22 2.18 0.093

Sapling location.Quad ratic 3 0.32 0.11 1.09 0.357

Sapling location.Cubic 3 1.36 0.46 4.59 0.004

Deviations 6 3.04 0.51 5.10 <0.001

Sapling location.Site.Species 15 1.53 0.10 1.03 0.430

Residual 149 14.77 0.10

Total 248 97.178


57

Pro

po

rtio

n w

ith c

ones


0.0

0.2

0.4

0.6

0.8

1.0

400 450 500 550 600

(c) Saplings with cones

Isolated

Dominant

Canopy (both spp.)

Random

0

1

2

3

4

5

6

7(a) Sapling height

0

2

4

6

8

10

12

14

Random Callitris canopy

Eucalyptus canopy

Dominant Isolated

(b) Sapling dbh

Sapling type

Heig

ht (m

)D

BH

(cm

)

Figure 3.4 (a) Height, (b) diameter at breast height (DBH) and (c) fruiting success of

Callitris sapling types: under Callitris or Eucalyptus canopies or random, dominant and

isolated in gaps (mean and standard error).


58

Table 3.2 Results of regression analyses of the density of recently dead Callitris

saplings and the combined density of saplings (alive and recently dead) in gaps and

under canopies.

Variable Value SE t P R2

Gap

Constant 0.10 0.07 1.36 0.178 0.22

Sapling density 0.10 0.02 4.15 0.0001

Canopy

Constant 0.01 0.04 0.16 0.87 0.41

Sapling density 0.14 0.02 5.52 <0.001

0.0

0.2

0.4

0.6

0.8

1.0

1.2 (a) Gap

Ln

(density+

1)

of

recently

dead

(no

.m-2

)

Density of live and recently dead (no. m-2)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 1 2 3 4 5 6 7

(b) Canopy

Figure 3.5 Linear regressions of densities of recently dead Callitris saplings with

combined sapling density (alive and recently dead) in (a) gaps and under (b) canopies

(95% CI).


59

Effects of sapling regeneration on canopy trees along the rainfall gradient

Eucalyptus canopy trees had 3-4 times larger canopy areas, were taller and had larger

DBHs than coexisting Callitris canopy trees (Z = -4.69, P < 0.001; Z = -4.00; P <

0.0001; Z = -4.69, P < 0.0001; Fig. 3.6).

Contrary to predictions recent mortality of Callitris canopy trees was relatively

consistent along the rainfall gradient, but was significantly and negatively associated

with the presence of Callitris sapling regeneration, which was predicted (Z = 2.02, P <

0.04; Fig. 3.7a). For Callitris canopy trees surrounded by regeneration, mortality was

not significantly correlated with mean annual rainfall (Z = -0.27, P < 0.79, r = -0.09),

density of Callitris or Eucalyptus canopy trees (Z = -0.54, P < 0.59, r = -0.21; Z = -1.36,

P < 0.17, r = 0.64), or the mean density of live and recently dead Callitris sapling

regeneration (Z = 1.09, P < 0.28, r = 0.52).

Eucalyptus canopy tree mortality was low and did not respond negatively to the

presence of Callitris sapling regeneration or positively to increasing rainfall, as

predicted by the hypotheses. Only two of the sampled 240 Eucalyptus canopy trees died

during the recent drought and these were not surrounded by Callitris sapling

regeneration. However, Eucalyptus canopy trees suffered leaf stress, defined as > 5%

leaf loss, which mostly ranged from 6 – 20%. The proportion of Eucalyptus canopy

trees with leaf stress was not significantly associated with the presence of Callitris

sapling regeneration (Z = 1.83, P < 0.07). This result is close to significance at P < 0.05

level, indicating some ambiguity. At the lower end of the rainfall gradient, Eucalyptus

surrounded by regeneration are more likely to be stressed than those that are not (Fig.

3.7b). For Eucalyptus canopy trees surrounded by regeneration the proportion of trees

with leaf stress was not significantly correlated with the densities of Callitris or

Eucalyptus canopy trees, the mean density of sapling regeneration or rainfall (Z = 1.23,

P < 0.22, r = 0.58; Z = -0.84, P < 0.40, r = -0.35; Z = -0.78, P < 0.43; r = -0.32; Z = -

1.49, P < 0.14; r = -0.64). These results indicate that at lower mean annual rainfall,

higher leaf stress in Eucalyptus canopy trees may be associated with the presence of

Callitris sapling regeneration.

The proportion of Eucalyptus canopy trees resprouting from the trunk was not

significantly affected by the presence of Callitris sapling regeneration (Z = 0.11, P <

0.91; Fig. 3.7c). Among Eucalyptus canopy trees surrounded by regeneration, the


60

proportion resprouting from trunk was significantly and positively correlated with

Callitris canopy tree density (Z = 2.17, P < 0.03, r = 1.00; Fig. 3.7d), but not

Eucalyptus canopy tree density, the mean density of sapling regeneration or rainfall (Z

= 1.02, P < 0.31, r = 0.49, Z = -0.51, P < 0.61, r = 0.26, Z = -0.89, P < 0.37, r = -0.37).

0

50

100

150 (a)

Heig

ht (m

)

0

5

10

15

20

25 (b)

Cano

py a

rea (

m2)

0

20

40

60

80

100

Callitris Eucalyptus

(c)

DB

H (cm

)

Figure 3.6 (a) Canopy area, (b) height and (c) diameter at breast height (DBH) of

Callitris and Eucalyptus canopy trees.


61

0

50

100

150

200

250

400 450 500 550 600

(d) Density of Callitris & Eucalyptus

Callitris (no regeneration)

Callitris (regeneration)

Eucalyptus (regeneration)

Eucalyptus (no regeneration)

0.0

0.2

0.4

0.6

0.8

1.0 (c) Eucalyptus resprouting f rom trunk

0.0

0.2

0.4

0.6

0.8

1.0(b) Eucalyptus with leaf stress

Eucalyptus (regeneration)

Eucalyptus (no regeneration)


0.0

0.2

0.4

0.6

0.8

1.0(a) Callitris

Callitris (regeneration)

Callitris (no regeneration)P

rop

ort

ion d

ead

Pro

po

rtio

nD

en

sity

(no

. ha

-1)

Figure 3.7 Trends in mortality and stress levels of canopy trees with and without

Callitris sapling regeneration along the rainfall gradient: (a) Callitris mortality; (b)

Eucalyptus with leaf stress (>5% leaf loss); (c) Eucalyptus resprouting from the trunk;

and (d) the density of Callitris and Eucalyptus canopy trees.


62

Discussion

The results indicate feedback between Callitris saplings and Eucalyptus and Callitris

canopy tree dynamics, which varied along the rainfall gradient. Saplings were less

dense, smaller and less likely to be reproducing under canopies than in the gaps, with

canopies reducing sapling density more at the higher end of the rainfall gradient.

Drought preferentially thinned saplings at higher density. In turn, the combined effects

of drought and sapling regeneration reduced the resilience of canopy trees. Callitris

canopy trees surrounded by dense sapling regeneration suffered higher mortality than

similarly placed Eucalyptus canopy trees, which suffered greater levels of stress at the

lower end of the rainfall gradient. These trends highlight how interactions among

climate, canopies and regeneration influence stand dynamics in this semi-arid system.

Canopy effects on Callitris sapling density along the rainfall gradient

As hypothesised, densities of Callitris saplings were lower under canopies than in gaps,

but the magnitude of this difference was relatively small and changed along the rainfall

gradient. Difference in sapling densities in gaps and under canopies were greater at

higher than lower rainfall. This implies increasing canopy competition as rainfall

increases. At high rainfall, competition for resources under canopies may have lowered

sapling survival more than in adjacent gaps (Fuentes et al. 1984; Belsky 1994). By

comparison, at the low end of the rainfall gradient, competitive effects of trees may

have been reduced by positive effects of canopies, such as increased humidity and

nutrients or reduced temperatures, wind and herbivory (Barnes & Archer 1999;

Pugnaire & Luque 2001; Bruno et al. 2003; Arrieta & Suarrez 2005). If canopies did

improve conditions for Callitris saplings at low rainfall, it was not enough to sustain

higher densities than in adjacent gaps. Thus the net effect of canopy trees on

regeneration was competitive rather than facultative.

Canopy effects on Callitris sapling growth and reproduction along the rainfall

gradient

As predicted growth rates of Callitris saplings were slower under canopies than in the

gaps, where there was also a greater probability of fruiting . Saplings under canopies

were short and without fruits, as were the majority of saplings surrounded by dense

sapling regeneration in gaps. Slow growth rates and negligible seed production of

similarly aged Callitris saplings in dense stands, have previously been recorded (Lacey


63

1972; Hawkins 1966; FCNSW 1988). By comparison, growth rates and reproductive

success were significantly greater for the less frequent but taller or isolated saplings in

gaps. These growth rate differences suggest a longer period of vulnerability for Callitris

saplings under canopies than in gaps. Slow development, leaves non-resprouters, like

Callitris, particularly vulnerable to disturbances like fire, since their survival is

dependent on seeds or thick bark which increases with plant size (Bowman et al. 1988;

Bond 1989; Bowman & Harris 1995). If fire penetrates the stand, the higher

flammability of litter under Eucalyptus compared with Callitris canopy trees (Scarff &

Westoby 2006) heightens this vulnerability (Clayton-Greene 1981).

Density, growth and reproduction of Callitris saplings under Callitris and Eucalyptus

canopies

Contrary to predictions, densities of Callitris saplings were similar beneath Eucalyptus

and Callitris canopies. This trend suggests comparable levels of competition by the

different canopy species on Callitris saplings. Indeed other studies have found that

during drought Eucalyptus and Callitris trees use similar amounts of water (Zeppel &

Eamus 2004).

In contrast and as hypothesised, growth rates of Callitris saplings under Eucalyptus

canopy trees were faster than those under Callitris canopy trees. This may be related to

higher nutrient litter under Eucalyptus because of higher leaf turnover or greater light

infiltration through the more open Eucalyptus canopies (Clayton-Greene 1981; Bowman

& Wilson 1988). Other studies have shown that nutrient levels and light availability can

influence the performance plants in semi-arid systems (Holmgren et al. 1997; Armas &

Pugnaire 2005). Further research is needed to determine the mechanisms underlying

trends in the density and growth of Callitris saplings under Eucalyptus and Callitris

canopy trees. Reproductive success of Callitris saplings under canopy trees was so low,

that comparison between Eucalyptus and Callitris canopy trees was not made.

Callitris sapling mortality

Mortality of Callitris saplings was related to sapling density during the drought. As

hypothesised, the density of recently dead Callitris saplings increased as the density of

saplings increased, resulting in 10 to 30% mortality. Density dependent mortality of

woody species, including Callitris has previously been observed during droughts (Lacey

1973; FCNSW 1988; Belsky & Blumenthal 1997). Drought is especially important as a


64

‘thinning agent’ for species with slow rates of self-thinning, like Callitris (Lacey 1972,

1973; FCNSW 1988), when other disturbances, like fire are absent (Lacey 1973;

Bowman et al. 1988; Fule & Covington 1998).

Effects of sapling regeneration on canopy trees along the rainfall gradient

Despite their ability to similarly reduce water usage during periods of water deficit

(Zeppel & Eamus 2005), Callitris canopy trees suffered higher mortality than

Eucalyptus canopy trees during the drought. As hypothesised, the mortality of Callitris

canopy trees not surrounded by sapling regeneration was lower (0 - 10%), than those

surrounded by sapling regeneration (0 - 25%). Dense Callitris regeneration has

promoted the death of both Callitris and Eucalyptus canopy trees in previous droughts,

but has not been quantified (Lacey 1973; FCNSW 1988; Allen 1998). By comparison,

the mortality of Eucalyptus canopy trees was low (<2%), irrespective of sapling

regeneration presence. In previous droughts, higher levels of mortality of a range

Eucalyptus species have been recorded in northern Australian savannas (Fensham

1998b; Fensham & Holman 1999; Fensham et al. 2005; Fensham et al. 2009).

Although Callitris canopy tree mortality was not greater at lower rainfall, as

hypothesised, a greater proportion of Eucalyptus trees showed signs of stress at lower

rainfall. These stressed Eucalyptus trees had greater leaf loss than and were resprouting

from the trunk. These stress indicators were associated with the presence of dense

sapling regeneration and higher Callitris canopy density, respectively. These Eucalyptus

canopy trees may be predisposed to mortality during future droughts, if their resprouting

ability becomes compromised (Attiwill & Clayton-Greene 1984). Indeed, it is possible

that this ability contributed to the higher survival of Eucalyptus than Callitris canopy

trees during this latest drought.

The effects of predicted changes in climate

Whilst Callitris sapling regeneration was common, Eucalyptus sapling regeneration was

infrequent in this semi-arid system. Eucalyptus species most commonly regenerate after

fires (Gill 1997) and since European management, fire frequency and intensity have

declined, often to the extent of exclusion (Noble 1997). Once dense patches of Callitris

become established, their low flammability can maintain this fire regime in low-

moderate, but not in extreme fire weather conditions (see Chapter 4). Despite predicted

increases in extreme fire-weather days (Lucas et al. 2007), fire frequency is unlikely to


65

increase because of predicted declines in herbaceous fuels, which limit fires in this

region, together with high levels of landscape fragmentation (Bradstock 2010).

In contrast, the predicted drought regime (Nicholls 2004; CSIRO & BOM 2007;

Hennessy et al. 2008), is likely to accentuate Callitris dominance, if the current fire

regime is maintained. As droughts increase in frequency and intensity, mortality rates of

Callitris sapling regeneration may increase. As gaps appear they are more likely to be

filled by Callitris than Eucalyptus seedlings, since Eucalyptus seedlings do not compete

well with dense Callitris regeneration (FCNSW 1988). Thus Callitris and Eucalyptus

sapling regeneration may be maintained at their current levels of dominance and

infrequency, respectively. As ageing canopy trees die, Eucalyptus are unlikely to be

replaced by conspecifics because of their infrequency. Consequently, the system may

shift further toward Callitris dominance, especially in the west where Eucalyptus

canopy trees were most stressed during this latest drought.

Conclusion

Interactions among climate, sapling regeneration and canopy trees influenced dynamics

in this semi-arid system. Feedback between sapling regeneration and canopy trees was

largely negative. Canopy trees moderated rather than excluded dense sapling

regeneration, allowing Callitris to persist at relatively high densities, 50 years after their

establishment. In turn, dense Callitris sapling regeneration increased mortality rates or

stress levels in canopy trees during drought. As drought severity increases and fires

continue to be excluded from these semi-arid systems, the fire-sensitive Callitris is

likely to maintain and perhaps increase its dominance.

Acknowledgments

An Australian Research Council (ARC) Linkage Grant between Charles Sturt

University (CSU) and NSW Department of Environment and Climate Change and

Water (DECCW NSW) provided the funding for this research (Project number

LP066829). I would like to thank Stephen Campbell and Alan Smith from State Forests

NSW for their assistance and Livia Tigwell, Andrew Cohn and John Warren for

assistance in the field. Terry Koen (DECCW NSW) and Simon McDonald (CSU)

assisted with statistical advice.


66

Chapter 4. Feedback between Callitris patches and fire intensity

67

Chapter 4

Feedback between patches of Callitris glaucophylla and fire intensity

influences Callitris survival in flammable Eucalyptus woodlands in

different fire-weather conditions

Resprouting Eucalyptus species following fire in the Eucalyptus matrix in extreme fire-weather conditions (top). A Callitris patch following fire in low to moderate fire-weather conditions (bottom). in Pilliga Conservation Area, north eastern NSW (Source: K. Ross, J.Cohn)


68

Chapter 4

Feedback between patches of Callitris glaucophylla and fire intensity

influences Callitris survival in flammable Eucalyptus woodlands in

different fire-weather conditions

Abstract

I examined how interactions between weather and vegetation influence spatial

heterogeneity in fire intensity, permitting the survival of fire-sensitive Callitris

glaucophylla within a matrix of highly flammable Eucalyptus species. The study was

undertaken in Pilliga State Forest, south eastern Australia. In areas burned under

extreme and low-moderate fire-weather conditions, fire intensity and survival of

Callitris were measured in the flammable Eucalyptus matrix and in Callitris dominated

patches. Pre-fire fuel characteristics were measured in an adjacent unburnt area.

Flammable fuel levels were lower in Callitris patches than in the matrix. In extreme

fire-weather conditions, Callitris patches did not reduce incoming fire intensity from the

matrix, since weather controlled fire intensity. However in low-moderate, fire-weather

conditions, Callitris patches reduced incoming fire intensity from the matrix, the

magnitude of which increased with increasing Callitris basal area. Survival of Callitris

was 4% and 54% in extreme and low-moderate fire-weather conditions, respectively.

Callitris survival was negatively related to char height and positively related to trunk

diameter. The relative importance of fuels and weather in determining fire intensity

varied, influencing Callitris survival. In low-moderate, fire-weather conditions, Callitris

survival was higher in patches dominated by Callitris, where flammable fuels and fire

intensities were lower than in the matrix. In extreme fire-weather conditions,

flammability differences were less important than weather, resulting in high fire

intensities in Callitris patches and very low Callitris survival. Thus, under low-

moderate, fire-weather conditions, patch structure created by the less flammable

obligate-seeding Callitris reduced fire intensity, enabling it to persist in a flammable

Eucalyptus matrix.


69

Introduction

In many plant communities, disturbance prevents competitive exclusion and promotes

species coexistence, as described by non-equilibrium theories of diversity (Bond & van

Wilgen 1996; Huston 2003). Coexistence of species of varied resilience and competitive

status may depend upon heterogeneity in disturbances, such as fire regimes (frequency,

intensity and season), across differing spatial and temporal scales (Cowling 1987;

Whelan 1995; Williams et al. 1994).

A fundamental question in fire ecology is – how do long-lived, obligate-seeders persist

in flammable landscapes (Bond & van Wilgen 1996)? Adaptations, such as, serotinous

cones and thick bark are major determinants of persistence (Agee 1998; Keeley &

Zedler 1998). Bark thickness is important for species that are weakly serotinous and

varies among species and increases with tree age (Bond & van Wilgen 1996; Keeley &

Zedler 1998). Interactions with fire intensity determine survival, with thicker barked

individuals surviving higher temperatures (Peterson & Arbaugh 1986).

Spatial variation in fire intensity is influenced by interactions among topography,

weather and fuels (Catchpole 2002; Boer et al. 2008). The relative importance of

weather and fuels has been debated but appears to vary greatly among ecosystems

(Bessie & Johnson 1995; Hely et al. 2001; Turner et al. 2003). In some ecosystems,

variations in fuel levels influence fire intensity under moderate weather conditions, but

fuels have little influence on fire intensity under extreme weather conditions (i.e. high

temperature and wind speed, and low humidity and soil moisture; Bessie & Johnson

1995; Turner et al. 2003). By contrast, in other ecosystems, variations in fuel levels

influence fire intensity across the entire range of weather conditions, including extreme

conditions (Hely et al. 2001).

Flammability varies greatly among vegetation types due to trait differences among

component species (Williamson & Black 1981; van Wilgen et al. 1990; Schwilk &

Ackerly 2001). Flammability is determined by a range of characteristics including fuel

type, biomass and spatial continuity. Low flammability may result from shedding of

lower branches and understorey suppression, both of which reduce the vertical

continuity of fuels, resulting in less damage to the canopy (Schwilk 2003; Knapp &

Keeley 2006). Leaf litter also affects stand flammability. Litter with a high surface-area

to volume ratio (e.g. needles) can pack more tightly, reducing aeration and increasing


70

moisture levels, thereby reducing flammability (Schwilk & Ackerly 2001; Scarff &

Westoby 2006). Patches of species with less flammable traits may therefore promote

their own survival and coexistence among competing flammable species by influencing

the spatial heterogeneity of fire intensity (Cowling 1987; van Wilgen et al. 1990;

Bowman 1998).

Whilst studies have examined how fire intensity affects the survival of long-lived, non-

serotinous, obligate-seeders (e.g. Rodrigo et al. 2004 2007), few have attempted to

examine survival in relation to the spatial variation of fuels in varying fire-weather

conditions (Bowman & Wilson 1988; Bowman & Panton 1993). No study appears to

have quantified interactions amongst flammability, fire intensity and obligate-seeder

survival at the patch scale under varying fire-weather conditions.

In Australian woodlands, flammable and resprouting Eucalyptus species often coexist

with less flammable, non-serotinous, obligate-seeding conifers of the genus Callitris

(Bowman & Harris 1995). Mixed forests dominated by these two genera provide an

ideal model system to examine two key questions that underlie the survival of long-

lived, obligate-seeders in flammable landscapes: (1) how do weather and vegetation

(fuel) patterns interact to influence spatial heterogeneity in fire intensity; and (2) to what

extent do these interactions permit the survival of patches of fire-sensitive species

(obligate-seeders) within a matrix dominated by highly flammable, fire-tolerant species

(resprouters)?

In this study I used this model system to investigate these two issues, by examining the

following hypotheses: (1) under extreme fire-weather conditions, patches of Callitris

will have little influence on the incoming fire intensity from the Eucalyptus matrix; (2)

under low-moderate fire-weather conditions, patches of Callitris of high basal area will

reduce the incoming fire intensity from the Eucalyptus matrix; (3) survival of Callitris

will be negatively related to fire intensity and positively related to tree size; and (4)

ground fuels will be of lower flammability in patches of Callitris than in the Eucalyptus

matrix, thereby reducing fire intensity in patches of Callitris.


71

Methods

Study area

The study was undertaken in the Pilliga Conservation Reserve and State Forest in north-

eastern New South Wales, south-eastern Australia. ‘The Pilliga’ holds the most

extensive stands of Callitris glaucophylla in Australia (270 000 ha; Thompson &

Eldridge 2005). The area experiences a quasi Mediterranean climate of warm, dry

summers and mild, moist winters (Gill & Moore 1990). The dominant trees include

non-serotinous, obligate-seeding C. glaucophylla J. Thomps. & L. Johnson and C.

endlicheri (Parl.) Bailey, and resprouting Eucalyptus crebra F. Muell., E. fibrosa

F.Muell., E. nubila Maiden & Blakely, and E. trachyphloia (F. Muell.) K.D. Hill &

L.A.S. Johnson (Binns & Beckers 2001). Dense patches of C. glaucophylla and C.

endlicheri, typically 0.01-0.25 km2 in size (hereafter ‘Callitris patches’) are scattered

throughout a more open matrix dominated by Eucalyptus with sparse Callitris (hereafter

the ‘Eucalyptus matrix’). Topography is flat to gently undulating.

On 29th

November 2006, during a lengthy drought (Australian Bureau of Meteorology

unpubl.), lightning ignited a number of fires in the Pilliga, one of which is the focus of

this study. Approximately 70 000 ha burnt in the first 24 hours in extreme fire-weather

conditions, and over the following 2 weeks, a further 27 000 ha burnt in low-moderate

fire-weather conditions (DECC NSW unpubl.). Fire-weather conditions are defined by

the Forest Fire Danger Index (FFDI), which is based on a drought factor and ambient air

temperature, wind speed and relative humidity (Noble et al. 1980). During extreme

conditions, the maximum temperature and wind speed reached 37oC and 57 km hr

-1, and

the minimum humidity was 12%. In the low-moderate fire-weather conditions

temperatures and wind speeds were generally lower and relative humidity higher and all

were more variable (Australian Bureau of Meteorology unpubl.).

Sampling was undertaken within the burnt area and data on pre-fire fuel characteristics

were collected from an adjacent unburnt area. Within the burnt area there were two fire-

weather zones, extreme and low-moderate. Sampling locations were determined from

1:50 000 scale aerial photos taken in 2003. Callitris patches were at least 1 ha and

dominated by C. glaucophylla and/or C. endlicheri, which have comparable architecture

and leaf characteristics inferring similar flammabilities (Harden 2000; Scarff &

Westoby 2006). Sampling occurred between September and November 2007 and was

restricted to areas of < 5o slope.


72

Fire intensity indicators and Callitris survival

Paired quadrats were placed in the direction of the fire, so that the incoming quadrat was

burnt before the outgoing quadrat (Fig. 4.1). Quadrat pairs were located across four

vegetation transitions: in the Eucalyptus matrix (both quadrats placed in the matrix);

across the Eucalyptus matrix – Callitris patch boundary (incoming quadrat in matrix

and outgoing in Callitris patch); in the centre of Callitris patches (both in Callitris

patch); and across the Callitris patch – Eucalyptus matrix boundary (incoming in

Callitris patch and outgoing in adjacent Eucalyptus matrix). Each transition pair was

replicated ten times in each fire-weather zone, giving 40 pairs in the extreme fire-

weather zone and 40 pairs in the low-moderate fire-weather zone. To minimise spatial

auto correlative effects, all paired locations were separated by at least 100 m.

Fire

E/E Ee/Ce

C/C

> 100 m

Eucalyptus matrix

Ce/Ee

Callitris patchin out

in outin out

in out

Fig. 4.1 Diagrammatic representation of the sampling design within the burned area.

Data were collected in quadrat pairs located across four vegetation transitions: in

Eucalyptus matrix (E/E); across Eucalyptus matrix – Callitris patch boundary (Ee/Ce;

‘e’ indicates an edge quadrat); in Callitris patches (C/C); and across the Callitris patch –

Eucalyptus matrix boundary (Ce/Ee). Paired quadrats were placed in the direction of the

fire, so that the incoming (in) quadrat was burnt before the outgoing (out) quadrat and

all pairs were separated by at least 100 m.

Fire direction was determined by a combination of means: (1) reports on prevailing

wind direction; (2) char heights were lower on the windward side of trees than the more

protected leeward side; and (3) shrubs often bent and remained fixed in the direction of

the fire.

Char height and post-fire stem tip diameter were measured as indicators of fire intensity

(Cain 1984; Moreno & Oechel 1989). Char height was measured as the maximum

height of charring on the bole (Sah et al. 2006), which provides a relative measure of


73

fire intensity (Cain 1984). Leaf scorch height, a direct measure of fire intensity

(Williams et al. 2003), was not used because the substantial distance to the first leaves

on many Eucalyptus would have resulted in imprecise measures of intensity in low-

moderate weather conditions. The ability of some bark types (e.g. stringybarks) to carry

flames more easily than others (McCarthy et al. 1999), and their interaction with wind

speed (Cheney 1981) was addressed by sampling Eucalyptus species with a range of

bark types at each site. Stem tip diameter is a measure of the minimum diameter of

branches remaining after the fire, with more intense fires leaving larger diameter

branches (Moreno & Oechel 1989). It is correlated with maximum surface temperatures

during a fire. Canopy scorch and survival were recorded as measures of fire severity

(Bond & van Wilgen 1996; Turner et al. 1999).

Char height, percent canopy scorch, survival and sizes of individuals (height and

diameter at breast height at 1.3 m, DBH) were measured on targeted size classes of

unconsumed Callitris and the Eucalyptus in the centre of each 10 m x 10 m quadrat

(two trees of each): small Callitris (3-15 cm DBH), large Callitris and Eucalyptus (> 15

cm DBH). Small Callitris established after 1952, and large Callitris originated from

1890 to 1911 (Austin & Williams 1988; Norris et al. 1991; Whipp 2009). Targeted size

classes were chosen to provide a consistent index of fire intensity. Stem tip diameters

were measured on Callitris with DBHs of 1-4 cm. Stem tip diameters were measured on

branches of at least 1 cm length and approximately 1.3 m from the ground. Five

branches were sampled on each of five Callitris in each quadrat. To determine basal

area and density of Callitris and Eucalyptus in each quadrat, the number of all trees in

each DBH size class was counted (0-5, 5.1-10, 10.1-20, 20.1-30, 30.1-40, 40.1-50, 50.1-

60, 60.1-70 cm). Sparsely distributed larger trees (> 20 cm DBH) were measured in a

larger quadrat (15 m x 15 m) centred on the main inner quadrat (10 m x 10 m).

Fuel characteristics

Ground fuel characteristics were sampled in an unburnt zone adjacent to the burnt zone,

with four quadrats one in each vegetation type (Eucalyptus matrix, Eucalyptus matrix

edge, Callitris patch edge, Callitris patch). To achieve comparable fuel characteristics,

unburned vegetation types were chosen in the same manner as the burned vegetation

types and fire history records for the unburned and the burned zones were similar.

Sampling occurred in November 2007 to approximate grass and herb conditions at the

time of the fire.


74

The following fuel characteristics were measured in four 1 m x 1 m quadrats, positioned

2 m inside each corner of a 10 m x 10 m quadrat: percent cover of bare ground, leaf

litter of Callitris and Eucalyptus, grass, herbs and shrubs; the number of all shrubs and

shrubs > 1 m tall; and the average height of shrubs, grasses and herbs. Litter depths

were measured in each quadrat and at six other fixed locations within each 10 m x 10 m

quadrat. Tree density was measured in DBH size classes (see above) to estimate

Callitris and Eucalyptus basal area and density. Ground fuel was collected in a 0.25 m x

0.25 m plot in the centre of each 1m x 1m quadrat. Fuel was sorted into total fine fuel

(litter, twigs < 6 mm diameter, bark), grasses, herbs, and twigs 6-20 mm diameter.

Samples were dried for 24 hrs at 70oC and weighed.


Basal areas of Callitris and Eucalyptus

Two factor Analyses of Variance (ANOVA) were used to examine whether Callitris

and Eucalyptus basal areas varied between the four vegetation types (Eucalyptus matrix,

Eucalyptus matrix edge, Callitris patch edge, Callitris patch) and two fire-weather

zones (low-moderate, extreme). Only data from the incoming quadrats were used, since

data from the outgoing quadrats were not independent. Basal area data were

transformed to correct for skewing and heterogeneity of variance. Tukey tests were used

for post-hoc pairwise comparisons (Quinn & Keough 2002).

To assess whether Callitris and Eucalyptus basal areas in each vegetation type in the

unburnt area were representative of the burnt area, Kruskal-Wallis tests and associated

post-hoc tests were used.

Fire intensity indicators

Linear regressions were used to examine the relationship between fire intensity

indicators in the incoming quadrat (‘Intensity Indicator In’) and the abundance of

Callitris and Eucalyptus in the outgoing quadrat, and fire intensity indicators in the

outgoing quadrat (‘Intensity Indicator Out’) within each fire-weather zone (low-

moderate, extreme) for all paired quadrats:

Intensity Indicator Out ~ Intensity Indicator In – Callitris Abundance Out + Eucalyptus

Abundance Out


75

Intensity indicators included Eucalyptus char height in the low-moderate fire-weather

zone and Callitris stem tip diameter in the extreme fire-weather zone. In both fire-

weather zones, measures of Callitris and Eucalyptus abundances included basal area

and density. Pearson’s correlation coefficient was used to check for collinearity between

predictor variables. Akaike Information Criterion (AIC) was used to compare models,

the best of which had the smallest AIC (Quinn & Keough 2002).

A Kruskal-Wallis test followed by post-hoc tests were used to compare the change in

magnitude of fire intensity indicators from the incoming to the outgoing quadrat for

each transition pair within each fire-weather zone.

Callitris survival

Logistic regressions (logit-link) were used to examine the survival of Callitris in

relation to char height on Callitris and Callitris size (DBH and height) in the low-

moderate fire-weather zone only (incoming quadrats), since most Callitris died in the

extreme fire-weather zone:

Callitris Survival ~ Char Height on Callitris * Callitris Size

Correlation tests were used to determine significant collinearity between predictor

variables, which precluded them from testing within the same model. A pseudo R2 was

calculated for each model (R2

= 1- (residual deviance/null deviance)) and AIC was

used to compare models (Veall & Zimmerman 1996). Kruskal-Wallis tests followed by

post-hoc tests were used to compare the proportion of Callitris which survived amongst

vegetation types. Only data from the incoming quadrats were used, since data from the

outgoing quadrats were not independent.


Kruskal-Wallis tests followed by post-hoc tests were used to examine the effects of

vegetation type on ground fuel characteristics, including shrub cover, shrub density,

Eucalyptus and Callitris leaf litter cover, litter depth, bare ground cover, shrub height,

fine fuel weight and total fuel weight (fine fuel weight, grass, herbs, twigs 6-20 mm

diameter). Given their sparseness, data on grass and herb cover were not analysed.


76

Results

Basal areas of Callitris and Eucalyptus

Callitris and Eucalyptus basal areas varied significantly between vegetation types, but

not between fire-weather zones. In the burnt area, Callitris basal area was significantly

greater in the Callitris patches and Callitris patch edges than in the Eucalyptus matrix

and Eucalyptus matrix edges, indicating a sharp boundary between the Eucalyptus

matrix and Callitris patches (F3, 71 = 25.33, P < 0.001; Fig. 4.2). Eucalyptus basal area

was significantly greater in the Eucalyptus matrix and Eucalyptus matrix edge than the

Callitris patch edge but not the Callitris patches (F3, 71 = 5.71, P = 0.0015), indicating a

relatively even distribution of Eucalyptus basal area. Callitris had similar mean basal

area in Callitris patches as did Eucalyptus in the Eucalyptus matrix (0.11-0.14 m2 100

m-2

).

Basal areas of Callitris and Eucalyptus were generally comparable between the burnt

and unburnt areas (where fuels were sampled). Only the basal area of Callitris

(incoming quadrats) was significantly greater in the unburnt than the burnt area in

Eucalyptus matrix edges and Callitris patches (X2

1 = 7.32, P = 0.0068, X2

1 = 5.0, P =

0.019; Fig. 4.2).

0

10

20

30

40

E b

E u

b

Ee b

Ee u

b

Ce b

Ce u

b

C b

C u

b

Eucalyptus

Callitris

Vegetation type (burned , unburned)

Basal a

rea (

m2

ha

-1)

Fig. 4.2 Basal areas of Callitris and Eucalyptus in vegetation types in the burned area

(b) and in the unburned area (ub) where fuel data were collected (E = Eucalyptus

matrix, Ee = Eucalyptus matrix edge, Ce = Callitris patch edge, C = Callitris patch).

Means and standard errors are given.


77


Char heights on Eucalyptus in the low-moderate fire-weather zone (0-8.5 m; Fig. 4.3a)

corresponded with Cheney’s (1981) low-moderate fire intensities (1.5-6 m, < 500-3000

kW m-1

). In the extreme fire-weather zone, most char heights (7-16 m; Fig. 4.3b)

corresponded with Cheney’s (1981) moderate to very high fire intensities (6-16 m,

3000-70000 kW m-1

). The ‘moderate’ values were an underestimation, since char

heights were limited by Eucalyptus heights. For this reason stem tip diameters were

used as indicators of fire intensity for analyses within the extreme fire-weather zone,

while char heights were used for analyses within the low-moderate fire-weather zone.

In the low-moderate fire-weather zone, average char heights on Callitris were lower

than those on adjacent Eucalyptus, especially below 3 m char heights on the Eucalyptus

(Fig. 4.3a). Stem tip diameters on Callitris mostly ranged from 2-4.5 mm in the extreme

fire-weather zone (Fig. 4.3b) compared to 0.5-2 mm in the low-moderate fire-weather

zone.


78

0

2

4

6

0 5 10 15 20

Char height on Eucalyptus (m)

Char

heig

ht

on C

alli

tris

(m

)

0

2

4

6

0 5 10 15 20

Ste

m tip

dia

mete

r o

n C

alli

tris

(m

m)

(a) Low-moderate f ire weather zone

(b) Extreme f ire weather zone

Fig. 4.3 Comparisons of fire intensity indicators: (a) maximum char heights on the

trunks of Callitris (> 15 cm DBH) versus maximum char heights on the trunks of

Eucalyptus (> 15 cm DBH) in the low-moderate fire-weather zone and; (b) stem tip

diameter on Callitris (1-4 cm DBH) versus char heights on Eucalyptus (> 15 cm DBH)

in the extreme fire-weather zone. Data are means for each quadrat.


79

There was only one significant correlation between a fire intensity indicator and

Callitris and Eucalyptus abundance measures used in the analyses of intensity and

Callitris survival. In the low-moderate fire-weather zone, the incoming char height was

negatively correlated with Callitris basal area in the outgoing quadrat (r = -0.34, t34 = -

2.12, P = 0.04). Since the correlation strength was relatively low, the variables were not

excluded from the analyses.

In the extreme fire-weather zone, linear regressions indicated that stem tip diameter in

the outgoing quadrat was linearly and positively related to stem tip diameter in the

incoming quadrat (Table 4.1, Fig. 4.4), indicating that variation in fuels had no

significant effect on fire intensity. Outgoing stem tip diameter was not significantly

affected by Callitris or Eucalyptus basal area or density in the outgoing quadrat.

Table 4.1 Models of fire intensity indicators in the outgoing quadrat as a function of

intensity indicators in the incoming quadrat and abundances of Callitris and Eucalyptus

in the outgoing quadrat for each fire-weather zone. Intensity Indicator Out ~ B0 + B1

Intensity Indicator In - B2 Callitris Abundance Out + B3 Eucalyptus Abundance Out.

Models are ordered from best to worst according to AIC.

Intensity Abundance Coefficients R2

AIC

indicator B0 B1 B2 B3

(a) Extreme fire-weather zone

Stem tip diameter Basal area 0.76 0.78 0.47 48

Stem tip diameter Density 0.76 0.78 0.47 48

(b) Low-moderate fire-weather zone

Char height Basal area 2.02 0.69 11.00 0.64 131

Char height Density 0.97 0.76 0.59 135

By contrast, in the low-moderate fire-weather zone, the better model indicated by AIC,

showed that char height in the outgoing quadrat was positively related to char height in

the incoming quadrat and negatively related to Callitris basal area in the outgoing

quadrat (Table 4.1). Thus char height decreased as Callitris basal area increased (Fig.

4.5). For example, an incoming char height of 5 m at the lowest Callitris basal area (0

m2 100 m

-2) was reduced to 3 m in the outgoing quadrat at the highest Callitris basal

area (0.2 m2

100 m-2

). In the second model, outgoing char height was positively related

to incoming char height but not significantly related to Callitris density. In neither

model did Eucalyptus abundance in the outgoing quadrat have a significant effect on the

outgoing char height.


80

0

1

2

3

4

5

0 1 2 3 4 5

Stem tip diameter In (mm)

Ste

m tip

dia

mete

r O

ut

(mm

)

Fig. 4.4 Stem tip diameter in the outgoing quadrat as a function of stem tip diameter in

the incoming quadrat in the extreme fire-weather zone. Stem Tip Diameter Out ~ 0.76 +

0.78 Stem Tip Diameter In (95% confidence intervals).

Char

heig

ht

Out

(mm

)

0 2

4

6

8 1

0

Fig. 4.5. Char height in the outgoing quadrat as a function of char height in the

incoming quadrat and Callitris basal area (BA) in the outgoing quadrat in the low-

moderate fire-weather zone. Char height was measured on the Eucalyptus. Char Height

Out ~ 2.02 + 0.69 Char Height In – 11.0 Callitris Basal Area Out.


81

The Kruskal-Wallis test and post hoc tests indicated a significant decline in char height

from the Eucalyptus matrix edge – Callitris patch edge transition (Table 4.2), coinciding

with an increase in Callitris basal area (Fig. 4.2). There were no other significant

changes in char height between vegetation transitions, despite a significant decline in

Callitris basal area across the Callitris patch edge – Eucalyptus matrix edge transition

(Fig. 4.2). In the extreme fire-weather zone there was no significant change in stem tip

diameter across any vegetation transition (Table 4.2).

Table 4.2 Results of the Kruskal-Wallis tests showing the median change in intensity

indicators from the incoming to the outgoing quadrats across four vegetation transitions

in two fire-weather zones: in Eucalyptus matrix (E/E); across Eucalyptus matrix –

Callitris patch boundary (Ee/Ce); in Callitris patches (C/C); and across the Callitris

patch – Eucalyptus matrix boundary (Ce/Ee). The median (Med.) change and mean

intensity indicator (standard error) for each incoming and the outgoing quadrat are also

given.

Vegetation Low-moderate fire-weather Extreme fire-weather

transition Char height (m) Stem tip diameter (mm)

Med. Mean (se) Med. Mean (se)

change change

E/E 0.15 5.3 (0.8)/5.6 (0.6) 0.23 3.2 (0.3)/3.0 (0.4)

Ee/Ce -1.65 4.4 (0.5)/2.5 (0.4) 0.18 3.0 (0.3)/3.1 (0.2)

C/C 0.40 2.0 (0.8)/2.8 (0.8) 0.11 2.5 (0.3)/2.6 (0.4)

Ce/Ee 0.75 3.8 (1.0)/5.0 (1.0) 0.15 3.0 (0.2)/3.1 (0.3)

X2 15.3 0.9

P 0.002 0.9

Callitris survival

In the extreme fire-weather zone, 4% (8/202) of Callitris individuals survived, all of

which were within one Callitris patch, near the edge of the extreme fire-weather zone.

In the low-moderate fire-weather zone, 54% (133/234) of Callitris individuals survived.

Here, Callitris survival was significantly greater in Callitris patches and Callitris patch

edges than in the Eucalyptus matrix and Eucalyptus matrix edges (X2

3 = 25.4, P < 0.001;

Fig. 4.6).

Callitris survival was best related to percent canopy scorch. Since 99% (239/241) of

Callitris individuals were killed by 100% canopy scorch, modelling was unnecessary.

This threshold in Callitris survival was reached in the low-moderate fire-weather zone.


82

In this zone, logistic regression modelling indicated that Callitris survival was

negatively related to char height and positively related to Callitris DBH and height

(Table 4.3, Fig. 4.7). Callitris DBH was a slightly better predictor of Callitris survival

than height, as indicated by AIC values and by the higher proportion of correctly

classified cases when fitted values were compared to actual values (Table 4.4). Most

Callitris died when they experienced char heights exceeding 1 m, with survival

restricted to trees > 25 cm DBH

0.0

0.2

0.4

0.6

0.8

1.0

E Ee Ce C

Vegetation type

Surv

ival (

pro

po

rtio

n)

Fig. 4.6. Proportional survival of the Callitris in each vegetation type in the low-

moderate fire-weather zone (E = Eucalyptus matrix, Ee = Eucalyptus matrix edge, Ce =

Callitris patch edge, C = Callitris patch). Means and standard errors are given.

Table 4.3 Models of Callitris survival in relation to char height on Callitris and

Callitris size in the low-moderate fire-weather zone. Models are ordered from best to

worst according to AIC. Survival ~ B0 + B1 Char Height + B2 Callitris Size + B3 Char

Height * Callitris Size.

Factors Coefficients R2 AIC

Intensity

Indicator Size B0 B1 B2 B3

Char height DBH -0.6 -3.5 0.3 0.58 147

Char height Height -2.5 -3.3 0.6 0.54 154


83

Surv

ival p

rob

ab

ility

00.5

1

.0

Fig. 4.7. Probability of Callitris survival with char height and diameter at breast height

(DBH) in the low-moderate fire-weather zone.

Table 4.4 Classification table comparing the actual and the fitted values for alive and

dead Callitris for the survival models.

Model Fitted

Survival~ Dead Alive Correct (%)

Char height + DBH Actual dead 83 15 85

Actual alive 7 128 95

Average 91

Char height + Height Actual dead 78 19 80

Actual alive 8 127 94

Average 87


84


Five of the nine fuel characteristics differed significantly among vegetation types (Table

4.5). Shrub cover was greater in the Eucalyptus matrix than in all other vegetation types,

and shrub density was lower in Callitris patches than in all other vegetation types.

Eucalyptus leaf litter cover was greater in the Eucalyptus matrix and Eucalyptus matrix

edges than in Callitris patches and Callitris patch edges; the converse was true for

Callitris leaf litter cover and as well the Eucalyptus matrix had lower cover than

Eucalyptus matrix edge. Litter depth was greater in the Eucalyptus matrix than in all

other vegetation types. There were no significant effects of vegetation type on bare

ground cover, shrub height, fine fuel weight (litter, twigs < 6 mm diameter, bark) and

total ground fuel weight (fine fuel, grass, herbs, twigs 6-20 mm diameter).

Table 4.5 Results of Kruskal-Wallis tests on the effects of vegetation type (Eucalyptus

matrix (E), Eucalyptus matrix edge (Ee), Callitris patch edge (Ce), Callitris patch (C)),

on ground fuel characteristics. Means of significant tests are given in ascending order

and underlining indicates no significant difference (ns = not significant).

Factor Means for vegetation types X2

P

E Ee Ce C

Shrub cover (%) 24.1 6.0 5.3 2.8 14.02 0.003

Shrub density (number m-2

) 5.5 2.3 4.7 1.0 8.41 0.040

Eucalypt leaf litter cover (%) 52.1 45.9 23.2 23.9 12.01 0.005

Callitris leaf litter cover (%) 0.7 14.6 58.6 67.2 44.07 <0.001

Litter depth (mm) 20.4 15.9 14.6 16.2 8.61 0.040

Bare ground cover (%) 15.63 6.84 ns

Shrub height (m) 0.52 2.60 ns

Fine fuel wt (kg ha-1

) 7786 1.95 ns

Total ground fuel wt (kg ha-1

) 17581 3.76 ns


85

Discussion

These results show that fire intensity was influenced by interactions between fire-

weather and vegetation structure, which in turn influenced Callitris survival. Under

extreme fire-weather, Callitris patches had very little effect on the incoming fire

intensity from the surrounding Eucalyptus matrix, resulting in a very low probability of

Callitris survival. By contrast, under low-moderate fire-weather, Callitris patches

reduced incoming fire intensity from the Eucalyptus matrix, resulting in a higher

probability of Callitris survival. Fewer shrubs and the greater cover of Callitris leaf

litter may have contributed to lower flammability in the Callitris patches. Thus patch

structure contributes to spatial heterogeneity of fire intensities in low-moderate fire-

weather, which enhances Callitris survival and coexistence with its more flammable

Eucalyptus competitor. This is especially important for the persistence of non-

serotinous Callitris, which are less likely to have seeds available post-fire than

serotinous Callitris.


In flat landscapes, fire intensity is influenced by interactions between weather and fuels

(Catchpole 2002; Boer et al. 2008), with their relative importance varying among

ecosystems (Bessie & Johnson 1995; Hely et al. 2001; Turner et al. 2003). In this study,

as hypothesised, fuels had less influence than weather on inferred fire intensity in the

extreme fire-weather zone. Fire intensity in outgoing quadrats, as measured by the

remaining stem tip diameter, was positively and linearly related to intensity in incoming

quadrats. Most Callitris patches experienced similar fire intensities to the Eucalyptus

matrix, despite lower flammability fuels. This result is consistent with a number of

studies, which found fuels to be less important in extreme fire-weather conditions

(Bessie & Johnson 1995; Turner et al. 2003). In at least two studies, the occurrence of

small patches of surviving obligate-seeders after crown fires, suggests that fuels have

some influence (Greene & Johnson 2000; Rodrigo et al. 2007).

As predicted in the low-moderate fire-weather zone, vegetation influenced fire intensity,

as measured by char height. This trend has been found in a range of other ecosystems

(Bessie & Johnson 1995; Turner et al. 1999; Knapp & Keeley 2006). In this study,

Callitris patches of more than 1 ha reduced the incoming fire intensity from the

Eucalyptus matrix within 10 m of patch edges. The magnitude of this abrupt decline in

intensity at the patch edge increased with increasing Callitris basal area, with a


86

maximum of 20 m2 ha

-1 having the greatest effect. In other studies, lower intensity fires

have been observed in clumps of non-serotinous, obligate-seeding Callitris intratropica

compared with the surrounding more flammable forest (Bowman et al. 1988; Bowman

& Panton 1993).

Since trees with larger diameters contribute significantly more to basal area, larger

diameter Callitris were more likely to lower flammability than smaller diameter

Callitris in low-moderate fire-weather conditions. Indeed, larger diameter trees are

usually taller and have fewer lower branches, reducing the vertical continuity of fuels

(Alexander et al. 2006; Thompson & Spies 2009). The largest Callitris in this study,

established from 1890 to 1911 and in the early 1950s (Austin & Williams 1988; Norris

et al. 1991; Whipp 2009). Whilst those from 1890 - 1911 period were characterised by

an absence of lower branches, those from the 1950s had no foliage on the lower

branches. The older of these cohorts (1890 – 1911) survived a fire in the early 1950s,

when they were approximately 50 years old, a similar age to the Callitris in this study,

which established in the early 1950s. This suggests that a fire free interval of at least 50

years allows Callitris patches to grow old and dense enough to reduce fire intensity in

low-moderate fire-weather conditions and survive. This is consistent with the

interpretation that weak serotiny, which characterises Callitris glaucophylla, is an

adaptation to facilitate inter-fire recruitment (Enright et al. 1998; Whelan et al. 1998).

Callitris in this study successfully recruit in the absence of fire, following above

average rainfall (Austin & Williams 1988).

Callitris survival

Differences in fire intensity have consequences for the survival of Callitris. Only 4% of

Callitris survived fire in the extreme fire-weather zone, since most of their foliage was

either scorched or burnt, and like most conifers, Callitris does not resprout following

this level of damage (Gill 1981; Bond & Midgley 2001; Russell-Smith 2006). Given

this susceptibility, regular high intensity fires have led to declines in non-serotinous

seeder persistence and shifts in favour of resprouters or serotinous seeders in a range of

ecosystems (e.g. Prior et al. 2007; Pausas et al. 2008).

By comparison, Callitris survival was relatively high, reaching 54% in the low-

moderate fire-weather zone. Callitris survival increased as char height, a surrogate for

fire intensity declined and Callitris diameter, a surrogate for bark thickness and tree


87

height increased. These trends are confirmation of the hypotheses. The survival of other

non-serotinous, obligate-seeder species are similarly affected by these interactions

(Bowman et al. 1988; Rodrigo et al. 2004 2007; Russell-Smith 2006). Callitris survival

was higher in Callitris patches, since they experienced lower fire intensities than in the

adjacent Eucalyptus matrix. When char heights on Callitris exceeded 1 m, most

Callitris suffered mortality, with survival restricted to trees > 25 cm DBH. This

preferential mortality or thinning of smaller size-classes is especially important for

obligate-seeder species not capable of self-thinning (Fule & Covington 1998; Russell-

Smith 2006), since it increases growth rates and shortens the time to maturity for the

survivors (Lacey 1972).


How do we account for the differences in fire intensity between the Eucalyptus matrix

and the Callitris patches in the low-moderate fire-weather zone? Other studies have

found greater fuel biomass beneath Eucalyptus species, largely due to contributions

from woody material and also grasses (Clayton-Greene 1981; Bowman & Wilson

1988). It was unlikely that grasses played a significant role in the drought coincident

fire in this study, since grasses were sparse when I sampled one year after the drought

had broken. I also found similar levels of ground fuels in the Eucalyptus matrix and in

the Callitris patches, which may be explained by the ubiquity of Eucalyptus across the

patch transition (Fig. 2) and differences in the specific leaf area, which influence leaf

longevity (Lusk et al. 2003; Prior et al. 2003). Leaves of higher specific area,

characterising Eucalyptus, have a shorter leaf life span on the tree, so are more likely to

make a larger contribution to leaf litter than the leaves of lower specific area

characterising Callitris (Lusk et al. 2003; Prior et al. 2003).

This study did however find differences in shrub cover and leaf litter characteristics

between the Eucalyptus matrix and the Callitris patches. Greater shrub cover in the

Eucalyptus matrix is likely to have increased intensity by increasing elevated fuel levels

(Cheney 1981; Alexander et al. 2006; Thompson & Spies 2009). Higher cover of

Callitris and Eucalyptus leaf litters, in their respective locations may have influenced

flammability through differential spatial packing (Schwilk & Ackerly 2001; Scarff &

Westoby 2006). Despite comparable fine fuel biomass, the needle-like leaves of the

Callitris resulted in tightly packed, shallower leaf litter in the Callitris patches than in

the Eucalyptus matrix, which in experiments by Scarff & Westoby (2006) resulted in


88

lower flammability. Correspondingly, at lower intensities, char heights were

consistently lower on Callitris than on Eucalyptus, although the influence of higher

shrub cover in the Eucalyptus matrix cannot be discounted. These trends are consistent

with the hypotheses on flammability differences between Callitris patches and

Eucalyptus matrix and agree with results from other studies on the relative litter

flammabilities of these coexisting functional types (Clayton-Greene 1981; Bowman &

Wilson 1988).

Summary

In a flat landscape, patch scale interactions between fuels and weather during fires

determine the fire intensity and thus the survival probability of non-serotinous, obligate-

seeding Callitris within a matrix of more flammable resprouting Eucalyptus. If Callitris

can establish patches of critical size and basal area during long fire free intervals,

vegetation–fire intensity feedbacks increase their chances of survival during low-

moderate fire-weather conditions. Thus, patch structure can influence spatial

heterogeneity of fire intensities and may be the key to the persistence of less flammable

species in a matrix of flammable species. In contrast, in extreme weather conditions,

where there is no such feedback, since weather has greater influence than fuel, the

survival of less flammable Callitris is very low. Predictions of an increase in the

frequency of extreme fire-weather days under global warming (Lucas et al. 2007),

suggest that fuel differences may be overridden by fire-weather in the future,

encouraging Eucalyptus to profit over Callitris.

Acknowledgements.

Australian Research Council (ARC) Linkage Grant between Charles Sturt University

(CSU) and NSW Department of Environment and Climate Change and Water (DECCW

NSW) provided the funding for this research (Project number LP066829). I would like

to thank NSW DECC staff at the Baradine office and Andrew and Katrina Cohn for

assistance in the field. Simon McDonald (CSU) assisted with statistical advice.

Chapter 5. Synthesis

89

Chapter 5.

Synthesis

Callitris glaucophylla and Eucalyptus species woodland, Merrimerriwa Range, central western New South Wales (Source: K. Ross)


90

Chapter 5.

Synthesis

This research contributes to a better understanding of the dynamics of dominant

woodland trees in relation to climate and disturbance regimes in south eastern Australia.

The focus was on Callitris glaucophylla and a range of Eucalyptus species, which vary

in abundance across their range of coexistence and beyond (Sivertsen & Clarke 2000;

Harris & Lamb 2001). Functional differences between Callitris and Eucalyptus

contribute to these trends. Callitris glaucophylla is a slow-growing, obligate-seeding

gymnosperm and Eucalyptus species are faster growing, resprouting angiosperms

(Clayton-Greene 1981; Bowman & Harris 1995; Bell & Williams 1997; Gill 1997). The

relative success of Callitris and Eucalyptus was expected to vary with different rates

and intensities of disturbances, which varied along rainfall gradients (Bellingham &

Sparrow 2000).

I examined the following:

(1) How demographic patterns of Callitris glaucophylla are associated with rainfall

and disturbances along rainfall gradients;

(2) How competitive feedback between Eucalyptus species and Callitris

glaucophylla trees and C. glaucophylla regeneration influences stand dynamics

along a rainfall gradient;

(3) How feedback between patches of Callitris glaucophylla and fire intensity

influences Callitris survival in flammable Eucalyptus woodlands in different fire-

weather conditions.

The study area was in south eastern Australia, extending from the temperate region in

the east to the arid region in the west (600 - 200 mm). The thesis was written in three

chapters, with each chapter addressing one of the above questions. Surveys of varying

scales were used to address each question (Fig. 5.1). Since the first question examined

climatic effects, this study encompassed the entire rainfall gradient. The second

question related to comparative competitive effects of Callitris and Eucalyptus, and


91

extended from the temperate to the semi-arid zone, where they coexist. The third

question was examined within a large area of woodland in the temperate zone, where

fires still occur regularly.

In this chapter I summarize the findings of the three chapters. I then integrate the

findings by defining geographic zones, within which demographic patterns of Callitris

and Eucalyptus are associated with particular suites of disturbance regimes (grazing,

fire, tree removal). I then make predictions of future dynamics and distributions of these

woodland dominants, as disturbance regimes change and shift with climate change.

Finally I outline directions for future research.

2

3

4

Figure 5.1 A satellite image of Australia showing the scale of each study, numbered by chapter

(Source: Geosciences, Australia)


92

Chapter 2

Demographic patterns of Callitris glaucophylla are associated with rainfall and

disturbances along rainfall gradients

An important and topical ecological question is - how will climate change influence

species distributions? To predict changes in species distributions, it is important to

determine the factors responsible for current demographic patterns (Pearson & Dawson

2003). A first step in this process is to identify associations between demographic

patterns and environmental factors using broad-scale surveys along climatic gradients

(Peters et al. 2007). Species that are widespread, long-lived and can be aged are ideal

and can indicate how other species with similar traits may behave under climate change.

Callitris is widespread, ranging from the arid to the temperate zone. It lives for

approximately 200 - 250 years, and can be aged, so is ideal for this challenge (Bowman

& Harris 1995).

In this chapter I examined how demographic patterns of Callitris glaucophylla were

associated with rainfall and disturbances along rainfall gradients.

The number of Callitris cohorts since 1950 was recorded in relation to mean annual

(200 - 600 mm) and seasonal (summer, uniform, winter) rainfall, edaphic factors

(topography) and disturbance regimes (tenure, vertebrate grazing, fire). Data were

collected along three transects, each within a seasonal rainfall zone spanning the annual

rainfall gradient.

A switch from recruitment success to failure occurred at 405 mm annual rainfall in the

south-west. Below 405 mm annual rainfall where winter or uniform rainfall

predominate, Callitris has failed to recruit since the mid to late 1800s. This has resulted

in ageing Callitris, which live for approximately 200 - 250 years (Bowman & Harris

1995), so the older trees are approaching senescence. Recruitment failure was

associated with farms which graze livestock and the presence of rabbits. Sheep and

rabbits are renowned for depressing or preventing the establishment of Callitris and a

range of other palatable woody species in Australia (Auld & Keith 2009). By

comparison, recruitment has occurred more often since the 1950s, both above 405 mm

annual rainfall and also below 405 mm annual rainfall where summer rainfall is

predominant. In these regions recruitment was most successful where Callitris occurred


93

on tenures other than farms, like roadsides and travelling stock routes, which have

historically lower grazing levels and where rabbits were less abundant. The absence of

fires may have also contributed to Callitris expansion, as it has with other fire sensitive

woody species (Belsky & Blumenthal 1997).

The study demonstrates how demographic patterns in Callitris were closely associated

with the spatial arrangement of disturbance regimes, dictated by regional land use, and

their relationship with rainfall along rainfall gradients. Low annual rainfall and high

levels of introduced grazing has led to ageing, contracting populations, whilst higher

annual rainfall with low levels of grazing and infrequent fire has led to younger,

expanding populations. As climate changes, rainfall, land use and disturbance regimes

will shift along climatic gradients, influencing the demography and distribution of

Callitris.

Chapter 3

Competitive feedback between Eucalyptus species and Callitris glaucophylla trees

and C. glaucophylla regeneration influences stand dynamics along a rainfall

gradient

Woody plant encroachment is a major issue in semi-arid regions throughout the world

(Scholes & Archer 1997; Asner et al. 2004), and the mechanisms triggering it are still

actively debated (Bond & Midgley 2000; Fensham et al. 2005; Sankaran et al. 2005).

Little attention has been given to the influence of canopy trees on woody encroachment

and how encroachment may in turn influence the survival of canopy trees.

In this chapter I examined how competitive feedback between Eucalyptus species and

Callitris glaucophylla trees and C. glaucophylla regeneration influences stand dynamics

along a rainfall gradient.

I sampled in six forests along a rainfall gradient during a drought. In each forest I

sampled saplings under canopies of Eucalyptus and Callitris and in adjacent gaps.

Canopy tree survival was assessed along two transects in each forest, one with and one

without regeneration.


94

The results indicate feedback between Callitris saplings and Eucalyptus and Callitris

canopy tree dynamics, which varied along the rainfall gradient. Under canopies,

saplings were less dense, smaller and less likely to be reproducing than those in gaps.

Thus canopies increased the vulnerability of Callitris saplings, which are dependent on

seeds or thick bark, to survive disturbances like fire (Bowman et al. 1988; Bowman &

Harris 1995). There were higher densities of Callitris saplings under canopies at the

lower than higher end of the rainfall gradient, but the net effect of canopy trees on

sapling regeneration along the rainfall gradient was competitive rather than facultative.

Callitris canopy trees suffered higher mortality than Eucalyptus canopy trees during the

latest drought, despite similar abilities to reduce water usage during periods of water

deficit (Zeppel & Eamus 2005). The mortality of Callitris canopy trees not surrounded

by Callitris sapling regeneration ranged from 0 to 10%, compared with a consistent

10% for those surrounded by Callitris sapling regeneration. By comparison, the

mortality of Eucalyptus canopy trees was low (< 2%), irrespective of Callitris sapling

regeneration presence. Despite high survival, Eucalyptus canopy trees showed signs of

stress at low rainfall, coincident with the presence of Callitris sapling regeneration and

higher densities of Callitris canopy trees.

This study shows that interactions among climate, sapling regeneration and canopy trees

influenced dynamics in this semi-arid system. Feedback between sapling regeneration

and canopy trees was negative, especially during drought. Canopy trees moderated but

did not exclude sapling regeneration, which in turn increased mortality rates of Callitris

or stress levels in Eucalyptus canopy trees during drought. As droughts increase in

frequency and severity, saplings and canopy trees may continue to thin (Nicholls 2004;

CSIRO & BOM 2007; Hennessy et al. 2008).

Chapter 4

Feedback between patches of Callitris glaucophylla and fire intensity influences

Callitris survival in flammable Eucalyptus woodlands in different fire-weather

conditions

A fundamental question in fire ecology is – how do long-lived, obligate-seeders persist

in flammable landscapes (Bond & van Wilgen 1996)? Spatial heterogeneity in fire


95

intensity, driven by interactions between weather and fuels or flammability may

contribute to species persistence (Whelan 1995; Boer et al. 2008).

In this chapter I examined how feedback between patches of Callitris glaucophylla and

fire intensity influenced Callitris survival in flammable Eucalyptus woodlands in

different fire-weather conditions.

The study was undertaken in Pilliga State Forest. In areas burned under extreme and

low-moderate fire-weather conditions, fire intensity and survival of Callitris were

measured in the flammable Eucalyptus matrix and in Callitris dominated patches. Pre-

fire fuel characteristics were measured in an adjacent unburnt area.

The results showed that fire intensity was influenced by interactions between fire-

weather and vegetation structure, which in turn influenced the survival of the obligate-

seeding, Callitris. Under extreme fire-weather conditions, Callitris patches had very

little effect on the incoming fire intensity from the surrounding Eucalyptus matrix,

resulting in a very low probability of Callitris survival (4%). By contrast, under low-

moderate fire-weather conditions, Callitris patches reduced incoming fire intensity from

the Eucalyptus matrix, resulting in a higher probability of Callitris survival (54%).

Callitris survival increased as char height declined and Callitris diameter increased.

This selective mortality or thinning of smaller size-classes is especially important for

obligate-seeders like Callitris, which are not capable of self-thinning (Fule & Covington

1998; Russell-Smith 2006).

Differences in shrub cover and leaf litter characteristics between the Eucalyptus matrix

and Callitris patches may explain lower fire intensity in the latter. Greater shrub cover

in the Eucalyptus matrix is likely to have increased fire intensity by increasing elevated

fuel levels (Cheney 1981; Alexander et al. 2006; Thompson & Spies 2009). Higher

covers of Callitris and Eucalyptus leaf litters in their respective locations, may have

influenced flammability, through differences in spatial packing (Schwilk & Ackerly

2001; Scarff & Westoby 2006). Despite comparable fine fuel biomass, the needle-like

leaves of Callitris result in tightly packed, shallower leaf litter in Callitris patches than

in the Eucalyptus matrix, which in experiments by Scarff & Westoby (2006) resulted in

lower flammability.


96

In this flat landscape, patch scale interactions between fuels and weather during fires

determined the fire intensity and thus the survival probability of less flammable Callitris

within a matrix of more flammable resprouting Eucalyptus. If Callitris can establish in

dense patches during long fire free intervals, vegetation–fire intensity feedbacks

increase their chances of survival during low-moderate fire-weather conditions. Thus,

patch structure can influence spatial heterogeneity of fire intensities and may be the key

to the persistence of less flammable species in a matrix of flammable species. In

contrast, in extreme fire-weather conditions, where there is no such feedback, since

weather has greater influence than fuel, the survival of less flammable Callitris is very

low.

The dynamics of woodland dominants in geographic regions

The aim of this section is to integrate the results from the three studies. The studies

indicated that the dynamics of Callitris and Eucalyptus were associated with climate

and disturbance regimes along climatic gradients. Disturbance regimes were broadly

determined by land use (e.g. farms, roadsides, state forest, national parks), the nature of

which varied along rainfall gradients. A useful integrative approach is to define

geographic regions, within which different patterns of Callitris and Eucalyptus

dynamics are associated with unique suites of disturbance regimes (grazing, fire, tree

removal). This spatial approach is used to illustrate and explain, firstly the current

dynamics and distributions of Callitris and Eucalyptus, and secondly to speculate on

those of the future, under predicted trends in climate in south eastern Australia.

Callitris has contracted in the region (Region 1) below 405 mm mean annual rainfall,

where rain falls predominantly in winter or uniformly throughout the year (Fig. 5.2a).

Callitris populations in this region are ageing, with the majority dating from the mid to

late 1800s. Regeneration failure of Callitris was associated with low annual rainfall,

farms with high levels of livestock grazing and the presence of rabbits.

In a region of expansion (Region 2), Callitris has recruited a number of times since the

1950s (Fig. 5.2a). Above 405 mm mean annual rainfall, irrespective of seasonal

rainfall, regeneration of Callitris was most successful on tenures with low to moderate

levels of livestock grazing (e.g. roadsides, travelling stock routes, State Forests and

National Parks) and where rabbits and fire were absent. Less successful Callitris

regeneration occurred on farms and/or where rabbits were present. A combination of


97

Sydney

CobarBroken Hill

200 600

Un

ifo

rm


0 200

Kilometres

N

400

Se

aso

na

l ra

infa

ll

Region 1. Callitris

contraction (winter &

uniform rainfall, farms,

rabbits)

Region 2. Callitris

expansion &

dominance (winter &

uniform rainfall,

roadsides , no rabbits, no fires)

Region 2. Callitris expansion

(summer rainfall, farms (west),

roadsides (east), few rabbits) Region 3. Callitris

& Eucalyptus

coexist (national

park, regular fires ,

few rabbits)

(a)

Sydney

CobarBroken Hill

200 600


0 200

Kilometres

N

400

Se

aso

na

l ra

infa

ll

Region 2.

Callitris expand

(slower) &

increase

dominance as Eucalyptus die

(drier)

Region 3.Callitris &

Eucalyptus coexist if

similar fire regime or

Eucalyptus dominate

if fires increase(similar climate to

now )

3.

Region 1.

Callitris extinct

(drier)

(b)

Region 2. Callitris expand slower in west (drier)

than in the east (similar climate to now)

?

?

?

Figure 5.2 Satellite images of New South Wales showing the dynamics of woodland

dominants in different geographic regions: (a) since 1950s; and (b) predicted under

climate change (Source: Geosciences, Australia).


98

successful Callitris regeneration, ringbarking of Eucalyptus canopy trees (Allen 1995),

and low levels of Eucalyptus regeneration, has led to a switch from Eucalyptus to

Callitris domination in State Forests in the south east (Lunt et al. 2006; see Chapter 3).

Below 405 mm mean annual rainfall in the summer rainfall region of the north west,

Callitris regeneration has been successful on farms, where rabbits were infrequent. This

suggests that rabbit infrequency, rainfall during the growing season (summer) or their

interactions have contributed to establishment success here, when it failed further south

in winter rainfall zone.

Callitris and Eucalyptus coexist, where low levels of introduced grazers and regular

fires occur in the Pilliga forests of the north east (Fig. 5.2a; Region 3). If the fire-

sensitive Callitris can form dense patches of regeneration between fires (i.e. ~ 50 years,

see Chapter 4) that are capable of reducing fire intensity from the incoming Eucalyptus

matrix, then larger Callitris can survive fires under low-moderate fire weather

conditions. Large changes in the fire regime are likely to affect dominance in this

region.

Predicted dynamics of woodland dominants in geographic zones under climate

change

Drier conditions are expected over most of south eastern Australia (Nicholls 2004;

CSIRO & BOM 2007; Hennessy et al. 2008; DECCW 2009). Droughts are expected to

increase in intensity and severity. A shift to summer dominant rainfall is not expected to

compensate for a decline in rainfall, especially in the south west. Only in the north east

might a predicted increase in summer rainfall compensate for increasing temperatures

(DECCW 2009). These climatic changes may influence Callitris establishment and

survival both directly through processes like desiccation, and indirectly by changing

disturbance regimes, like grazing and fire (Gworek et al. 2007; Boulant et al. 2009). In

this section I predict changes in the dynamics of Callitris and Eucalyptus, under

predicted climate change in each geographic region.

The local extinction of Callitris is possible in the south west, as opportunities for

regeneration decline and ageing Callitris trees senesce (Region 1; Fig. 5.2b). Predicted

declines in the populations of rabbits and livestock as pasture biomass declines with

higher temperatures (Scanlan et al. 2006; Gunasekera et al. 2008) may not be enough to

compensate the direct effects of drought on Callitris seedling recruitment. This south


99

west region may also expand further east if grazing by livestock and other introduced

herbivores displace unprofitable crop lands (Dale 1997; Dunlop & Brown 2008;

Garnaut 2010). Refuges from grazing (e.g. fenced roadsides, travelling stock routes)

will become increasingly important for the regeneration of Callitris, as the combined

effects of drought stress and grazing challenge survival thresholds in this region

(Dunlop & Brown 2008).

Callitris dominance is predicted to increase in the south east (Region 2; Fig. 5.2b), as

ageing Eucalyptus trees senesce and are replaced by Callitris, since Eucalyptus saplings

are infrequent. Their infrequency may be related to competition from dense Callitris

regeneration (FCNSW 1988), and also fire exclusion, which reduces establishment

opportunities (Bell & Williams 1997; Gill 1997). Fire frequency is unlikely to increase

because of predicted declines in herbaceous fuels, which limit fires in this region

together with high levels of landscape fragmentation (Bradstock 2010). Drier

conditions, fire exclusion and low-moderate grazing levels are likely to shift the system

further toward Callitris dominance, especially in the west of this region where

Eucalyptus canopy trees were most stressed during the latest drought (see Chapter 3).

In the north, changes in Callitris regeneration is predicted to vary from east to west. In

the north east, Callitris regeneration may continue at similar rates, if climate and

disturbance regimes remain relatively unchanged (Region 2; Fig. 5.2b). In the north

west, where the climate is predicted to be similar to the south west, opportunities for

regeneration may similarly decline (Region 2). But unlike the south west, this region

has younger Callitris cohorts to replace ageing Callitris trees, at least in the short term.

As temperatures increase, there is potential for Callitris to extend its range in the east,

where low temperatures are currently believed to limit seed germination (FCNSW 1988;

Fig. 5.3b). This expansion will also depend on disturbance regimes. If the potential for

fire increases in the more easterly ecosystems, as wet fuels dry (Bradstock 2010), the

expansion of Callitris will depend on the altered fire regime.

Dominance of Callitris will most likely to be determined by future fire regimes in the

Pilliga forests in the north east (Region 3; Fig. 5.2b). If predicted increases in

temperatures are matched by increases in summer rainfall (DECCW 2009) and the

current fire regime persists, coexistence of Callitris and Eucalyptus may continue.


100

However, if rising temperatures reduce fuel moisture increasing fire potential

(Bradstock 2010), the system may shift toward Eucalyptus dominance. Intervals shorter

than 50 years may not be sufficient for Callitris to form the critically dense patches

capable of reducing incoming fire intensity from a Eucalyptus matrix during low-

moderate fire weather conditions (see Chapter 4).

Further work

Determination of causal relationships between Callitris dynamics and the associated

variables identified in this research requires smaller scale experimental studies. An area

of obvious interest is below 405 mm mean annual rainfall, where recruitment is

associated with interactions between seasonal rainfall (summer, winter) and grazing

(sheep, rabbits). Disentangling and thus determining the relative importance of these

variables on recruitment success will increase our ability to predict Callitris dynamics

and distributions, as drivers change and shift with climate.

Eucalyptus vary in abundance across their range in south eastern Australia. In this study

low levels of Eucalyptus regeneration were common. Determining the extent of

regeneration success and failure and the variables associated with these patterns will

increase our understanding of Eucalyptus dynamics and advise targeted smaller scale

research. This will increase our ability to predict Eucalyptus dynamics and distribution,

as drivers change and shift with climate. Because Eucalyptus are functionally different

from Callitris, drivers of demographic patterns may also differ.

Persistence of obligate-seeders in flammable systems is dependent on survival during

fire and post-fire regeneration. Little is known of post-fire seedling dynamics of

Callitris at sites which have experienced different fire intensities and how competition

with Eucalyptus seedlings determines the subsequent patterns of dominance. For

example, are fruiting cones (and seeds) on Callitris consumed during high intensity

fires, so that post-fire colonisation can only occur from areas, such as edges, which have

burned at lower intensity? If so, it may be difficult for Callitris to recolonise this area

before the next fire, resulting in Eucalyptus domination. This research would enhance

our ability to predict the likelihood of obligate-seeder persistence in flat landscapes,

where fires of mixed intensities occur regularly.


101

Adaptive management is one approach that may be taken to address further work on

Callitris and Eucalyptus. Adaptive management is defined as ‘the integration of science,

management and learning within a collaborative framework’ (Jacobson et al. 2006). In

practice, adaptive management actions are experiments designed to gain information

(Holling 1978), and management decisions are reviewed regularly rather than following

a rigid experimental protocol (Lee 1999). By addressing the previous hypotheses on

Callitris and Eucalyptus regeneration success across the landscape, setting up

appropriate experiments to address these hypotheses and monitoring the results is likely

to contribute to disentangling the effects of climate and disturbances. For example,

results from this research suggest that recruitment failure of Callitris below 405 mm

mean annual rainfall in the winter rainfall zone is related to grazing by herbivores,

particularly rabbits. Seasonal timing of rainfall may also have some influence.

Establishing exclosures (exclude rabbits, exclude sheep/cattle/kangaroos, exclude all,

exclude none) to monitor Callitris recruitment success at sites below and above 405 mm

mean annual rainfall in the winter and summer rainfall zones, will help us understand

the relative contribution of grazing and seasonal rainfall to recruitment success. The

monitoring results would be fed interactively into the adaptive management approach to

inform further research and management (Schreiber et al. 2004).

Summary

It is clear that the relative success of Callitris and Eucalyptus was associated with

climate and disturbance regimes along climatic gradients. Prescribing different

disturbance regimes along the climatic gradient will contribute to the persistence of

these functional types. At the higher end of the rainfall gradient, a regime of regular

mixed intensity fires and low levels of grazing by introduced herbivores is likely to

contribute to the coexistence of Eucalyptus and Callitris. At the lower end of the rainfall

gradient where Callitris is in decline, reducing the levels of introduced grazers is likely

to contribute its persistence in this region, especially if droughts increase in intensity

and frequency and opportunities for establishment decline.

References

102

References

Ackerly, D. (2004). Functional strategies of chaparral shrubs in relation to seasonal

water deficit and disturbance. Ecological Monographs, 74(1), 25-44.

Ackerly, D. D. (2003). Community assembly, niche conservatism, and adaptive

evolution in changing environments. International Journal of Plant Sciences,

164(3), S165-S184.

Ackerly, D. D., & Reich, P. B. (1999). Convergence and correlations among leaf size

and function in seed plants:a comparative test using independent contrasts.

American Journal of Botany, 86(9), 1272-1281.

Agee, J. K. (1998). Fire and pine systems. In D. M. Richardson (Ed.), Ecology and

biogeography of Pinus (pp. 193-218). Cambridge: Cambridge University Press.

Alexander, J. D., Seavy, N. E., Ralph, C. J., & Hogoboom, B. (2006). Vegetation and

topographical correlates of fire severity from two fires in the Klamath-Siskiyou

region of Oregon and California. International Journal of Wildland Fire, 15(2),

237-245.

Allcock, K. G., & Hik, D. S. (2004). Survival, growth and escape from herbivory are

determined by habitat and herbivore species for three Australian woodland

plants. Oecologia, 138, 231-241.

Allen, M. (1995). The landscape archaeology of two western New South Wales forests.

Unpublished Master of Letters, University of New England, Armidale.

Allen, M. R. (1998). Forest history projects for State Forests of New South Wales: case

studies of three cypress pine forests in the Lachlan and Bogan River

Catchments, Forbes Forestry Distict on Back Yamma, Euglo South and Strahorn

State Forests. Sydney: State Forests of New South Wales.

Andrzejewska, L., & Gyllenberg, G. (1980). Small herbivore subsystem. In A. I.

Breymeyer & G. M. van Dyne (Eds.), Grasslands, Systems analysis and Man

(ist ed., pp. 201-268). Cambridge: Cambridge University Press.

Araújo, M. B., & New, M. (2007). Ensemble forecasting of species distributions.

Trends in Ecology & Evolution, 22(1), 42-47.

Armas, C., & Pugnaire, F. I. (2005). Plant interactions govern population dynamics in a

semi-arid plant community. Journal of Ecology, 93(5), 978-989.

Arno, S. F., & Gruell, G. E. (1986). Douglas-fir encroachment into mountain grasslands

in southwestern Montana. Journal of Range Management, 39(3), 272-276.

References

103

Arrieta, S., & Suárez, F. (2005). Spatial patterns of seedling emergence and survival as

a critical phase in holly (Ilex aquifolium L.) woodland recruitment in Central

Spain. Forest Ecology and Management, 205(1-3), 267-282.

Asner, G. P., Elmore, A. J., Olander, L. P., Martin, R. E., & Harris, T. (2004). Grazing

sysytems, ecosystem responses, and global change. Annual Review of

Environment & Resources, 29(1), 261-C-264.

Attiwill, P. M., & Clayton-Greene, K. A. (1984). Studies in gas exchange and

development in a subhumid woodland. Journal of Ecology, 72, 285-294.

Auld, T. D., & Keith, D. A. (2009). Dealing with threats: integrating science and

management. Ecological Management and Restoration, 10, s79-s87.

Austin, M. P., & Williams, O. B. (1988). Influence of climate and community

composition on the population demography of pasture species in semi-arid

Australia. Vegetatio, 77, 43-49.

Barnes, P. W., & Archer, S. (1999). Tree-shrub interactions in a subtropical savanna

parkland: competition or facilitation? Journal of Vegetation Science, 10(4), 525-

536.

Beerling, D. J., & Woodward, F. I. (1996). Palaeo-ecophysiological perspectives on

plant responses to global change. Trends in Ecology & Evolution, 11(1), 20-23.

Bell, D., T. (2001). Ecological response syndromes in the flora of southwestern Western

Australia:fire resprouters versus reseeders. The Botanical Review, 67(4), 417-

440.

Bell, D. T., & Williams, J. E. (1997). Eucalypt ecophysiology. In J. Williams & J.

Woinarski (Eds.), Eucalypt Ecology:Individuals to Ecosystems (pp. 168-196).

Cambridge: Cambridge University Press.

Bellingham, P. J., & Sparrow, A. D. (2000). Resprouting as a life history strategy in

woody plant communities. Oikos, 89, 409-416.

Belsky, A. J. (1994). Influences of Trees on Savanna Productivity: Tests of Shade,

Nutrients, and Tree-Grass Competition. Ecology, 75(4), 922-932.

Belsky, A. J., & Blumenthal, D. M. (1997). Effects of Livestock Grazing on Stand

Dynamics and Soils in Upland Forests of the Interior West. Conservation

Biology, 11(2), 315-327.

Benson, D., & McDougall, L. (1998). Ecology of Sydney plant species: Part 6.

Dicotyledon family Myrtaceae. Cunninghamia, 5, 808-987.

References

104

Benson, J. S., Allen, C. B., Togher, C., & Lemmon, J. (2006). New South Wales

vegetation classification and assessment:Part 1 plant communities of NSW

western plains. Cunninghamia, 9(3), 383-450.

Bertness, M. D., & Callaway, R. M. (1994). Positive interactions in communities.

Trends in Ecology and Evolution, 9, 191-193.

Bessie, W. C., & Johnson, E. A. (1995). The relative importance of fuels and weather

on fire behaviour in subalpine forests. Ecology, 76(3), 747-762.

Binns, D., & Beckers, D. (2001). Floristic patterns in the Pilliga. In J. Dargavel (Ed.),

Perfumed Pineries:Environmental History of Australia's Callitris Forests (pp.

104-110). Canberra: Australian National University.

Boer, M., Sadler, R., Bradstock, R., Gill, A., & Grierson, P. (2008). Spatial scale

invariance of southern Australian forest fires mirrors the scaling behaviour of

fire-driving weather events. Landscape Ecology, 23(8), 899-913.

Boland, D. J., Brooker, M. I. H., Chippendale, G. M., Hall, N., Hyland, B. P. M.,

Johnston, R. D., et al. (2006). Forest Trees of Australia (5th ed.). Collingwood:

CSIRO Publishing.

Bond, W., J. , & Midgley, G., F. . (2000). A proposed CO2 controlled mechanism of

woody plant invasion in grasslands and savannas. Global Change Biology, 6(8),

865-869.

Bond, W. J. (1989). The tortoise and the hare:ecology of angiosperm dominance and

gymnosperm persistence. Biological Journal of the Linnean Society, 36, 227-

249.

Bond, W. J., & Midgley, J. J. (2001). Ecology of sprouting in woody plants:persistence

niche. Trends in Ecology and Evolution, 16(1), 45-51.

Bond, W. J., Smythe, K.-A., & Balfour, D. A. (2001a). Acacia species turnover in space

and time in an African savanna. Journal of Biogeography, 28, 117-128.

Bond, W. J., & van Wilgen, B. W. (1996). Fire and Plants. London: Chapman & Hall.

Boulant, N., Garnier, A., Curt, T., & Lepart, J. (2009). Disentangling the effects of land

use, shrub cover and climate on the invasion speed of native and introduced

pines in grasslands. Diversity and Distributions, 15, 1047-1059.

Boulant, N., Kunstler, G., Rambal, S., & Lepart, J. (2008). Seed supply, drought, and

grazing determine spatio-temporal patterns of recruitment for native and

introduced invasive pines in grasslands. Diversity and Distributions, 14(5), 862-

874.

References

105

Bowman, D. M. J. S. (1998). Tansley Review No. 101. The impact of aboriginal

landscape burning on the Australian biota. New Phytologist, 140(3), 385-410.

Bowman, D. M. J. S., & Harris, S. (1995). Conifers of Australia's dry forests and open

woodlands. In N. J. Enright & S. Hill Robert (Eds.), Ecology of the Southern

Conifers (1 ed., pp. 252-270). Melbourne: Melbourne University Press.

Bowman, D. M. J. S., & Latz, P. K. (1993). Ecology of Callitris glaucophylla

(Cupressaceae) on the MacDonnell Ranges, Central Australia. Australian

Journal of Botany, 41, 217-225.

Bowman, D. M. J. S., Price O,, Whitehead, P. J., & Walsh, A. (2001). The 'wilderness

effect' and the decline of Callitris intratropica on the Arnhem Land Plateau,

northern Australia. Australian Journal of Botany, 49, 665-672.

Bowman, D. M. J. S., & Panton, W. J. (1993). Decline of Callitris intratropica

R.T.Baker & H.G. Smith in the Northern Territory: implications for pre- and

post-European colonization fire regimes. Journal of Biogeography, 20, 373-381.

Bowman, D. M. J. S., & Wilson, B. A. (1988). Fuel characteristics of coastal monsoon

forests, Northern Territory. Journal of Biogeography, 15, 807-817.

Bowman, D. M. J. S., Wilson, B. A., & Davis, G. W. (1988). Response of Callitris

intratropica R.T.Baker & H.G. Smith to fire protection, Murgenella, Northern

Australia. Australian Journal of Ecology, 13, 147-159.

Bradstock, R. A. (2010). A biogeographic model of fire regimes in Australia: current

and future implications. Global Ecology and Biogeography, 19(2), 145-158.

Bradstock, R. A., & Cohn, J. C. (2002). Fire regimes and biodiversity in semi-arid

mallee ecosystems. In R. A. Bradstock, J. E. Williams & A. M. Gill (Eds.),

Flammable Australia:the fire regimes and biodiversity of a continent (pp. 238-

258). Cambridge: Cambridge University Press.

Breiman, L., Freidman, J. H., Olshen, R. A., & Stone, C. J. (1993). Classification and

regression trees. New York: Chapman & Hall Inc.

Brodribb, T., & Hill, R. S. (1998). The photosynthetic drought physiology of a diverse

group of southern hemisphere conifer species is correlated with minimum

seasonal rainfall. Functional Ecology, 12(3), 465-471.

Brooker, M. I. H., & Kleinig, D. A. (1990). Field Guide to Eucalypts. Southwestern and

Southern Australia (Vol. 2). Sydney: Inkata Press.

Brookhouse, P., Nicholson, D., Green, D., & Lord, B. (1999). Large Pilliga fires and the

development of fire management and fire suppression strategies. In I. Lunt (Ed.),

Proceedings of the Australian Bushfire Conference.

References

106

Brown, J. R., & Archer, S. (1999). Shrub invasion of grassland: recruitment is

continuous and not regulated by herbaceous biomass or density. Ecology, 80(7),

2385-2396.

Bruno, J. F., Stachowicz, J. J., & Bertness, M. D. (2003). Inclusion of facilitation into

ecological theory. Trends in Ecology & Evolution, 18(3), 119-125.

Bureau of Meteorology. (2006). Special climate statement 9:An exceptionally dry

decade in parts of southern and eastern Australia:October 1996-September

2006. Melbourne: Bureau of Meteorology.

Bureau of Meteorology. (2010a). Major seasonal rainfall zones of Australia Retrieved

1/7/2010, 2010

Bureau of Meteorology. (2010b). Maps of rainfall percentages. Melbourne: Bureau of

Meteorology.

Bureau of Rural Sciences. (2009). 2001/02 Land uses of Australia, V.3. 2010

Burgman, M. A., & Lamont, B., B. . (1992). A stochastic model for the viability of

Banksia cuneata populations: environmental, demographic and genetic effects.

Journal of Applied Ecology, 29(3), 719-727.

Burrows, D. M., & Burrows, W. H. (1992). Seed production and litter fall in some

eucalypt communities in central Queensland. Australian Journal of Botany,

40(3), 389-403.

Burrows, W. H. (2002). Seeing the wood(land) for the trees-an individual perspective of

Queensland woodland studies (1965-2005). Tropical Grasslands, 36, 202-217.

Calamassi, R., Rocca, G. D., Falusi, M., Paoletti, E., & Strati, S. (2001). Resistance to

water stress in seedlings of eight European provenances of Pinus halapensis

Mill. Annals of Forest Science, 58, 663-672.

Callaway, R. M., & Walker, L. R. (1997). Competition and facilitation: a synthetic

approach to interactions in plant communities. Ecology, 78(7), 1958-1965.

Camarero, J. J., & Gutierrez, E. (2007). Response of Pinus uncinata recruitment to

climate warming and changes in grazing pressure in an isolated population of the

Iberian system (NE Spain). Arctic, Antarctic and Alpine Research, 39, 210-217.

Castro, J., Zamora, R., Hodar, J. A., & Gomez, J. M. (2005). Alleviation of summer

drought boosts establishment success of Pinus sylvestris in a mediterranean

mountain: an experimental approach. Plant Ecology, 181(2), 191-202.

Catchpole, W. (2002). Fire properties and burn patterns in heterogeneous landscapes. In

R. A. Bradstock, J. E. Williams & A. M. Gill (Eds.), Flammable Australia: the

References

107

fire regimes and biodiversity of a continent (pp. 49-75). Cambridge: Cambridge

University Press.

Cheney, N. P. (1981). Fire behaviour. In A. M. Gill, R. H. Groves & I. R. Noble (Eds.),

Fire and the Australian Biota (pp. 151-175). Canberra: Australian Academy of

Science.

Chesterfield, C. J., & Parsons, R. F. (1985). Regeneration of three tree species in arid

south-eastern Australia. Australian Journal of Botany, 33, 715-732.

Clarke, P. J. (2000). Plant population processes in temperate woodlands in eastern

Australia. In R. J. Hobbs & J. Yates Colin (Eds.), Temperate Eucalypt

Woodlands in Australia:Biology,Conservation, Management and Restoration

(pp. 248-270). Sydney: Surrey Beatty and Sons.

Clarke, P. J. (2002). Habitat insularity and fire response traits:evidence from a

sclerophyll archipelago. Oecologia, 132, 582-591.

Clarke, P. J., & Knox, J. E. (2002). Post-fire response of shrubs in the tablelands of

eastern Australia:do existing models explain habitat differences? Australian

Journal of Botany, 50, 53-62.

Clayton-Greene, K. A. (1981). The autecology of Callitris columellaris F.Muell. and

associated Eucalyptus spp. in south-eastern Australia. Unpublished PhD thesis,

University of Melbourne, Melbourne.

Clayton-Greene, K. A., & Ashton, D. H. (1990). The dynamics of Callitris

collumellaris/Eucalyptus albens communities along the Snowy River and its

tributaries in south-eastern Australia. Australian Journal of Botany, 38, 403-432.

Colls, K., & Whitaker, R. (2001). The Australian Weather Book. Sydney: Reed New

Holland Pty Ltd.

Cooke, B. D. (1977). Factors limiting the distribution of wild rabbit in Australia. Paper

presented at the Exotic species in Australia-their establishment success,

Adelaide.

Covington, W. W., & Moore, M. M. (1994). Southwestern ponderosa forest structure-

changes since Euro-American settlement. Journal of forestry (USA), 92, 39-47.

Cowling, R. M. (1987). Fire and its role in coexistence and speciation in Gondwanan

shrublands. South African Journal of Science, 83, 106-112.

Cox, S. J., Sivertsen, D. P., & Bedward, M. (2001). Clearing of native woody vegetation

in the New South Wales northern wheatbelt:extent,rate of loss and implications

for biodiversity. Cunninghamia, 7(1), 101-155.

References

108

CSIRO. (2008). Special climate statement 16:Long term rainfall deficiencies continue

in southern Australia while wet conditions dominate the north. Canberra

[www.bom.gov.au/climate/current/statements/scs16.pdf]: CSIRO.

CSIRO & BOM. (2007). Climate change in Australia: technical report 2007. Canberra:

CSIRO, Bureau of Meterology and Australian Greenhouse Office in partnership

with the Australian Climate Change Program.

Cunningham, G. M., Mulham, W. E., Milthorpe, P. L., & Leigh, J. H. (1992). Plants of

Western New South Wales. Sydney: Inkata Press.

Curtin, F. K. (1987). Notes on the establishment of cypress pine regeneration: Forestry

Commission of NSW.

Dale, V. H. (1997). The relationship between land-use change and climate change.

Ecological Applications, 7(3), 753-769.

Dale, V. H., Joyce, L. A., McNulty, S., Neilson, R. P., Ayres, M. P., Flannigan, M. D.,

et al. (2001). Climate Change and Forest Disturbances. (Cover story).

Bioscience, 51(9), 723.

Davis, S. D., Ewers, F. W., Wood, J., Reeves, J. J., & Kolb, K. J. (1999). Differential

susceptibility to xylem cavitation among three pairs of Ceanothus species in the

Transverse Mountain Ranges of southern California. Ecoscience, 6, 180-186.

DECCW NSW. (2009). NSW State of the Environment 2009. Sydney: NSW Dept of

Environment, Climate Change and Water.

Denham, A. J., & Auld, T. D. (2004). Survival and recruitment of seedlings and suckers

of trees and shrubs of the Australian arid zone following habitat management

and the outbreak of Rabbit Calicivirus Disease (RDC). Austral Ecology, 29(5),

585-599.

Denny, M. (1992). Historical and ecological study of the effects of european settlement

on inland NSW. Sydney: Nature Conservation Council of New South Wales.

Diaz, S., Noy-Meir, I., & Cabido, M. (2001). Can grazing response of herbaceous plants

be predicted from simple vegetative traits? Journal of Applied Ecology, 38(3),

497-508.

Dunlop, M., & Brown, P. R. (2008). Implications or climate change for Australia's

national reserve system: a preliminary assessment. Canberra: Report to

Department of Climate Change.

Eldridge, D. J., & Greene, R. S. B. (1994). Microbiotic soil crusts - a review of their

roles in soil and ecological processes in the rangelands of Australia. Australian

Journal of Soil Research, 32(3), 389-415.

References

109

Enright, N. J., & Lamont, B. B. (1992). Recruitment variability in the resprouting shrub

Banksia attenuata and non-sprouting congeners in the northern sandplain heaths

of southwestern Australia. Acta Oecologica, 13(6), 727-741.

Enright, N. J., Marsula, R., Lamont, B. B., & Wissel, C. (1998). The ecological

significance of canopy seed storage in fire-prone environments: a model for non-

sprouting shrubs. Journal of Ecology, 86(6), 946-959.

Facelli, J. M., & Springbett, H. (2009). Why do some species in arid lands increase

under grazing? Mechanisms that favour increased abundance of Maireana

pyramidata in overgrazed chenopod shrublands of South Australia. Austral

Ecology, 34(5), 588-597.

FCNSW. (1988). Notes on the Silviculture of Major NSW Forest Types.10. Cypress

Pine Types. Sydney: Forestry Commission of NSW.

Fensham, R. J. (1998a). The grassy vegetation of the Darling Downs, south-eastern

Queensland, Australia. Floristics and grazing effects. Biological Conservation,

84(3), 301-310.

Fensham, R. J. (1998b). The influence of cattle grazing on tree mortality after drought

in savanna woodland in north Queensland. Australian Journal of Ecology, 23(4),

405-407.

Fensham, R. J. (2008). Leichhardts maps: 100 years of change in vegetation structure in

inland Queensland. Journal of Biogeography, 35(1), 141-156.

Fensham, R. J., Fairfax, R. J., & Ward, D. P. (2009). Drought-induced tree death in

savanna. Global Change Biology, 15(2), 380-387.

Fensham, R. J., & Holman, J. E. (1999). Temporal and spatial patterns in drought-

related tree dieback in Australian savanna. Journal of Applied Ecology, 36(1),

1035-1050.

Florence, R. G. (1996). Ecology and Siviculture of Eucalypt Forests. Melbourne:

CSIRO.

Foden, W., Midgley, G. F., Hughes, G., Bond, W. J., Thuiller, W., Hoffman, M. T., et

al. (2007). A changing climate is eroding the geographical range of the Namib

Desert tree Aloe through population declines and dispersal lags. Diversity &

Distributions, 13(5), 645-653.

Foran, B. D. (1986). The impact of rabbits and cattle on an arid calcareous shrubby

grassland in central Australia. Vegetatio, 66, 49-59.

Fox, M. D. (1991). The natural vegetation of the Ana Branch-Mildura 1:250000 map

sheets. Cunninghamia, 2(3), 443-493.

References

110

Fuentes, E. R., Otaiza, R. D., Alliende, M. C., Hoffmann, A., & Poiani, A. (1984).

Shrub clumps of the Chilean matorral vegetation: structure and possible

maintenance mechanisms. Oecologia, 62(3), 405-411.

Fulé, P. Z., & Covington, W. W. (1998). Spatial patterns of Mexican pine-oak forests

under different recent fire regimes. Plant Ecology, 134(2), 197-209.

García, D., Zamora, R., Hódar, J. A., & Gómez, J. M. (1999). Age structure of

Juniperus communis L. in the Iberian peninsula: Conservation of remnant

populations in Mediterranean mountains. Biological Conservation, 87(2), 215-

220.

Garnaut, R. (2010). Climate change and the Australian agricultural and resource

industries. Australian Journal of Agricultural and Resource Economics, 54(1),

9-25.

GenStat. (2009). The guide to GenStat Release 12. Part 2: Statistics. Waterhouse Street,

Hemel Hempstead, Hertfordshire HP1 1ES, UK: VSN International

Gill, A. M. (1981). Adaptive responses of Australian vascular plant species to fires. In

A. M. Gill, R. H. Groves & I. R. Noble (Eds.), Fire and the Australian Biota.

Canberra: The Australian Academy of Science.

Gill, A. M., & Bradstock, R. A. (1992). A national register for the fire response of plant

species. Cunninghamia, 2, 653-660.

Gill, A. M., & Moore, P. H. R. (1990). Fire intensities in eucalyptus forests of

southeastern Australia. Paper presented at the International Conference on

Forest Fire Research, Coimbra, Portugal.

Gill, M. (1997). Eucalypts and fires:interdependent or independent? In J. Williams & J.



Graz, F. P. (2008). The woody weed encroachment puzzle: gathering pieces.

Ecohydrology, 1(4), 340-348.

Greene, D. F., & Johnson, E. A. (2000). Tree recruitment from burn edges. Canadian

Journal of Forest Research, 30(8), 1264-1274.

Grime, J. P. (1979). Plant Strategies and Vegetation Processes. Chichester: John Wiley

and Sons.

Grime, J. P. (2001). Plant Strategies, vegetation processes and ecosystem properties.

Chichester: John Wiley and Sons.

Grubb, P. J. (1977). The maintenance of species-richness in plant communities:the

importance of the regeneration niche. Biological Review, 52, 107-145.

References

111

Guarín, A., & Taylor, A. H. (2005). Drought triggered tree mortality in mixed conifer

forests in Yosemite National Park, California, USA. Forest Ecology and

Management, 218(1-3), 229-244.

Guisan, A., & Thuiller, W. (2005). Predicting species distribution: offering more than

simple habitat models. Ecology Letters, 8(9), 993-1009.

Gunasekera, D., Tulloh, C., Ford, M., & Heyhoe, E. (2008, 13 June 2008). Climate

change:opportunities and challenges in Australian agriculture. Paper presented

at the Facing climate change:reserach on adaptation og agrosystems, University

of Sydney.

Gurevitch, J., Scheiner, S. M., & Fox, G. A. (2006). The Ecology of Plants. Sunderland,

MA: Sinauer Associates, Inc.

Gutiérrez, J. R., Holmgren, M., Manrique, R., & Squeo, F. A. (2007). Reduced

herbivore pressure under rainy ENSO conditions could facilitate dryland

reforestation. Journal of Arid Environments, 68(2), 322-330.

Gworek, J. R., Vander Wall, S. B., & Brussard, P. F. (2007). Changes in biotic

interactions and climate determine recruitment of Jeffrey pine along an elevation

gradient. Forest Ecology and Management, 239(1-3), 57-68.

Haase, P., Pugnaire, F. I., Clark, S. C., & Incoll, L. D. (1996). Spatial patterns in a two-

tiered semi-arid shrubland in southeastern Spain. Journal of Vegetation Science,

7(4), 527-534.

Hanson, P. J., & Weltzin, J. F. (2000). Drought disturbance from climate change:

response of United States forests. The Science of The Total Environment, 262(3),

205-220.

Harden, G. J. (2000). Flora of New South Wales (Vol. 1). Sydney: New South Wales

University Press.

Harden, G. J. (2002). Flora of New South Wales (Vol. 2). Sydney: New South Wales

University Press.

Harrington, G. N. (1991). Effects of soil moisture on shrub seedling survival in a semi-

arid grassland. 1338-1149.

Harrington, G. N., Dawes, G. T., & Ludwig, J. A. (1981). An analysis of the vegetation

pattern in a semi-arid Eucalyptus populnea woodland in north-west New South

Wales. Australian Journal of Ecology, 6, 279-287.

Harris, M., & Lamb, D. (2001). Recent disturbances to Callitris forests in Southern

Queensland. In J. Dargavel (Ed.), Perfumed Pineries:Environmental History of

References

112

Australia's Callitris Forests (pp. 125-135). Canberra: Australian National

University.

Hattersley, P. W. (1983). The distribution of C3 and C4 grasses in Australia in relation to

climate. Oecologia, 57(1/2), 113-128.

Hawkins, P. J. (1966). Seed production and litterfall studies of Callitris columellaris.

Australian Forest Research, 2, 3-16.

Heckathorn, S. A., McNaughton, S. J., & Coleman, J. S. (1999). C4 plants and

herbivory. In R. F. Sage & R. K. Monson (Eds.), C4 Plant Biology (pp. 285-

312). London: Academic Press.

Hely, C., Flannigan, M., Bergeron, Y., & McRae, D. (2001). Role of vegetation and

weather on fire behavior in the Canadian mixedwood boreal forest using two fire

behavior prediction systems. Canadian Journal of Forest Research, 31(3), 430-

441.

Henery, M. L., & Westoby, M. (2001). Seed mass and seed nutrient content as

predictors of seed output variation between species. Oikos, 92(3), 479-490.

Hennessy, K., Fawcett, R., Kirono, D., Mpelasoka, F., Jones, D., Bathols, J., et al.

(2008). An assessment of the impact of climate change on the nature and

frequency of exceptional climatic circumstances: Commonwealth of Australia.

Hennessy, K., Lucas, C., Nicholls, N., Bathols, J., Suppiah, R., & Ricketts, J. (2005).

Climate change impacts on fire-weather in south-east Australia: CSIRO

Atmospheric Research

Australian Government Bureau of Meterology.

Hilbert, D. W. (1987). A model of life history strategies of chapparal shrubs in relation

to fire frequency. In J. D. Tenhunen, F. M. Catarino, O. L. Lange & W. C.

Oechel (Eds.), Plant Response to Stress:Functional Analysis in Mediterranean

Ecosystems. (pp. 597-606). Berlin: Springer-Verlag.

Hill, K. D. (1998). Pinophyta. In Flora of Australia, vol. 48: Ferns, Gymnosperms and

Allied Groups (pp. 545-596). Melbourne: CSIRO, Australia.

Hill, R., S. (1995). Conifer origin, evolution and diversification in the southern

hemisphere. In N. J. Enright & S. Hill Robert (Eds.), Ecology of the Southern

Conifers (1 ed., pp. 10-29). Melbourne: Melbourne University Press.

Hill, R. S., & Brodribb, T. J. (1999). Turner Review No. 2. Southern conifers in time

and space. Australian Journal of Botany, 47, 639-696.

References

113

Hodgkinson, K., Harrington, G. N., & Miles, G. E. (1980). Composition, spatial and

temporal variability of the soil seed pool in a Eucalyptus populnea shrub

woodland in central New South Wales. Australian Journal of Ecology, 5, 23-29.

Hodgkinson, K. C. (1979). The shrubs of poplar box (Eucalyptus populnea) lands and

their biology. The Rangeland Journal, 1(4), 280-293.

Hodgkinson, K. C. (1991). Shrub recruitment response to intensity and season of fire in

a semi-arid woodland. Journal of Applied Ecology, 28, 60-70.

Hodgkinson, K. C., & Harrington, G. N. (1985). The case for prescribed burning to

control shrubs in eastern semi-arid woodlands. Australian Rangeland Journal,

64-74.

Holling, C.S. 1978. Adaptive environmental assessment and management. John Wiley,

London.

Holmgren, M., Lopez, B. C., Gutierrez, J. R., & Squeo, F. A. (2006b). Herbivory and

plant growth rate determine the success of El Nino Southern Oscillation-driven

tree establishment in semiarid South America. Global Change Biology, 12(12),

2263-2271.

Holmgren, M., & Scheffer, M. (2001). El Niño as a window of opportunity for the

restoration of degraded arid ecosystems. Ecosystems, 4(2), 151-159.

Holmgren, M., Scheffer, M., & Huston, M. A. (1997). The interplay of facilitation and

competition in plant communities. Ecology, 78(7), 1966-1975.

Holmgren, M., Stapp, P., Dickman, C. R., Gracia, C., Graham, S., Gutierrez, J. R., et al.

(2006a). Extreme climatic events shape arid and semiarid ecosystems. Frontiers

in Ecology and the Environment, 4(2), 87-95.

Hua, Q. (2009). Radiocarbon:a chronological tool for the recent past. Quaternary

Geochronology, 4, 378-390.

Huntly, N. (1991). Herbivores and the dynamics of communities and ecosystems.

Annual Review of Ecology & Systematics, 22, 477-503.

Huston, M. (1979). A general hypothesis of species diversity. The American Naturalist,

81-101.

Huston, M. (2003). Understanding the effects of fire and other mortality causing

disturbances on species diversity. In I. Abott & N. Burrows (Eds.), Fire in

Ecosystems of south-west Western Australia: Impacts and Management (pp. 37-

70). Leiden, The Netherlands: Backhuys Publishers.

Insightful Corp. (2007). S-Plus 8 Guide to Statistics. Seattle, Washington: Insightful

Corporation.

References

114

IPCC. (2007). Climate change 2007: The physical basis. Contributions of working

group 1 to the fourth assessment report of the intergovernmental panel on

climate change. New York.

Jacobson, S.K., Morris, J.K., Sanders, J.S., Wiley, E.N., Brooks, M., Bennetts,

R.E., Percival, H.F. & Marynowski, S. 2006. Understanding Barriers to

Implementation of an Adaptive Land Management Program.

Conservation Biology 20: 1516-1527.

Johnston, T. N. (1969). The effect of sheep and rabbit grazing on regeneration of white

cypress pine. Australian Forest Research, 4, 3-12.

Jump, A. S., Mátyás, C., & Peñuelas, J. (2009). The altitude-for-latitude disparity in the

range retractions of woody species. Trends in Ecology & Evolution, 24(12), 694-

701.

Keeley, J., E. , & Zedler, P. H. (1978). Reproduction of chaparral shrubs after fire: a

comparison of sprouting and Seeding strategies. American Midland Naturalist,

99(1), 142-161.

Keeley, J. E., Hopkins, A. J. M., & Lamont, B. B. (1986). Resilience of mediterranean

shrub communities to fire. In B. Dell (Ed.), Resilience in Mediterranean type

ecosystems (pp. 95-112). Dordrecht, The Netherlands: W.Junk Publishers.

Keeley, J. E., & Zedler, P. H. (1998). Evolution of life histories in Pinus. In D. M.

Richardson (Ed.), Ecology and biogeography of Pinus (1 ed., pp. 219-249).


Keith, D. (2004). Ocean Shores to Desert Dunes:the Native Vegetation of New South

Wales and the ACT. Sydney: Department of Conservation (NSW).

Keith, D. A. (1991). Coexistence and species diversity in upland swamp vegetation:the

roles of an environmental gradient and recurring fires. Unpublished PhD,

University of Sydney, Sydney, Australia.

Keith, D. A., Akcakaya, H. R., Thuiller, W., Midgely, G. F., Pearson, R. G., Phillips, S.

J., et al. (2008). Predicting extinction risks under climate change: coupling

stochastic population models with dynamic bioclimatic habitat models. Biology

Letters, 4, 560-563.

Knapp, E. E., & Keeley, J. E. (2006). Heterogeneity in fire severity within early season

and late season prescribed burns in a mixed-conifer forest. International Journal

of Wildland Fire, 15(1), 37-45.

References

115

Knox, K. J. E., & Clarke, P. J. (2005). Nutrient availability induces contrasting

allocation and starch in resprouting and obligate seeding shrubs. Functional

Ecology, 19, 690-698.

Kraaij, T., & Ward, D. (2006). Effects of rain, nitrogen, fire and grazing on tree

recruitment and early survival in bush-encroached savanna, South Africa. Plant

Ecology, 186(2), 235-246.

Krassilov, V. A. (1978). Late Cretaceous gymnosperms from Sakhalin and the terminal

Cretaceous event. Palaeontology, 21, 893-905.

Kruger, F. J. (1983). Plant community diversity and dynamics in relation to fire. In F. J.

Kruger, D. Mitchell & J. Jarvis (Eds.), Mediterranean-type ecosystems: the role

of nutrients (pp. 446-472). Berlin: Springer-Verlag.

Kruger, F. J., & Bigalke, R. C. (1984). Fire in fynbos. In P. V. Booysen & N. M.

Tainton (Eds.), Ecological effects of fire in South African ecosystems

(pp. 67-114). New York: Springer-Verlag.

Lacey, C. J. (1972). Factors influencing occurrence of cypress pine regeneration in

New South Wales. Sydney: Forestry Commission of NSW.

Lacey, C. J. (1973). Silvicultural characteristics of white cypress pine. Sydney: Forestry

Commission of NSW.

Lamont, B. B., Groom, P. K., & Cowling, R. M. (2002). High leaf area mass per area of

related species assemblages may reflect low rainfall and carbon isotope

discrimination rather than phosphorous and nitrogen concentrations. Functional

Ecology, 16, 403-412.

Landsberg, J., Lavorel, S., & Stol, J. (1999). Grazing response groups among

understorey plants in arid rangelands. Journal of Vegetation Science, 10, 683-

696.

Landsberg, J., & Stol, J. (1996). Spatial distribution of sheep, feral goats and kangaroos

in woody rangeland paddocks. Rangeland Journal, 18(2), 270-291.

League, K., & Veblen, T. (2006). Climatic variability and episodic Pinus ponderosa

establishment along forest-grassland ecotones of Colorado. Forest Ecology and

Management, 228, 98-107.

Lee, K.N. 1993. Compass and gyroscope: integrating science and politics for the

environment. Island Press, Washington D.C.

Letnic, M., Tamayo, B., Dickman, C. R., & Kelt, D. A. (2005). The responses of

mammals to La Nina (EL Nino Southern Oscillation)-associated rainfall,

References

116

predation and wildfire in central Australia Journal of Mammalogy, 86(4), 689-

703.

Lindsay, A. D. (1967). Forest types of the New South Wales Cypress Pine Zone (No.

Technical report No. 8). Sydney: Forestry Commission of NSW.

Llorens, L., Peñuelas, J., Beier, C., Emmett, B., Estiarte, M., & Tietema, A. (2004).

Effects of an experimental increase of temperature and drought on the

photosynthetic performance of two Ericaceous shrub species along a north–

south European gradient. Ecosystems, 7(6), 613-624.

Lloyd, A. H. (1997). Response of tree-line populations of foxtail pine (Pinus

balfouriana) to climate variation over the last 1000 years. Canadian Journal of

Forest Research, 27, 936-942.

Loehle, C. (2000). Strategy space and the disturbance spectrum:a life-history model for

tree species coexistence. The American Naturalist, 156(1), 14-33.

Loehle, C., & LeBlanc, D. (1996). Model-based assessments of climate change effects

on forests: a critical review. Ecological Modelling, 90(1), 1-31.

López, B. C., Holmgren, M., Sabaté, S., & Gracia, C. A. (2008). Estimating annual

rainfall threshold for establishment of tree species in water-limited ecosystems

using tree-ring data. Journal of Arid Environments, 72(5), 602-611.

Lucas, C., Hennessy, K., Mills, G., & Bathols, J. (2007). Bushfire weather in

southeastern Australia: recent trends and projected climate change impacts.

Sydney: The Climate Institute of Australia.

Luck, G. W. (2007). The relationships between net primary productivity, human

population density and species conservation. Journal of Biogeography, 34(2),

201-212.

Ludwig, J., Tongway, D. J., Freudenberger, D., & Noble, J. (Eds.). (1997). Landscape

Ecology Function and Management: Principles from Australia's Rangelands.

Collingwood: CSIRO, Australia.

Lunney, D. (2001). Causes of the extinction of native mammals of the Western Division

of New South Wales: an ecological interpretation of the nineteenth century

historical record. The Rangeland Journal, 23(1), 44-70.

Lunt, I., & Spooner, P. G. (2005). Using historical ecology to understand patterns of

biodiversity in fragmented agricultural landscapes. Journal of Biogeography, 32,

1859-1873.

Lunt, I. D. (1995). European management of remnant grassy forests and woodlands in

south-eastern Australia. Victorian Naturalist, 112, 239-249.

References

117

Lunt, I. D. (2005). Effects of Stock Grazing on Biodiversity Values in Temperate Native

Grasslands and Grassy Woodlands in SE Australia:A Literature Review.

Canberra: Environment ACT.

Lunt, I. D. (2005). Technical Report 18. Effects of Stock Grazing on Biodiversity Values

in Temperate Native Grasslands and Grassy Woodlands in SE Australia:A

Literature Review. Canberra: Environment ACT.

Lunt, I. D., Jones, N., Spooner, P. G., & Petrow, M. (2006). Effects of European

colonization on indigenous ecosystems:post-settlement changes in tree stand

structures in Eucalyptus-Callitris woodlands in central New South Wales,

Australia. Journal of Biogeography, 33, 1102-1115.

Lusk, C. H., & Matus, F. (2000). Juvenile tree growth rates and species sorting on fine-

scale soil fertility gradients in a Chilean temperate rain forest. Journal of

Biogeography, 27(4), 1011-1020.

Lusk, C. H., Wright, I., & Reich, P. B. (2003). Photosynthetic differences contribute to

competitive advantage of evergreen angiosperm trees over evergreen conifers in

productive habitats. New Phytologist, 160, 329-336.

McCarthy, G.J., Tolhurst, K.G. & Chatto, K. (1999). Overall Fuel Hazard Guide. In:

Research Report No. 47, Department of Natural Resources and Environment,

Melbourne.

McDonald, R. C., Isbell, R. F., Speight, J. G., Walker, J., & Hopkins, M. S. (1990).

Australian Soil and Land Survey:Field Handbook (2nd ed.). Melbourne: Inkata

Press.

McKinney, M. L., & Lockwood, J. L. (1999). Biotic homogenization: a few winners

replacing many losers in the next mass extinction. Trends in Ecology &

Evolution, 14(11), 450-453.

McPherson, G. R., Wright, H. A., & Wester, D. B. (1988). Patterns of shrub invasion in

semiarid texas grasslands. American Midland Naturalist, 120(2), 391-397.

Meentemeyer, R. K., Moody, A., & Franklin, J. (2001). Landscape-scale patterns of

shrub-species abundance in California chaparral – The role of topographically

mediated resource gradients. Plant Ecology, 156(1), 19-41.

Meserve, P. L., Kelt, D. A., Milstead, W. B., & Gutierrez, J. R. (2003). Thirteen years

of shifting top-down and bottom-up control. Bioscience, 53(7), 633-646.

Meyer, S. E., & Pendleton, B., K. (2005). Factors affecting seed germination and

seedling establishment of a long-lived desert shrub (Coleogyne ramosissima:

Rosaceae). Plant Ecology, 178(2), 171-187.

References

118

Miller, E. A., & Halpern, C. B. (1998). Effects of environment and grazing disturbance

on tree establishment in meadows of the central Cascade Range, Oregon, USA.

Journal of Vegetation Science, 9(2), 265-282.

Moreno, J. M., & Oechel, W. C. (1989). A simple method for estimating fire intensity

after a burn in Californian chapparral. Acta Oecologica, 10(1), 57-68.

Mortlock, W., & Lloyd, M. (2001). Floradata-A Guide to Collection, Storage and

Propagation of Australian Native Plant Seed. Searchable Database February

2001. Canberra: Australian Centre for Mining Environmental Research

(Brisbane), Australian National Botanic Gardens, CSIRO Forestry and Forest

Products, Greening Australia Limited.

Murphy, B. P., & Bowman, D. M. J. S. (2007). Seasonal water availability predicts the

relative abundance of C3 and C4 grasses in Australia. Global Ecology &

Biogeography, 16(2), 160-169.

Myers, K., Parer, I., Wood, D., & Cooke, B. D. (1994). The rabbit in Australia. In H. V.

Thompson & C. M. King (Eds.), The European Rabbit (pp. 108-157). Oxford:

Oxford University Press.

Nicholls, N. (1992). Historical El Nino/Southern Oscillation variability in the

Australasian region. In H. F. Diaz & V. Markgraf (Eds.), El Nino: Historical

and paleo aspects of the Southern Oscillation (pp. 151-172). Cambridge:

Cambridge University Press.

Nicholls, N. (2004). The changing nature of Australian droughts. Climatic Change,

63(3), 323-336.

Niinemets, U. (2001). Global-scale climatic controls of leaf dry mass per area, density

and thickness in trees and shrubs. Ecology, 82(3), 453-469.

Noble, I. R., Bary, G. A. V., & Gill, A. M. (1980). McArthur's fire-danger meters

expressed as equations. Australian Journal of Ecology, 5, 201-203.

Noble, I. R., & Gitay, H. (1996). A functional classification for predicting the dynamics

of landscapes. Journal of Vegetation Science, 7, 329-336.

Noble, J. C. (1997). The delicate and noxious scrub. Canberra: CSIRO.

Noble, J. C. (2001). Regulating Callitris populations: tale of two pineries. In J. Dargavel

(Ed.), Perfumed Pineries:Environmental History of Australia's Callitris Forests

(pp. 173-183). Canberra: Australian National University.

Noble, J. C., & Vines, R. G. (1993). Fire studies in mallee (Eucalyptus spp.)

communities of western New South Wales: grass fuel dynamics and associated

weather patterns. The Rangeland Journal, 15(2), 270-297.

References

119

Norris, E. H., Mitchell, P. B., & Hart, D. M. (1991). Vegetation changes in the Pilliga

forests: a preliminary evaluation of the evidence. Vegetatio, 91, 209-218.

Noss, R. F., Franklin, J. F., Baker, W. L., Schoennagel, T., & Moyle, P. B. (2006).

Managing fire-prone forests in western United States. Frontiers in Ecology and

the Environment, 4(9), 481-487.

Noy-Meir, I. (1973). Desert ecosystems: environment and procedures. Annual Review of

Ecology & Systematics, 4, 25-51.

Ojeda, F. (1998). Biogeography of seeder and resprouter Erica species in the Gape

Floristic Region-Where are the resprouters? Biological Journal of the Linnean

Society, 63(3), 331-347.

Ojeda, F., Brun, F., G. , & Vergara, J., J. . (2005). Fire, rain and the selection of seeder

and resprouter life-histories in fire-recruiting, woody plants. New Phytologist,

168(1), 155-165.

Parer, I. (1987). Factors influencing the distrubtion and abundance of rabbits

(Oryctolagus cuniculus) in Queensland. Proceedings of the Royal Society of

Queensland, 98, 73-82.

Parker, D., & Lunt, I. D. (2000). Stand structure and recruitment patterns in Callitris-

Eucalyptus woodlands in Terrick Terrick National Park, Victoria. Victorian

Naturalist, 117, 207-213.

Paula, S., & Pausas, J. G. (2006). Leaf traits and resprouting ability in the

Mediterranean basin. Functional Ecology, 1-7.

Pausas, J. G., & Bradstock, R. A. (2007). Fire persistence traits of plants along a

productivity and disturbance gradient in mediterranean shrublands of south-east

Australia. Global Ecology & Biogeography, 16(3), 330-340.

Pausas, J. G., Bradstock, R. A., Keith, D. A., Keeley, J. E., & GCTE. (2004). Plant

functional traits in relation to fire in crown-fire ecosystems. Ecology, 85(4),

1085-1100.

Pausas, J. G., Carbó, E., Caturla, R. N., Gil, J. M., & Vallejo, R. (1999). Post-fire

regeneration patterns in the eastern Iberian Peninsula. Acta Oecologica, 20(5),

499-508.

Pausas, J. G., Llovet, J., Rodrigo, A., & Vallejo, R. (2008). Are wildfires a disaster in

the Mediterranean basin? A review. International Journal of Wildland Fire,

17(6), 713-723.

References

120

Pearson, R., G. , & Dawson, T., P. . (2003). Predicting the impacts of climate change on

the distribution of species: are bioclimate envelope models useful? Global

Ecology & Biogeography, 12(5), 361-371.

Peters, D. P. C., Bestelmeyer, B. T., Herrick, J. E., Fredrickson, E. L., Monger, H. C., &

Havstad, K. M. (2006). Disentangling complex landscapes: new insights into

arid and semiarid system dynamics. Bioscience, 56(6), 491-501.

Peterson, D. L., & Arbaugh, M. J. (1986). Postfire survival in Douglas-fir and lodgepole

pine:comparing the effects of crown and bole damage. Canadian Journal of


Pierce, L. L., Running, S. W., & Walker, J. (1994). Regional-scale relationships of leaf

area index to specific leaf area and leaf nitrogen content. Ecological

Applications, 4(2), 313-321.

Platt, W. J., Evans, G. W., & Rathbun, S. L. (1988). The population dynamics of a long-

lived conifer (Pinus palustris). The American Naturalist, 131(4), 491-525.

Pook, E. W. (1985). Canopy dynamics of Eucalyptus maculata Hook. III Effects of

drought. Australian Journal of Botany, 33(1), 65-79.

Pook, E. W., Costin, A. B., & Moore, C. W. E. (1966). Water stress in native vegetation

during the drought of 1965. Australian Journal of Botany, 14, 257-267.

Porteners, M. F. (1993). The natural vegetation of the Hay Plain:Booligal-Hay and

Deniliquin-Bendigo 1:250 000 maps. Cunninghamia, 3(1), 1-122.

Porteners, M. F., Ashby, E. M., & Benson, J. S. (1997). The natural vegetation of

Pooncarie 1:250 000 map. Cunninghamia, 5(1), 139-231.

Prior, L. D., Bowman, D. M. J. S., & Brook, B. W. (2007). Growth and survival of two

north Australian relictual tree species, Allosyncarpia ternata (Myrtaceae) and

Callitris intratropica (Cupressaceae). Ecological Research, 22(2), 228-236.

Prior, L. D., Eamus, D., & Bowman, D. M. J. S. (2003). Leaf attributes in the seasonally

dry tropics: a comparison of four habitats in northern Australia. Functional

Ecology, 17, 504-515.

Pugnaire, F. I., & Luque, M. T. (2001). Changes in plant interactions along a gradient of

environmental stress. Oikos, 93(1), 42-49.

Pulsford, I. F., Banks, J. C. G., Hodges, L., & Good, R. B. (1992). Land use history of

the white cypress pine forests in the snowy valley, Kosciusko National Park. In

P. Grenier (Ed.), Les Alpes Australiennse Review De Geographie Alpine (Vol.

Tome LXXX, pp. 86-104). Grenoble: Institut de Geographie Alpine.

References

121

Pye, M. G., Gadek, P. A., & Edwards, K. J. (2003). Divergence, diversity and species of

the Australasian Callitris (Cupressaceae) and allied genera: evidence from ITS

sequence data. Australian Systematic Botany, 16(4), 505-514.

Quinn, G. P., & Keough, M. J. (2002). Experimental design and data analysis for

biologists (1st ed.). Cambridge: Cambridge University Press.

Read, J. (1995). Recruitment characteristics of the white cypress pine (Callitris

glaucophylla) in arid South Australia. The Rangeland Journal, 17(2), 228-240.

Rebertus, A. J., Williamson, G. B., & Moser, E. B. (1989). Longleaf pine pyrogenicity

and turkey oak mortality in Florida xeric sandhills. Ecology, 70(1), 60-70.

Reich, P. B., Ellsworth, D. S., Walters, M. B., Vose, J. M., Gresham, C., Volin, J. C., et

al. (1999). Generality of leaf trait relationships:a test across six biomes. Ecology,

80(6), 1955-1969.

Reich, P. B., Walters, M. B., & Ellsworth, D. S. (1997). From tropics to tundra:global

convergence in plant functioning. Proceedings of National Academy of Sciences,

USA, 94, 13730-13734.

Reid, N., & Landsberg, J. (2000). Tree decline in agricultural landscapes:what we stand

to lose. In R. J. Hobbs & J. Yates Colin (Eds.), Temperate Eucalypt Woodlands

in Australia:Biology,Conservation, Management and Restoration (pp. 127-166).

Sydney: Surrey Beatty and Sons.

Richardson, D. M., & Bond, W. J. (1991). Determinants of plant distribution: evidence

from pine invasions. The American Naturalist, 137(5), 639-668.

Roderick, M. L., Berry, S. L., & Noble, I. R. (2000). A framework for understanding the

relationship between environment and vegetation on the surface area to volume

ratio of leaves. Functional Ecology, 14, 423-437.

Rodrigo, A., Quintana, V., & Retana, J. (2007). Fire reduces Pinus pinea distribution in

the northeastern Iberian peninsular. Ecoscience, 14(1), 23-30.

Rodrigo, A., Retana, J., & Pico, F. X. (2004). Direct regenerationis not the only

response of mediterranean forests to large fires. Ecology, 85(3), 716-729.

Russell-Smith, J. (2006). Recruitment dynamics of the long-lived obligate seeders

Callitris intratropica (Cupressaceae) and Petraeomyrtus punicea (Myrtaceae).

Australian Journal of Botany, 54(5), 479-485.

Sankaran, M., Hanan, N. P., Scholes, R. J., Ratnam, J., Augustine, D. J., Cade, B. S., et

al. (2005). Determinants of woody cover in African savannas. Nature,

438(7069), 846-849.

References

122

Sankaran, M., Ratnam, J., & Hanan, N. (2008). Woody cover in African savannas: the

role of resources, fire and herbivory. Global Ecology and Biogeography, 17(2),

236-245.

Scanlan, J. C., Berman, D. M., & Grant, W. E. (2006). Population dynamics of the

European rabbit (Oryctolagus cuniculus) in north eastern Australia: Simulated

responses to control. Ecological Modelling, 196(1-2), 221-236.

Scarff, F. R., & Westoby, M. (2006). Leaf litter flammability in some semi-arid

Australian woodlands. Functional Ecology, 20, 745-752.

Scheffer, M., Carpenter, S., Foley, J. A., Folke, C., & Walker, B. (2001). Catastrophic

shifts in ecosystems. Nature, 413(6856), 591-596.

Scholes, R. J., & Archer, S. R. (1997). Tree-grass interactions in savannas. Annual

Review of Ecology and Systematics, 28(1), 517-544.

Schreiber, E.S.G., Bearlin, A.R., Nicol, S.J. & Todd, C.R. 2004. Adaptive

management: a synthesis of current understanding and effective

application. Ecological Management & Restoration 5: 177-182.

Schulze, E., Kelliher, F. M., Korner, C., Lloyd, J., & Leuning, R. (1994). Relationships

among maximum stomatal conductance, ecosystem surface conductance, carbon

assimilation rate, and plant nitrogen nutrition: a global ecology scaling exercise.

Annual Review of Ecology and Systematics, 25(1), 629-662.

Schulze, E. D., Williams, R. J., Farquhar, G. D., Schulze, W., Langridge, J., Miller, J.

M., et al. (1998). Carbon and nitrogen isotope discrimination and nitrogen

nutrition of trees along a rainfall gradient in northern Australia. Functional Plant

Biology, 25(4), 413-425.

Schwilk, D. W. (2003). Flammability is a niche construction trait: canopy architecture

affects fire intensity. The American Naturalist, 162(6), 725-733.

Schwilk, D. W., & Ackerly, D. D. (2001). Flammability and serotiny as

strategies:correlated evolution in pines. Oikos, 94, 326-336.

Schwinning, S., & Sala, O. E. (2004). Hierarchy of responses to resource pulses in arid

and semi-arid ecosystems. Oecologia, 141(2), 211-220.

Schwinning, S., Sala, O. E., Loik, M. E., & Ehleringer, J. R. (2004). Thresholds,

memory, and seasonality: understanding pulse dynamics in arid/semi-arid

ecosystems. Oecologia, 141(2), 191-193.

Scott, J. A. (1992). The natural vegetation of the Balranald-Swan Hill area.

Cunninghamia, 2(4), 597-652.

References

123

Shinneman, D. J., & Baker, W. L. (2009). Historical fire and multidecadal drought as

context for pinon juniper woodland restoration in western Colorado. Ecological

Applications, 19(5), 1231-1245.

Sivertsen, D., & Clarke, P. J. (2000). Temperate woodlands in New South Wales:a brief

overview of distribution, composition and conservation. In R. J. Hobbs & J.

Yates Colin (Eds.), Temperate Eucalypt Woodlands in

Australia:Biology,Conservation, Management and Restoration (pp. 6-16).

Sydney: Surrey Beatty and Sons.

Smith, R. E., van Wilgen Brian, W., Forsyth, G. G., & Richardson, D. M. (1992).

Coexistence of seeders and resprouters in a fire-prone environment:the role of

ecophysiology and soil moisture. In W. van Wilgen Brian, D. M. Richardson, F.

J. Kruger & H. J. van Hensbergen (Eds.), Fire in South Africa Mountain Fynbos

(pp. 108-122). Berlin: Springer-Verlag.

Sparrow, A. D., & Bellingham, P. J. (2001). More to resprouting than fire. Oikos, 94(1),

195-197.

Specht, R. L. (1981). Responses to fires of heathlands and related shrublands. In A. M.

Gill, R. H. Groves & I. R. Noble (Eds.), Fire and the Australian Biota.

Canberra: Australian Academy of Science.

Specht, R. L., & Specht, A. (1999). Australian Plant Communities:dynamics of

structure, growth and biodiversity. Melbourne: Oxford University Press.

Squeo, F. A., Holmgren, M., Jimenez, M., Alban, L., Reyes, J., & Gutierrez, J. R.

(2007). Tree establishment along an ENSO experimental gradient in the

Atacama desert. Journal of Vegetation Science, 18(2), 195-202.

Steel, G. D., & Torrie, J. H. (1960). Principles and Procedures of Statistics with Special

Reference to Biologivcal Sciences. London: McGraw-Hill Book Company, Inc.

Suding, K. N., & Hobbs, R. J. (2009). Threshold models in restoration and

conservation: a developing framework. Trends in Ecology & Evolution, 24(5),

271-279.

Tabachnick, B. G., & Fidell, L. S. (2007). Using Multivariate Statistics (5th ed.).

Boston: Pearson Education Inc.

Thomas, C. D., Cameron, A., Green, R. E., Bakkenes, M., Beaumont, L. J., Collingham,

Y. C., et al. (2004). Extinction risk from climate change. Nature, 427(6970),

145-148.

References

124

Thompson, J. R., & Spies, T. A. (2009). Vegetation and weather explain variation in

crown damage within a large mixed-severity wildfire. Forest Ecology and

Management, 258(7), 1684-1694.

Thompson, W. A., & Eldridge, D. A. (2005). White cypress pine (Callitris

glaucophylla):a review of its roles in landscape and ecological processes in

eastern Australia. Australian Journal of Botany, 53, 555-570.

Thuiller, W., Araujo, M. B., & Lavorel, S. (2003). Generalized models vs. classification

tree analysis: predicting spatial distributions of plant species at different scales.

Journal of Vegetation Science, 14(5), 669-680.

Tilman, D. (1988). Plant Strategies:Dynamics and Structure of Plant Communities.

Princeton: Princeton University Press.

Tiver, F., & Andrew, M. H. (1997). Relative effects of herbivory by sheep, rabbits,

goats and kangaroos on recruitment and regeneration of shrubs and trees in

eastern South Australia. Journal of Applied Ecology, 34(4), 903-914.

Trabaud, L. (1991). Fire regimes and phytomass growth dynamics in a Quercus

coccifera garrigue. Journal of Vegetation Science, 2(3), 307-314.

Turner, M. G., Baker, W. L., Peterson, C. J., & Peet, R. K. (1998). Factors Influencing

Succession: Lessons from Large, Infrequent Natural Disturbances. Ecosystems,

1(6), 511-523.

Turner, M. G., Romme, W. H., & Gardner, R. H. (1999). Prefire heterogeneity, fire

severity, and early postfire plant reestablishment in subalpine forests of

Yellowstone National Park, Wyoming. International Journal of Wildland Fire,

9(1), 21-36.

Turner, M. G., Romme, W. H., & Tinker, D. B. (2003). Surprises and lessons from the

1988 Yellowstone fires. Frontiers in Ecology and Environment, 1(7), 351-358.

van de Koppel, J., Rietkerk, M., & Weissing, F. J. (1997). Catastrophic vegetation shifts

and soil degradation in terrestrial grazing systems. Trends in Ecology &

Evolution, 12(9), 352-356.

van Dyne, G. M., Brokington, N. R., Sczocs, Z., Duek, J., & Ribic, C. A. (1980). Large

herbivore subsystem. In A. I. Breymeyer & G. M. van Dyne (Eds.), Grasslands,

Systems analysis and Man (ist ed., pp. 270-516). Cambridge: Cambridge

University Press.

Van Wagner, C. E. (1973). Height of crown scorch in forest fires. Canadian Journal of


References

125

van Wilgen, B. W., Higgins, K. B., & Bellstedt, D. U. (1990). The role of vegetation

structure and fuel chemistry in excluding fire from forest patches in the fire-

prone fynbos shrublands of South Africa. Journal of Ecology, 78, 210-222.

Veall, M. R., & Zimmermann, K. F. (1996). Pseudo-R2 measures for some limited

dependent variable models Journal of Economic Surveys, 10, 241-259.

Veblen, T. T. (1992). Regeneration dynamics. In D. C. Glenn-Lewin, R. K. Peet & H.

Veblen Thomas (Eds.), Plant Succession: Theory and Prediction (ist ed., pp.

152-187). London: Chapman & Hall.

Venables, W. N., & Ripley, B. D. (2002). Statistics and Computing (4th ed.). New

York: Springer-Verlag.

Vesk, P., A. , Leishman, M., R. , & Westoby, M. (2004). Simple traits do not predict

grazing response in Australian dry shrublands and woodlands. Journal of

Applied Ecology, 41(1), 22-31.

Vesk, P. A., & Westoby, M. (2003). Drought damage and recovery-a conceptual model.

New Phytologist, 160, 7-14.

Vilagrosa, A., Cortina, J., Gil-PelegrÃn, E., & Bellot, J. (2003). Suitability of drought-

Preconditioning techniques in mediterranean climate. Restoration Ecology,

11(2), 208-216.

Walker, B. H. (1993). Rangeland ecology: understanding and managing change. Ambio,

22(2/3), 80-87.

Wardell-Johnson Grant, W., Williams, J. E., Hill, K. D., Cumming, R., & Woinarski, J.

(1997). Evolutionary biogeography and contemporary distribution of eucalypts.

In J. Williams (Ed.), Eucalypt Ecology:Individuals to Ecosystems (pp. 92-128).


Warner, R. R., & Chesson, P. L. (1985). Coexistence mediated by recruitment

fluctuations:a field guide to the storage effect. The American Naturalist, 125(6),

769-787.

Watson, I. W., Westoby, M., & Holm, A. M. (1997). Continuous and episodic

components of demographic change in arid zone shrubs:models of two

Eremophila species from Western Australia compared with published data on

other species. Journal of Ecology, 85, 833-846.

Wells, P. M., & Hill, R. S. (1989). Fossil imbricate-leaved Podocarpaceae from Tertiary

sediments in Tasmania. Australian Systematic Botany, 2, 387-423.

Westoby, M. (1999). The LHS scheme in relation to grazing and fire. Proceedings of

6th International Rangeland Congress, Townsville, 2, 893-896.

References

126

Westoby, M., Walker, B., & Noy-Meir, I. (1989). Opportunistic management for

rangelands not at equilibrium. Journal of Range Management, 42(4), 266-274.

Whelan, R. J. (1995). The ecology of fire. Cambridge: Cambridge University Press.

Whelan, R. J., De Jong, N. H., & Von Der Burg, S. (1998). Variation in bradyspory and

seedling recruitment without fire among populations of Banksia serrata

(Proteaceae). Australian Journal of Botany, 23, 121-128.

Whipp, R. K. (2009). Historical vegetation change in relation to forest management in

the Pilliga State Forests of northern NSW, Australia. . Unpublished PhD thesis,

Charles Sturt University, Albury.

Wiegand, K., Jeltsch, F., & Ward, D. (2004). Minimum recruitment frequency in plants

with episodic recruitment. Oecologia, 141(2), 363-372.

Williams, J., & Woinarski, J. (1997). Eucalypt Ecology:Individuals to Ecosystems.


Williams, J. E., & Brooker, I. H. (1997). Eucalypts:an introduction. In J. Williams & J.



Williams, J. E., Whelan, R. J., & Gill, A. M. (1994). Fire and environmental

heterogeneity in southern temperate forest ecosystems: implications for

management. Australian Journal of Botany, 42(2), 125-137.

Williamson, G. B., & Black, E. M. (1981). High temperature of forest fires under pines

as a selective advantage over oaks. Nature, 293(5834), 643-644.

Wilson, A. D. (1990). The effect of grazing on Australian ecosystems. Proceedings of

Ecological Society of Australia, 16, 235-244.

Wright, C. S., & Agee, J. K. (2004). Fire and vegetation history in the Eastern Cascade

Mountains, Washington. Ecological Applications, 14(2), 443-459.

Wright, I. J., Reich, P. B., & Westoby, M. (2001). Strategy shifts in leaf physiology,

structure and nutrient content between species of high- and low-rainfall and

high- and low-nutrient habitats. Functional Ecology, 15, 423-434.

Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z., Bongers, F., et al.

(2004). The worldwide leaf economics spectrum. Nature, 428, 821-827.

Zavala, M. A., & Zea, E. (2004). Mechanisms maintaining biodiversity in

Mediterranean pine-oak forests: insights from a spatial simulation model. Plant

Ecology, 171, 197-207.

References

127

Zeppel, M., & Eamus, D. (2008). Coordination of leaf area, sapwood area and canopy

conductance leads to species convergence of tree water use in a remnant

evergreen woodland. Australian Journal of Botany, 56(2), 97-108.

How are the dynamics of woodland dominants influenced by ...

Documents

Transcript of How are the dynamics of woodland dominants influenced by ...