Wildlife, Fire and Future Climate: A Forest Ecosystem Analysis

By CSIRO PUBLISHING

The conservation of Earth's forest ecosystems is one of the great environmental challenges facing humanity in the 21st century. All of Earth's ecosystems now face the spectre of the accelerated greenhouse effect and rates of change in climatic regimes that have hitherto been unknown. In addition, multiple use forestry – where forests are managed to provide for both a supply of wood and the conservation of biodiversity – can change the floristic composition and vegetation structure of forests with significant implications for wildlife habitat.

Wildlife, fire and future climate: a forest ecosystem analysis explores these themes through a landscape-wide study of refugia and future climate in the tall, wet forests of the Central Highlands of Victoria. It represents a model case study for the kind of integrated investigation needed throughout the world in order to deal with the potential response of terrestrial ecological systems to global change. The analyses presented in this book represent one of the few ecosystem studies ever undertaken that has attempted such a complex synthesis of fire, wildlife, vegetation, and climate.

Wildlife, fire and future climate: a forest ecosystem analysis is written by an experienced team of leading world experts in fire ecology, modelling, terrain and climate analysis, vegetation and wildlife habitat. Their collaboration on this book represents a unique and exemplary, multi-disciplinary venture.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Jul 1, 2002
• ISBN: 9780643099852

Book preview

Wildlife, Fire &

Future Climate

A Forest Ecosystem Analysis

Wildlife, Fire &

Future Climate

   A Forest Ecosystem Analysis

Brendan Mackey, David Lindenmayer,

Malcolm Gill, Michael McCarthy

and Janette Lindesay

© 2002 Brendan Mackey, David Lindenmayer, Malcolm Gill, Michael McCarthy and Janette Lindesay

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

National Library of Australia Cataloguing-in-Publication entry

Wildlife, fire & future climate: a forest ecosystem analysis.

Bibliography.

Includes index.

ISBN 0 643 06756 6 (paperback).

ISBN 0 643 09004 5 (eBook).

1. Forest ecology – Victoria – Central Highlands Region. 2. Forest fires – Victoria – Central Highlands Region. 3. Climatic change – Victoria – Central Highlands Region. I. Mackey, Brendan.

333.7509945

Available from

CSIRO Publishing

150 Oxford Street (PO Box 1139)

Collingwood VIC 3066

Australia

Front cover photo by Bill Incoll

Set in Minion, Frutiger

Cover design by Melissa Gibson

Typeset by Desktop Concepts P/L

Printed in Australia by Brown Prior Anderson

Contents

Preface and Acknowledgements

Chapter 1: Forest Refugia

Introduction

Forest Conservation: A Burning Question

Chapter Outline and Overview of Methods

Chapter 2: Defining Forest Refugia for Arboreal Marsupials

Introduction

Origin of the Term ‘Refugia’

Threatening Processes

Habitat Requirements of Arboreal Marsupials

Importance of Forest Structure to Habitat

Some Working Hypotheses

Conclusions

Chapter 3: Local Extinction or Persistence of Mountain Ash due to Fire Regimes

Introduction

Mortality of Trees

Estimating Mean Fire Interval: Methods

Persistence of Mountain Ash

Estimating Mean Fire Interval: Results

Discussion

Conclusions

Chapter 4: Environmental Controls on Vegetation Structure and Fire Regimes

Introduction

Overview of Methods

Site-based Plot Analyses

Mapped Vegetation Analysis

Spatial Models of Potential Refugia

Discussion and Conclusions

Chapter 5: Climate and Fire in the Central Highlands of Victoria

Introduction

Overview of Dominant Climatic Processes

Climate Record

Linking Weather and Fire

Future Climate Scenarios

Discussion and Conclusions

Chapter 6: Future of Refugia in Mountain Ash Forests of the Central Highlands of Victoria

Introduction

Do Refugia Exist?

Implications for Forest Management

New Silvicultural Practices in Mountain Ash Forests

Policy Implications

Implications for Modelling and Data Shortages

Conclusions

Appendix 1: Spatial Database Development

Introduction

Digital Elevation Model (DEM)

Terrain and Hydrological Attributes (TAPES_G)

Compound Terrain Index (TWI)

Solar Radiation Attributes (SRAD)

Meso-scaled Climatic Data

Site-based Field Data

Appendix 2: Spatial Climate Models

Fitting Climate Surfaces

Data Quality Considerations

Using the Climate Surfaces

References

Index

Preface and Acknowledgements

This book represents the collaborative efforts of a research team over a three-year period. Different team members took a leadership role in undertaking and documenting the research reported in each of the book’s six chapters. In addition, the research in each chapter was undertaken in association with different research assistants:

•  Brendan Mackey and David Lindenmayer were the lead authors for Chapter 1;

•  Brendan Mackey and David Lindenmayer were the lead authors for Chapter 2 in association with Jocelyn Bell;

•  Michael McCarthy and Malcolm Gill were the lead authors for Chapter 3 in association with Sebastian Lang and Peter H.R. Moore;

•  Brendan Mackey and Michael McCarthy were the lead authors for Chapter 4 in association with Ian Mullen and David Lindenmayer. Statistical analysis of the site data in Chapter 4 was undertaken in association with Ross B. Cunningham and Christine F. Donnelly;

•  Janette Lindesay, Malcolm Gill, Michael McCarthy and Brendan Mackey were the lead authors for Chapter 5 in association with Aidan Heerdegan and Peter Moore.

•  Brendan Mackey and David Lindenmayer were the lead authors for Chapter 6.

Figures 4.1 and 4.2 are reprinted from ‘Factors affecting stand structure in forests – are there climatic and topographic determinants?’ by D.B. Lindenmayer, B.G. Mackey, I.C. Mullen, M.A. McCarthy, M.A. Gill, R.B. Cunningham and C.F. Donnelly, Forest Ecology and Management, 123 1999, pages 55–63, with permission of Elsevier Science. Chapter 3 is based in part on material from McCarthy et al. (1999) and is reprinted with permission of Elsevier Publications.

The core research presented in this book was resourced by a grant from the Australian Greenhouse Office. We are most grateful for its generous support. Andrew Taplin from Environment Australia helped catalyse the original project this book was based on and provided excellent support in his role as departmental liaison officer. We are also grateful for critical comment on an earlier version of this text received from Jann Williams and Dan Mckenney. However, responsibility for the content of this book, including any material that addresses policy and management implications, lies entirely with the authors. We are grateful to the Centre for Plant Biodiversity Research, CSIRO Plant Industry, for a summer scholarship award that enabled Sebastian Lang to assist in this research.

Thanks are due to Clive Hilliker for his technical assistance in laying out many of the graphs and tables. At various times during the life of this project, Gemma Waldendorp, Lauren Jones, Ryan Incoll and Maria Arnold provided much-valued production assistance. Ian Mullen and Aidan Herdeegan took a lead role in producing the figures and tables in Chapters 4 and 5. Ian Mullen was also responsible for most of the spatial database development reported in Appendix 1, while Aidan Herdeegan undertook the climatic spatial interpolation work reported in Appendix 2. We acknowledge with thanks the invaluable advice on spatial interpolation received from Michael Hutchinson. Thanks also to John Gallant for sharing his expertise in digital terrain analysis.

The research reported in this book owes much to the lifelong dedication of David Ashton and his pioneering work on the ecology of ash forests. David Ashton has spent more than 50 years studying Mountain Ash forests in Central Victoria, and it is difficult to adequately convey the extent and impact of his scientific contribution. Our research has benefited greatly from the many insights and suggestions he has so willingly shared. We warmly acknowledge here his contributions to both our work and Australian ecology.

CHAPTER 1

Forest Refugia

INTRODUCTION

Many of the world’s forest ecosystems are facing a range of external pressures including the (human-driven) enhanced greenhouse effect and forestry practices. The former threatens the resilience of forest ecosystems (sensu Hollings 1996), while the latter tends to produce forests that are structurally and floristically less complex and more uniform than unmanaged forests (Lindenmayer & Franklin 1997). In this book we examine the significance of environmental heterogeneity and temporal variability for the conservation of arboreal marsupials in the face of possible future climates. The focus of our investigations was the role of refugia in protecting fauna populations from inappropriate fire regimes, and the potential impact on refugia of global climate change.

The forest ecosystems of the Central Highlands of Victoria were selected as the case study to examine environmental heterogeneity, refugia and future climate in relation to the species of arboreal marsupials that occur there – species of significant conservation and forest management concern. The Central Highlands of Victoria lie about 120 km north-east of the city of Melbourne and cover approximately 1° of latitude and 1° of longitude (Figure 1.1). The region experiences mild, humid winters with occasional periods of snow. Summers are generally cool. The long-term average annual precipitation ranges from 600–1600 mm, with annual mean maximum temperatures ranging from 8° to 20°C, and temperature minimums from 4° to 10°C (see Chapter 5). The landscape comprises mountainous terrain, with steeply sloping valleys of over 1000 m relief. The region is therefore environmentally heterogeneous in terms of both climatic and topographic gradients.

This visually spectacular landscape is further characterised by a complex matrix of structurally and taxonomically diverse vegetation types, especially tall wet Eucalyptus forest ecosystems, including Mountain Ash (Eucalyptus regnans F. Muell), Alpine Ash (Eucalyptus delegatensis R. Baker) and Shining Gum (Eucalyptus nitens Maiden). These are interdigitated with patches of cool temperate Myrtle Beech (Nothofagus cunninghamii [Hook] Oerst.) rainforest and Acacia species (Mueck 1990). Collectively, these tall wet Eucalyptus forests are referred to as montane ash forests. Mountain Ash forests predominate among the tall wet Eucalyptus forests of the Central Highlands of Victoria. As discussed in Chapter 3, individuals of this tree species often die when their leafy canopies are killed or scorched by fire, and the species usually regenerates only from seed shed following fire. Stand-replacing fires can lead to the growth of even-aged stands, which are common in these landscapes. This is unusual, as most Australian Eucalyptus forests are dominated by multi-species and multi-aged stands.

Arboreal marsupial populations of the Central Highlands of Victoria have been the focus of considerable research since the early 1990s. Of particular relevance is a suite of projects focused on arboreal marsupials, such as Leadbeater’s Possum (Gymnobelideus leadbeateri McCoy), that investigated their biology, population dynamics and habitat requirements (reviewed by Lindenmayer 2000). These studies established quantitative relationships between the distribution and abundance of arboreal marsupials and vegetation structure and floristics, patch size and shape, and the spatial configuration of suitable habitat patches. This extensive background of habitat information was a major factor influencing our choice to focus on these fauna in this book.

Figure 1.1: Location of the Central Highlands of Victoria study area. Also shown are the Maroondah and O’Shannassy catchments where much of the analysis in later chapters was focused.

The results of the empirical analyses presented in this book cannot necessarily be readily extrapolated beyond the study region, partly because of the unique ecological features that characterise the Central Highlands of Victoria. Nonetheless, all terrestrial ecosystems are subject to the laws of physics, chemistry, biology and ecology (although it can be argued that the science of ecology is characterised by empirically based generalisations rather than universally held ‘laws’). By definition, all forest ecosystems are characterised and dominated by a canopy of trees, which control sub-canopy ecosystem processes and are subject to disturbance regimes. Therefore, in addition to their unique ecological characteristics, the forest ecosystems of the Central Highlands of Victoria share with all other forest types many similar biophysical processes and land use pressures. Hence, the concepts that underpin this study, and the methodology we developed and applied, are directly relevant to many forested landscapes ecosystems elsewhere in Australia and around the world. For example, workers in the tall wet forests of western North America will find a particular resonance with the challenges facing the forests of the Central Highlands of Victoria, and with the results of our investigations.

FOREST CONSERVATION: A BURNING QUESTION

Throughout the world, forest conservation is a controversial issue. About 50% of Earth’s forests have been destroyed, often to make way for agriculture. Western Europe, the eastern United States and much of China and India were once dominated by forest ecosystems (WCMC 1998). Australia has lost about half its forests since the arrival of Europeans (Resource Assessment Commission 1992). Deforestation remains a critical problem in many developing nations, such as Brazil and the Democratic Republic of Congo, that contain the bulk of the remaining tropical forests (Fearnside 1987, 2001).

Even when forests are protected in national parks and nature reserves there is ongoing debate about how those remaining ecosystems should be managed. Gorshkov (1995), for example, argued that prohibiting any kind of modern land use activity is a prerequisite to ensuring the proper ecological functioning of forest ecosystems. Other political commentators have argued that the ecological integrity of forest ecosystems can only be guaranteed through active human intervention, specifically timber harvesting and prescribed burning (Tuckey 2001). The question of what constitutes ‘good’ forest conservation management largely hinges on the extent to which anthropogenic activities ‘mimic’ or diverge from natural disturbance regimes (Hunter 1994; Lindenmayer & McCarthy 2002).

The Eucalyptus-dominated forest ecosystems of Australia evolved over 60 million years, following the fragmentation of Gondwanaland and Australia’s subsequent slow drift northward. As the climate became warmer and drier, the cool temperate Gondwanic forests gave way to the contemporary Eucalyptus- and Acacia-dominated vegetation formations (White 1994). On mainland Australia, Gondwanic genera such as Nothofagus and Eucryphia still occur, but usually as small islands within extensive stands of sclerophyll forest (Read & Brown 1996). Forest-dependent fauna, such as the endemic species of Australian arboreal marsupials, also evolved during the last 20 million years (Archer & Gothelp 1991). As discussed in Chapter 2, in Australia’s Eucalyptus-dominated forests, specific habitat relationships often exist between the fauna and vegetation structure and composition.

The oldest reliable records of human occupation in Australia date between 30 000 and 50 000 years before present (Roberts et al. 1994). Clearly, Australian forest species evolved long before the evolution of humans, let alone their arrival in Australia. However, in the absence of humans, forest ecosystems are still subject to natural disturbances, including climate change, fire, earthquake and storms (Franklin et al. 2000). The last two can be locally important in some landscapes and cause great physical change to a site. Fires are an integral component of many forest ecosystems, although the ecosystems experience widely different fire regimes. Climate change potentially affects forest ecosystems throughout the world. Indeed, climatic conditions that affect the growth of forest trees and plants and related ecosystem processes have been demonstrated to change substantially over time scales ranging from decades to millennia (Moss & Kershaw 2000). Given that forest ecosystems can be conceived as dynamic ecological entities, why be concerned about the impact of human land use activities – are they not just another in a list of external perturbations? From this perspective, it is valid to ask whether timber harvesting and forest management could be conducted in ways that resemble, or at least are more convergent with, natural disturbances.

What is often important is not the impact of a single disturbance but the pattern of events over a period of time. Hence, we are interested in not the significance of a single disturbance event but in defining the disturbance regime (Gill 1975; Agee 1993; Ashton & Martin 1996a, 1996b). Of interest are the impacts of fire regimes on vegetation and, in turn, the habitat of arboreal marsupials. However, this kind of simple linear thinking is misleading and must be avoided. Fire, climate and vegetation are interdependent phenomena and should not be examined linearly or in isolation from each other.

The precise characteristics of these interdependencies are discussed in later chapters, but there are some salient points to mention here. A forest fire will only ignite and spread if certain physical and biological conditions are met – these relate to prevailing weather conditions and the availability of fuel (McArthur 1967). For example, it is impossible for a surface fire to ignite and spread if conditions are sodden. It is important to note that what we call ‘fuel’ is the dead biomass (leaves, bark, woody debris) produced by plant photosynthesis that has yet to decompose. Both the type of plant species and the litter that occur in a location, together with the rate of photosynthesis and biomass production and decomposition, are a function of prevailing climatic conditions. Therefore, the dominant species composition in a forest and the rates of production and decomposition can change when climate regimes shift (Mackey & Sims 1993).

While there are logical connections between climate, fire, vegetation and fauna habitat, the precise nature of these relationships and how they unfold in a given landscape remains poorly understood. Ongoing empirical investigations into case study landscapes are needed. Although ecological investigations are traditionally restricted to small spatial scales, there is increasing recognition of the need to investigate ecological pattern and process across space/time scales, including at the landscape scale (Allen & Hoekstra 1992; Wiens 1999). This is particularly the case when investigating the conservation status and viability of forest-dependent fauna. Of interest here is the landscape-wide pattern of potentially suitable habitat and its occupancy by populations of the target species. Given that over the last 60 million years most of Australia’s forest ecosystems have experienced some kind of fire regime, various hypotheses have been proposed to explain the persistence of wildlife populations.

One hypothesis suggests that the persistence of arboreal marsupial populations (in the face of the perturbations caused by fire regimes) is due to the presence of refugia (although this hypothesis has not been formally defined in the scientific literature, to our knowledge, it is part of the lexicon of forest management science). The Oxford Dictionary (Hawker 1996) defines a refuge as a shelter from pursuit or danger. In the context of this book, ‘refugia’ are locations within a landscape that protect a population of arboreal marsupials from the effects of inappropriate fire regimes. Such refugia are considered particularly critical to the persistence of fauna in forest ecosystems that experience intense, stand-killing fires (fires of such intensity that all the trees in a stand are killed; also called stand-replacing fires). The existence (if proven) of such refugia would have significant implications for forest management due to their special conservation value.

There is long-standing public concern in Australia and throughout the world about forest and biodiversity conservation, particularly for populations of arboreal marsupials (Lindenmayer 1996). Significant public funds have been invested in reviews of the forest sector. One of the most recent was the Regional Forest Agreement Process (Commonwealth of Australia 1999). A significant component of this process was a review of the comprehensiveness, adequacy and representativeness of the forest reserve system. ‘Comprehensiveness’ and ‘representativeness’ refer to the extent to which classes of biodiversity are protected within the forest reserve system. The former term refers to the percentage reserved of broad vegetation associations and the latter term the percentage reserved of the variation within vegetation associations.

The criterion of adequacy is concerned with the efficacy of the reserve system in the long-term conservation of, among other things, viable populations of arboreal marsupials. The adequacy of a reserve system is commonly assessed using attributes such as the size, shape and connectivity of the reserved areas. However, if refugia locations are critical to the long-term persistence of animal populations in a landscape, their current and possible future locations must be factored into the reserve design process. Although it has been recognised that the existence of refugia is a potentially critical issue in the adequacy of conservation reserves, not until this study were data and analytical methods available to enable this factor to be included within conservation assessments. For example, Baker (1992) noted that very few studies in any ecosystem examined the quantitative aspects of factors influencing the patch mosaic (sensu Forman 1995).

Any analysis of fire, vegetation, habitat and, in turn, the existence of refugia, demands consideration of the effects of global climate change induced by anthropogenic perturbation of the global carbon cycle – the so-called greenhouse problem. Recent reviews published by the Intergovernmental Panel on Climate Change (IPCC 2001) have established the scientific basis for projected changes over the coming decades in both temperature and precipitation regimes, particularly in relation to the likelihood of increased climatic variability and frequency of extreme events. Given the (albeit poorly defined) relationships between climate, fire, vegetation and wildlife habitat, any study of refugia would be foolish to ignore the actual and potential impacts of greenhouse-forced climate change.

Unfortunately, investigating ecological phenomena at the landscape scale, including climate change effects, is not straightforward. Even in the absence of accelerated greenhouse effects, climate is inherently variable in space and time, so distinguishing between current and future patterns of variability, and their significance at the landscape scale, is problematic. Nonetheless, we had little choice but to accept this analytical challenge given the overarching importance of current and future climatic regimes to landscape ecosystem pattern and process.

In summary, we were faced with a set of interrelated problems. First, can the existence of refugia for arboreal marsupials be empirically demonstrated? Given this, how are they related to fire regimes, and to what extent are they a function of environmental heterogeneity and temporal variability? Addressing these questions required examining variation in fire intensity and between-fire interval, the relative controls of climate and topography on fire regimes, and how these interact to influence the spatial patterning of vegetation. The next step was to take scenarios for future climate and investigate how they might affect the potential occurrence of refugia.

Although the Central Highlands of Victoria was selected because of the quantity and quality of information on the region, significant data gaps remain. Indeed, a generic problem in landscape-wide ecological studies is the lack of observational data about fundamental processes over significant time periods (this is actually a general problem for ecology, see Maurer 1999). There are clearly limits in our ability to untangle the ecological complexity of landscapes in data-sparse regions. A critical phase of our investigations