Wildlife, Fire and Future Climate

A Forest Ecosystem Analysis

Brendan B. Mackey, David D. Lindenmayer, Malcolm M. Gill, Michael M. McCarthy, Janette J. Lindesay, Brendan Mackey, David Lindenmayer, Malcolm Gill, Michael McCarthy, Janette Lindesay

• Publisher: CSIRO PUBLISHING
• Year: 2002
• ISBN:
  9780643099852
• Topic: Biowissenschaften
• Subtopic: Biologie

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.

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 c...