Download Fire Management for Biodiversity

F I R E AND
B I R DS
Fire Management for Biodiversity
Compiled by
Penny Olsen and Michael Weston
Supplement to Wingspan, vol. 15, no. 3,
September 2005
2
Fire and Birds
Fire and Birds
3
‘Fire is a force in terrestrial ecosystems that is equalled only by the impact of humans as they transform,
fragment and degrade the Earth’s vegetation… Regardless of the changes wrought, fire is a central process in
creating and maintaining ecosystem patterns and processes in most, if not all, Australian terrestrial ecosystems.’
John Woinarski & Harry Recher 1997
Fire is part of Australians’ heritage—it has and continues to shape much of our natural world. Historically,
a fine-scale mosaic of different aged burns predominated, but since the early 20th century there has been
an increase in hotter, extensive fires, more threatening from both a human and biodiversity perspective.
Such inappropriate fire regimes, including the exclusion of fire, threaten over 50 bird species and subspecies
and, since European settlement, have contributed to the extinction of two species. Yet inappropriate
management of fire is one of the most tractable of the major threats to Australia’s wildlife. To some extent
the old adage ‘fight fire with fire’ applies. Used well, fire is an effective, economical tool for land managers.
Reintroduction of occasional fire into some landscapes, and return to a finer mosaic of burning, will
not prevent wildfires; it may, however, reduce their impact, by maintaining fire-dependent habitat and
protecting fire-sensitive birds. There will always be uncertainty, but by striving for practical and effective fire
management objectives we can better manage the risks to both human life and property, and biodiversity.
Fire regime—the pattern of fire in a particular area over a number of years, including seasonality,
frequency, interval between fires, intensity, extent and patchiness
Above: Black Kites at a grassfire in the Kimberley, Western Australia, catching animals flushed by the smoke and flames. Photo by Dennis Sarson, Lochman Transparencies
Left: The habitat of the Southern Emu-wren is vulnerable to wildfires in the dry seasons, but also threatened by deliberate burning for pasture. Photo by Nicholas Birks
KEY POINTS
• Fire is an essential, natural force in the
Australian landscape. Every year the
continent has a large number of wildfires,
caused by lightning or humans. Most are
not regarded as disasters.
• Biodiversity loss is associated with high
fire frequency, intense broadscale fire,
and fire exclusion, all of which tend to
homogenise the landscape. Increasing
loss of habitat heterogeneity from the
landscape is the single most damaging
effect of poor fire management.
• Some 50 bird taxa and their habitats
are currently threatened by inadequate,
ill-informed fire management.
• The response of birds to fire is influenced by
their life history characteristics and habitat
needs. Hence, individual species may vary
in their response to fire, and particular
fire regimes (including fire exclusion) are
essential to the survival of some species.
• The remaining populations of birds
most threatened by inappropriate fire
management occur in coastal heath and
thickets, mallee and northern grasslands
and woodlands—often areas of low
human population.
• Land managers are seeking to incorporate
ecological requirements into their
management decisions and actions, and
most States and Territories now require
land managers to manage fire in fire-prone
vegetation, and to use fire in ecologically
sustainable ways.
• Fire management can generally be
incorporated into management plans
in such a way that neither biodiversity
conservation nor other goals, particularly
protection of human life and property, are
sacrificed.
• It is important to identify regimes known
or suspected to cause biodiversity and
threatened species loss and manage the
land to avoid these regimes.
• Maintaining a mosaic of habitats of
different fire ages is recommended for
large area fire management, but where
habitats are small, fragmented and
isolated, the options are fewer and the
issues more complex.
• Across the north, regulation of exotic
pasture grasses is a high priority for
fire management and biodiversity
conservation.
• Fire management must be ongoing and
adequately resourced.
• There is a need for greater awareness,
research and understanding of the role of
fire in natural systems.
• In the absence of complete knowledge
to guide management, it is prudent to
apply the precautionary principle (that
is, pursue the fire regime that is likely to
be lowest risk in terms of environmental
damage) and adaptive management.
• There is a need for a set of simple, flexible
guidelines to help land managers manage
fire for biodiversity conservation.
OVERVIEW
Birds to Watch
For several bird species an inappropriate fire
regime is the only major threat to their survival.
Without continued implementation of effective
fire regimes, these species are likely to move
closer to extinction in the near future. They
include: the Ground Parrot, Noisy Scrub-bird,
Eastern Bristlebird, Western Bristlebird and
Western Whipbird. The Partridge Pigeon,
Malleefowl, Golden-shouldered Parrot,
Black-eared Miner and, possibly, the Night
Parrot are also at great risk from inappropriate
fire regimes, among other threats.
Inappropriate fire regimes are the
greatest threat to Australia’s birds
after direct human destruction and
alteration of habitats. Potentially, they
are also one of the most manageable of
threats. The introductory pages of this
supplement summarise the key points
from a diffuse and often contradictory
literature on the interaction between
fire and birds. They are complemented
by stand-alone narratives authored by
ecologists concerned that we are
often failing in our management of
fire. While there is certainty that current
fire regimes are often damaging to the
land and its wildlife, there is still much
to learn about effective fire management
to conserve biodiversity. Our hope is
that this report will lead to greater
awareness of the issues, implementation
of bird and biodiversity-friendly fire
management regimes, and refinement
in knowledge of appropriate burning
practices.
Australia has long been a dry,
fire-prone country and for some
60,000 years people have used fire to
manipulate the continent’s vegetation
and wildlife. So commonplace is fire
that much of the Australian biota has
adjusted to occasional burning.
‘You got to burn the country… not leave
him until he really dry, that means you
destroy everything… cook it… Soon as wet
season stop… grass starting to dry out, you
burn grass then… so you don’t burn all the
animals… if you burn it [then], some have
a place to hide, but if you leave him, and
you burn the grass you kill everything, that’s
no good.’
Musso Harvey of the Yanyuwa
people of the Gulf of Carpentaria
(quoted in Baker 2003)
‘Fire, grass, kangaroos and human
inhabitants seem all dependent on each
other for existence in Australia, for any one
of these being wanting, the others could no
longer continue.’
Sir Thomas Mitchell, Journal of
an Expedition into the Interior
of Australia (1848)
In many habitats, fire histories
and Indigenous explanations reveal a
pre-European landscape of small burnt
patches of various stages of fire succession,
maintained by a system of burning which
controlled the build-up of fuel, and reduced
the incidence of large, intense fires. Since
European occupation the magnitude and
rate of change in human-fire interactions
has increased markedly. In the early
days after settlement, fire exclusion and
suppression was a common practice in
more highly developed areas, but deliberate
burning was used in many outlying
areas to reduce fire hazards or maintain
grazing values.
Contemporary fire regimes have
tended to be either of two extremes:
intense, extensive and uncontrolled wildfire
mainly late in the driest season; or land
seldom burnt. This results in loss of firesensitive species, habitats and patchiness in
intensively burnt areas, or wood thickening
and loss of savanna, granivorous species and
general biodiversity in areas seldom burnt.
Near centres of human population, at the
urban and agricultural interface, Australian
authorities have adopted broad area
burning to reduce fuel loads, and protect
human interests; however, fire may escape,
and arson is also common. Over time these
policies, mistakes and misdemeanours lead
to a loss of heterogeneity and biodiversity
across the landscape.
Until quite recently, European
fire management has been directed at
protection of human life and property.
Now, biodiversity management is seen
as important and there has been a shift
towards its inclusion in legislation,
4
Fire and Birds
Fire and Birds
Responses of birds to a single
fire event
The effects of a fire on bird communities can be
difficult to predict. They vary with fire intensity,
season and extent; the characteristics of the site,
climatic variation before and after; and time
elapsed since fire. Species differ in their response
to fire and individual species vary in their response.
In general, the Australian bush and its wildlife
can cope with single fire events, but over time
frequent, widespread fires change the character of
the landscape and its inhabitants, to the detriment
of fire-sensitive species.
Before, during and after
Flame Robin. Photo by Peter Merritt
Large logs can act as micro fire-breaks, adding
desirable heterogeneity to the habitat, and they
offer refugia for animals in their moist interior.
Regrowth in the early succession stages after fires,
particularly in the south of the country (here Tiajara,
NSW), attracts small insect-eaters such as Scarlet and
Flame Robins. Photo by Graeme Chapman
Some definitions
Backburning: a fire that is intentionally lit to
consume fuel in the path of a wildfire.
Bushfire or wildfire: a general term used to
describe a fire in native vegetation.
Prescribed, controlled, fuel reduction,
or planned burning: a fire lit under specified
(i.e. relatively safe) environmental and
weather conditions to burn over a
predetermined area, at a time, intensity
and rate of spread necessary to achieve
particular management objectives.
Fire-break: any natural or constructed discontinuity in fuel used to break up, stop or
control the spread of a fire.
Fire age: the time elapsed since the last fire.
Fire intensity: the amount of energy released
per unit length of the fire front, usually
expressed as kilowatts (kW) per metre.
Fire regime: the history of fire in a particular
area, comprising the frequency, intensity,
season of burning (i.e. the time of year a
fire occurs) and the type.
Type: refers to whether the fire is above-ground
or below-ground (i.e. a peat fire), but as
almost all fires in Australia are aboveground this component is usually irrelevant.
Fire succession: the series of stages that
vegetation goes through in its recovery
post-fire.
Fuel load: the amount of combustible material,
especially easily ignitable material.
Hazard reduction burn: a fire intentionally lit to
reduce the amount of fuel, and consequently
reduce the risk posed by wildfires. Fuel is any
material, such as bark, leaf litter and living
vegetation, which will burn.
Trittering: mowing and mulching which
involves pulverising the leaf litter so that it
becomes compacted and burns more slowly
with a reduced flame height.
Adapted from the NSW National Parks & Wildlife
Service Fire Management Manual (2001)
policy and management (see ‘Bushfire
management in the Northern Territory’
p. 16, ‘Tarawi Nature Reserve’ p. 5).
Recent legislation in most Australian States
and Territories not only requires that fire
is managed by land managers in fire-prone
vegetation, but that it is used in ecologically sustainable ways. The importance of
periodic prescribed fire (controlled burns)
to increase or maintain biodiversity is now
widely accepted.
Insights from Indigenous experience
can help inform the development of
effective fire management practices. Sadly,
much of this knowledge has been lost or
eroded. More importantly, landscapes are
now substantially modified from those
managed by Indigenous Australians—by
weed invasion, introduced grasses, grazing,
clearing and drainage; and reduced,
fragmented habitats. These factors
can dramatically alter fire impacts on
landscapes and ecosystems.
‘I am confident that the indiscriminate
burning of bush, which is the concomitant of
all farming and grazing operations, is by a
long way the major cause of the disappearance
of many of our rarer birds.’
Edwin Ashby (1924)
The environmental impacts of fire on
natural ecosystems are not well understood
due to the complexity of the problem,
paucity of research, and megadiversity of
the Australian landscape. There are few
detailed long-term studies of the effects
of fire on avian communities, but there
is sufficient information on fire effects
from a broad cross-section of Australian
habitats to identify patterns of response
to individual fires (see ‘Responses of birds
to fire’ p. 5) and predict likely long-term
effects. Depending on the severity of the
fire, the amount of vegetation burned and
killed, and the weather during the recovery
period, most avian communities recover
rapidly following single fires regardless of fire
intensity. However, even single fires may pose
a significant threat to species with restricted
distribution, limited reproductive potential,
poor dispersal ability and/or narrow habitat
requirements. Birds persisting in fragmented
habitats are particularly at risk because the
whole fragment can burn and, post-fire, small
fragments may not be recolonised by all of
their original species.
Nonetheless, far more threatening
across a broad range of avian communities,
are increases in fire frequency and extent,
on the one hand, and fire exclusion on
the other. These tend to homogenise the
landscape, and offer fewer options for
survival and recolonisation.
Detrimental fire regimes have
contributed to the extinction of two
of the three bird species, and three
of the four subspecies, lost since
European colonisation. Inappropriate
fire management is now a factor in the
threatened status of over 50 nationally
recognised threatened bird species (see table
p. 6). Of those threatened species whose
relationships with fire have been welldocumented, almost all show a preference
for less frequent, less extensive fires than
have occurred in recent decades.
Though the task is daunting, the
costs of inaction are great. Fire can be an
economical way to manage large tracts
of land, on a fine scale, with benefits
to biodiversity and human needs. The
challenge will be to identify and develop
the appropriate fire regimes to maintain
fire-dependent habitats and protect the
fewer fire-sensitive habitats. The single
greatest need for land managers incorporating fire into their management is a
useful tool to determine appropriate fire
regimes (see ‘To burn or not to burn’ and
‘Towards a set of biodiversity-friendly fire
management principles’ p. 6).
‘There was a sound as of thunder, mingled with
the crashes of trees and the wild cries of legions of
birds of all kinds; which fell scorched and blackened
and dead to the ground.’
W. Howitt (1856) describing the
1851 Victorian bushfires
The immediate response to fire by birds depends on
the extent, speed and intensity of the fire; habitat;
and species of bird. It is likely that many birds take
action well in advance, and depart or seek shelter.
Some birds are overcome by the heat, smoke
or flames, and their nests may be destroyed. More
mobile species tend to fare better, though some
terrestrial species take refuge in wet areas, burrows
and logs.
Other birds, including some raptors,
insectivores and seed-eaters, are attracted to fire
and its charred results, moving in from adjacent
habitats or drawn from far and wide by the
smoke. This is particularly so in the north, where
Black and Brown Falcons, Black and Whistling
Kites, Australian Bustards, woodswallows and
Tree Martins often arrive in large numbers to feed
on flushed, killed or injured animals. Granivores,
including Gouldian Finches, exploit seeds exposed
by the fire. After a relatively short time they move
on, sometimes to the next fire. By contrast, in
temperate areas a more successional colonisation of
the burnt areas occurs and the initial post-fire scene
is often one with few birds.
5
Maintaining fire-free habitat
fragments: Helmeted Honeyeater
A number of colonies of the critically
endangered Helmeted Honeyeater (west
Gippsland subspecies, also known as the
Yellow-tufted Honeyeater) were destroyed
by the Ash Wednesday fires—for example,
those at Cardinia and Cockatoo Creeks. Thus,
the remaining colonies, in streamside lowland
swamp forest at Yellingbo, are maintained
as essentially fire-free, although the Ranger
has trialled small-scale burning to control the
introduced grass Phalaris.
BRUCE QUIN, Department of Sustainability
and Environment, Victoria
The first year
‘In the first year after the fire, the avifauna of burnt
habitats differs from that which develops later, or
which was present before the burn.’
John Woinarski & Harry Recher (1997)
The reduction of food and cover for some resident
bird species leads to starvation, predation or
emigration. Species suited to more open habitats,
mainly seed-eaters and insectivores, may move in
temporarily, from adjacent areas or further afield,
to take advantage of a flush of insects and seeds
generated by the quickly responding growth of
grasses and flowering annuals. Examples include:
Straw-necked Ibis; Torresian Crow; butcherbirds;
Red-tailed Black-Cockatoo; Galah; quail; nightjars;
Red-backed Kingfisher; Partridge Pigeon and
several finch species in the north; Scarlet and Flame
Robins; and Buff-rumped Thornbills in the south.
These species move on or their numbers thin out as
regeneration progresses.
Cumulative impacts:
the Black-eared Miner
The Black-eared Miner numbers less than
450 breeding individuals. Its habitat is mallee
that has been unburnt for 20 or more years,
where it reaches highest densities 40 or more
years since burning. Much of its former habitat
in the Murray Mallee has been cleared for
agriculture. Yellow-throated Miners have
colonised man-made clearings within intact
mallee and are interbreeding with the
Black-eared Miners, swamping most
populations except at Bookmark Biosphere
Reserve. Hybridisation and fire are now the
major threats to the species’ survival.
MICHAEL CLARKE,
La Trobe University, Victoria
Later years
‘The pattern and rate of change in species
abundance following fire differs according to
habitat, floristics, fire history, climate, patch size
and isolation and patchiness of the fire.’
John Woinarski & Harry Recher (1997)
The land area of Australia burned in the last five
years. Map prepared by Cameron Yates and Brian
Lynch (NT Department of Infrastructure Planning
and Environment), from information supplied by
WA DOLA.
As regeneration proceeds, species present
before the fire begin to return or build up; most
populations recover within three to ten years, while
a few take decades to return to pre-fire levels.
Periods of low rainfall may slow the recovery of
vegetation and return of birds. In habitats typically
slow to regenerate, such as old-growth forest, their
characteristic species—e.g. hollow tree nesters
and canopy feeders—may still be absent or less
abundant 50 years after an intense fire that kills
canopy trees.
FIRE SEASONS
Fires in the mallee can create flushes of food
for the Malleefowl, but it needs long-unburnt
country to breed. Photo by Lynn Pedler
DARWIN
TENNANT CREEK
PORT HEADLAND
MT ISA
ALICE SPRINGS
Tarawi Nature Reserve fire management plan
BRISBANE
GERALDTON
KALGOORLIE
PERTH
SYDNEY
CANBERRA
ADELAIDE
MELBOURNE
Winter and spring
Spring
Summer
Spring and summer
Summer and autumn
HOBART
The timing of the main fire seasons reflects the
different weather patterns. Note the tendency for
summer/autumn to be the danger period in the
southern States, and winter/spring in the north.
Source: www.bom.gov.au
Tarawi Nature Reserve is in the semi-arid mallee of New South Wales. The key elements of the Reserve’s
fire management plan put into practice much of the new knowledge in conservation science.
General conservation goal: to manage fire to minimise the risk of any native species becoming extinct.
Objectives:
• ensure consecutive fires are at least 20 years apart in any one area;
• ensure a range of post-fire ages are present in the reserve (at least 50% of each mallee community
should be >40 years old) to promote patchiness in wildfires, although a scale of patchiness is not
specified due to lack of knowledge;
• determine the biodiversity values of long-unburnt mallee (>70 yrs);
• burn strategically to contain the spread of wildfires that may violate the thresholds for fire
frequency, post-fire age diversity or patchiness requirements.
Extracted from Keith et al. (2002)
6
Fire and Birds
Fire and Birds
Towards a set of biodiversity-friendly fire management principles
South-eastern Red-tailed
Black-Cockatoos and Fire
The endangered South-eastern Red-tailed BlackCockatoo is a dietary specialist; it eats only the
seeds of three tree species—Brown Stringybark,
Desert Stringybark and Buloke. Supplies of these
seeds are patchy in space and time, and flocks tend
to concentrate on the seed stocks that can be most
efficiently harvested. There is mounting evidence
that cockatoo breeding success is related to the
availability of fresh stringybark seed crops.
The stringybark forests of south-western
Victoria are highly flammable and the heathy
understorey requires fire to maintain its floristic
diversity. However, fires that burn or scorch the
canopy result in reduced seed production for up to
nine years. Therefore, the widespread burning of
stringybark forests through the twentieth century
may have been a cause of population decline in
the Red-tailed Black-Cockatoo. With this in mind,
a moratorium on burning was imposed in 1989.
However, there are now good ecological and asset
protection reasons for resuming planned burns.
In an attempt to resolve this impasse, the
Department of Sustainability and Environment,
with the support of the recovery team, has
trialled methods of burning the understorey while
minimising canopy scorch. This is a tall order in
stringybark forests, where the fibrous bark tends
to carry flames upwards. However, given the right
conditions of fuel moisture and weather, the trials
are suggesting that ecological burning aims, or asset
protection aims, can be met whilst minimising the
impact on future crops of stringybark seed.
PETER MENKHORST AND JIM MCGUIRE, Department
of Sustainability and Environment, Victoria
Australia is a large and diverse country, hence
there are many views and requirements for
biodiversity-friendly fire management. The
following are some of several recommendations
in the literature:
• Encourage precautionary fire management.
• Be aware that fire regimes need to be local and
focus on particular objectives.
• Wherever possible develop guidelines and
prescriptions for landscapes and biological
communities, not individual plants and
animals.
• However, where there are species or groups of
species that are susceptible to decline under
certain fire regimes, fire management regimes
should be framed around their needs.
• Aim for fire management targets that are
ranges rather than optima; for example, rather
than choosing the best time for one species
pick a range of times and fire intervals that
cover several species.
• Avoid fire regimes known or suspected to result
in loss of biodiversity in the same or related
ecosystems.
• Be aware that fire frequency is a key element of
fire management—with few exceptions both
fire exclusion and short intervals between fires
over a broad area are damaging to biodiversity.
• Similarly, broadscale, intense fires are
damaging to heterogeneity of habitats
•
•
•
•
•
•
•
•
across the landscape, and hence to
biodiversity values.
Hence, aim to reduce the extent of land/
proportion of vegetation community burned
in single-fire events, no matter what time of
year that happens.
Use fire at a small scale to promote spatial
and temporal variability in fire regimes across
the landscape, that is, a patchwork of areas of
different fire-ages and histories.
Understand that fire-generated patchiness
is good but must be at the right scale (for
example, in relation to the species being
managed or the size of the remnant).
Develop prescriptions to limit the extent and
spatial invariability of fires by controlling fire
behaviour, rather than by imposing artificial
exclusion zones and intervals between fires.
Accept that there will always be gaps in
knowledge, which introduce an element of
uncertainty in the decision-making process.
Allow for unplanned fires.
Climate change means that the country is
likely to be entering a period of hotter, drier
conditions, which will produce more extreme
fire events, which will need to be factored
into management.
Apply adaptive management, including
monitoring, to inform decisions and
improve management.
Threatened birds and fire regimes. This table presents some simple rules for bird-friendly fire management1.
Local conditions and threatened species should be used as a finer guide to management.
To burn or not to burn
Fire regimes result from complex interplay
between biotic (e.g. fuel loads) and abiotic factors
(e.g. weather conditions), increasingly overlaid
with human decisions in support of a variety of
land uses, including biodiversity conservation.
Land managers must weigh up these factors to
estimate the threat from uncontrolled fire and the
desirability of a controlled burn.
contours
slope
fire history
vegetation
land cover
fuel loadings
McArthur
Fire Model
Main
habitat
Threatened species for which
inappropriate fire management is a threat2
Current general fire problem
for bird conservation
Fire management for bird
conservation
Coastal and
sub-coastal
shrublands
Orange-bellied Parrot CE
Ground Parrot (eastern) V
Ground Parrot (western) CE
Rufous Scrub-bird (northern) V
Noisy Scrub-bird V
Southern Emu-wren (Fleurieu Peninsula) CE
Southern Emu-wren (Eyre Peninsula) V
Eastern Bristlebird (northern) CE
Eastern Bristlebird (southern) E
Western Bristlebird V
Rufous Bristlebird (Otways) V
Western Whipbird (western heath) V
Too frequent burning;
too coarse a mosaic.
Pattern: Fine mosaic of areas of
different fire-ages with a bias towards
retention of older fire ages.
Intensity: Low.
Mallee
Malleefowl V
Mallee Emu-wren V
Black-eared Miner E
Western Whipbird (eastern) E
Fire management (prescribed
burns or Aboriginal burning
regimes) has been much
reduced so that wildfires are
too hot and extensive.
Pattern: Mosaic of areas of different
fire-ages with a bias towards
retention of older fire ages. Fire
frequency at most every 40 years in
any particular area.
Intensity: Low.
Temperate
eucalypt
open forests
Red-tailed Black-Cockatoo (south-eastern) E
Glossy Black-Cockatoo (Kangaroo Island) E
Helmeted Honeyeater CE [riparian]
Forty-spotted Pardalote E
Fire exclusion.
Pattern: Mosaic of fire-ages across
the landscape with a bias towards
retention of older fire ages.
Intensity: Flexible use of a broad
range of fire regimes.
Tropical and
sub-tropical
forests/
rainforest
Southern Cassowary V
Red Goshawk V
Black-breasted Button-quail V
Masked Owl (Tiwi Islands) E
Fire exclusion permits spread
of fire-sensitive species.
Pattern: Mosaic of fire-ages across
the landscape with a bias towards
retention of older fire ages.
Intensity: Occasional hot burns.
Tropical and
sub-tropical
savanna
woodlands
and
grasslands
Buff-breasted Button-quail E
Partridge Pigeon (western) V
Golden-shouldered Parrot E
Paradise Parrot Ex
Night Parrot CE
Hooded Robin (Tiwi Island) V
Crested Shrike-tit (northern) E
Gouldian Finch E
Star Finch (Cape York) E
Fires now too hot, extensive
and frequent, or, alternatively, too infrequent.
Pattern: Fine-scale mosaic of
fire-ages across the landscape,
with between 10 and 30% of
landscape burned each year.
Intensity: Occasional hot burns.
Wetlands
Australian Bittern (Australasian) V
Burning of habitat to create
pasture.
No prescribed burns.
standard fuel &
weather conditions
Hazard
structures
roads and trails
residential
Risk
revegetation
residential
structures
constructed
values
fauna values
flora values
natural
values
Risk/Hazard
matrix
Values
Threat
One approach to the prediction of the threat of fire.
Source http://www.planning.sa.gov.au
1Based
on information in Woinarski (1999) www.deh.gov.au/biodiversity/publications/technical/fire/
from Garnett & Crowley (2002). Categories of threat in decreasing order of severity: Ex = Extinct;
CE = Critically Endangered; E = Endangered; V = Vulnerable; subspecies name in brackets. Twenty-two other species
are listed as Near Threatened, and inappropriate fire regimes are among their major threats.
2Extracted
7
LIVING WITH FIRE—BIRDS
IN NORTHERN AUSTRALIA
Australia’s north is a landscape of fire, and some naive
ideal of fire suppression and exclusion will neither
work nor suit most bird species
by John Woinarski
Climatic conditions, vegetation, social factors and human population
dispersion cause fire to be a very different phenomenon in northern
Australia to that elsewhere in the continent. The north is monsoonal,
with a strongly seasonal climate marked by a long, almost rainless,
warm dry season (typically from about April to October), and a strongly
contrasting hot wet season.
Fire is an inevitable part of this landscape. The dense, tall grasses grow
rapidly over the short wet season, and then dry out over the long dry season,
becoming highly flammable. Without human intervention, this vegetation
burns readily with ignition from lightning strikes, a characteristic feature of
the first thunderstorms that mark the transition from the dry to the wet.
Humans have long been moulding this landscape. For the Indigenous
people fire has been a crucial part of life and their main management tool
for 40,000 years or more. Such application of fire was far from reckless.
The consequences of ill-advised fire management were severe: vital food
sources could be eliminated through inappropriate fires, and fires that
spread to burn neighbouring clan estates could invite retribution. In parts
of northern Australia this tradition continues. In these areas, the application
of fire may be a highly skilled and carefully considered operation, carried
out diligently throughout the year. Typically, this traditional management
involves many small ‘cool’ fires, producing an intricate network: a fine-scale
mosaic of unburnt patches and patches burned at different times through
the year. Often, the more fire-sensitive components of the landscape—such
as rainforest patches—were explicitly and deliberately protected from fire,
typically by burning around their margins. These elements were valued
because they produced important foods, particularly yams, in traditional
diets; and the maintenance of such fire-sensitive elements in a clan estate
was a sign of proper management.
Over the last century, traditional Indigenous fire management in
most of northern Australia has broken down. Indigenous people were
dispossessed from much of the landscape. Even in lands that maintained
Indigenous ownership, population dispersion and lifestyles changed. In the
decades around the Second World War, most of the population moved off
their clan estates to live at mission stations and other townships. In some
areas, this dynamic has been reversed through the ‘outstation movement’,
beginning in the 1970s. But, over most of northern Australia, traditional
burning has been disrupted and replaced by very different fire regimes.
Below left: With less than 50 known individuals the Eastern Bristlebird (northern
subspecies), of Cooloola National Park and the Conondale Ranges, Queensland,
is critically endangered. Its main threat is changed fire regimes, with fire now too
frequent, destroying the species’ tussock grass habitat, or too infrequent, allowing
shrubs to become too dense. Below: Fires remove dense ground cover to expose
seeds on which the finches feed, but may also destroy wet season food species,
and nests. Here a Masked Finch collects charcoal for its nest (presumed to help
keep the nest dry and sanitised). Photos by Graeme Chapman
Wildfire at Kakadu National Park, Northern Territory. Photo by Raoul Slater
Fuelling fire
The current fire regimes vary somewhat between tenures. On pastoral
lands, which occupy about half of northern Australia, fire is generally
unwanted and suppressed—livestock consume most of the grassy fuels,
so that fire behaviours and intensities are now different to those that
prevailed over tens of thousands of years. Partly because of the reduced
incidence of fires, vegetation in many pastoral areas is changing, often
with increases in the density and extent of woody shrubs and decrease in
grassland areas (‘vegetation thickening’). Fires on other tenures are now
characterised by an increased incidence of late dry season burns, and these
are typically more intense and extensive. The previous fine mosaic of
patchy burnt and unburnt areas has been replaced by broad-scale conflagrations that homogenise the landscape. Far more often than previously,
these fires engulf the fire-sensitive elements of the landscape.
Superimposed on this change in fire regimes is the consequence of
the increasing and largely uncontrollable spread of African and South
American pasture grasses, deliberately introduced by pastoralists and
pasture scientists to transform (‘improve’) the landscape. This insidious
set is marked by their ability to dominate almost every environment in
northern Australia, including wetlands (Para Grass, Olive Hymenachne),
tropical open forests and savanna woodlands (Gamba Grass, Mission
Grass, Guinea Grass) and semi-arid woodlands (Buffel Grass). The
replacement of native understoreys by introduced pastures is itself an
ecological wound but, worse, these grasses also greatly alter fire regimes.
Largely because they are voracious consumers of soil nutrients, exotic
grasses produce fuel loads that are far greater than native grasses (up to
seven times the biomass), and they typically cure later in the dry season.
This inevitably means that fires fuelled by exotic, invasive pastures
are more intense (up to 10 times hotter) than fires with native grass
understoreys. Whereas previously fires in northern Australia burned the
grassy understorey and rarely affected the tree layer, exotic grasses are now
fuelling crown fires that kill the tree layer.
Currently, fires burn about 30–50% of northern Australia each year.
Many areas have been burnt every year over the last decade. The scale and
recurrence of fires is vastly different to southern Australia. In extent, the
regular fires of northern Australia dwarf the bushfires in south-eastern
Australia, such as Ash Wednesday and those in the south-east highcountry in 2003. For example, in five days of October 2004, a fire in
the Sturt Plateau and northern Tanami Desert of the Northern Territory
burned out an area greater in size than Tasmania, but didn’t even make
headlines in the local newspapers.
8
Fire and Birds
Fire and Birds
9
Long-term impacts on birds
Above: The threatened White-throated Grasswren occurs only on and around the Arnhem Land escarpment
where frequent hot fires are causing spinifex to be replaced by annual sorgum. Photo by Graeme Chapman
Below right: Partridge Pigeons (here the yellow-faced Kimberley race) persist in areas where a mosaic of
fire ages has been maintained; the western subspecies is threatened, but the eastern subspecies is faring
better where traditional burning regimes have continued. Photo by Jiri Lochman, Lochman Transparencies
Short-term impacts on birds
So, how do birds in northern Australia fare
with fire, and how have recent changes in
fire regimes and management affected birds?
There is substantial variation among
the bird species of northern Australia in their
immediate responses to fire, and in their
longer-term responses to fire regimes. Many
bird species are attracted to burning and
recently burnt areas. The most conspicuous of
the immediate responders are raptors—particularly Whistling Kites, Black Kites and Brown
Falcons—which hunt, sometimes in very
large numbers, among the flames of the fire
front for fleeing large invertebrates and small
vertebrates. These birds may not only benefit
directly from fires, but may also be firebugs
themselves—there are many reports of kites
spreading fires by snatching up burning sticks
and dropping them ahead of the flames. This
attraction of raptors to fires was used by some
Aboriginal hunters, who built elaborate rock
hides in which they lit smoky fires and waited
to grab or spear hawks that came close to
investigate. Other birds attracted to the fire
front include woodswallows and swifts which,
like the raptors, may make regional-scale
movements tracking the fires.
Even the relatively slow and cool
burns of the early dry season may kill some
invertebrates and small vertebrates, and
many birds are attracted to the carrion in the
immediate aftermath of fires. These include
butcherbirds, Australian Bustards, crows, ibis,
Magpie-larks and kingfishers. As with the
raptors, some of these birds may undertake
regional-scale movements from one recently
burnt area to another. Fires provide benefit
not only for carrion-eaters, but also to a wide
range of terrestrial species. When unburnt,
the dense tall grasses of northern Australia
may be a major impediment to foraging for
many bird species. The cool fires of the early
dry season remove this barrier, but typically
do not greatly reduce the density of invertebrates or seeds lying dormant on the soil
surface. Thus, recently burnt areas attract a
very wide range of seed-eating birds, typically
including Red-tailed Black-Cockatoos,
Galahs, Cockatiels, Little Corellas, Peaceful
Doves, Diamond Doves, Partridge Pigeons
(or their near equivalent, Squatter Pigeons,
in Queensland), and a multitude of finches.
The open ground layer makes for easier
foraging for insectivores and carnivores, such
as Hooded Robins and butcherbirds.
Of course, not all is for the best. Certain
species, such as Red-backed Fairy-wrens and
some quails, need the dense grass layer for
shelter and nesting; and will suffer greatly
increased predation rates in recently burnt
areas. In a delicate balancing act, some bird
species need both burnt and unburnt areas.
For example, in the early dry season, Masked
Finches and Partridge Pigeons nest on or
near the ground. Fires at such times are likely
to destroy many nests, and, for Partridge
Pigeons at least, nest predation rates are likely
to be much lower where their ground nests
are at least partly sheltered under grass cover.
But these birds will struggle to access enough
food over the course of the dry season if
their territories are completely unburnt. The
optimum fire regime for such species is one
of very fine-scale intricate burning, where
each year part of the territory or home range
is burnt and part unburnt. Such a regime
requires fire patches at about a hectare
scale. This pattern of burning is no longer
prevalent in northern Australia. Rather, the
current regime is characterised by burnt and
unburnt areas in patches of tens to hundreds
of square kilometres.
Birds respond not only in the short-term to
individual fires, but also over the longer-term
to the patterned history of burning over many
years. In parts of temperate Australia (such as
in the mallee, coastal heaths and mountain
forests), this is a sharply etched response, as
there may be a conspicuously contrasting
vegetation succession over the multi-decade
interval between fires. The vegetation response
to fire regimes in northern Australia is typically
more nuanced. The relatively few long-term
studies—that typically describe experimental
situations where great efforts must be made to
exclude fires for any length of time—suggest that
there is relatively little, or slow, change in plant
species composition with increasing time since
fire. However, protection from fire does lead to
a substantial structural change in vegetation.
When an area is unburnt for five or more years,
instead of the very simple vegetation profile of
trees and grass that characterises most northern
Australian forests, a diverse shrubby mid-storey
develops, and continues to increase in height,
cover and diversity with increasing duration since
fire. Correspondingly, with increasing shading,
the grass layer diminishes. These vegetation
changes inevitably favour some bird species
and disadvantage others. Many of the shrubby
plants produce (very tasty) fleshy fruits, and the
increasingly dense mid-storey provides good
nesting sites and an increased foraging resource
for some insectivorous birds. Birds that occur in
greater abundance in such relatively long-unburnt
areas include White-throated Honeyeater, Dusky
Honeyeater, White-gaped Honeyeater, Northern
Fantail, Weebill, Lemon-bellied Flycatcher, and
Bar-shouldered Dove. Conversely, birds such as
butcherbirds, kingfishers and Red-backed Fairywrens, that prefer the simpler structure of trees
and grass, are markedly less abundant in areas
long untouched by fire.
Over the long period, fire regimes change
not only the vegetation profile within any patch,
but also the borders between—in doing so they
alter the relative extent of different vegetation
Left: A change in fire regime threatens the
eastern Spinifex Pigeon. Photo by Jonathan
Munro, www.wildwatch.com.au
Below: Late dry season fires which burn the
fringes of gallery rainforest and paperbark
thicket can destroy large trees with holes for the
Rufous Owl and its prey species, including the
Brush-tailed Possum. Photo by Jonathan Munro,
www.wildwatch.com.au
types. In northern Australia, frequent fires will
generally favour expansion of grasslands, and
reduction in fire frequency will favour increases
of woody vegetation. On pastoral lands of
Cape York Peninsula, a reduced frequency
of fires has led to invasion of grasslands by
paperbarks (Melaleuca), substantially reducing
the area of suitable habitat for the endangered
Golden-shouldered Parrot. Conversely, the
grass-wrens of sandstone escarpments in
northern Australia (the Black, White-throated
and Carpentarian) are now all threatened by a
fire regime that is too frequent, and eliminates
the large old clumps of spinifex (Triodia)
that these species require. Acacia-dominated
woodlands (the ‘pindan’ of the Kimberley and
lancewood in the Northern Territory and parts
of central Queensland) are also being eroded by
the increasing frequency, intensity and extent of
fires, to the detriment of birds such as Hooded
Robin and Grey-crowned Babbler. Other
fire-sensitive vegetation in northern Australia
includes rainforests and heathlands, and these
may also be under some threat from increasingly
frequent fire.
Accepting a fiery future
Over tens of thousands of years, the fire
regimes of northern Australia were actively
and consistently managed. This produced
some (not necessarily ideal) sort of
equilibrium and reasonable stability in the
vegetation patterning, and hence bird species
composition. Over the last century (in many
areas, over the last few decades), that longestablished fire regime has come to a halt,
and in its place is an erratic, inconsistent
and far less knowledgeable management
of fire. This change has destabilised that
equilibrium. As a result, some species will
increase and some will decrease. Some of the
decreasers are likely to suffer at least regional
extinctions.
The conservation and management
challenge in northern Australia is profound.
It is a landscape of fire, and some naive ideal
of fire suppression and exclusion will neither
work nor suit most bird species. Fortunately,
there remain some areas where traditional
Aboriginal management of fire is practised,
and we still have the opportunity to learn
from this expertise. One priority is to provide
support for the nascent ranger schemes that
have sprung up on some Aboriginal lands,
to help get Aboriginal people back onto and
managing their estates.
Another priority is to develop far better
regulation of the use of exotic pasture grasses.
Currently, there are almost no limitations on
the deliberate spread of these weedy species
on pastoral lands. Rather, it is typically
encouraged, and there is a largely passive
acceptance by most regulatory authorities of
the grasses’ inexorable spread from pastoral
properties to neighbouring Aboriginal and
conservation lands. Another priority is to
improve understanding of the role of fire in
northern Australia. Bushfires in southern
Australia tend to be infrequent, ‘destructive’
and shocking: for a short time at least they
concentrate the mind. In northern Australia
they are so much a part of the landscape that
they are accepted blithely. Their environmental impact is far more subtle and gradual
than in the south, and hence we are less
driven to think about their role. But ecology
and conservation can be undermined as
much by repeated subtle and insidious factors
as by occasional showy episodes.
John Woinarski works as Principal Research
Scientist for the Northern Territory’s Department
of Infrastructure, Planning and Environment,
and is engaged in a broad range of research
projects including wildlife survey, conservation
planning, ecological studies of threatened animal
species and the impacts of fire and pastoralism.
Fire and Birds
Fire and Birds
MORNINGTON: A MODEL FOR FIRE MANAGEMENT IN
AUSTRALIA’S TROPICAL SAVANNAS
The magic of satellite imagery helps guide biodiversity-friendly fire management on a vast Kimberley land holding
by Steve Murphy, Sarah Legge and Nat Raisbeck-Brown
Declining seed-eaters
Mornington is home to one of the last remaining
large populations of Gouldian Finches. Once
spread right across the northern third of the
continent in flocks of thousands, in the last few
decades a precipitous decline has restricted the
finch to tiny pockets, in numbers of mere tens
and hundreds. Many other species of seed-eating
(granivorous) birds have suffered a similar fate. It
is thought that the main reason for these declines
is the exacerbation of a natural food shortage
early in the wet season.
The onset of the annual monsoon causes
widespread germination of seed reserves
that are the staple of most tropical grassland
granivores during the dry season. When rains
come, perennial grass species are the first to
respond with seed, and the time this takes to
occur is the ‘Achilles heel’ of these granivores.
If early seeding perennials are delayed, become
scarce or drop out of the system completely,
Gouldian Finches and other granivores may
be forced to fly long distances in search of
food. This, combined with the physiological
stress of starvation, may be enough to tip
them over the edge.
Ongoing research at Mornington,
involving telemetry and indices of health,
has shown that Gouldian Finches switch
to spinifex as soon as the superabundant
sorghum starts to germinate. In other parts
of their distribution, they switch to Cockatoo
Grass or Ribbon Grass. All of these plant
species are sensitive to fire, and evidence
suggests that some, particularly spinifex,
require long fire-free intervals to set seed.
Also, being unpalatable to stock, spinifex is
often burnt repeatedly by graziers in pasture
improvement programs.
As is often the case, ecological patchiness
or heterogeneity, especially in relation to fire
history, is the key to maintaining large areas
of seeding grasses for Gouldian Finches in the
early wet season.
Mornington: recent fire history
Areas burned on Mornington in three consecutive
years (2001, 2002 and 2003). Note that this is a
see-through map so that it is possible to see all
years at once, with the effect that the darker
the stippling, the more frequently the area has
burnt. It is clear that some areas experienced
large fires over multiple years, whereas others
areas experienced smaller fires and overall more
heterogeneity. Differences like these are due
mainly to the effects of topography and other
geographical happenstance.
In terms of heterogeneity Mornington is
somewhat lucky. In some areas, creeks and rocky
topographical features break-up the landscape so
that, in the absence of hands-on management,
fires tend to burn heterogeneously by default. In
other areas, however, fairly uniform landscapes
mean that unimpeded fires tend to burn large
areas, creating little in the way of habitat heterogeneity. Hence, when left unchecked, large areas
of Mornington burn too frequently and too
homogeneously (see map at left).
AWC acquired Mornington in
2001. Prior to this, very little active land
management meant that large areas burned
late in the year (when fires are most intense,
and can cause most ecological damage)
even though the total number of fires was
small at this time (see graph top right). The
negative effects of hotter, more damaging
fires compounded the negative effects of
Prescribed burning at Mornington.
Photo by Steve Murphy
homogeneity. This pattern not only occurred
on Mornington, but also was evident
throughout the region (see map centre right).
The Mornington fire management
paradigm
Given that fires late in the year are often
started by lightning strikes, it is easy to accept
them as natural, and therefore tempting to
think that they may not present a problem
ecologically. While this may be true in a vast
landscape that is pristine, the reality is that
truly pristine and ecologically intact habitats
nowadays are rare; instead they are highly
fragmented and separated by areas no longer
suitable to maintain and support complete
species assemblages. This is true even for
tropical savannas that are superficially intact.
Thus, land managers can no longer afford
to risk losing habitat patches in single fire
events because re-colonisation is impeded by
fragmentation, to the point where species can
be lost from patches forever.
Our long-term aim at Mornington is to
reduce the area burned in single fire events.
We can achieve this via a program of wet
and early dry season fuel reduction burning
and physical fire-breaks around particularly
important areas (e.g., wet season food areas
for Gouldian Finches). To help us with
this challenging task, sophisticated satellite
imagery and interpretation allow us to look at
past burning patterns, keep track of current
fires, and determine if management actions are
achieving our aims.
Eye in the sky: fire mapping via satellite
Fire information for the Kimberley is recorded from satellite imagery in
two forms:
(1) ‘Hot Spots’ show the location of current fires and are mapped from
satellite imagery up to 8 times a day (see map below right). Hot spots are
accurate to about 1.1 km and give a rough guide to the location and extent
of the fire front. To show the direction of travel, different symbols show
hotspots that were detected 0–6, 6–12, 12–24 and 24–48 hours, and 2–7
days, previously. Land managers can be alerted instantly and automatically
to the presence of a fire within a user-specified area via email (see the North
Australia Fire Information website: http://138.80.201.250/nafi/init.jsp).
(2) ‘Fire Affected Areas’ (FAAs or fire scars) are mapped from satellite
imagery every 7–9 days and show the area burned or affected by fire (see
map centre right). The accuracy of the fire scar maps derived from satellite
imagery is dependent on the image resolution. FAA maps are used by
local government agencies and land managers in the Kimberley region to
assess what, when, and how often areas have been burnt since 1993. They
are used to assess broadscale fire regimes and as a guide in aerial control
burning during preventative burning programs in the early dry season.
Currently, there are four types of imagery used to map FAAs in the
Kimberley, each with a different resolution (see table). On Mornington,
FAA maps (see maps left and centre right) are derived from NOAA,
MODIS and Landsat Quicklooks. The fire scar information is used to
determine aspects of fire history such as fire extent and frequency, which
helps us determine which areas have been burnt too often, and which
have not been burnt for a long time. Together with information about
floristics and other factors (for example, wet season feeding areas for the
Gouldian Finch), this helps us to decide which areas to protect from fire,
and which may require burning. Maps that show the time of year a fire
occurred (early or late dry season) are good indicators of fire intensity.
Both of these types of fire map are vital for operations on Mornington,
and should be used wherever possible to inform biodiversity-friendly fire
management programs right across Australia’s tropical savannas.
Studies in recent years have indicated that northern Australia is on
the edge of a biodiversity crisis, which almost certainly relates to changes
in the way fires now burn across the landscape. While the general cause
is widely accepted (that is, homogenisation of the landscape), the specific
ecological interactions remain unclear. However, this lack of knowledge
should not preclude active management within a set of broadly defined
and relatively simple guidelines, for example, reducing the size of
individual fires and generally increasing the fire-free interval within
localised vegetation communities. Thankfully, the visual tools to help us
guide and assess these management objectives are freely available.
The size of the
largest fires recorded
12
during the dry season
at Mornington
600
10
(columns, legend
8
on left axis),
400
superimposed on
6
the average number
4
of fires per month
200
(line, right axis),
2
1999–2003. The
0
0
largest recorded
early dry
mid dry
late dry
build-up
fire occurred during
(mar–may)
(june–july)
(aug–sept)
(oct–dec)
the late dry season
(August-September), but the average number of fires per month on the property
was lowest during this period. In other words, individual fires were more extensive
in the late dry season.
800
14
Average no.
At over 3,200 km2 Mornington Wildlife
Sanctuary in the central Kimberley is a big
place. Although its size is fairly typical of
holdings in the tropical savannas of Australia,
it is atypical because it is being managed for
the conservation of biodiversity rather than
for cattle grazing. Mornington is owned and
managed by the non-government, non-profit
Australian Wildlife Conservancy (AWC),
which owns 12 other ecologically important
properties around Australia. Mornington
protects a diverse set of ecosystems that are
under-represented in government National
Park estate. Sandstone escarpments, blacksoil plains, moist rocky gullies, streamside
vegetation, spinifex ridges and open savannas
support a diverse assemblage of plants and
animals. Not surprisingly, fire—or equally
important its absence—is the dominant
ecological force affecting all of these habitats.
11
When dry season food resources disappear Gouldian Finches switch to spinifex seed;
because spinifex doesn’t set seed until many years after fire, too frequent burning
can destroy this critical resource. Photo by Steve Murphy
Km2
10
Area burned by time
of year for 1999
over Mornington
(boundary shown
in purple) and the
surrounding region.
Late dry season fires
tend to burn more
extensively than those
lit earlier.
Summary of the imagery available to map areas affected by fire
Satellite Image Type
Resolution
Smallest fire scar that can be detected
NOAA
1.1 km
2 x 2 km
MODIS
250 m
500 x 500 m
Landsat
30 m
60m x 60 m
Landsat Quicklooks
120 m
240m x 240 m
Cost to user
Frequency of fire scar information Area covered
Free
9 days
Australia
Free
7 days
Australia’s tropical savanna
(Cairns to Broome)
>$1000
15 days
150 km
Free
15 days
150 km
Steve Murphy and Sarah Legge are ecologists with
the Australian Wildlife Conservancy. They are based at
Mornington, in the Kimberley Region of Western Australia.
Nat Raisbeck-Brown is a Geographical Information
Systems analyst who works for the Kimberley Regional
Fire Project, based in Broome.
A Hot Spot and Fire Affected Area map for Mornington Wildlife Sanctuary. These
maps are updated regularly and are freely available to land managers from the
North Australia Fire Information service http://www.firenorth.org.au/nafi/init.jsp
Fire and Birds
How hot?
There is one more contributing factor, and
that is the type of fuel available to burn. In
most northern environments, fires travel
through the grass layer only. Changes in fire
regimes through the drier savanna woodlands
of the Top End of the Northern Territory
have promoted heavy fuel loads of the annual,
native sorghum. In places, African Gamba
Grass introduced for cattle grazing has led
to even heavier fuel loads. Repeated fires in
pastures dominated by these heavy grass loads
can eventually eliminate all canopy trees.
In wetter areas, fuel loads have probably
been reduced. Cattle collect in these areas as
the dry season progresses. The heavier the
grazing pressure, the lower the fuel loads, and
the less effective fires are at maintaining open
vegetation. And the less grass there is, the easier
it is for trees to invade. Some ecologists believe
that the wet sclerophyll forests of northern
Queensland will never burn hot enough while
cattle are allowed to graze in them.
MANAGING PERCEPTIONS
Can a change in attitude towards fire and its management ameliorate
environmental problems in Australia’s north?
by Gabriel Crowley and Peter Thompson
One of the most striking features of northern
Australia in the late dry season is the amount
of recently burnt country. Each year millions
of hectares of Cape York Peninsula, northwest
Queensland, Top End of the Northern Territory
and Western Australia’s Kimberley region
are burnt. The 2004 season was particularly
severe (see box p. 14), but in most years up
to one-quarter of northern Australia is burnt.
Unsurprisingly, fire management is one of the
biggest environmental issues in the north. John
Woinarski’s article (see pp. 7–9) describes how
failure to manage fire affects the environment
and its inhabitants. Intense fires mean tree loss,
and extensive dry season fires not only render
vast areas of habitat unsuitable for fire-sensitive
birds, but also prevent the lighting of fires needed
later in the year for habitat maintenance. Tropical
fires are also one of Australia’s biggest producers
of greenhouse gas.
A state of indifference
Effective fire management requires an
understanding of complex environmental
issues, adequate resources, and cooperation
between land managers across the north. The
problem is not simply one of too much fire,
but a failure to manage fire. As in southern
Australia, not enough importance has been
placed on preventing wildfires for effective
management to be instituted or adequately
funded. But in contrast to southern Australia,
the apathy is more likely to result from
familiarity than from infrequent experience.
To many northern Australians, fire is no
more than an inconvenience that brings
days of unpleasant smoke and blackens
the countryside. As the intensity of heat is
generally far lower than in the south, people
are rarely killed or even injured by bush-fires.
A different north
Clearly, without effective fire management,
the north Australian landscape will gradually
change. In drier areas, where fires are frequent
and intense, savanna woodlands could become
treeless. Heath plants, which can take years to
produce seed, may disappear. In wetter areas,
where early fires are ineffective at retarding
invading trees and shrubs, and the grass is
heavily grazed, reduced fire intensity will
continue to allow wet sclerophyll forests and
grasslands to disappear under rainforests and
ti-tree woodlands. These vegetation changes
will advantage some common birds and other
Economic costs are usually restricted to loss
of pastures for a few weeks to a few months,
depending on how late the wet season rains set
in. The country will green up again, so why
worry? That apathy is beginning to change.
Good fire, bad fire: shifting perceptions
The push for change has not come from people
who have suffered heavy losses, but from a real
concern for the environment, and recognition of
the impact that removal of the Indigenous people
has had on the land. In 1969, Rhys Jones caused
a stir by insisting that far from being firebugs,
by burning wherever they went, Aboriginal
Australians effectively managed the country. But
it was not until 1985 that it was demonstrated
why this was good for the land. Chris Haynes,
intending to spend time convincing a group in
Arnhemland to burn less, found instead that
the more burning they did, the less country got
burnt. Fires lit soon after the wet season were
small and patchy. As the country dried out, the
fires spread a little further, and as the people
moved about more, networks of fires produced
fire-breaks. Late in the dry season the country
was tinder dry, but the mosaic of burnt and
unburnt areas stopped any fires that were lit from
travelling far. Soon came the appreciation that
the more work put into burning country early
in the dry season, the more country remained
unburnt into the next wet.
This appreciation underpinned fire
management on Kakadu National Park from
the late 1980s. But it has not always been
successful. Proper fire management requires
planning and resources, substantial financial
and time commitments, and has to be done
properly every year. But in this country, most
environmental management is funded poorly
or briefly, with the assumption that problems
needing recurrent funding can’t be fixed, so aren’t
worth the effort.
Expanding rainforests and thickening
woodlands
In Queensland, a different picture was emerging. It
was found that grasslands and wet sclerophyll forests
were being invaded by more fire-sensitive, fire
retarding ti-tree woodlands and rainforests. Similar
changes have now been observed in other parts of
northern Australia. Important habitats are being
lost because of ineffective fire regimes. Rather than
disproving claims of too much fire, these changes
indicate that lack of fire and too much fire can and
do co-exist. In the past, mosaics that stopped fire
spreading uncontrollably also left pockets of fuel
right until the first lightning storms. Those pockets
tended to be in places too wet to burn earlier in
the year. On Cape York Peninsula, long grasses in
drainage depressions typically won’t burn before
September. By then, they should be surrounded
by a network of fire-breaks. If not, a fire sweeping
through the country in October will consume such
grassland along with the rest of the country. But, if
protected, these grasslands will burn readily as the
first storms arrive. The closer the fires to the wet
season, the more grasses are favoured over trees, so
the grasslands are maintained.
wildlife, but specialised species such as the
Golden-shouldered Parrot and Yellow-bellied
Glider will continue to decline.
Failure to manage
Three different sectors control most of the
lands of the north: conservation, pastoralism
and Indigenous land use. The reasons these
groups aren’t controlling fire as well as they
might comes down to lack of understanding,
motivation, cooperation and capacity.
One would think that loss of pasture,
and potential loss of stock would encourage
pastoralists to limit the amount of country
burned, and it does. But pastoralists are
reluctant to burn green, nutritious grass at the
start of the dry season, and would rather risk
the loss of crisp dry fodder at a time when rain
may be just around the corner. Besides, the
best of the sleek, fat cattle have been sold by
the middle of the dry season, and most of the
bony beasts that collect around water holes at
the end of the year seem to get through, even
if fires remove the last skerrick of grass.
National Parks managers, on the other
hand, are not worried about turning off a
profit, so should be able to burn at whatever
time of year is appropriate. But that time
has to fit in with other Park duties, such as
visitor management, toilet cleaning, road
maintenance, meetings and paperwork, and the
availability of staff and of helicopters to get to
remote areas. Funding cycles may mean plans
can’t be made until well into the financial year,
when burning programs need to start in April
and continue un-interrupted until the end of
July. Critically, greater political kudos is gained
from buying properties for National Parks, than
13
for managing them well; budget allocations
reflect this.
Many hope that the simplest answer is a
return to Indigenous burning patterns. But in
the vast expanses of traditional lands across the
north, mosaic burning persists mainly in places
where the traditional owners have never left
the land: where hunting, fishing and collecting
bush tucker are not just part of the lifestyle, but
an integral part. Returning to country does not
often mean returning to constant broadscale use
of country. Much of the country is too remote
to get to, and the fires lit around outstations
and along access tracks are inadequate to pull
up wildfires. A more important problem for
traditional owners is gaining access to the land,
and deriving a means of earning a living on it.
Top left: Across northern Australia, the Red-backed
Fairy-wren favours dense ground vegetation, which
can be destroyed by frequent, widespread fires.
Photo by Graeme Chapman
Centre: The aftermath of fire in tropical woodland
near the Drysdale River, Western Australia,
habitat of the endangered Crested Shrike-tit
(northern subspecies). The species feeds on insects
under the bark, which is damaged by the too
frequent, hot fires that are now more prevalent
than under Indigenous burning regimes.
Photos by Graeme Chapman
Below: Grazing by cattle, and introduced pasture
grasses, have changed the ground rules for fire
management over much of the north. The Australian
Bustard prefers the early successional stages after
fire; often moving to newly burnt grasslands and
spinifex. Photo by Michael Todd
14
Fire and Birds
A landscape of fire
In 2004, fires burned 25% of Australia’s tropical savanna region. Most of the
early season fires were lit as part of dedicated land management, of which a
major objective is to stop the spread of later fires. But this burning was not
enough. Much of the Kimberley (regions NK and CK), Arnhemland (ARC, ARP and
CA), Sturt Plateau (STU), NT Gulf (GFU and GUC) and Cape York Peninsula (CYP
and GUP) were burnt in extensive late dry season wildfires. This pattern is similar
year in year out, and the consequent loss of trees and of plants dependent on
fire for seeding is a particular concern in parts of the Northern Territory.
Areas that weren’t burnt tend to be more intensively managed, with
smaller block sizes and fuel sparse on the ground by the end of the year.
These include parts of Queensland that are subject to vegetation thickening,
attributed to a combination of overgrazing and inadequate use of fire.
Map compiled from: (i) Moderate Resolution Imaging Spectroradiometer (MODIS) mapping
undertaken by the Bushfires Council of the Northern Territory and the Cooperative
Research Centre for Tropical Savannah Management, and (ii) Advanced Very High
Resolution Radiometer (AVHRR) satellite sensor data, containing 250 m and 1.1 km, onground pixel resolution, mapped by the Remote Services Unit of the Department of Land
Information, Western Australia. Explanations for the abbreviated names of the bioregions
can be found at: http://www.deh.gov.au/parks/nrs/ibra/
Kimberley Fire Project
The Fire Control Team (FCT) project (www.kimberley-fire-project.com.au)
has been operating since January 2003 under the Kimberley Regional Fire
Management Project. The FCT has built relationships with two Aboriginal
language groups in the Fitzroy Valley to develop fire management teams
that can assist landholders with on-ground preventative burning. The teams
are now looking to strengthen the project through a Business Plan that takes
into account cultural boundaries.
Fire and Birds
15
Good management: putting fire on the agenda
The problems of the different sectors may vary. But many of the
solutions are the same. In remote areas, good fire management needs
adequate resources and people with the necessary knowledge and skills.
Across the north there are many projects trying to develop cooperation
between all landholders. These have recently combined into The North
Australia Fire Project of the Cooperative Research Centre for Tropical
Savanna Management (http://savanna.cdu.edu.au/information/
fire_knowledge_project.html). The project has four elements. The first
is helping with immediate fire management by providing up-to-date
information on where fires are, and assisting with putting in fire-breaks.
The second is building the capacity of communities to plan and manage
by providing jobs in fire management and training (in everything from
incident management to use of computers, geographic positioning
systems and geographic information systems). The third is to provide
access to information required for long-term planning, such as the
response of plants and animals, fire behaviour and cattle management,
and to develop appropriate fire strategies to guide planning. The fourth
is to provide information about fire through dedicated websites.
Seeing the fire problem as more than just an environmental issue is
already having positive results. Rural fire brigades are providing training
for jobs in natural resource management tasks that may include fencing,
weed and feral animal control, as well as fire management. Cooperation
between landholders in the Kimberley in Western Australia has resulted
in the traditional owners being invited on to cattle stations to undertake
fire management (see box below left). The Cape York Peninsula
Sustainable Fire Project has been successful at convincing increasing
numbers of landholders to undertake preventive fire management,
assisting them with the information, skills and resources to do so.
And as mosaic burning is also an effective means of reducing carbon
emissions, there is the prospect for substantial funds to be committed
towards fire planning and management well into the future.
Despite the 2004 fire season being one of the worst in recent years
for the north, cooperative programs for managing fire are starting to
show results. Land managers are beginning to appreciate just how much
effort needs to go into fire management, and it will be some years before
effective systems are in place. The most important thing now is for the
momentum to be maintained.
Gabriel Crowley and Peter Thompson are with the Cooperative Research
Centre for Tropical Savanna Management’s North Australia Fire Project, and
Peter is employed by the Cape York Peninsula Development Association.
Below and inset: The grassy habitat of the Golden-shouldered Parrot needs
a hot burn to stop encroachment by dense woodland formerly held at by
Indigenous burning practices. Photos by Michael Todd
FIRE AND ITS MANAGEMENT
AT NEWHAVEN RESERVE:
TAKING A REGIONAL
APPROACH
Recent fires have returned Newhaven to a mosaic
of fire ages
by Grant Allan
In the past there have been extensive fires in the Newhaven Reserve
region, notably in the mid 1970s, associated with the last prolonged
period of above-average rainfall in central Australia. There were also
several fires within and surrounding the Reserve in 1985. However,
by 1999 the effect of previous fires was no longer obvious. The threeyear period of above-average rainfall from 1999 to 2001 created
uniformity in the fuel load and a high fire risk throughout the region.
Consequently, fire management was identified as a priority for the new
management team at Newhaven.
The sequence of Landsat satellite image maps from 2001 to 2004
illustrates a series of fires that have caused dramatic changes to the landscape
of the Reserve. From 1999 to mid 2002, most fires started beyond the
Newhaven boundary and burned a relatively short distance into the Reserve.
Newhaven became an island of fuel. Subsequent fires, mostly started as
roadside ignitions within Newhaven, changed that situation.
As a result of these recent fires, the pattern and distribution of fire
ages within Newhaven changed from a relatively uniform mature state
to a variety of recently-burnt early successional states. Currently, there
are very few long unburnt patches of vegetation within the Reserve and
all are small.
Birds Australia recently received a small EnvironmeNT grant
from the Northern Territory government for Rachel Paltridge of
Desert Wildlife Services, Alice Springs, to prepare a fire management
strategy for Newhaven. The strategy will be closely linked to the
Vegetation Map of Newhaven prepared in 2003. It will be important
to include a regional perspective in the strategy, as many fires burn into
Newhaven from beyond its boundaries. The strategy will require both
an active burning program as well as an ability to suppress fires when
necessary. An additional challenge will be to ensure the strategy includes
appropriate sensitivities to the local bird populations, and addresses the
use of fire to control invasive weeds such as Buffel Grass.
Grant Allan, Bushfires Council of the
Northern Territory, Alice Springs
Fire can spread from
cars lit and left to burn
by the roadside; here at
Newhaven, but not an
uncommon event across
northern Australia.
Photo by Don Royal
Kurungal Fire Team use a station map to plan fire management on Bohemia
Downs in the Kimberley of Western Australia. Photo by Will Philippiadis
Landsat satellite image maps for Newhaven (at centre): May 2001; October
2001, January 2004; November 2004. Recent fires are red, which gradually
fade to orange then yellow; the salt lake is blue and white; the main road is
black and white; tracks are green-blue; and tenure boundaries are orange.
16
Fire and Birds
17
Fire and Birds
BUSHFIRE MANAGEMENT IN THE NORTHERN TERRITORY
Fire management issues in
Central Australia
White-breasted Woodswallows (left) and Crimson
Chats (below) are attracted in great numbers to
northern Australian fires, to hawk for insects flushed
by the heat and smoke. Photos by Greg Holland and
Leon Keasey, www.birdphotos.com.au and Dave
Watts, Lochman Transparencies, respectively
Reprinted courtesy of Grant Allan,
Bushfires Council of the Northern Territory
www.ipe.nt.gov.au/whatwedo/bushfires/index.html
Spinifex grasslands cover most of Uluru-Kata Tjuta National Park
(UKTNP). While the Anangu traditional owners and park managers (Parks
Australia North [PAN]) have undertaken cool season, small patch burns for
20 years, intense summer wildfires periodically burn vast areas of the park
and surrounding spinifex landscapes, generally following extended periods of
above-average rainfall. Commencing in 1987, CSIRO, PAN and traditional
owners studied the distribution of vertebrates across the park’s habitats, and
how animal distribution and abundance changed in response to rainfall
and fire. Periodic monitoring of eight permanent sites has continued to the
present, and 13 years’ data from one site are presented here to illustrate the
effects of fire and rainfall pulses on a spinifex bird community.
Site 5 is a Soft Spinifex landscape, with scattered Desert Oak and
sparse shrubs. A north-south road bisects the site, and the eastern half (East)
last burned in 1985, two years prior to the first fauna survey, whereas the
western half (West) last burned in 1976. Bird abundance was low initially
(densities about one bird ha-1). In spring 1988, more than 250 mm of rain
fell and a dramatic increase in bird abundance resulted, largely due to the
immigration of nomadic species such as Budgerigar, Crimson Chat and
Masked Woodswallow (see graph top right). Despite further substantial falls
of rain over the rest of 1988 and early 1989, abundance declined.
In 1992 a wildfire burned West, reducing it from mature spinifex
habitat to an early successional state. At the same time, eight years after
last burning, East was approaching a mature state. Following a substantial
rainfall (more than 100 mm) over summer 1994–1995, a dramatic
increase in bird abundance was recorded during the autumn 1995 survey.
When the graph (below right) is examined closely, it can be seen
that within the first three-year period post-fire bird abundance was
always greater on whichever half was more recently burnt: that is, East
in the 1987 and 1988 surveys, and West the 1994 and 1995 surveys.
Whilst observations from a single site may be equivocal, similar results
were obtained at other spinifex-dominated sites, reinforcing the notion
that early successional states support higher abundances than mature and
maturing spinifex habitats. Particularly following heavy rainfall, early
600
500
400
400
300
200
200
100
0
0
Bird Abundance
800
700
Julian Reid is based at CSIRO Sustainable Ecosystems and ANU Centre
for Resource and Environmental Sciences, Canberra.
Bird Abundance vs Time Since Fire, East
900
600
• Typically, spinifex grassland bird communities have few species and
these are low in abundance. Less than 10 species appear to be resident
at Site 5, and most of these (Singing Honeyeater, Yellow-throated
Miner, Black-faced Woodswallow, Pied Butcherbird) are associated
with emergent shrubs and trees, such as Desert Oak, rather than
spinifex itself.
• Bird communities differ in abundance and composition in response to
fire in spinifex grasslands. However, the magnitude and timing of these
responses are not predictable, and are largely dependent on ensuing
major rainfall events. Following heavy rains, immigrant nomadic
species can numerically dominate spinifex habitats and be particularly
abundant in recently burnt areas, within three years of last fire, when
a burst of ephemeral plant growth provides diverse food resources
(nectar, seed, invertebrates).
• Of the (loosely) sedentary species that feed and nest at ground and nearground level, fairy-wrens and the Striated Grasswren favour mature
spinifex grasslands, while the Banded Whiteface occurs patchily in
early successional habitats. The presence of habitat of an appropriate
successional state does not guarantee a species’ occupancy.
Preliminary management recommendations:
• Given the patchy distribution of bird species of conservation
significance, and the capability of severe wildfire to reduce extensive
tracts of spinifex landscapes to the one successional state, it is prudent
to pursue an active patch burn strategy, as has been adopted at
UKTNP. This form of management can limit the extent of some
wildfires and will promote finer-scaled mosaics of different fire ages
than result from the action of wildfires alone.
• Wildfires should not be excluded from these landscapes—indeed they
probably cannot be excluded from any but small areas. Patches burned
by summer wildfires, in combination with cool season prescriptive
burns, can be expected to increased vegetation heterogeneity across
the landscape, and such mixes ought to promote a greater bird species
diversity at the larger spatial scale.
6mo RF
Abund E
Abund W
900
800
700
600
500
400
300
200
100
0
0.0
Bird Abundance vs Time Since Fire, West
Abundance
800
ut
9
Sp 9
r9
9
Land managers need to plan for fire, both for
its use and its risk. Fire may not be a regular
feature within any specific part of the landscape,
but it is important to be prepared. Active fire
management is required, as is the appropriate use
and timing of fire.
The easiest job is to monitor spinifex areas
and burn patches to create a mosaic of fire ages.
It is also important to manage fuel loads in highrisk areas, especially along main travel routes
where uncontrolled ignitions are most likely. It
is possible to use fire as well as grading, slashing
and grazing programs to reduce fuel loads.
It is more difficult to be prepared for seasons
of above-average rainfall and widespread fuel
loads. However, it is important to plan to use
fire, for habitat management or control of woody
thickening, and be ready to respond when season
and rainfall conditions combine to create an
opportunity to burn.
What are the lessons?
3.0
6.0
9.0
Years since last fire
12.0
15.0
900
800
700
600
500
400
300
200
100
0
0.0
3.0
6.0
9.0
12.0
Years since last fire
There was no consistent relationship between bird abundance and time since last fire in
spinifex grasslands—the highs in bird abundance were rainfall driven.
A
Large wildfires are undesirable for most plants
and animals, especially if there are no unburnt
island refuges within the burnt area. This
includes both domestic stock and native animals.
What we can do
successional habitats experience a burst of ephemeral plant growth that can
attract a long list of nomadic bird species, for example, Little Button-quail,
Australian Bustard, raptors, songlarks and White-winged Triller.
Birds of the spinifex are ruled by rainfall and fire
by Julian Reid
9
Sp 7
r9
7
Threats
Spinifex grassland burns near Hall’s Creek, Western Australia. Photo by Graeme Chapman
FIRE AND BIRDS IN
SPINIFEX GRASSLANDS
A
ut
The non-spinifex communities are a diverse
range of vegetation communities. They include
grasslands on cracking clay soils, open woodlands
with a mix of grasses and acacias, and some areas
of acacia woodlands with minimal understoreys.
Fuel loads in these areas respond more rapidly
to rainfall relative to the spinifex areas, but cattle
and other grazing animals, including termites,
readily consume the grass. Therefore fuel loads
only build to high levels across the landscape
after extended periods of widespread rain, when
growth can far exceed consumption.
The mountain ranges are complex areas,
with non-spinifex communities dominating the
flatter portions of the landscape and spinifex
dominating the hills and ridges. Within the
ranges are innumerable gorges and gullies,
which are both important scenic locations for
parks and tourism as well as refuges for animals
and fire sensitive plant species.
The Striated Grasswren favours mature spinifex. Photo by Lynn Pedler
A
ut
9
Sp 4
r9
4
A
ut
95
Climate is the main determinant of the
occurrence of fires in central Australia. During
the last century periods of above average rainfall
have occurred in cycles of 25 to 30 years.
Historical accounts of extensive wildfire seasons
show a direct correlation with these wetter
years rather than with periods of drought.
Nevertheless, fires do occur at other times, when
local conditions contribute to the accumulation
of fuel and enable ignitions to spread and
develop into major fires.
Historically, the patterns of fire activity are
associated with three distinct components of
the landscape, with links to broad vegetation
communities. ‘Spinifex communities’ are most
regularly and extensively impacted upon by fire.
‘Non-spinifex vegetation communities’ have
the least frequent incidence of fires. ‘Mountain
range areas’, with a mix of spinifex and nonspinifex vegetation, have their own unique fire
management issues. Interestingly, the separation
roughly equates to areas of Aboriginal land,
pastoral land, and conservation land, respectively, but during periods of above-average
rainfall each landscape component can be
severely affected by wildfires.
Spinifex communities occur on the
extensive areas of sandplains and dunefields,
typical of desert regions. Spinifex is very
flammable and fuel loads steadily accumulate
as the hummocks increase with time. The rate
of accumulation is variable in association with
rainfall, and the interval between fires—limited
by the accumulation of fuel—can be as long as
20 years.
Rainfall previous 6 months (mm)
Current fire regimes
The time of the fire in relation to soil moisture
and rainfall can also increase the impact of fires.
Hot fires in dry times mean the post-fire recovery
of vegetation and associated food resources is
very slow. Uncontrolled ignitions in times of
widespread fuel loads, especially during recent
years, have caused large wildfires. The fires have
damaged infrastructure and fences, destroyed
pasture for grazing animals, and directly killed
cattle and other animals.
Fires can also reduce the extent of fire
sensitive vegetation communities, such as
Mulga. The fires kill mature Mulga plants and
increase the opportunity for spinifex to invade
the Mulga community.
A relative new feature of the fire regimes
of central Australia is Buffel Grass. It was
introduced as a pasture grass, but has taken over
most alluvial areas and is continuing to spread
into other habitats. Buffel Grass can recover
quickly from a fire and will out compete other
plants to become dominant.
The rapid build-up of Buffel Grass fuel has
increased the fire frequency in many areas and
long-lived woody species, such as River Red
Gums, Corkwoods and Beefwoods, are suffering
from the frequent fires.
There are many areas in central Australia
where it is desirable to exclude fire. These include
sheltered gullies in range country, and long-lived
fire sensitive communities such as Mulga and
Lancewood. Protecting these areas from fire is a
difficult task.
A
ut
8
Sp 6
r8
A 6
ut
8
Sp 7
r8
A 7
ut
8
Sp 8
r8
A 8
ut
8
W 9
in
8
Sp 9
r8
A 9
ut
90
Although fire is not generally perceived as
being associated with desert landscapes, in
arid Australia fire occurs on a regular basis and
is an important driving force in the ecology.
It is therefore critical that arid zone land
managers plan for the use of fire and be ready
to manage its impact.
Total bird abundance recorded on 1-km transects on the East (shown as x–x) and West (o--o) halves of Site 5, Uluru Kata Tjuta National Park following fires. Rainfall is shown as bars.
15.0
Fire and Birds
19
Far left: Western Whipbird (male). Photo by Graeme Chapman
Left: Western Bristlebird. Photo by Simon Nevill, Lochman Transparencies
Noisy Scrub-birds, Western Ground Parrots, Western Bristlebirds
and Western Whipbirds—all known to be fire sensitive and listed as
threatened species—occur on the south coast of Western Australia,
often in close proximity to each other.
Curiously, these birds occur in an environment in which fire is
a prominent natural process. Summer lightning strikes ignite intense
wildfires, a regular occurrence on the south coast of Western Australia.
For about the last 50,000 years this has been overlain by Aboriginal
burning, apparently frequent but spatially variable. Interestingly, the
intensive Aboriginal fire regimes allowed scrub-birds, bristlebirds,
Ground Parrots and whipbirds to persist in this fire prone environment.
Then, about 100 years ago, another major change occurred—Europeans
started fragmenting the natural vegetation and made profound changes
to existing fire regimes.
Generally, fire became less frequent but more severe and more
extensive, and re-colonisation of burnt areas was hampered by fragmentation of the habitat. This process has been a major factor in the
decline in scrub-birds, bristlebirds, Ground Parrots and whipbirds, and
undoubtedly some other species. It has also resulted in major challenges
for conservation managers.
Managing Mt Manypeaks
One area that constitutes a particular challenge in fire management is
Mt Manypeaks, to the east of Albany and just east of Two Peoples Bay. A
census in 2001 showed that this area contained about 55% of the global
population of the Noisy Scrub-bird, more than 25% of the population
of the mallee heath subspecies of the Western Whipbird and, on the
lower slopes, significant numbers of the Western Bristlebird. In addition,
small numbers of the Critically Endangered Western Ground Parrot were
known from the flats to the north. The area also supports threatened
mammals and plants, as well as invertebrates of high conservation
significance. For some of these species, optimal fire regimes are unknown;
on 31 December. Fire fighting resources were immediately mobilised
and, because of the high conservation values, suppression of the fire
was given a very high priority. Over 160 personnel, including CALM
staff from as far away as Geraldton (over 700 km to the north), were
involved with the suppression activities over the ensuing days. Four
water bombers, two spotter planes and one helicopter were also
involved. Unfortunately the high fuel loads on the southern slopes and
lack of ground access made operations very difficult. It took several
days to establish secure boundaries around a larger area and this made it
possible to contain the fire through back-burning. Populations of Noisy
Scrub-birds and other threatened species to the east and west of the Mt
Manypeaks study area were then protected, as well as private property to
the north. The fire was under control by 6 January, although pockets of
vegetation associated with peat were still smouldering in early February.
The impact of the wildfire on the Noisy Scrub-bird was substantial
(see map below). While a few would have undoubtedly managed to
escape to neighbouring areas, it is expected that almost all of the birds
in the 427 territories counted in the Manypeaks study area in 2001
have perished. Furthermore, the loss in this one event of 4500 ha within
approximately 13,000 ha of optimal habitat, in addition to approximately 4000 ha of habitat that has been burnt by wildfires since 2001,
means that there are now only around 4500 ha of optimal scrub-bird
habitat left in the Albany area.
NOISY SCRUB-BIRD POPULATION INDEX
Mt Manypeaks
450
427
400
350
Note: 1997 figures based on estimates
from partial survey of Mt Manypeaks
300
250
156
150
100
0
319
223
200
50
322
60
4
4
12 10 15
81
100
26 32
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
Fire management must be tailored to local values
and constraints
by Allan Burbidge, Sarah Comer and Alan Danks
for others, requirements differ between species. As distributions of these
species are highly overlapping, management of fire is not straightforward.
While some of the lower slopes and adjacent plain have burnt in recent
years, the mountainous part was last burnt in 1979 and in recent years
have carried high fuel loads. The Department of Conservation and Land
Management (CALM), which is responsible for the management of this
area, and the South Coast Threatened Birds Recovery Team (SCTBRT),
were concerned about the inherent risks to very high nature conservation
values posed by these high fuel loads. However, the density of threatened
birds in the area and their conflicting requirements made simple solutions
difficult. And any fuel reduction program in scrub-bird habitat would
result in reductions in the numbers of individuals. Issues such as these
were discussed at a series of workshops in 2002–2003, which ultimately
developed a strategy aimed at creating a mosaic of fuel ages in the area over
time. Many people contributed, including SCTBRT members, members
of other relevant recovery teams and various fire management specialists.
In 2004, as part of this fire management strategy, the SCTBRT endorsed a
prescription burn that was to be carried out in winter 2005.
However, in late December 2004, before the strategy could be
implemented, a lightning strike ignited the southern slopes of the
mountain. This was to develop into a major challenge for fire fighters,
and result in significant losses of Noisy Scrub-birds.
A local fisherman reported a thin column of smoke in the area
Population Index (No. Singing Males)
THREATENED BIRDS AND
WILDFIRE IN SOUTH-WEST
WESTERN AUSTRALIA
The recent wildfire has highlighted the importance and historical
successes of the scrub-bird translocation program. This was aimed at
decreasing the susceptibility to fire by spreading the population across the
wider landscape. Mt Manypeaks was the site of the first translocation of
Noisy Scrub-birds in 1983. The founder group of 31 birds (18 males and
13 females), was released in 1983 and 1985. The growth in this area to
a population index of 427 singing males in 2001 (see graph below) was
quite remarkable, and was achieved through a policy of fire exclusion. The
population was continuing to increase, as indicated by a survey of 800 ha
on the eastern end of Mt Manypeaks in 2004. However, the recent wildfire
has reduced total numbers to roughly what they were ten years ago.
The Western Whipbird (western heath subspecies) population was
also impacted by the wildfire. The area burned contained habitat for
approximately one-quarter of the total population. In conjunction with
wildfires in 2003, the habitat for approximately one-third of the western
heath subspecies of whipbird has been affected.
Year
Above: The Noisy Scrub-bird is restricted to a few small, isolated populations
in south-western Australia; its decline in large part due to inappropriate
burning regimes—it needs dense thickets unburnt for 40 years
or more. Photo by Graeme Chapman
Left: 2001 Noisy Scrub-birds territories in the Albany management zone
(shaded grey) and fire history 2000–2004
20
Fire and Birds
Fire and Birds
21
BIRDS AND FIRE IN THE
MALLEE HEATHLANDS
OF NGARKAT
Lack of information on ‘hot spots’—havens for wildlife
during times of stress—hinders fire management
by David C. Paton, Daniel J. Rogers,
Matthew J. Ward and Jody A. Gates
be different from the next. It follows that,
for optimum conservation outcomes, fire
management for a given site must be tailored to
the local values and constraints, and based on
good knowledge of the response to fire shown
by individual species.
In the case of Mt Manypeaks, everyone
agrees that the area is far too valuable for
the ‘do nothing’ option to be acceptable, as
it will simply result in further extensive and
damaging fires. In the next few years there is
an opportunity to establish a managed fire
regime that minimises the risk of extensive
fires affecting the whole of the mountain
area. Of course this means the area will
support fewer Noisy Scrub-birds and Western
Whipbirds than it did immediately prior
to the December 2004 wildfire, but the
population should be more sustainable.
This, in turn, may benefit other threatened
species such as the Western Bristlebird.
It is likely that a number of strategies will
be utilised, in combination with public participation to increase awareness of the values of
the mountain and the practical difficulties
in management. Some extra slashed breaks
have been put in place in strategic locations
in recent years, but more physical fire-breaks
are not likely to be an option, partly because
of the terrain, and partly because of the
increased risk of the spread of Phytophthora
(which destroys many plant communities on
the south coast), and increased fire ignition
risk. Aerial ignition under moist conditions,
which allows the ridges to burn but not the
gullies, could be used to provide a mosaic
of fire ages. This approach is the one most
Clockwise from top left:
Fire burning through thick heath towards the
Mt Manypeaks ridge in the first few days of the
2004 fire. Photo by Alan Danks
The heathlands of Mt Manypeaks before the 2004
fire. Photo by Sarah Comer
Members of the South Coast Threatened Birds
Recovery Team—John Blyth, Allan Burbidge and
PhD student Abby Berryman—visit Manypeaks
after the fire. Photo by Alan Danks
Waterbombers were used in an attempt to slow
progress of the fire front during the initial stages
of the wildfire. Photo by Alan Danks
How does mallee-heath birdlife respond to fires?
Unlike heathland vegetation, most heathland birds have no special
adaptations to cope with fire. They either flee the fire by moving to
unburnt areas, or they perish. One only needs examine recently burnt
likely to provide for long-term persistence of
the diverse conservation values of the area,
although this will mean lower numbers of
Noisy Scrub-birds.
Managing threatened birds on Western
Australia’s fire prone south coast is challenging
and complex. As part of their commitment
to this task, the South Coast Threatened
Birds Recovery Team and local CALM staff
will continue to design and implement fire
management strategies and monitoring of
outcomes, to improve the conservation status
of threatened plants and animals in the region.
Allan Burbidge, Sarah Comer and
Alan Danks are ecologists with the
Department of Conservation and Land
Management, Western Australia.
This diagram is a composite, with six study sites, last burnt in 1978, contributing to abundance
estimates beyond 15 years since the fire. The other 26 sites contributed primarily to the earlier
time since fire categories. The patterns shown are truncated in the sense that they are limited
to heaths that are no more than 26 years post-fire, so some care is required in interpreting
post-fire patterns. Most of Ngarkat’s mallee heathland birds do occupy older heaths, up to
50 years post-fire at least.
1
RP
AM
CAL
SBT
TCH
VFW
0.8
0.6
0.4
0.2
0
0
5
15
10
TSF (years)
20
Hot fires are considered a threat to the patchy mallee shrub habitat of the
Red-lored Whistler. Photo by Lynn Pedler
areas in Ngarkat to appreciate this—immediately after fire there are no
typical heathland birds present on burnt areas. In many cases the adjacent
unburnt areas are already occupied and so there are few places available
to act as refuges. Counts in unburnt areas adjacent to burnt areas in
Ngarkat typically show no increase in abundance even after extensive
fires—clearly, few birds survive by moving to adjacent unburnt areas.
In contrast, mallee-heath vegetation responds immediately
post-fire. Some species sprout from epicormic buds protected from the
heat of the fire by bark (Eucalyptus), soil (Allocasuarina, Leptospermum)
or dense foliage (Xanthorrhoea). Other plant species subsequently
re-establish populations from germination of seeds protected in woody
cones during the fire’s passage with the seeds shed immediately after the
fire (Banksia) or germinate from seed banks in the soil. Although this
vegetation response may be rapid, for most of the birds there is a period
of several to many years before the structure and productivity of these
regenerating heaths are once again adequate to support them, the time
varying between species. Thus, fires in these mallee heaths result in the
temporary removal of suitable habitat for most species of birds.
Many patterns of recovery
The great variation in annual changes in winter (July–August) density of
six species of birds over time since fire (TSF). Density is given relative to the
maximum density recorded for that species. RP = Richard’s Pipit, AM = Australian
Magpie, CAL = Rufous Fieldwren, SBT = Slender-billed Thornbill, TCH = Tawnycrowned Honeyeater, VFW = Variegated Fairy-wren.
Proportion density (0–1)
On the other hand, Western Bristlebirds
colonised the Manypeaks ridge after the 1979
fire, but as the vegetation increased in height,
they disappeared from the area. Following
the recent fire, presumably they will be early
colonisers and will do well again for a period
of time. After every fire there are both positive
and negative impacts.
One positive outcome of the Mt
Manypeaks fire is that we now have a ‘blank
slate’ with respect to planning for future
fire management. Monitoring populations
of scrub-birds, whipbirds and bristlebirds
in 2005 will give us a good picture of the
current population status of all of these
species in the Albany Management Zone.
In coming years, it will allow us to study
their responses to wildfire in a way that
hasn’t previously been possible. Post-fire
development of both vegetation and leaflitter invertebrates (food for all three bird
species) will also be monitored.
Results of such studies on threatened birds
are being used to underpin future management
decisions aimed at avoiding similar drastic
population declines on the south coast in the
future. Exactly what this management will be at
Manypeaks is not yet clear, but it will continue
to involve ornithologists, fire specialists and
local area managers in collaboration, to develop
practical options that have specific conservation
aims, and to monitor the effect of management
actions. An important principle here is that
fire management must be site-specific. The
conservation values and physical environment
of one site are different from the next. Also,
the preferred fire regime for one species will
Ngarkat is an extensive area (2,670 km2) of mallee heath vegetation in
South Australia that abuts Victoria’s Big Desert (see map p. 22). The area
has been set aside as a reserve in part because it supports a range of rare
and threatened species, including Mallee Emu-wrens, Western Whipbirds,
Red-lored Whistlers, Striated Grasswrens and Slender-billed Thornbills.
Also present are Tawny-crowned Honeyeaters, Purple-gaped Honeyeaters,
Rufous Fieldwrens, Shy Heathwrens, Inland and Buff-rumped Thornbills,
Variegated and Superb Fairy-wrens, and Southern Scrub-robins.
The area experiences frequent fires that are started by lightning
during late spring and summer. Beginning in December 1990 there
have been four extensive fires that burned an estimated 582 and
658 km2 (January 1999), 274 km2 (November 2002) and 200 km2
(January 2005), respectively (see map p. 22). Despite being a large park,
over the last 15 years more than a third of it has consisted of vegetation
in early (less than 10 years) post-fire succession, and over 60% is less
than 20 years post-fire. Some of the central areas of Ngarkat have
experienced up to six fires in the last 60 years and only a limited area
(15%) has not been burnt in that time.
Each year since 1990, the numbers of birds using these mallee heaths
were counted along two 1.5 km long (x 100 m wide) strip transects at 32
sites, spaced at least 3 km apart. During that time, 26 sites have been burnt
at least once, with six of these burned twice and one site burned three times,
at least partially. These counts provide a unique opportunity to investigate
how the bird populations recover following fire.
25
Understanding the patterns of re-colonisation of burnt habitat by
heathland birds, and the factors that influence them, are critical to
managing these mallee heaths. Simply recording when a species is
present on regenerating heaths is unlikely to be an adequate measure of
successful re-colonisation. In the initial years of occupation, the birds
that are present may not be able to breed successfully. Ideally, one needs
to establish when regenerating heaths allow a species to reproduce.
Nevertheless, the abundances of species on regenerating heaths with
time since fire are likely to be indicative of its habitability.
Our data illustrate that Richard’s Pipit, an open country species
prominent on recently burnt heaths, is an early successional species (see
graph at left). Other species take longer to re-establish on regenerating
heaths. Some, such as the Slender-billed Thornbill and Rufous
Fieldwren, are more abundant on heaths around five years post-fire,
whereas Variegated Fairy-wrens are more prominent about 15-years
post-fire, but all of these species are present in older heaths as well.
The most striking pattern in the abundances of most species is
that they fluctuate, sometimes greatly, from one year to the next. Thus,
there are no simple patterns of changes in abundance post-fire, and
factors other than time since fire are also important. For example, the
low numbers of Tawny-crowned Honeyeaters ten years post-fire are due
largely to a period of drought from 1999–2002, nine to 12 years since
the 1990 fire, and the high numbers 13–16 years post fire are due to
high counts of this species during a wetter period (1992–1994) on sites
last burned in 1978.
22
Fire and Birds
Fire and Birds
The actual timing and extent of recolonisation are not just a function of the
regenerating heaths reaching some level of
suitability, but are also likely to be influenced
by the area of heath that is burnt relative
to the unburnt heaths that remain. If large
areas are burnt, the capacity of the remaining
unburnt areas to produce sufficient surplus
individuals to restock the burnt areas with
birds is reduced—that is, the larger the fire,
the slower the rate of re-colonisation.
Post-fire environmental conditions in
combination with the proximity of older
heaths that act as sources of recruits are also
likely to influence the speed of recovery.
For example, droughts affect reproductive
performance and survival of birds, and
hence the ability of the unburnt heaths
to provide recruits for re-colonisation of
regenerating heaths. Dry periods also affect
the growth and survival of heathland plants.
So droughts dampen the recovery of these
mallee heathlands and their bird populations.
The pattern of post-fire recovery, therefore,
is unlikely to be identical after every fire for
either the plants or the birds.
Are mallee heathland birds evenly
distributed over Ngarkat?
Map of the ever-changing mosaic of major fire scars at Ngarkat between 1945 and 2005.
Some sites support more bird species
than others, regardless of time since
fire: mean species richness for four
sites last burned in 1978 (open bars)
and four sites last burned in 1990
(solid bars).
Species Richness
25
20
15
10
5
0
OO
AA
BB
SS
SITES
QQ
JJ
DD
EE
Data show the mean number of species
recorded (+ s.e.) during duplicate counts
of birds along two 1.5 km x 100 m transects
at each site in winter over seven years
(1996–2002 inclusive).
The abundance and performance of the
biological assets within Ngarkat are almost
certainly patchily distributed. Particular
areas—somewhat paradoxically known as
‘hot spots’—are likely to be much more
important for supporting biota than others,
especially during periods of drought or stress.
This fundamental concept is supported by
the results of monitoring birds, mammals and
selected plants at Ngarkat since 1990. For
example, some sites consistently support more
bird species than others and this is independent
of time since fire (see graph at left). Greater
numbers of species are consistently recorded
in some 1 ha (100 m) sections along each
transect, while other sections consistently
support few species (see maps below).
Distribution of species richness along the north western transect at Site OO for winter 1991, 1995 and 2000, respectively. The size of each spot is proportional to the
number of species detected during duplicate counts for each 100 m section of the 1500 m transect. The darker patches on the image show where the heath vegetation
is taller and denser and includes more eucalypts; these areas are on the south-east faces of dunes or in depressions and consistently support more bird species relative
to other sites in the same time period.
1991
1995
2000
For Ngarkat the hot spots appear to be
centred on areas that receive higher moisture
(e.g. near the bases of the eastern/southern
faces of sand-hills). These moister areas
appear to be more frequent in the western
and southern parts of Ngarkat, where annual
rainfall is higher, and the terrain hillier. The
sizes of hot spots are not well defined and vary
from less than a hectare to tens of hectares.
Other taxa show similar patterns with
respect to their distribution, abundance and
performance. For example, pygmy-possums
occupy more habitat areas following several
years of good conditions, but in droughts
their distributions retract to moister areas.
Floral performances of key plants like Banksia
ornata (number of inflorescences produced,
sizes of inflorescences and nectar production)
are higher in moister areas. Although floral
production for this species is depressed in
drought, the plants in the moister areas always
perform better than conspecifics living in
poorer areas (e.g. interdunal flats) and have
lower rates of plant mortality.
Simple fire management models and
strategies that ignore these patterns of distribution place at risk the biological assets of
a region. If the hot spots were all burnt in
one fire, then drought-sensitive taxa have no
refuges during droughts, and hence greatly
reduced ability to survive in the landscape.
What do we need to know to manage
fires and birds in the mallee heathlands
of Ngarkat?
In fire-prone heaths like those of Ngarkat
there are two key components to fire
management: (1) reducing the risks of
extensive areas being burnt in single fires;
and (2) protecting key biological assets from
fire, particularly when much of Ngarkat is in
early post-fire succession, as is currently the
case. This is all the more so because extensive
areas of the adjoining Big Desert are also in
early post-fire succession: a consequence of
an extensive fire in 2002–2003 that burned
more than 1800 km2 of mallee heathland.
Implementing such a management strategy
requires, as a minimum, knowledge of the
locations of the key biological assets, and how
these vary in space and time.
Ngarkat supports a range of threatened
and vulnerable bird species, and these are
important assets. Slender-billed Thornbills,
despite showing a strong numerical response to
heaths during early stages of regeneration, are
not likely to be threatened by fire since they are
widely distributed and can use a wide range of
post-fire heath communities. Our knowledge
of the ecological requirements of Mallee
Emu-wrens (see p. 24), Red-lored Whistlers
and Western Whipbirds in Ngarkat is poor by
comparison but in general, fires, particularly
extensive fires, are considered a threat.
Lack of knowledge hampers good fire
management
As for most of temperate Australia, there is an
urgent need for detailed ecological studies on
key fauna to inform and develop sensible fire
management in Ngarkat. This lack of detailed
information on key assets, and how their
performances vary across the landscape and
under different prevailing conditions, and not
fire per se, are the biggest threats to birdlife in
these mallee heath systems. While we lack this
basic knowledge, key assets are at risk, and our
management of fire will be grossly inadequate
and negligent.
David C. Paton, Daniel J. Rogers and
Matthew J. Ward are based at the School of
Earth and Environmental Sciences, University
of Adelaide, and Jody A. Gates is with the
Department of Environment & Heritage
at Berri, South Australia.
Continued overleaf
23
Top left and right: Variegated Fairy-wren and
Rufous Fieldwren. Photos by Lynn Pedler
Above: Burnt (1 month after fire) and unburnt
(12 years since fire) heathlands in Ngarkat
Conservation Park show the stark contrast in
habitat for birds.
Below: Ngarkat Conservation Park often render
extensive areas of heathlands temporarily inhospitable
for heathland birds. Extensive fires like this in
November 2002 threaten the long-term survival of
heathland birds. Photos supplied by David Paton
24
Fire and Birds
Fire and Birds
25
Birds and Fire in the Mallee Heathlands of Ngarkat continued
BIRDS, FIRE AND NADGEE
For Australia’s south-east forest and heathland
birds, not all fires are a disaster, but too frequent
burning by land managers can be
by Harry F. Recher
Distribution and status of Mallee Emu-wrens in Ngarkat.
The figure shows sites where Mallee Emu-wrens were recorded in the 1990s but are no
longer present (red spots), sites where they have been recorded in the last 2 years (green
spots), and sites where they were recorded in the 1990s but have not been confirmed as
still present (yellow spots). The grey background shows areas that have been burnt since
November 1997, where the habitat is currently not suitable for Mallee Emu-wrens.
Mallee Emu-wren. Photo by Graeme Chapman
Mallee Emu-wrens: a species in peril
Within Ngarkat the distribution of Mallee Emu-wrens is linked to dense low
vegetation provided primarily by patches of Triodia scariosa, often in association
with dense patches of Allocasuarina pusilla, and sometimes Xanthorrhoea
australis. This habitat is not evenly distributed over the park. For example the
larger patches of dense Triodia scariosa are largely restricted to the bases of
southern to eastern facing dunes; areas that are likely to get more moisture and
be more productive (better growth of plants, higher invertebrate abundances).
Over the last 15 years, combinations of fires and droughts have greatly
reduced the distribution and abundance of Mallee Emu-wrens in Ngarkat
(see map). These declines are not just due to areas being burnt, because the
species has also disappeared from heaths that have not been burnt, and those
disappearances have coincided with droughts. Less than 10% of over 100
locations where Mallee Emu-wrens were detected in the 1990s are known to
be currently occupied, and then only by small numbers of birds (less than 10).
Recent extensive fires have rendered much of Ngarkat temporally unsuitable
for these birds, and even when these heaths become suitable structurally it is
hard to imagine how the few, small and fragmented populations which remain
can re-colonise the now extensive areas of regenerating heaths in Ngarkat and
the Big Desert. A fire that burns the few areas currently supporting birds would
be tragic. There is clearly a premium to protect the remaining populations of
this species from further fires, to give them a chance to recover their distribution and abundance by re-colonising areas that are currently unoccupied.
In December 1972, wildfire burned all but 100 ha or so of the Nadgee
Nature Reserve and much of the adjoining State Forests on the far south
coast of New South Wales. Birds were caught up in a firestorm by the
thousands and dropped into the sea, to wash up in windrows along the
beaches. By any description, the Nadgee fire was an extreme event. In all
probability, more than 500,000 birds were killed in Nadgee alone, along
with tens of thousands of mammals, reptiles and frogs and countless
millions of insects, spiders and other invertebrates. With the fire still
burning in logs and hollow trees, a Pallid Cuckoo was found feeding on
caterpillars in the blackened moors, and a group of Southern Emu-wrens
survived in one of the tiny patches of unburnt vegetation along the Nadgee
River, but there wasn’t much else in the way of living birds in the aftermath.
With wildfires like that, it is no wonder that the majority
of Australians and the national media regard bushfires with fear.
Uncontrolled fires are described as ‘destroying’ the bush, as ‘a holocaust’,
and as so devastating that they have to be prevented almost regardless of
cost. Yet, within months of the fire, Nadgee was green again and birds
had returned. After a few years there was no evidence that any species of
plant or animal present before the fire had been lost from the Reserve.
Some birds at Nadgee may have flown over the flames and dropped
down onto the burnt country behind the fire front. I’ve seen birds do this
during a controlled burn in the Brisbane Waters National Park (BWNP)
just north of Sydney. Although fuel loads in the BWNP heath and
woodland were high, and flames soared above the tallest trees and shrubs,
birds moved easily in front of and above the fire to appear minutes later
on the blackened vegetation to feed on less fortunate insects and small
lizards. Despite a careful search after the flames had died, we found no
dead birds, but the burn was only 20 ha or so in area. As best we could tell
over the following years, all the banded birds whose home ranges included
the burnt patch survived. With this small fire, it was easy for birds to shift
short distances and find suitable habitat in which to feed and shelter. This
could not have happened to any extent at the Nadgee fire.
Some birds at Nadgee no doubt escaped the flames by flight and
returned from outside the reserve as the fires died and the vegetation
regrew or, like the Pallid Cuckoo, found food and shelter in the burnt
landscape. However, the rate of recovery of the heath and forest birds
was too rapid to be accounted for by the few survivors within the reserve
itself. Nadgee had to be recolonised by birds moving in from unburnt
country north and south of the reserve.
The source of birds colonising Nadgee has always been a matter of
great curiosity for me, but the answer will probably never be known. Not
that it isn’t important, but what really matters is that the birds, plants and
other animals recovered and recovered rapidly despite the intensity and
extent of the blaze. If this tells us one thing about the ecology of fires and
birds in Australia, it is that a single fire, regardless of how hot and big it
is, need not be a devastating event. I never thought of the Nadgee fire as
a ‘holocaust’, despite seeing all the death it brought. Instead, I liked to
describe the 1972 fire as a ‘Grand Natural Experiment’—a chance event
that allowed us to study how wildfire affected the Australian fauna.
At first my research with Dan Lunney from the then NSW National
Parks and Wildlife Service focused on the effects of the fire on a community
of small mammals we had been studying at Nadgee since 1968. However,
in 1979 I learned that some people in the Service thought the heaths at
Nadgee needed to be burnt. Their reasoning, apart from a prevailing notion
that uncontrolled wildfires needed to be prevented, was that the heaths had
not been burnt for seven years and the regrowth vegetation was no longer
suitable for Ground Parrots; an iconic and threatened species restricted
to shrubby, low heathlands along the coast and in high country on the
escarpment. Their reasoning made no sense to me, as during our routine
sampling of small mammals on the coastal moors at Nadgee we never failed
to find Ground Parrots and other heath birds in abundance. I suspected
that the alleged demise of Ground Parrots at Nadgee had more to do with
a ‘burning’ need to reduce fuel loads to protect people and their property
than any demonstrated need to manage heath vegetation for parrots. With
the help of the Field Observers Club and staff from the Australian Museum,
I therefore conducted a survey of heath birds at Nadgee. The survey was
done in the spring of 1979 and demonstrated that not only were Ground
Parrots abundant at Nadgee, but so were a full complement of other heath
birds, including the Eastern Bristlebird, another threatened species. As far as
I knew, the plans to burn the heaths were suspended.
Curiously, in 1980, the Nadgee heaths did burn, but not at the
hand of the Service. Human carelessness in forest bordering the reserve
led to a fire in October, which affected most of Nadgee. Unlike the
1972 fire, the 1980 fire, although extensive, was not an extreme event
and much of the reserve remained unburnt or only lightly scorched.
However, all but half of one of the coastal moors did burn intensely.
Only the southern half of Impressa Moor escaped burning in 1980 and
I decided to monitor the recovery of the heath avifauna on the northern,
burnt half of the moor using the southern half as an ‘unburnt control’.
Remember, the whole moor burned in 1972.
Recovery of birds on Impressa Moor from the 1980 fire was as rapid
as after the 1972 fire, but now we had census data to better describe the
pattern of colonisation by birds. In the first year after the fire, Impressa
Moor was ‘invaded’ by large numbers of quail, bronzewing pigeons and
finches responding both to the seed released after the fire from the burnt
heath plants and then to the seed produced by the abundant growth of
native grasses. Forest birds were also prominent in the first 12 to 18 months
and were obviously taking advantage of the abundance of insects feeding
on the luxuriant, and no doubt nutritious, plant growth. However, within
two years, the original heathland avifauna had re-established itself and the
numbers of seed-eaters and forest birds had diminished. As the heath grew
and the vegetation became taller and denser, with fewer patches of bare
soil, birds like the Australian (Richard’s) Pipit declined and disappeared.
By 2000, numbers of Tawny-crowned Honeyeaters—another species which
forages extensively on the ground—appeared to have fallen, as perhaps had
the numbers of Ground Parrots and Striated Fieldwrens, especially in the
southern half of the moor where the vegetation was now 28 years post-fire.
Also in 2000, several pairs of bristlebirds had established territories: one in
post-1972 vegetation and two others in post-1980 vegetation.
It is still too early to be sure of all the changes that have taken place
on Impressa Moor since the 1972 and 1980 fires. What is certain is
that, after nearly 30 years, both the vegetation and the birds associated
with it continue to change in abundance and species composition.
Even the sequence of two intense fires within a decade had not led to
the loss of any species of bird, plant or other animal from the moor.
These observations do lead to some questions about the management of
vegetation, birds and fire in Australia’s heaths, forests and woodlands.
First, it is clear that an intense wildfire need not be a disaster for
the fauna or the vegetation. Even a second fire relatively soon after the
first may not lead to the loss of species, so long as there is a source of
colonists nearby. Second, if fire is to be used to manage the birds or
other fauna of a heathland or forest, who decides which species are most
important? Had Impressa Moor been burnt in 1979 as planned, the
reduction in the height and density of the vegetation would have
favoured pipits and Tawny-crowned Honeyeaters. A controlled burn
in 1979 would probably have neither advantaged nor disadvantaged
the local Ground Parrot population, but it would probably have
disadvantaged the Eastern Bristlebirds. Bristlebirds appear to need
fairly dense and tall thickets that, at least at Nadgee, take some years to
develop after fire. Perhaps, what bristlebirds really require is a fire mosaic
of relatively mature vegetation of different ages post-burning. These
questions remain to be answered, but I can be confident that repeated
fuel reduction or hazard reduction burning at relatively short intervals
(e.g., 4 to 6 years) can only change the Australian bush to the detriment
of a large part of its birdlife.
Harry F. Recher is Emeritus Professor of Environmental Management at
Edith Cowan University, Joondalup, Western Australia.
Top: The Richard’s Pipit is an early successional species in heathland,
arriving soon after fire reduces the height and density of the vegetation.
Photo by Dave Watts, Lochman Transparencies
Below Left: Dense, floristically diverse heaths provide ideal habitat for the
Eastern Bristlebird: plenty of cover and protection from predators, places to
nest and insect prey. The heath shrubs depend on fire to stimulate seed
release and germination. Photo by Graeme Chapman
Below: The same heath after fire. Bristlebirds are typically absent in such
areas immediately following a bushfire. As the vegetation cover returns so
do the Bristlebirds, provided there are neighbouring populations from
which the birds can recolonise. Photo by Graeme Chapman
26
Fire and Birds
Fire and Birds
TOWARDS ECOLOGICALLY SUSTAINABLE FIRE MANAGEMENT
Left: To breed, Glossy Black-Cockatoos
are dependent on tree holes in large,
old trees, alive or dead. Salvage
logging in the wake of a fire can
remove such trees from the landscape
for decades. Photo by Lynn Pedler
More research is vital; in the meantime, a key management objective should be to ensure that, on both small and
large scales, the same fire regime is not applied uniformly everywhere
by David Lindenmayer
Fire is a natural influence on the majority of Australian terrestrial
ecosystems. Yet, inappropriate fire regimes are second only to land
clearing as a threat to avifauna in Australia: they have contributed to
the extinction of two species and three subspecies of Australian birds
and threaten more than 50 others. In south-western Australia, approximately 45% of mammals and birds may be negatively influenced by
altered fire regimes. The impacts are not limited to vertebrates; inappropriate fire regimes threaten almost 20 plant species.
It is perhaps unsurprising that the impacts of fire on the vast
majority of Australian organisms remain poorly known. A major
and sustained research effort is required to better understand such a
key environmental process, particularly given the number of species
threatened by fire. Sustaining interest in the research effort is also
critical because where there is public concern about fires and fire
impacts it is usually short-lived following major fire years.
Identifying appropriate fire regimes
Attempts to control or prevent high-intensity fire attract considerable
public attention, and there are conflicts between interests, in part
because of ongoing expansion of human settlements into areas of
flammable and fire-prone native vegetation. The most appropriate fire
regime will depend on the objectives of management and the characteristics of the ecosystem. Management may involve both the suppression
of unwanted (wild)fires and ignition of prescribed fires. Objectives
will vary depending on the proximity of people and property and the
relative importance of conservation and water production. This is
complicated by the fact that few (if any) areas have just one economic or
ecological value.
Even if biodiversity conservation is the primary management
objective, differences between vegetation communities and variation
in the fire response of individual elements of the biota (plants and
animals) mean that there are no simple recipes. Given such complexity,
what can be done? One approach is to vary fire regimes between
and within landscapes, creating a range of conditions. Therefore, if
unsuitable habitats are created in one area, there will be other places
where a species can survive. This is termed risk-spreading. Large parts of
some landscapes such as the edges of National Parks bordering grazing
properties or urban developments might be subject to frequent lowintensity prescribed burns (as is now done by many organisations such
as the New South Wales Department of Environment and Conservation
and Parks Victoria). Others more remote from human infrastructure
may be burned less frequently and/or less of the area may be burned. Yet
others may remain unburned by prescribed fire (although wildfire may
still occur).
Two of many key variables influencing the impacts of prescribed fires
on biota are fire frequency and the proportion of a landscape burned (i.e.
fire extent). Unfortunately, very little is known about how frequent and
extensive prescribed fire must be to reduce the risk of wildfire, or about
the impacts on biodiversity of varying prescribed fire frequencies and
sizes. Such uncertainty highlights the need to record actions and then
monitor the response of plants and animals. Observations would be fed
back to land managers so that practices could be improved. This is called
adaptive management or learning by experiment and monitoring. Despite
its potential to improve biodiversity outcomes, it has rarely been used by
resource management agencies.
The prescribed burning example outlined above is highly
simplified. Landscapes are not homogenous—some areas such as the wet
gullies in valley floors are less likely to burn than other places. Patchy
burns of varying intensity are not easy in practice. The timing and
intensity of burning add further complexity. Such variation may provide
opportunities for ‘natural ecological experiments’ and a feedback process
to inform management.
Perhaps the best-informed perspective on fire management comes
from Malcolm Gill (1999, p. 47), who believes that the management of
fire for biodiversity conservation should:
‘... aim at achieving suitable proportions of landscape with a variety of
times-since-fire stages within appropriate intensity levels at appropriate
times of the year and within appropriate frequency range.’
This is an important objective, but it is not easy to achieve because
using fire as a management tool is influenced by many practical
constraints, including (in many cases) a limited time window during
which prescribed fires can be applied safely, and limited financial and
human resources for fire management. Where prescribed burning is
difficult or may have negative impacts on some biota, it may be better
to use other forms of management, such as grazing. In the case of the
threatened Eastern Bristlebird—a species whose habitat requirements
encompass multi-layered heathland vegetation—fire exclusion is the most
appropriate conservation strategy. However, this may not be possible in
some of the areas that support some of the last major populations of the
species (such as the Jervis Bay area and Booderee National Park on the
south coast of New South Wales) because of nearby human infrastructure.
Right: Fire in Moreton Island National
Park. Photo by Raoul Slater
Given the need for fuel management in these areas, strategic slashing
of vegetation has been recommended. In other cases, where prescribed
burning is essential, the direction of fire fronts should be planned to
provide escape routes for animals, given the limited movement ability
of the Eastern Bristlebird. Many of these recommendations have been
embraced as part of the management of Eastern Bristlebird habitat, such
as that in Booderee National Park. Nevertheless, in late 2003, a major
wildfire burned much of the complex multi-layered heathland habitat
occupied by the Eastern Bristlebird in Booderee National Park.
David Lindenmayer is a Research Professor at the Centre for Resource and
Environmental Studies, The Australian National University, Canberra,
and is coordinating a major large-scale ‘natural fire experiment’ at Booderee
National Park.
Acknowledgments
Much of this paper is drawn from the textbook Practical Conservation
Biology (2005, CSIRO Publishing), co-written with Professor Mark
Burgman, and many of the insights come from the Booderee National
Park project, which is a collaboration between The Australian National
University, Department of Environment and Heritage, Wreck
Bay Aboriginal Community, Canberra Ornithologists Group, and
Earthwatch Institute. Chris MacGregor, Damian Michael, Mason
Crane, Martin Fortescue, Nick Dexter, Matt Hudson and Scott
Surridge play pivotal roles in that ongoing project.
Post-fire vegetation management: the case against
‘salvage’ logging
Many ecologists view natural disturbances such as wildfires as key ecosystem
processes rather than ecological disasters that require human repair—which
is the way they are typically portrayed by the media. Contrary to public
perception and media reports, major disturbances can sometimes aid ecosystem
restoration by recreating some of the structural complexity and landscape
heterogeneity lost through previous management such as intensive forestry.
For example, major wildfires can generate significant volumes of dead trees
and logs that provide important habitat for birds and other organisms, but are
depleted by traditional forestry practices.
In some forests, salvage logging operations following fires are used
to harvest fire-damaged timber and pulpwood. These activities can have
significant negative impacts on ecosystems and undermine many of the
ecological benefits of major disturbances. Ecological impacts of salvage
operations can manifest in at least two key ways.
First, there may be major impacts on ecosystem processes. Impaired
quality and quantity of water supplied from salvage-logged catchments has
been documented in several regions. Second, the removal of large quantities
of biological legacies (such as fire-damaged trees) threatens the persistence of
some taxa. For example, extensive salvage harvesting after the 1939 wildfires
in southern Australia produced a long-lasting shift in cavity-tree resources that
was manifested at patch, landscape and regional scales. In that case, salvage
harvesting of large, hollow, burned living and dead trees continued for two
decades in the wet forests of the Central Highlands of Victoria, and removed
critical habitat for cavity-using vertebrates.
The impacts of salvage harvesting indicate a need for large areas, such as
national parks, to be exempt from such activities. In addition, wherever salvage
harvesting continues, carefully formulated policies are needed to ensure the
retention of dead trees, live trees, logs and islands of undisturbed or partially
disturbed vegetation.
27
Fire and Birds
Fire and Birds
FIRE AND FEATHERS: MANAGING FIRE-SENSITIVE BIRDS
ON A FLAMMABLE CONTINENT
We are only just beginning to understand and manage fire in all its complexity
by Elizabeth Tasker and Jack Baker
Fire is vital to the functioning of Australian ecosystems, and is likely
to affect the birds in most habitats. Understanding how fire may affect
birds will become increasingly important for conservation in a landscape
where habitats are more and more fragmented and degraded, and where
climate change is likely to result in very different patterns of fire in the
future. Australian researchers have been at the forefront of a revolution
in our understanding of fire ecology over the past twenty years. Applying
this improved knowledge to management of birds and their habitats is
likely to be critical to the survival of several threatened species, such as
the Eastern Bristlebird and Ground Parrot, as well as to the viability of
many more common species across the landscape.
Fire ecology
Until quite recently, most research on the impacts of fire on birds in
Australian environments has been on single fire events. Because the
response of animals and plants varies according to an almost infinite
combination of the factors, and to a large degree every fire is unique, it
has been difficult to build generalisations or rules that can be used to
guide management decisions.
An alternative way of looking at fire ecology is proving more useful.
It entails looking at the cumulative impacts of a series of fires, or, if a
particular fire is of interest, examining it in context, especially in light
of the fires that have gone before it. Documenting and understanding
these ‘fire regimes’, as they are termed, was stimulated by seminal work by
Malcolm Gill of CSIRO in the mid-1970s. Over the last two decades, it
has led to major improvements in our understanding.
The key elements of a fire regime consist of fire frequency, intensity and
season. Particular regimes may lead to population declines, and even local
extinctions. The basic difference between considering a fire in isolation and
considering it in context is illustrated by the following example. In many
areas of north-eastern New South Wales and south-eastern Queensland
cattle grazing in eucalypt forests is accompanied by very frequent burning
in winter by graziers, to promote an open grassy understorey. Any single fire
taken on its own is so low in intensity that it has little noticeable impact on
the native animals and plants present. However, when taken together, over
many decades, these fires cause major changes in the vegetation structure
and composition, in turn affecting which animals can survive in these areas.
It is now widely recognised that plants and animals are not ‘adapted
to fire’ per se, though they often do possesses characteristics or attributes
that enable them to survive particular fire regimes, or to avoid or escape
a given fire. This is clearly illustrated by the Heath Banksia, which is
dependent on fire to release seed from its woody cones, but is eliminated
from sites burned too often. For nectar-feeding birds, the loss of Heath
Banksias and other proteaceous shrubs could be critical.
Fire regimes and fire-sensitive birds
Some birds are recognised as particularly vulnerable to the impacts of a
fire event or fire regime (as are some other animals and plants). Typically,
fire-sensitive birds are ground-dwelling, cover-dependent, poor fliers, poor
dispersers, low in fecundity, or some combination of these. In other words,
they lack the characteristics possessed by many other birds that would allow
them to avoid, or more rapidly recover from, the effects of fire. In line with
our current understanding of fire regimes a more up-to-date and informative
terminology could be to specify whether a species is ‘fire-interval sensitive’,
‘fire-intensity sensitive’, and so on. For example, birds that are fire-intensity
sensitive or live in a limited or isolated area can be drastically affected by even
a single intense fire, which may result in the loss of an entire population.
There are relatively few studies of birds in which the impacts of fire
regimes have been considered, although these are becoming more common.
An emerging theme is that frequently burnt habitats may not provide the
food, cover or nesting resources required by some species. For example,
Woinarski and Recher noted that the Rufous-crowned Emu-wren and the
Striated Grasswren are restricted to long-unburnt hummock grasslands. In
forest environments, frequent fire has been shown to disadvantage species
that favour dense, shrubby understorey (such as the Golden Whistler),
or thick leaf litter (such as the Pilotbird), while birds that favour open
understorey may increase. Grey-crowned Babblers are less abundant in
frequently-burnt roadside verges compared to unburnt verges, and this has
been attributed to the lack of shrub cover in the former situation.
One of the most comprehensive studies on the effect of a series of fires
on birds was carried out by Rowley, Brooker and Russell on the Splendid
Fairy-wren. They found that repeated fires reduced the density, productivity
and survival of individuals for some years after each fire, and populations
thus became dependent on recruitment from outside the area.
Flexible fire management
The challenge with fire management is to maximise the conservation of
biodiversity while protecting human life and property and maintaining
natural resource values (e.g. timber) where appropriate. Until recently,
fire management has generally been ‘prescriptive’, that is applying a
predetermined action according to a fixed set of rules. For example,
prescribing the burning of an area of habitat for a threatened species every
15 years, based on the observation that elsewhere
the maximum numbers of this bird were found
in vegetation of this (post-fire) age. From the
point of view of local operational management
this may be easy, but there are substantial
drawbacks. Prescriptions that are applicable to
a particular species in one region at a particular
point in time are unlikely to be ‘portable’ to
another time, place, species or vegetation type.
Fixed prescriptions may have unexpected
outcomes for the target species by impacting on
its habitat in the longer term, and may negatively
affect other species. Prescriptive management
also ignores the unpredictability of fire, and the
need for population monitoring, arguably one
of the simplest and most important conservation
management tools available. Often we simply
don’t know enough about a species to predict
how it will respond.
In contrast, ‘adaptive’ management involves
constant review and updating of plans and the
actions that are proposed, based on how the fire
history of an area has changed since the last plan
(e.g. a wildfire may have come through the site)
and on research or monitoring showing how the
target species are responding to management
actions. Adaptive management takes into
account that there is insufficient knowledge of
how most of the bird species we want to manage
will respond to fire, and the need to manage for
uncertainty. It also aims to minimise the risk
of extinction, rather than maximise densities, a
subtle distinction which allows other factors such
as habitat maintenance to play a greater role in
planning.
The following case studies demonstrate
the value of population monitoring and
adaptive management for two fire-sensitive
bird species, the Ground Parrot and the
Eastern Bristlebird, in Barren Ground
Nature Reserve, on the escarpment behind
Wollongong in New South Wales.
The Ground Parrot
The Ground Parrot occurs in coastal southeastern and south-western Australia and is
listed as threatened nationally as a result
of significant habitat loss and increasingly
fragmented populations. It survives in its
fire-prone habitats by being a strong flier and
having a high reproductive output, but is
fire-sensitive because it is ground-dwelling,
cover-dependent and probably a poor
disperser. In general, it requires habitat with
a minimum fire-age of one year for feeding
and three years for breeding, and will go
locally extinct if there is insufficient time for
a breeding population to re-establish between
fires. Despite the Ground Parrot being much
studied, there remains considerable debate
about its response to fire and which fire
management is appropriate. It continues
to be regarded by some as a species which
is dependent on frequent fire to maintain
suitable habitat, even though evidence now
indicates that this is not necessarily so.
At Barren Grounds approximately 280
ha burned in a bushfire in January 1983. A
study was immediately set up by the bird
observatory wardens, Richard and Pat Jordan,
to census Ground Parrots each year in an 80
ha plot (Redbank Gully) and see how their
numbers changed following the fire. Since
then, each year in early spring the parrots
are counted as they are flushed by a line of
‘beaters’ walking through the heath.
In accordance with various predictive
models current at the time, the expectation
was that the population would initially rebuild,
then gradually decrease over time to zero at
about 12 years post-fire. The models assumed
that Ground Parrot habitat needed to be burnt
quite often in order to keep it suitable. Various
proponents had prescribed burning every 6–8,
8–10 or 12 years at Barren Grounds. However,
before the prescribed burns could be carried
out a wildfire burned much of the habitat,
and they were subsequently abandoned.
Monitoring since 1983 has demonstrated that
the parrots have persisted for almost twice as
long as was predicted, in fact they reached
maximum density 13 years after the fire
(see graph below).
29
Clockwise from far left:
Frequent low intensity fires can cause major changes to
eucalypt forests by eliminating the shrub understorey
and favouring an open, grassy groundcover unsuited
to many of the original birds. This site has been grazed
by cattle and intentionally burnt at regular intervals for
about a century. Despite this hazard reduction burning,
and low fuel loads, the site still burns intensely in
extreme weather. Photos by Pavel German
A backburn is commonly used to finish off a prescribed
burn, however, if fire-sensitive, endangered and
relatively immobile animals, such as Eastern Bristlebirds,
are present they may be trapped between the two fires.
In such areas a mechanical method of fuel removal may
be preferable. Photo by Pavel German
A low intensity fire, such as this lit by a grazier in
north-eastern New South Wales, may appear relatively
harmless. However, such fires may not be hot enough
to stimulate seed release and/or germination in many
native plant species; if repeated at frequent intervals
they cause major changes in the species composition
and structure of the vegetation. Photo by Pavel German
The Ground Parrot—critically endangered in the west
and vulnerable in the east—lives in heathland, a habitat
that has been fragmented, making it more vulnerable
to fire. Photo by Dave Watts
Since 1998, 15 years post-fire, the
density of parrots recorded in the census
has fluctuated considerably. To complicate
the pattern, recent listening (aural) surveys
have detected more parrots than the annual
censuses (see graph below). Nevertheless,
the overall figures suggest a decline in the
population which, taken in conjunction
with what is known about the ecology of
the vegetation on which the parrot depends,
indicates that a strategic and carefully planned
fire may now be appropriate.
Ground Parrot densities at Redbank Gully, an 80 ha site, at Barren Grounds following a bushfire in January 1983.
Ground Parrot densities at Redbank Gully, Barren Grounds
5
Parrots per 10 ha
28
4
3
Parrot density
from flushing census
2
Parrot density
from calling survey
1
0
0
5
10
15
Years after the 1983 fire
20
25
30
Fire and Birds
Fire and Birds
FURTHER READING
Fire is the main threat to the few, isolated, remaining populations of the
endangered Eastern Bristlebird (southern subspecies), considered to be a firesensitive bird but with habitat that is dependent on fire for its maintenance.
Photo by Graeme Chapman
The Eastern Bristlebird
The Eastern Bristlebird is a classic fire-sensitive bird and is found only in
three widely separated populations in eastern New South Wales. The two
largest populations are in the central part of the State—one in the vicinity
of Jervis Bay and the other centred on Barren Grounds.
In long-unburnt habitat at Barren Grounds, Jervis Bay and Nadgee,
the birds have bred and persisted at high densities for many years. In the
three decades to 2002, these habitats were characterised by long intervals
between fires, and by fires that burned relatively small areas. Surveys in
five different areas of Barren Grounds Reserve, each of which was burnt
at a different time, show a clear pattern for bristlebird densities to increase
from almost zero immediately following a fire to stable levels of around
two birds per 5 ha approximately ten years post fire (see graph below). In
December 2003 a wildfire burned more than half of the bristlebird habitat
at Jervis Bay. David Bain from the University of Wollongong has been
monitoring the recovery of the population, and early results show that
intensely burnt areas have suffered the greatest loss of bristlebirds, while
in unburnt areas their density has increased. He has also found that the
proximity of refuges to burnt areas has been important for survival.
Paradoxically, although the Eastern Bristlebird (like the Ground
Parrot) is fire-sensitive, it depends on habitats that are very flammable and
fire-dependent, and have many plant species that need fire to stimulate
flowering, seed production, seed release and/or germination. So, it is
important to determine the range of fire regimes under which not only
bristlebirds do well, but also the vegetation on which they depend. Recent
guidelines produced by the Department of Environment and Conservation (NSW) suggest maximum and minimum intervals between fires
that will maintain the diversity within different broad vegetation types.
Such intervals could be used as a starting point to guide fire management
for this species, modified by feedback from monitoring of bird numbers
in the areas being managed. We are currently undertaking a review and
synthesis of this information for the bristlebird.
Given its rarity and highly fragmented habitat—often there will not
be adjacent populations from which birds can recolonise following a major
fire—management must take into account the extremely fire-sensitive
nature of the Eastern Bristlebird. In the event of unplanned fires, protection
of areas with important concentrations of bristlebirds, as well as fire refuges,
should be a priority. In fuel-reduction zones around human property that
are in bristlebird habitat, strategic slashing and/or trittering should be used
instead of frequent burning to reduce fuel loads. If prescribed burns do need
to be carried out, ameliorative measures can be adopted. The usual practice
to finish off a prescribed burn is to carry out backburning. However, such a
practice is likely to block the escape route for Eastern Bristlebirds and cause
high mortality. As an alternative, where it is intended to use a burnt area to
stop a prescribed burn, this area could be burnt several days previously, or
mechanical fuel reduction methods used instead.
Eastern Bristlebird densities between 1992 and 1998 at five areas in Barren
Ground Nature Reserve where the most recent fires had occurred, in 1991, 1988,
1983, 1980 and 1979.
Density of birds per 5 ha
3
Practical application of adaptive management for firesensitive birds
To list hard and fast rules for the management of fire-sensitive birds
would be contrary to the principles of adaptive management. Conservation managers often need to make decisions about fire without having
detailed information about the fire-sensitive populations they are
managing, and are often faced with small and/or declining populations
in fragmented habitats. In such cases, the following generalisations may
be useful guidelines:
1) The extent of a single fire should be small relative to the total
amount of habitat;
2) Fire should not be prescribed at a fixed interval in any given area;
3) The need to apply fire to create appropriate habitat for a species
should be reviewed in terms of population monitoring in the area,
and in terms of long-term maintenance of the habitat;
4) Plan to protect refuge areas for fire-sensitive species.
We believe there are three key tools to implement successful fire
management of fire-sensitive birds:
• Monitoring of the population of concern;
• Documentation of the fire history of the site;
• Application of what is known from the field of vegetation ecology
to manage the food resources and habitat of the birds over time.
Monitoring may involve considerable resources—the Ground
Parrot study involved approximately 50 volunteers each year for 3 hours,
repeated over 21 years, giving a total of more than 3,000 ‘volunteer
hours’ (not including travel time or costs). However, less labourintensive alternatives may be available. The Ground Parrot has such a
distinctive call that novices can be trained to recognise its call in less
than an hour, making aural surveys feasible.
Fires are now mapped for all national parks and reserves in New
South Wales, using a Geographical Information System, and this easily
allows the fire history of an area to be analysed and regularly updated,
and for sophisticated fire planning to be carried out.
Finally, use of guidelines, such as those produced by scientists
from the Department of Environment and Conservation on the fire
frequency thresholds of different vegetation types, coupled with what
is known of the habitat requirements of species, such as the Ground
Parrot and Eastern Bristlebird, will allow greatly improved management,
understanding, and—ultimately—conservation of fire-sensitive birds
across the landscape.
Liz Tasker and Jack Baker work in the Biodiversity Conservation Science
Section of the Department of Environment and Conservation, New South
Wales. Liz is with the Fire Ecology Unit and has a PhD in fire ecology from
the University of Sydney. Jack is a former warden of Barren Grounds Bird
Observatory, where he first met Eastern Bristlebirds and Ground Parrots.
2
1991
1988
1983
1980
1979
1
Acknowledgments
0
0
5
10
Years after fire
15
20
31
Many volunteers assisted with the annual Ground Parrot census.
The NSW National Parks and Wildlife Service and Barren Grounds
Bird Observatory facilitated its organisation.
Ashby E 1924 Notes on extinct or rare Australian birds, with suggestions as to some
of the causes of their disappearance. Part 1. Emu 23: 178–183.
Baker J & Whelan RJ 1994 Ground Parrots and fire at Barren Grounds, New South
Wales: A long-term study and an assessment of management implications. Emu
94: 300–304.
Baker J 2000 The Eastern Bristlebird: cover-dependent and fire-sensitive. Emu 100:
286–298.
Baker R 2003 Yanyuwa classical burning regimes, Indigenous science and crosscultural communication. Pp. 198–204 in G. Cary, D. Lindenmayer & S. Dovers
(eds) Australia Burning: Fire Ecology, Policy and Management Issues. CSIRO
Publishing; Melbourne.
Bowman DMJS 1998 The impact of Aboriginal landscape burning on the Australian
biota. New Phytologist 140: 373–381.
Bradstock RA, Williams JE & Gill AM (eds) 2002 Flammable Australia. The Fire
Regimes and Biodiversity of a Continent. Cambridge University Press; Melbourne.
Braithwaite RW & Estbergs J 1987 Fire-birds of the Top End. Australian Natural
History 22: 299–302.
Brooker LC & Brooker MG 1994 A model for the effects of fire and fragmentation
on the population viability of the Splendid Fairy-Wren. Pacific Conservation Biology
1: 344–358.
Burbidge AH 2003 Birds and fire in the Mediterranean climate of south-west
Western Australia. Pp. 321–347 in I Abbott and N Burrows (eds) Fire in Ecosystems
of South-west Western Australia: Impacts and Management. Backhuys; Leiden.
Burrows N 2003 Using and sharing Indigenous knowledge. Pp. 205–218 in G
Cary, D Lindenmayer & S Dovers (eds) Australia Burning: Fire Ecology, Policy and
Management Issues. CSIRO Publishing; Melbourne.
Cary G, Lindenmayer DB & Dovers S (eds) 2003 Australia Burning. Fire Ecology,
Policy and Management Issues. CSIRO Publishing; Melbourne.
Crowley GM & Garnett ST 2000 Changing fire management in the pastoral lands
of Cape York Peninsula of northeast Australia, 1623 to 1996. Australian Geographic
Studies 3: 10–26.
Dovers, S 2003 Uncertainty, environment and fire: research and policy issues. Pp.
41–50 in G Cary, D Lindenmayer & S Dovers (eds) Australia Burning: Fire
Ecology, Policy and Management Issues. CSIRO Publishing; Melbourne.
Dyer R, Jacklyn P, Partridge I, Russell-Smith J & Williams R (eds) 2001 Savanna
Burning: Understanding and Using Fire in Northern Australia. Tropical Savannas
CRC; Darwin.
Ecological Society of Australia. ESA Position Statement: The Use of Fire in Ecosystem
Management. www.ecolsoc.org.au
Fraser F, Lawson V, Morrison S, Christopherson P, McGreggor S & Rawlinson M
2003 Fire management experiment for the declining Partridge Pigeon, Kakadu
National Park. Ecological Management and Restoration 4: 93–101.
Garnett S & Crowley G 2002 The Action Plan for Australian Birds 2000.
Environment Australia, Canberra & Birds Australia, Melbourne.
Gill AM 1999 Biodiversity and bushfires: an Australia-wide perspective on plantspecies changes after a fire event. Pp. 9–53 in AM Gill, JCZ Woinarski & A York
(eds) Australia’s Biodiversity—Responses to Fire. Environment Australia Biodiversity
Technical Paper 1.
Harrington G & Sanderson K 1994 Recent contraction of wet sclerophyll forest
boundary in the wet tropics of Queensland due to invasion by rainforest. Pacific
Conservation Biology 1: 319–327.
Haynes CD 1985 The pattern of munwag: Traditional Aboriginal fire regimes in
north-central Arnhemland. Proceedings of the Ecological Society of Australia 13:
203–214.
Howitt W 1856 Black Thursday. Household Words, London, 10 May 1856, pp.
388–395.
Jones R 1969 Fire-stick farming. Australian Natural History 16: 224–228.
Keith D, Williams J & Woinarski J 2002 Fire management and biodiversity conservation: key approaches and principles. Pp. 401–425 in R Bradstock, J Williams
& AM Gill (eds) Flammable Australia. The Fire Regimes and Biodiversity of a
Continent. Cambridge University Press; Cambridge.
Kenny B, Sutherland L, Tasker E & Bradstock R 2004 Guidelines for Ecologically
Sustainable Fire Management. NSW National Parks and Wildlife Service;
Hurstville.
Lindenmayer DB, Foster D, Franklin JF, Hunter M, Noss R, Schiemegelow F &
Perry D 2004 Salvage harvesting after natural disturbance. Science 303: 1303.
Loyn R 1997 Effects of wildfire on birds in far eastern Victoria. Pacific Conservation
Biology 3: 221–234.
Morton A, Presland G & Gibson M (eds) 2004 Fire Symposium Special Issue 2004.
Victorian Naturalist 121 (3).
O’Neill AL, Head LM & Marthick JK 1993 Integrating remote sensing and spatial
analysis techniques to compare Aboriginal and pastoral fire patterns in the East
Kimberley, Australia. Applied Geography 13: 67–85.
Pedler L 1997 Fire and Striated Grasswrens in Uluru-Kata Tjuta National Park. Pp.
91–93 in LG Woodcock (ed) Proceedings of ‘Back to the Future’ Natural Resources
Research Workshop, Uluru-Kata Tjuta National Park, 2729 August 1997. AGSO
Record 1997/55. AGSO; Canberra.
Press AJ 1988 Comparisons of the extent of fire in different land management
systems in the top end of the Northern Territory. Proceedings of the Ecological
Society of Australia 15: 167–175.
Reid J 1997 Uluru fauna surveys: baseline, monitoring, long-term research and
new lines of inquiry. Pp. 94–109 in LG Woodcock (ed) Proceedings of ‘Back to
the Future’ Natural Resources Research Workshop, Uluru-Kata Tjuta National Park,
27–29 August 1997. AGSO Record 1997/55. AGSO; Canberra.
Richards SA, Possingham HP & Tizard J 1999 Optimal fire management for
maintaining community diversity. Ecological Applications 9: 880–892
Rossiter NA, Setterfield SA, Douglas MM & Hutley LB 2003 Testing the grass-fire
cycle: alien grass invasion in the tropical savannas of northern Australia. Diversity
and Distributions 9: 169–176.
Russell EM & Rowley I 1993 Demography of the cooperatively breeding Splendid
Fairy-wren, Malurus splendens (Maluridae). Australian Journal of Zoology 41:
475–505.
Russell-Smith J 2002 Pre-contact Aboriginal and contemporary fire regimes of
the savanna landscapes of northern Australia: patterns, changes and ecological
consequences. In J. Russell-Smith, R. Craig, A.M. Gill, R. Smith & J. Williams
(eds) Australian Fire Regimes: Contemporary Patterns and Changes Since European
Settlement. www.deh.gov.au/soe/techpapers/fire/
Russell-Smith J, Craig R, Gill AM, Smith R & Williams J 2002 Australian fire
regimes: contemporary patterns (April 1998–March 2000) and changes since
European settlement. Australia: State of the Environment Second Technical Paper
Series (Biodiversity), Series 2. Department of the Environment and Heritage http:
www.deh.gov.au/soe/techpapers/fire/index.html
Russell-Smith J, Whitehead PJ, Cook GD & Hoare JL 2003 Response of Eucalyptusdominated savanna to frequent fires: lessons from Munmarlary, 1973–1996.
Ecological Monographs 73: 349–375.
SEQFBC 2002 Fire and Biodiversity Monitoring Manual: A Guide for Local Field
Officers and Land Managers. Southeast Queensland Fire and Biodiversity
Consortium & Greening Australia Queensland.
Sharp BR & Bowman DMJS 2004 Net woody vegetation increase confined to
seasonally inundated lowlands in an Australian tropical savanna, Victoria River
District, Northern Territory. Austral Ecology 29: 667–683.
Smith G 2004 Distribution and Habitat Requirements of Mallee Emu-wren (Stipiturus
mallee) within the 1990 Fire Scar of Ngarkat Conservation Park: Towards a Habitat
Suitability Model. B.Sc. (Hons) thesis, University of Adelaide.
Smith LH 1988 Life of the Lyrebird. Heinemann, Richmond.
Ward MJ & Paton DC 2004 Responses to fire of Slender-billed Thornbills Acanthiza
iredalei hedleyi in Ngarkat Conservation Park, South Australia. 1. Densities, group
sizes, distribution and management issues. Emu 104: 157–168.
Whelan RJ 2002 Managing fire regimes for conservation and property: an Australian
response. Conservation Biology 16: 1659–1661.
Whitehead PJ, Bowman DMJS, Preece N, Fraser F & Cooke P 2003 Customary
use of fire by Indigenous peoples in northern Australia: its contemporary role in
savanna management. International Journal of Wildland Fire 12: 415–425.
Williams RJ, Woinarski JCZ & Andersen AN 2003 Fire experiments in northern
Australia: lessons for ecology, management and biodiversity conservation. International Journal of Wildland Fire 12: 391–402.
Woinarski J & Recher H 1997 Impact and Response: a review of the effects of fire on
the Australian avifauna. Pacific Conservation Biology 3: 183–205.
Woinarski JCZ 1999 Fire and Australian birds: a review. Pp. 55–111 in AM Gill,
JCZ Woinarski & A York (eds) Australia’s Biodiversity—Responses to Fire, Plants,
Birds and Invertebrates. Biodiversity Technical Paper No. 1. Environment Australia;
Canberra.
Woinarski JCZ, Brock C, Fisher A, Milne D & Oliver B 1999 Response of birds and
reptiles to fire regimes on pastoral land in the Victoria River District, Northern
Territory. The Rangeland Journal 21: 24–38.
Woinarski JCZ, Risler J & Kean L 2004 The response of vegetation and vertebrate
fauna to 23 years of fire exclusion in a tropical Eucalyptus open forest, Northern
Territory, Australia. Austral Ecology 29: 156–176.
32
Fire and Birds
Kites and firefighters attracted to a Northern Territory grass fire. Photo by Rob Gray
CONTENTS
Acknowledgements
Key Points
2
Overview
3
Living with Fire—
Birds in Northern Australia
by John Woinarski
7
Mornington: A Model for
Fire Management in Australia’s
Tropical Savannas
by Steve Murphy, Sarah Legge and
Nat Raisbeck-Brown
10
Managing Perceptions
by Gabriel Crowley and Peter Thompson
12
Birds Australia thanks the R.E. Ross Trust for their generosity in funding this important publication.
We are also particularly grateful to the following individuals who made time in their busy schedules to
write the narratives that are the foundation of this report: Grant Allan, Jack Baker, Allan Burbidge,
Sarah Comer, Gabriel Crowley, Alan Danks, Jody Gates, Sarah Legge, David Lindenmayer,
Steve Murphy, David Paton, Nat Raisbeck-Brown, Harry Recher, Julian Reid, Daniel Rogers, Liz Tasker,
Peter Thompson, Matthew Ward and John Woinarski. Mike Clarke, Peter Menkhorst, Jim McGuire
and Bruce Quin contributed boxed text. Sarah Comer, Alan Danks, Leon Keasey and Greg Holland
(www.birdphotos.com.au), Jonathan Munro (www.wildwatch.com.au), Steve Murphy, David Paton,
Lynn Pedler, Will Philippiadis, Don Royal and Raoul Slater kindly allowed free use of their photographs.
Lastly, we are grateful to Allan Burbidge and the other authors of the various accounts, Peter
Menkhorst, Stephen Garnett, Barry Baker, who reviewed all or part of the report, Sophie Knezic for her
capable copy editing and guidance of production, and Andrea Williamson for her keen attention to design.
Birds Australia welcomes new members
and volunteers
Fire and its Management at
Newhaven Reserve:
Taking a Regional Approach
by Grant Allan
15
Bushfire Management in the
Northern Territory
by Grant Allan
Dedicated to the study, conservation and enjoyment
of native birds and their habitat
16
Fire and Birds in Spinifex Grasslands
by Julian Reid
17
Since 1901 Birds Australia (Royal Australasian
Ornithologists Union) has worked for the
conservation of Australasia’s birds and their
habitats. Its reserves and observatories protect bird
habitat and offer unique research and birdwatching
opportunities. As an independent, not-for-profit
organisation, Birds Australia relies on the financial
support of companies, trusts and foundations,
and private individuals, and the involvement of
thousands of volunteers in its projects.
18
Birds and Fire in the Mallee
Heathlands of Ngarkat
by David C. Paton, Daniel J. Rogers,
Matthew J. Ward and Jody A. Gates
21
Birds, Fire and Nadgee
by Harry F. Recher
24
Towards Ecologically Sustainable
Fire Management
by David Lindenmayer
The Threatened Bird Network links volunteers
with recovery efforts for more than 25 threatened
species. Contact Chris Tzaros (03 9882 2622;
[email protected])
JOIN TODAY!
By joining Birds Australia,
you help Australia’s wild birds and their habitats.
Title
First Name
Surname
Address
415 Riversdale Road,
Hawthorn East, Victoria 3123
Tel: (03) 9882 2622; Fax; (03) 9882 2677;
Email: [email protected]
Web site: www.birdsaustralia.com.au
Phone ((BH)
Email
Please accept my enclosed cheque for $68
$108 (family*)
26
Fire and Feathers: Managing
Fire-Sensitive Birds on a
Flammable Continent
by Elizabeth Tasker and Jack Baker
28
Further Reading
31
$50 (concession)
or $87 (family concession)
or debit my Bankcard
Visa
payable to ‘Birds Australia’
Mastercard
Funding for this report was generously provided
by The R.E. Ross Trust.
Front cover: One of 14 Black Falcons attracted to
a stubble fire at Riverton, South Australia, along
with two Australian Hobbies, four Brown Falcons
and a few Black Kites, within 20 minutes of
ignition. Photo by Nicholas Birks
Expiry Date
Layout and design: P.A.G.E. Pty Ltd
Reproduction and printing:
Courtney Colour Graphics
The views expressed in this supplement are not
necessarily those of Birds Australia or its Council.
Date
/
✂
Threatened Birds and Wildfire in
South-West Western Australia
by Allan Burbidge, Sarah Comer
and Alan Danks
The Atlas of Australian Birds is a long-term,
nationwide, volunteer-based bird-monitoring
project that welcomes new Atlasers.
Contact Andrew Silcocks (03 9882 2622;
[email protected])
Signature
/
/
SCIA.1
Birds Australia, 415 Riversdale Rd, Hawthorn East, Vic. 3123
Tel: (03) 9882 2622; Fax: (03) 9882 2677;
Email: [email protected]
*Family covers two members at the same address. Membership runs for 12 months.
Please enquire for rates for overseas members, and subscriptions to Emu.
Copyright: Reproduction in whole or part may only occur with the written permission of the Editor.
ISSN: 1036-7810