Fire, landscape pattern
and biodiversity in the
Otway Ranges
In 2008, DEPI (then DSE) introduced a landscape-scale mosaic burning program in Victoria with the intention of creating and maintaining spatially heterogeneous environments, which theory suggests should have a positive impact on biodiversity (the pyrodiversity/biodiversity hypothesis - see over). This research project, commissioned in 2009, has been investigating how the nature of these mosaics, which are a function of vegetation type and post-fire growth stage, influences biodiversity. In particular, the project was designed to test theoretical relationships between plant and animal diversity and landscape spatial pattern, and assess how other processes (such as herbivory and predation) might interact with fire-mediated effects. Results from this study provide a framework for understanding fire and biodiversity impacts, which can then be integrated within the developing Risk Landscape planning framework, to better understand trade-offs through scenario modelling.
The pyrodiversity/biodiversity hypothesis
It is widely assumed that fire regimes that vary over space and time benefit biodiversity conservation, as ecological theory predicts positive associations between species diversity and environmental heterogeneity. Heterogeneous landscapes should support higher species diversity than homogeneous landscapes by providing a range of resources for multiple species with differing ecological requirements. Mosaic burning aims to break up large (forested) landscapes by providing a mix of fire (at appropriate frequencies, seasons, intensities and scales) to create a more extensive range of vegetation age-classes (i.e. increased landscape heterogeneity). It is assumed that mosaic burning will also help reduce the size, severity and impact of large-scale fire events. (For more information see Di Stefano & York 2012.)
Variation in geology, soils, topography and rainfall generates spatial variation in vegetation communities across the landscape. This variation occurs at a range of spatial scales, with heterogeneity further influenced by fire, both planned and unplanned, which over time creates a shifting mosaic of vegetation age classes.
Lorne
Anglesea Damp Scrub
Forby Heathland Tall Mixed Treed Swamp Broad Vegetation Communities Damp Scrub Foothills Forest Forby Forest Heathland Tall Mixed Forest
Wet Forest
Ecological Vegetation Divisions (EVDs)
+ Fire History
EVD & Growth Stage
Damp Scrub Adol
Damp Scrub Juv
Damp Scrub Mature Damp Scrub Stasis Forby Adol Forby Juv Forby Mature Heathland Adol Heathland Juv Heathland Mature Tall Mixed Adol Tall Mixed Juv Tall Mixed Mature Tall Mixed Stasis Treed Swamp Mature
Landscape-scale
Loc
al
(patc
h)
sc
ale
Our study
T
he study was conducted in a 59,000-ha section of the Otway Ranges (the Great Otway National Park and Forest Park). The Otways are particularly suited to testing the environmental heterogeneity hypothesis as the landscape exhibits varying levels of heterogeneity with respect to two important design variables: vegetation community and time since fire. The project was designed to be implemented in three phases (see Di Stefano, Christie & York 2011).Phase 1: Mosaic study
In 2010, thirty-six 100-ha plots (land mosaics) were
established across a range of vegetation communities; with each mosaic varying in its spatial pattern (and complexity) as a function of vegetation type and growth stage. Five survey sites were established within each mosaic, each sampling a representative vegetation type/growth stage combination.
Biodiversity data were collected through systematic survey in 2010-11 to enable a better understanding of relationships between landscape pattern and species diversity, abundance, behaviour and related processes. We evaluated the separate effects of (i) fire-mediated spatial pattern, which can be
influenced by land managers, and (ii) spatial pattern in vegetation type, which remains relatively static. In this area vegetation types are characterised by distinct structural and floristic elements, so you would expect them to support different animal species. Within this framework, fire age-class (growth stage) reflects successional states, with each potentially supporting different assemblages. Spatial patterns in both these elements should therefore influence overall diversity at landscape scales. Fires in the Otway Ranges (including mosaic burning) over time will change existing landscape pattern; resampling of the land mosaics will allow further testing of the heterogeneity/diversity hypothesis.
Phase 2: Experimental burns
Based on an operational-scale planned burn, a smaller-scale manipulative experiment was established in 2010 (Henderson Creek catchment), using planned fire to better understand links between fire, diversity and associated processes. Using a before-after control-impact (BACI) design, with paired catchments, the project investigated the short-term impact of planned burning on landscape heterogeneity and associated biodiversity at the scale of a single burn block. Subsequently, a second experiment was established in 2012 (Breakfast Creek catchment), primarily to examine relationships between planned fire, predation (foxes and cats) and mammal populations.
Phase 3: Monitoring and evaluation
As planned burning activities (including mosaic burning) are undertaken in this landscape over time, changing the nature of the land mosaics, re-sampling of study sites will allow further testing of the heterogeneity/diversity hypothesis, and establishment of links between fire, habitat change and species distribution and abundance. As part of DEPI’s long-term monitoring program, resampling of land mosaics is planned for 2015/6.
How do heterogeneity in vegetation types and
post-fire age classes contribute to plant diversity at the
landscape scale?
Plant diversity
JANET COHN
H
eterogeneity in environmental factors and disturbances are important drivers of plant diversity within landscapes. Given that fire regimes are often manipulated to maintain plant diversity at the landscape scale, disentangling fire effects from environmental effects is not well researched, but nevertheless important at this scale. Along a gradient of decreasing productivity (rainfall and soil fertility) from moist forests to sclerophyllous heathlands, both prescribed burning and wildfires have created a mosaic of post-fire age classes in a range of vegetation types, providing an opportunity to investigate both sources of heterogeneity on plant diversity. Our survey units were mosaics (approx. 100 ha) of varying heterogeneity in the numbers of vegetation types, post-fire age classes and their combinations.We found that plant diversity increased as the number of vegetation types in mosaics increased and as mean annual rainfall increased from heathlands to moist forests (see below). There was also a weak (non-significant) positive relationship between plant
diversity and the number of post-fire age classes in mosaics. The effect of post-fire age-classes on plant diversity was much stronger, however, when each vegetation type was considered separately, with species composition more sensitive to post-fire age class in the moist forests at higher productivity compared with sclerophyllous heathlands at lower productivity. The effectiveness of mosaic burning in maintaining plant species diversity in environmentally diverse landscapes is therefore best assessed at the small scale, within vegetation types, where the influence of environmental factors is minimised.
The effect of age classes on plant diversity was much stronger when each vegetation type was considered separately, with species composition more sensitive to age classes in the moist forests.
“
”
Responses of plant diversity to vegetation type, age class and rainfall
Plan
t species tur
no
ver
Are germination cues in heath and forest different?
M
any plant species are reliant on regeneration from a soil-stored seed bank after fire in flammable systems. Germination of seeds in the soil can be triggered by a number of cues including smoke and temperature. Ground fuels (e.g. fine or woody litter, grasses, herbs) influence soil temperatures more than elevated fuels (e.g. shrubs). Characteristics of ground fuels, like type, amount, and moisture content determine theresidence time of burning, which in turn determines the soil temperature during fires. Coarse woody debris (CWD; >10 cm diameter) burns slowly compared with fine fuel (<6 mm diameter) and results in greater soil temperatures. Forests have greater amounts of CWD than heath because of the contributions from dense tall trees. So, we hypothesised that high soil temperatures may be more important for germination in forested than shrubby systems. We collected soil samples in heath and foothill forest and subjected the samples to a range of treatments including low heat (LH; 45 °C), high heat (HH; 100 °C), smoke (S), low heat + smoke and (LHS) high heat + smoke (HHS), control (C).
0 10 20 30 C LH LHS S HH HHS C LH LHS S HH HHS Heath Forest N o. speci es (m ean, se) 0 10 20 30 C LH LHS S HH HHS C LH LHS S HH HHS Heath Forest N o. speci es (m ean, se) 0 10 20 30 C LH LHS S HH HHS C LH LHS S HH HHS Heath Forest N o. speci es (m ean, se) 0 10 20 30 C LH LHS S HH HHS C LH LHS S HH HHS Heath Forest N o. speci es (m ean, se) 0 10 20 30 C LH LHS S HH HHS C LH LHS S HH HHS Heath Forest N o. speci es (m ean, se)
Germinants were identified and counted over a four month period in a glass house.
Preliminary analyses indicate that the number of germinant species was greater after high heat (+/- smoke) in both communities, but also after low heat (+ smoke) and smoke in heath; low heat had no effect. The composition of germinant species was different in the control (no treatment) compared with smoke and high heat (+/- smoke) in both communities. However, high heat may be singularly more important for the germination of a greater proportion of species in the forest. In particular, Tetrarrhena juncea (wire grass), a common flammable forest fuel, is dependent on high temperature for germination. Lower temperatures experienced during patchy, prescribed burns may reduce its abundance, and generally reduce the flammability of this system. This may have consequences for coexisting species with similar germination requirements. Whilst this research suggests that high heat may be more important for the germination of species in the forest, ongoing analyses will allow a more definite answer.
Forest
Heath
Number of germinant species in soil samples subjected to
different treatments
Bird diversity and function
HOLLY SITTERSDoes patchy, planned fire enhance ecosystem
function?
A
nimal species diversity is often associated with time since disturbance, but the effects of disturbances such as fire on functional diversity are unknown. Functional diversity measures the range, abundance and distribution of trait values in a community, and links changes in species composition with the consequences for ecosystem functions, such as seed dispersal, pollination and nutrient cycling. Improved understanding of the relationship between time since fire (TSF) and functional diversity is required given that the frequency and severity of fire are expected toincrease. We examined responses of bird functional diversity to TSF and two direct measures of habitat variation. We predicted that functional diversity would respond negatively to TSF, a surrogate measure of habitat variation, because recent fire was prescribed, and was typically patchy and of low severity. Patchy fire was expected to create variable habitat and provide greater opportunities for species to partition resources. Further, we anticipated a positive response of functional diversity to direct measures of plant diversity and vegetation structural heterogeneity.
We surveyed birds across a 70-year
chronosequence spanning four vegetation types (heathland, dry forest, foothills forest and wet forest). Six bird functional traits (body mass, clutch size, food type, foraging behaviour, foraging location and nest form) were used to derive
measures of functional richness and evenness, and we investigated the effects of TSF, plant diversity and structural heterogeneity on species richness and the functional diversity indices. We used a regression tree method to identify traits associated with species more common in young vegetation.
Functional richness was negatively associated with TSF in all vegetation types, suggesting that recent prescribed fire generates variable habitat and provides greater opportunities for resource partitioning. Species richness was not significantly associated with TSF, and is probably an unreliable surrogate for functional diversity in fire-prone systems. A positive relationship between functional
Use of patchy fire to break up large expanses of mature vegetation will enhance ecosystem function and resilience, while protecting species reliant on old vegetation.
“
Relationships between species
richness, functional richness and
time since fire
Response of functional evenness to structural
heterogeneity
evenness and structural heterogeneity was common to all vegetation types, suggesting that fine-scale (10s of metres) structural variation can enhance ecosystem function and resilience. Species more common in young vegetation were primarily linked by their specialist diets, indicating that ecosystem services such as seed dispersal and insect control are enhanced in more recently burnt vegetation. Buff-rumped Thornbill and Superb Fairy-wren were among the seven species most common in young vegetation, all of which build dome-shaped nests. This nest type offers better camouflage and shelter against predation than more open nests.
Our data indicate that patchy, prescribed fire sustains functional diversity, and we suggest that controlled use of patchy fire to break up large expanses of mature vegetation will enhance
ecosystem function and resilience, while protecting species that rely on old vegetation.
Buff-rumped Thornbill
Mammals, movement and fire
MATT SWAN
Insights from the Henderson Creek burn
experiment
P
lanned burns are predominantly patchy.Providing a mosaic of burnt and unburnt patches is often a stated aim of planned burning. This is important for biodiversity conservation because unburnt patches are assumed to provide habitat for a range of species. However, many questions remain about the role of unburnt patches for wildlife populations. For example, do unburnt patches simply allow species to survive fire, or do they sustain populations in the longer term so that they can recolonise burnt areas as vegetation structure develops? What shape and arrangement of unburnt patches would benefit certain species? How do
different species with differing requirements and habitat preferences utilise unburnt patches?
In order to answer these questions we used a planned burn as an experimental treatment, focussing on the role of unburnt gullies in providing habitat for ground-dwelling mammals. We
conducted surveys of mammal species before and after planned fire in both the Henderson Creek burn block (burnt in March 2012 as part of DEPI’s fuel reduction burning program) and at a nearby unburnt control site). Use of an experimental framework allowed us to infer causal relationships more readily than observational study alone.
The two small mammal species responded differently to the fire. Population sizes of Agile Antechinus increased in unburnt gullies. Bush Rat abundance remained the same in unburnt gullies and decreased sharply on burnt slopes.
“
The two small mammal species that were studied responded differently to the fire. Population sizes of Agile Antechinus increased in unburnt gullies, suggesting that some individuals moved their home ranges from burnt to unburnt patches in the short term after fire. Bush Rat abundance remained unchanged in unburnt gullies and decreased sharply on burnt slopes. The differences between these species are most likely due to their differing use of habitat and social structures. Antechinus are partly arboreal and this fire was very low severity, only affecting ground vegetation. Bush Rats primarily live on the ground and are sensitive to changes in understory vegetation. Furthermore Bush Rats are highly territorial and individuals in gullies may not be tolerant of individuals shifting their home ranges after fire.
GPS tracking of Swamp Wallabies before, during and after fire in the Henderson burn block showed that the fire had very little effect on this species. Individuals contracted their home ranges to unburnt gullies during fire but soon returned to their original home ranges, only avoiding severely burnt areas in the short term.
This experiment has shown that different species respond to fire and the particular spatial arrangement of unburnt patches depending on their home range size, habitat preferences and life history requirements.
Swamp Wallaby
Agile Antechinus
Small mammal abundance in unburnt gullies before and
after fire
Out of the fire, into the fox’s mouth
BRONWYN HRADSKY
Invasive predator and native prey responses
to planned burning
Control Impact(UB) Impact(B)
Pr (occurrence) 0.00.2
0.4 0.6 0.8
1.0 before after
Control Impact(UB) Impact(B)
Pr (occurrence) 0.00.2
0.4 0.6 0.8 1.0
Red fox Vulpes vulpes
Feral cat Felis catus
Control Impact(UB) Impact(B)
Pr (occurrence) 0.0 0.2 0.4 0.6 0.8 1.0 before after
Control Impact(UB) Impact(B)
Pr (occurrence) 0.0 0.2 0.4 0.6 0.8 1.0 Red fox Vulpes vulpes Feral cat Felis catus
Fire effects on invasive predator occurrence
Background
Invasive predators such as red foxes (Vulpes vulpes) and feral cats (Felis catus) pose a major threat to biodiversity globally. Yet little is known about how their impacts on native fauna are affected by other ecological processes such as fire. Fire typically creates open, less complex habitats, which may increase the vulnerability of ground-dwelling mammals to predators. Greater predation of native fauna by invasive species after fire could have serious implications for conservation, particularly given the extensive use of planned fire as a land management tool and predicted increases in wildfire under climate change.
Methods
To examine the effects of fire on foxes, feral cats and their native prey, we conducted a before-after control-impact study of a planned fire from January to July 2012. The Breakfast Creek burn block was a 1,190-ha area designated for burning, while the control block was an equivalent area about 10 km away with broadly similar topography and vegetation. We surveyed both blocks before and after the fire, deploying motion-sensing cameras to determine invasive predator and native mammal occurrence, collecting fox scats to analyse diet, and conducting structure surveys to assess vegetation. We examined the effects of the fire at the scale of the burn block and also compared burnt and unburnt patches within the burn.
Results
• In the unburnt forest, understorey cover was positively related to the occurrence of small- and medium-sized native mammals such as bush rats (Rattus fuscipes), swamp rats (R. lutreolus), long-nosed bandicoots (Peremeles nasuta) and short-beaked echidnas (Tachyglossus aculeatus). Understorey cover was negatively associated with the occurrence of swamp wallabies (Wallabia bicolor), feral cats and foxes.
Fire effects on fox diet
10%
BEFORE
55%
18%
n = 64
8%
AFTER
54%
29%
n = 50
• The fire created large strips of burnt vegetation interspersed with unburnt gullies, and reduced understorey cover in burnt areas by more than 80%.
• Foxes and feral cats used the burn block more intensely after the fire, particularly the burnt parts of the forest (see facing right). Meanwhile, the occurrence of small- and medium-sized native mammals declined in the same areas, and wallaby occurrence remained unchanged. • Fox diet changed after the fire, with consumption
of medium-sized native mammals more than doubling, and consumption of large native mammals declining by nearly 50% (see above). Changes in feral cat diet after fire need further investigation.
Management implications
The effects of the fire on habitat suitability, invasive predator occurrence and fox diet were likely to negatively impact small- and medium-sized native mammals but favour large native mammals such as wallabies.
Interactions between threats to native fauna can greatly increase the risk of species extinction. At present, management of fire and invasive predators is rarely coordinated. If fox and feral cat predation limits native mammal recovery after fire, integrated management of these processes could greatly benefit biodiversity.
Integrated management of fire and invasive predators could greatly benefit conservation.
“
”
Further experiments are needed to compare native mammal responses to fire in the presence and absence of invasive (and native) predators. The value of differently sized unburnt patches within burn blocks as refuges for native fauna also needs further investigation.
A more integrated understanding and
management of fire and invasive predators may be essential for conserving biodiversity in flammable ecosystems.
T
he Fire, Landscape Pattern and Biodiversity project is being undertaken by the Fire Ecology and Biodiversity Research Group in the Department of Forest and Ecosystem Science at the University of Melbourne (Creswick campus). For more information about their research see: http://science.unimelb.edu.au/fire-ecology-biodiversityThe project has been actively supported by the Department of Environment and Primary Industries and Parks Victoria, with ad-ditional assistance from the Collaborative Research Network, the Holsworth Wildlife Research Endowment and Birdlife Australia.
Student projects
Carolina Galindez Silva (PhD current): Fire, resources and faunal behavioural responses Bronwyn Hradsky (PhD current): Introduced predator responses to fire
Matt Chick (PhD current): The seasonal influence of fire and climate on the species composition of heathland across a precipitation gradient
Holly Sitters (PhD 2014): The influence of fire on forest birds at multiple scales
Matt Swan (PhD 2014): Interactions between fire, environmental heterogeneity and ground-dwelling mammals
Gloria Maikano (MSc 2014): Relative effects of heat and smoke on germination of seeds from heathy woodland and foothills forest soils
Andrew Heaver (PhD 2013): The effects of fire on bark habitats and associated biodiversity
Craig Mildwaters (BSc Hons 2013): Predicting the effects of prescribed burning on native ground dwelling mammals Amber Fordyce (BSc Hons 2013): The impact of prescribed fire on microhabitat selection of the bush rat, Rattus fuscipes John Loschiavo (BSc Hons 2012): Comparing field-based and remote-sensed methods for mapping a prescribed burn Francene O’Connor (MSc 2011): Characterisation of ant assemblages within the surface litter layer of a fire-prone landscape Erin Steel (BSc Hons 2011): The effects of fire and landscape heterogeneity on small mammals of the Otway Ranges, Victoria
Relevant publications
Cohn, J. S., Di Stefano J., Christie F., Cheers G. & York A. (in review) How do heterogeneity in vegetation types and post-fire age-classes contribute to plant diversity at the landscape scale? Forest Ecology and Management.
Di Stefano, J., Christie, F. & York, A. (2011) Fire, Landscape Pattern and Biodiversity. Project Establishment Report. September 2011. Forest and Fire Ecology Group, Department of Forest and Ecosystem Science, University of Melbourne.
Di Stefano, J. & York, A. (2012) Relationships between disturbance regimes and biodiversity: background, issues and approaches for monitoring. Fire and Adaptive Management Report No. 91. Victorian Government Department of Sustainability and Environment, Melbourne.
Sitters, H., Di Stefano, J., Christie, F.J., Swan, M. & York, A. (in review) Bird functional diversity decreases with time since distur-bance: does patchy fire enhance ecosystem function? Ecological Applications.
Sitters, H., Di Stefano, J., Christie, F.J., Sunnucks, P. & York, A. (in review) Bird diversity increases after patchy prescribed fire: implications from a before-after-control-impact study. International Journal of Wildland Fire.
Sitters, H., Christie, F., Di Stefano, J., Swan, M., Collins, P. & York, A. (2014) Associations between occupancy and habitat struc-ture can predict avian responses to disturbance: implications for conservation management. Forest Ecology and Management.
331, 227-236.
Sitters, H., Christie, F., Di Stefano, J., Swan, M., Penman, T., Collins, P. & York, A. (2014) Avian responses to the diversity and configuration of fire age classes and vegetation types across a rainfall gradient. Forest Ecology and Management. 318, 13-20. Swan, M., Gallindez-Silva, C., Christie, F., York, A. and Di Stefano, J. (in review) Patchy fires and fauna: contrasting responses of
small mammals to topographical refugia. Austral Ecology.
Swan, M., Christie, F., Sitters, H., York, A. & Di Stefano, J. (in review) Predicting faunal fire responses in heterogeneous landscapes: the role of habitat structure. Ecological Applications.
Swan, M., Di Stefano, J., Christie, F., Steel, E., Sitters, H. & York, A. (in review) Predicting ground-dwelling mammal responses to environmental heterogeneity in a fire prone landscape. Landscape Ecology.
Swan, M., Di Stefano, J., and Christie, F. (2014) Comparing the effectiveness of two types of motion-camera for surveying ground-dwelling mammals. In 'Camera Trapping: Wildlife Management and Research.' (Eds. P.D. Meek, P.J.S. Fleming, A.G. Ballard, S.C. Banks, A.W. Claridge, J.G. Sanderson and D.E. Swann). (CSIRO: Melbourne)
Swan, M., Di Stefano, J., Christie, F., Steel, E., & York, A. (2014) Detecting mammals in heterogeneous landscapes: Implications for biodiversity monitoring and management. Biodiversity and Conservation. 23, 343-355.