PROGRAM B © BUSHFIRE CRC LTD 2006

Patchiness of prescribed burns in dry sclerophyll

forests in south-eastern Australia
Penman, T.D., Kavanagh, R.P and
Forest Science Centre, Department of Primary Industries, New South Wales
Binns, D.L
Native Forest Division, Forest NSW, New South Wales

Introduction Results
The dry sclerophyll forests of south-eastern Australia are among •Fires burn in a patchy manner on both the coupe and plot scale
the most fire-prone forest communities in the world. Prescribed (see Figures 2 and 3 below)
burns are commonly used in these forests to reduce future fire •Moisture affects the extent of fire at both the coupe (aspect) and
risk in order to provide protection to people and assets. plot scale (distance from stream).
•Recent fire history (i.e. time since fire and coverage of the last
There has been considerable debate over the ecological impacts fire) influences fire behaviour at coupe and plot scale
of frequent prescribed burning however few data are available for •Ridge sites are unlikely to burn within 5 years of a fire and gully
dry sclerophyll forests. We examined the factors which influence sites are unlikely to burn within 10 years (Figure 4).
spatial variability of repeated low-intensity fires under two Figure 2: Percentage of coupes burnt
burning regimes for a 17 year period from 1988 to 2005 in the 100
a) Logged sites
100
b) Unlogged sites

Eden Burning Study Area (EBSA). 90

LF
LR
90 UF
UR
80 80
Percent of coupe burnt

Percent of coupe burnt

70 70

Methods
60 60

50 50

The study was conducted in the EBSA which is located in the 40 40

30 30
Yambulla State Forest, 29km south west of Eden, New South 20 20

Wales. The EBSA is comprised of 18 experimental coupes in a 10 10

0 0
randomized block design (Figure 1). 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Figure 3: Percentage of plots burnt

a) Logged sites b) Unlogged sites

Figure 1: The Eden Burning Study Area 100 100

LF
90 90
LR UF
UR
80 80
Percentage of plot burnt

Percentage of plot burnt

70 70

60 60

50 50

40 40

30 30

20 20

10 10

0 0
1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Year
Likelihood of burns
1.0

Figure 4: Predicted probability of

plots burning for gully, mid-
0.8

slope and ridge sites. Black

Probability of burning

0.6

line indicates a 50:50 chance of

burning.
0.4
0.2

Gully site
Mid slope site
Ridge site
0.0

5 10 15 20

Discussion Time since burn (years)

The results confirm anecdotal reports that prescribed burning in

Burns were attempted in the frequent burn sites in 1990, 1992,
1996, 2001, 2003 and 2005 and in the routine burn sites in 1992,
1996, 2001 and 2005.
Within each coupe, 18 permanently marked plots were used to
assess the extent of burning. Each plot was assessed for burn
coverage within three months of a burn. Using Generalized
Estimating Equations (GEE’s) we examined the variables
(environmental and historical) that affect how a site will burn at
the coupe and plot scale.
dry sclerophyll forests is patchy, and quantify the degree of
patchiness at both the site and the coupe scale. The
heterogeneity of these fires is likely to minimize the ecological
impacts. Unburnt areas provide refuges for fire sensitive species,
while colonizing species are able to exploit the newly burnt areas.
Gullies provides a longer term refuge site, as they are less likely
to burn over longer time periods regardless of management intent.
These results suggest that the ecological impacts of high
frequency, low intensity fires are likely to be far lower than is often
predicted based on assumptions of homogeneous landscapes and
uniform burning. The results of this study will be used in
combination with empirical vegetation response data to further
investigate and test this conclusion.

Contact details: trentp@sf.nsw.gov.au

PO Box 100,
Beecroft NSW 2119