for Fencing Products on Cleared Farmland
in Southern WA
By Chris Robinson and Tim Emmott
Greening Australia (WA)
A number of
Melaleuca species can be used in the con-
struction of brushwood fencing in Australia. These species
are often collectively referred to as "broombush". The most
common across southern Australia is
originally collected by Robert Brown at Port Lincoln in South
Australia in 1802. It is a hardy, bushy shrub that grows to
4m tall, with little or no main trunk. It has many long, thin,
vertically orientated branches which are strong and hard,
topped with a canopy of long thin hook-tipped leaves. It is
widespread in the drier woodlands, mallee and heath, where
it is generally slow growing. Most forms have the capacity to
resprout from an underground woody stem (lignotuber) after
fire, drought or physical damage.
The stems have been much favoured for the building of
fences and other structures around domestic buildings, par-
ticularly in south and eastern Australia. A small domestic
brushwood fencing industry has been operating for several
decades using broombush material harvested from the wild.
However, as wild harvesting is restricted by conservation
authorities in Western Australia, South Australia and
Victoria, the expansion of the brushwood industry is con-
strained. The supply to WA of ready made fencing panels
and bundles of brushwood, has recently been supplement-
ed by wild harvest from western NSW, with small quantities
harvested from native stands on private property in WA.
The harvesting of broombush from natural stands is either
restricted or illegal in most Australian states. The long term
sustainability of the industry relies on the development of
broombush plantations that will complement and eventually
replace the wild harvesting. An increase in the available sup-
ply of broombush would provide opportunities to expand
local and export markets.
Dryland salinity is an enormous economic, environmental
and social threat in the agricultural region of WA. Being a
long-lived perennial species, broombush plantations can
help to prevent or alleviate salinity, protect waterways, bol-
ster biodiversity in remnant vegetation and provide an
opportunity to diversify farm income
Broombush is a myrtaceous shrub belonging to a group of
species which comprise the
Melaleuca uncinata complex.
Recent field work and taxonomic study of the
uncinata complex (Lepschi) has shown that the complex
includes at least eleven species. Observations have shown
that several species can co-occur at any given location in
the wild. Many of the species are capable of resprouting
(from lignotuber: a swollen woody stem at ground level) after
the aerial stems are removed by physical damage (such as
cutting), drought or fire. True
M. uncinata occurs from near
Munglinup in WA, east to at least the Eyre Peninsula in SA.
In WA there are another ten closely related species.
However, not all of these species of broombush can be con-
sidered suitable for construction of brushwood fencing.
The preferred material for brush fencing typically has long,
thin straight stems (at least 1300mm long) with a persist-
ent smooth, dark bark. Species with short, crooked stems
or papery bark are not as desirable. However, the conditions
under which the broombush species are grown will affect
the growth rate and form which may determine usefulness
for manufacture of brushwood fencing. In WA, the species
most useful for brushwood fencing are
M. uncinata and two
M. atroviridis, although the generally slower grow-
M. hamata (widespread), M. osullivanii (Swan coastal
M. concreta (northern wheatbelt) may also be
Investors seeking to grow broombush for commercial pur-
poses must ensure they have the best species for their spe-
cific site conditions. The origin of seed used by nurseries to
produce seedlings (or seed used for direct seeding) should
be verified as suitable. If in doubt, contact an industry rep-
resentative or your local Greening Australia (WA) or CALM
The major commercial use for broombush is brushwood
fencing. The durability of brushwood fencing is highly
dependent on the quality of construction. Traditionally, in SA
and Victoria, bundles were cut from wild plants and trans-
ported to the cities where skilled tradesmen would con-
struct fences on site. This involved densely packing the
stems in a supporting frame of uprights and horizontal wires
that were neatly trimmed and capped with steel or brush-
wood. It is not uncommon, especially in rural areas to see
amateurishly constructed fences that sag, and have visible
gaps with stems falling out. In Western Australia, brush-
wood fencing can often be seen in some of the newer, more
exclusive residential developments.
As brushwood fence packing is not a traditional craft in WA,
the fences are usually constructed using panels, which have
been packed, wired and trimmed to consistent specification
in a factory. The panels are transported to the construction
site and erected on a sturdy concealed steel framework.
Brushwood may also be used in gazebos, pergolas and
gates, and other decorative landscape structures.
Sufficient experience already exists to suggest that
Melaleuca species can be commercially cultivated to pro-
duce brushwood fencing material. With good management
on favourable sites, brushwood should be profitable and
may provide opportunities for regional value adding.
Selecting species best suited to specific site conditions is
critical to the commercial viability of a broombush planta-
tion. Broombush species grow naturally in a wide range of
conditions from cracks in rocks, to shallow rocky soil on hill-
tops to deep mid-slope sands and the saline margins of
The opportunity exists to cultivate broombush (Melaleuca uncinata and related species) on cleared
agricultural land in southern WA, for the sustainable production of brushwood fencing material. Broombush
can be grown on land that is not highly productive for conventional agriculture, including sites that are
marginally saline and seasonally waterlogged.
This information aims to provide an overview of suitable species, production and establishment techniques
and opportunities associated with producing broombush for brush fencing.
will result from better growing conditions. It is important to
ensure adequate depth of soil for any broombush plantation
will be mid to lower slopes where there is a depth (at least
250 mm) of sand, sandy loam or gravely loam over clay with
run on of fresh water. Sandy and more transmissive soil
allows rapid penetration of rainfall through to the plant roots
and clay at a deeper level allows accumulation of moisture
at depth, which will be available to the deeper roots over
summer. Some species are tolerant to more saline condi-
tions, but higher levels of salinity will retard growth and
hence harvesting rotation will be longer. The fresher the run
on of water the better. Sites experiencing prolonged water-
logging should be avoided unless mounding can ameliorate
Broombush species grow naturally over a wide variation in
rainfall. Commercial broombush plantations are probably
best suited to 350 to 600 mm annual rainfall zones in
southern WA. Individual site characteristics (especially soil
depth), combined with annual rainfall, will determine growth
rates and time to commercial harvest.
Mid slopes - duplex soil
Melaleuca uncinata is lignotuberous and grows naturally on
shallow red loams over granite and is one of the most desir-
able forms of broombush. It will grow well on mid slope
duplex soils, deeper loams and sandy loams as long as
there is plenty of fresh water run on and the soil is not less
than 300 mm. Shallow soils over granite can be very soggy
in winter, but will dry out over summer causing stress and
stunted growth or plant death.
The lignotuberous form of
Melaleuca atroviridis, grows high
in the landscape in the northern central wheatbelt, and is
ideally suited to deep, pale yellow, acidic sands
Sandy flats along saline watercourses
The larger, upright form of
Melaleuca atroviridis is wide-
spread through the central and southern wheatbelt, has
ideal stems for brushwood production and grows vigorously
on shallow sands adjacent to saline river courses. It is ide-
ally suited to the sandy margins of saline flats. Mounding
may be desirable to improve drainage and growth. It may
also grow on more saline soils but cannot be expected to
grow with comparable vigour. The tree form is non-ligno-
tuberous, but will sprout from the trunk if cut well above
Swampy Coastal plain
Melaleuca osullivanii is native to the coastal plain between
Perth and Busselton and is sometimes used as capping on
brushwood panels. It grows naturally on grey sands over clay
in seasonally waterlogged areas. As wild trees they may
have an unruly, straggly form but dense cultivation in ideal
conditions should produce commercial stems.
spread across the wheat belt, occurring in a diverse range
of soil types from shallow rocky soil to clay and sandy clay
loams. In good conditions it may produce good stem form
for brushwood, although its growth rate may not match other
species. Considerable variation in plant form occurs
between different localities, necessitating care in selection
of seedlots for brushwood production. It may be used where
additional landcare objectives and the desire to use local
species may outweigh commercial growth rates.
M. concreta which grows from the the lower Murchison River
district southwards to the Cataby-Regans Ford district, can
also produce fine brushwood material if provided with suit-
able growing conditions.
A minimum of 5 hectares is suggested for a commercial
plantation. This should ensure that the producer will have
a sufficient quantity of material (at least one semi trailer
truck load of 20 tonnes) at harvest to be useful to manu-
facturers, and that there will be a substantial investment
incentive to ensure good management. Bigger plantations,
100 hectares or over may allow sufficient annual turnover
from staggered planting and harvesting to constitute a
stand alone enterprise.
Broombush species are somewhat susceptible to grazing by
stock, particularly in the initial years of establishment.
Plantations should be fenced to avoid physical damage and
opportunistic grazing. Broombush plantations will be better
suited to block type plantings, where specific soil types can
be targeted, and plantations easily fenced. Wide belt plant-
ings, can be considered along creek lines or existing fence
lines. Alley type systems may not be suitable, as extensive
fencing would be required if stock are to graze between the
Consideration should be given to accessibility for harvesting
work of vehicle access lines, adequate firebreaks and suffi-
cient turning space for operations such as spraying. Very
steep or rocky sites may not be suitable.
Remnant Melaleuca concreta Wongan Hills, WA
There are two main methods for establishing broombush;
nursery raised seedlings and direct seeding. Different site
preparation methods are required for each establishment
method. Direct seeding offers the potential for lower estab-
lishment costs, however reliable direct seeding methods for
broombush have not yet been proven. The most appropriate
method to use will depend on site conditions, available
equipment and finance and the acceptable level of risk to
Individual site characteristics influence site preparation
methods. Heavy grazing and spray topping in the year prior
to establishment can be effective in reducing weed burden
and accumulating soil moisture for sites where seedlings or
direct seeding techniques are to be used.
For establishment using seedlings, planning is required to
ensure seed of suitable provenance is available to your pre-
ferred nursery. Nurseries will need to have seed by
November-early December to ensure suitably advanced
seedlings are available for the planting season in the fol-
lowing year (June - August). When planted, seedlings must
have a well developed root ball and ideally be at least
Weed control in the year previous to planting will allow soil
moisture to accumulate.
On lighter textured soils, machine planting in a one pass
rip/scalp/plant operation is efficient and effective.
Scalping, where the surface layer of the soil containing weed
seed is mechanically removed, can provide effective weed
control in the year of planting. Alternatively, sites can be
rip/scalped and then hand planted. On sites prone to heavy
weed infestations, the application of a knockdown herbicide
2-4 weeks prior to scalping is recommended. However,
some weed cover between scalped lines may be desirable
on sites at risk of wind erosion.
On heavy textured soils, or sites prone to seasonal water-
logging, ripping and mounding is recommended. April is gen-
erally a suitable month to construct mounds, while the site
is relatively dry. Ensure that the mounding equipment
includes a roller to flatten lumps on the mound. A smooth,
flat surface will greatly improve chemical weed control.
Following the first weed germination (typically around late-
May), apply a knockdown/residual herbicide mix. A mixture
of glyphosate and simazine (at recommended rates) has
been found to be highly effective for controlling a wide range
of weed species on many sites. A second knockdown herbi-
cide treatment may be required just prior to planting if addi-
tional weed germination occurs. Care must be taken when
planting into ground previously sprayed with pre-emergent
herbicide as soil moisture and level of organic matter will
determine location and activity of the chemical.
Aim to plant seedlings between late June and early August.
Upland, free draining sites are suited to earlier planting,
while seasonally wet sites can be planted later in the year.
Ensure that nursery seedlings comply with recommended
quality specifications and are healthy and well-watered
immediately prior to planting. It is advisable to ensure that
the moisture filled root ball is planted at least 20mm below
Optimal seedling densities for the production of brush fenc-
ing material have not been fully researched. High stocking
densities will encourage more upright growth, whilst lower
stocking densities on less favourable sites will allow max-
imisation of individual plant height. For establishment using
the economic attractiveness of planting
seedlings will diminish substantially as stocking density is
A target establishment density of 2,000 to 5,000 plants per
hectare is recommended to achieve a workable compromise
between growth rates, plant form and establishment costs.
It is possible that stocking densities could be increased fur-
ther on prime sites, however the increased seedlings costs
may offset any productivity gains. Establishment costs using
seedlings, inclusive of all required operations, are typically
in the range of 50c-80c per planted seedling.
To achieve a stocking density of 5,000 stems per hectare,
plantations can be established with rows 4 metres apart,
and seedlings 500mm apart within the rows. At this densi-
ty, the planted cost will be between $2500 and $4000 per
Direct seeding offers scope for the establishment of high
density broombush plantings at relatively low cost. However,
fine seeded species such as broombush have proven to be
difficult to establish reliably using conventional direct seed-
ing methods. Improved methods are being actively investi-
Three year old brushwood plantation, Central Wheatbelt
kg of seed per hectare. Unselected seed generally costs
between $400 and $600 per kg. Specially selected seed
may cost over $1000 per kg. Therefore, if seeding at 0.5 kg
per hectare, seed will cost up to $500 per hectare. Higher
seeding rates can be used, which is still cheaper than buy-
ing nursery grown seedlings.
Some soil types and sites will be more suited to direct seed-
ing than others. Direct seeding will be less effective in fine
textured, dry soils, which can blow away with the seed or be
blown over and bury germinating seedlings. Sites with
uncontrolled heavy weed burden should not be considered
for direct seeding and deeper sandy sites may dry out too
fast for young seedlings to establish.
Direct seeding is most likely to succeed in weed free sites
where the soil is well structured with larger particle size,
which can hold seed in place, retain moisture and will not
equipment which seeds into a scalped bed from which the
topsoil and associated weed seed bank has been removed
by a scalping blade. A number of direct seeding machines,
designed to be towed by a medium sized tractor, are avail-
able for purchase or hire in WA. Press wheels are important
to provide good soil to seed contact. Residual herbicides
should be used with great care if sprayed in the year of
establishment. Soils with high organic matter that lock the
pre-emergent near the surface can be scalped off.
Seedlings emerging from direct seeding can be susceptible
to attack by redlegged earth mites, insects or vertebrate
pests. Regular monitoring is warranted to provide early
detection of pest activity. Control measures, such as the
application of registered insecticides, may be required.
Successful direct seeding can result in many seedlings ger-
minating per metre of rip line in spring, and establishment
of at least 10,000 per hectare. An ideal direct seeding
result would provide at least two plants per metre of rip line
(evenly distributed) at age one year. Row spacing at this den-
sity will determine overall plants per hectare. If insufficient
seedlings establish, nursery seedling planting may be
required in the second year to achieve stocking densities.
Second Year Weed Control
The growth of broombush will be improved if weeds are con-
trolled in the second year of establishment. This can be
achieved by spraying between rows with a Simazine based
residual herbicide at the break of season. If some weeds
have already emerged, a low rate (<250g per hectare active
ingredient) of a Glyphosate based herbicide can be added to
the inter row spray application.
A number of other selective herbicides may be suitable for
the control of specific problem weeds.
The number of plants per hectare, management, time to
first harvest and site conditions will determine the yield.
Observations in trial plantations and revegetation in WA indi-
cate variation in potential yields, but time to first harvest is
expected to be about five years. One private brushwood
fence building company promoting plantation establish-
ment, estimates 30 tonnes per hectare of brush is achiev-
able at the first harvest. However, bigger plants in roadside
revegetation have produced 25 kg each, which at a low den-
sity of 1000 per ha would produce 25 tonne/ha.
The prospectus of a private investment company's brush-
wood project forecasts a yield of 6 kg of brush material per
plant at first harvest at age five years, which equates to 48
tonnes per hectare. This estimate was made from extensive
sampling of established brushwood in the northern wheat-
kets for between $120 and $300 per tonne, which results
in a gross return of between $3,000 and $7,500 per
hectare at first harvest for a mid range yield of 25 tonnes
per hectare. The price paid per bundle will depend on the
quality of the brush in terms of the stem length, thickness,
degree of branching, leafiness, nature of the bark and cur-
rent market demand.
The time between the first and the second harvest may be
shorter, due to the plants having an established lignotuber
and root system, but will depend on annual rainfall as
stored groundwater will have been used by the growth to ini-
tial harvest. To date, the ability of different broombush
species to resprout under a regular harvesting regime has
not been well researched.
Eight month old direct seeded Melaleuca hamata
Bundles of wild harvested broombush
Broombush is traditionally harvested by hand using a
machete or shoulder mounted mechanical brushcutter and
then bound in bundles of 25kg to 35 kg. The establishment
of bigger plantations could facilitate the development of
mechanical harvesting and bulk handling of brushwood.
Harvesting should be carried out when there is sufficient
soil moisture to sustain resprouting, such as late winter or
early spring. Late autumn harvesting may also be consid-
ered but resprouting may be slow through winter. Harvesting
in summer may result in drought deaths.
To ensure survival of plantations after establishment, own-
ers need to ensure that they are adequately protected from
fire by careful site planning and establishment and mainte-
nance of firebreaks. It may be practical to spread larger
plantings into smaller discrete plantings to protect from the
potential of total loss to fire. A minimum firebreak of 20
metres of bare earth should be considered between block
plantings of five hectares. Local authorities and fire servic-
es can also advise on firebreak requirements.
Leaf chewing insects, such as spring beetles, are another
potential hazard capable of reducing plant productivity.
Insecticide treatments may be required for sites that are
REGIONAL VALUE ADDING
The profitability of broombush will be affected by distance
from markets. Unprocessed brush material has a low bulk
density, which makes it poorly suited to efficient, low cost
transportation. The prospect of value adding close to the
plantations may be attractive and feasible. Most broombush
sold in WA today is marketed as ready made panels which
could be constructed in regional factories and transported
as a much higher value end-product.
Some commercial panels are 1.8x 2.2 metres and require
about 100 kg. Assuming a yield of 25 tonnes per hectare of
brush produced in 5 years, this could be value added to 250
new panels. If each panel retails for around $125, a retail
value of $31,250 per hectare after five years could be
Government control Regulation
In WA, the Department of Conservation and Land
Management is responsible for protecting and conserving
the natural flora and through its Wildlife Protection Branch
ensures that any commercial harvesting of native species is
conducted in a sustainable manner. Currently the depart-
ment is not issuing licences to harvest broombush from
crown land due to concerns of overcutting and limited abili-
ty of natural stands to recover after harvesting. Harvesting
from wild plants on private land is technically possible; how-
ever harvested material cannot be sold without licence from
the Department of Conservation and Land Management.
Licences to harvest from wild plants on private land will not
be issued without inspection to determine if the proposed
harvest area is likely to recover. Currently there is very little
broombush supplied from private land in WA.
A licence to harvest broombush cultivated on private land
(previously cleared farmland) is obtainable from the
Department of Conservation and Land Management.
Landholders and investors considering growing broombush
for sale should discuss the matter of licencing for harvest
and sale with their local Conservation and Land
Current Commercial Planting
In recent years in WA a broombush industry based on culti-
vated plantations has begun to develop. Brushwood
Australia, a company based in Jandakot and involved in the
manufacture and construction of brushwood fences, has
been actively involved in promoting cultivation of broombush
to wheatbelt landowners. The Rewards Group Limited,
based in West Perth, has raised prospectus capital and
established commercial plantings on behalf of investors in
the central wheatbelt. There are also numerous small pri-
vate plantings. Some revegetation consultants can supply
broombush seed and direction on establishment and broom-
bush form selection.
Daniel Huxtable from the Department of Conservation
and Land Management Revegetation Systems Unit - for
providing useful information on all aspects of the indus
Stephen Darley of Brushwood Australia - for providing
cessing and fence construction;
Geoff Cockerton of Landcare Services - who was a great
ment and the brushwood industry; and
Wayne O'Sullivan - who gave freely of his knowledge of
Brushwood Australia. Brushwood Fencing. Kwinana, 2001.
Bulman, P, P Beale and A Knight. "Growing Broombush for
profit and land protection" Bulletin Jan 1998
Primary Industries and sources. South Australia,
ic study of Melaleuca uncinata complex. Centre for
Plant Biodiversity Research, Commonwealth
Government, Canberra, 2004.
Greening Australia (WA) Albany Office
08) 9892 8486
Greening Australia (WA) Northam Office
(08) 962 12400
This information is provided in good faith, and was consid-
ered correct at time of printing.
Greening Australia (WA) recommends that readers carefully
evaluate the accuracy, currency, completeness and rele-
vance of the material in this document for their purposes.
This document is not a substitute for independent profes-
sional advice and readers should obtain any appropriate pro-
fessional advice relevant to their particular circumstances.