All harvest of Brushwood in Western Australia currently comes from natural stands
on private land, mostly in the northeastern Wheatbelt. While there has been some
limited trial harvesting of plantationgrown Brushwood in Western Australia, there
are no available data on its commercial acceptability at this stage. Initial verbal
feedback from two manufacturers was not encouraging. It is anticipated that proposed
trials of machine harvesting and test panel construction in late 2008 will provide some
muchneeded information on this point. Evaluation of existing plantation resources is
complicated by the mixed nature of the early plantings. Nevertheless it is possible to
make a broad assessment of the potential for each species of Brushwood growing in
The assessment set out below is based on the published work of Brophy et al (2006)
who studied foliar oil content, and Craven et al (2004), who redefined the taxonomic
status of the Melaleuca uncinata complex in Western Australia.
Table 1. Commercial Potential of Western Australian Brushwood Species
Brushwood Panels Oil Extraction
On the basis of the work of Craven et al it appears that only two local species,
a result, these two species have been the focus of grower plantings in the last two or
However, the commercial potential of Brushwood is more complex than this table
would indicate. Within the most favoured species, M.atroviridis, there are two
recognised forms: a lower slope variety that has poorer form and does not resprout
after cutting, and an upper slope form that has better form and does resprout after
cutting. There is anecdotal evidence that within the upper slope variety, plants from
certain sites have superior stem form and growth rate. It is unclear whether the same
applies to M.hamata. For both species this is a critical issue that needs urgent further
Another factor to be considered is the nature of the bark. Some varieties of
Brushwood have hard, durable bark, which is sought after by brushwood
manufacturers. Others have softer bark that does not meet the durability requirement.
Since the commercial acceptability of a plantation is critically dependent on its height,
stem straightness and low incidence of forks, the benefit of investing in the highest
quality genetic material is apparent. A cooperative approach to identify the best
provenance and secure the germplasm in a seed production area, so that all future
plantings use superior quality seed, will be an important factor in creating a plantation
resource in Western Australia that is suitable for commercial exploitation.
In view of the variation in the growth rate and form of current farm plantings of
Brushwood, a set of standards to allow a grower to assess the value of a plantation
and improve resource estimates would be useful. The following table has been
developed to fulfil this need. It is a draft set of key attributes that needs to be field
tested and discussed with potential buyers, since it is the potential buyer who has the
final word on the acceptability of a plantation for brushwood manufacture. One
manufacturer has already endorsed the proposed standard.
The standard is intended for use with a plantation crop that has grown to a height of
about 1.6/1.8 m and assumes that the stem will be cut at about 20 cm above the
ground. For Brushwood greater than 1.8 m in height, the assessment needs to start at
the third fork from the tip to comply with the standard. Cutting would occur at the
Table 2. Proposed Assessment Scheme for PlantationGrown Brushwood
More than 1.5 m
1.2 to 1.5 m
Less than 1.2 m
from 30 cm above
of 10 stems
kinks in 5 out of 10
bends in more than
5 out of 10 stems
Stem Diameter at
Generally 10 mm
to 15 mm
than 15 mm
(less than 20
No more than 1 out
of 10 stems
13 out of 10 stems
More than 3 out of
Wide Angle forks
in lower 30 cm
More than 1 out of
A score of 8, with no individual attribute score of 3, is suggested as the maximum
acceptable for brushwood manufacture. An example of how a grower could use the
table is given in Table 3 below. It is intended that the assessment would be made on,
say, 6 representative sample plots of 10 trees in a row, distributed across the range of
growth of the plantation.
Table 3. Sample assessment of a plantation against the proposed Brushwood standard
Stems generally less than 10 mm
Acute angle forks
Wide angle forks
Extraction of oil from the foliage of Melaleuca species, known as tea tree oil, is a
commercial business in Queensland, although it has not developed to any extent in
Western Australia. The value of a species for commercial oil extraction depends on
the chemical composition of the oil and there is an existing international standard,
ISO 4730, for tea tree oil. The table in Appendix 2 compares the oil contents of the
WA species against this standard. The oil content data for the Western Australian
species are taken from Brophy et al, 2006. The table indicates that some species have
potential for use for oil extraction, but the data are highly variable for each species
where there was more than one sample.
No Western Australian species entirely complies with the ISO standard, although
M.hamata comes closest. For commercial development, it would be necessary to
carry out a rigorous selection process and propagate genotypes with the desired oil
composition. It may be possible to select a genotype of M.hamata that has both
desirable characteristics for brushwood and also compliance with the ISO oil
standard. This would require significant resources and costs, as well as some years to
Such a variety would offer the possibility of an integrated brushwood/oil extraction
operation that would be more economically viable than either activity alone.
Oil composition is, of course, only part of the information that is needed. The yield by
weight of the foliage is also a critical factor. Research on M.alternifolia in
Queensland has shown that there is significant variation in key commercial traits at
the provenance, family and individual tree levels. Narrow sense heritability was
shown to be high for oil concentration, moderate for some oil types and low for
growth parameters (Doran et al, 1996).
Given that there is already such a process in train in Queensland with the preferred
species for tea tree oil (M.alternifolia), it is doubtful that this will be a viable avenue
for development in Western Australia at this stage, unless new uses are found for the
dominant components of local species, or the market for ISO4730 standard tea tree oil
Discussions were held with two local commercial distiller firms as to their interest in
utilising any of the local species for oil extraction. Neither was interested at this stage,
but the possibility should still be kept in mind, as market requirements could change
in the future. Kalannie Distillers is fully occupied with oil production from oil mallee
plantations and Paperbark Essential Oils at Harvey is fully occupied with other
species. However, the latter does have an active research and development program
and could well evolve a product that uses Melaleuca oil in the future. The Baldivis
brushwood factory is also examining the potential for oil extraction from several local
species, including Melaleuca.
The map presented below shows the approximate location of current Brushwood
plantations, comprising both private grower and MIS plantings, based on the
inventory data provided. According to Trees Midwest and Trees South West staff,
there are no other significant plantations. It can be seen that the plantings have been
well concentrated in the central Wheatbelt with a smaller resource in the southern
Wheatbelt. It is desirable that further plantings remain concentrated in the central
zone, at least while the industry is in a developmental stage.
Figure 4. Location of Recent Brushwood Plantings
In making an estimate of potential Brushwood resources from plantations, several
aspects need to be considered:
there is considerable morphological variation within one species.
The growth rate.
The average green weight of a harvested plant.
The commercial acceptability of the plantation, in terms of stem straightness,
The weight per plant at harvest is a key factor in calculating a yield projection. The
only published data are contained in the prospectus for an MIS company (Rewards,
2003), which assumes an average weight per plant of 6 kg. This estimate appears to
have been derived from sampling natural stands of Brushwood in the northern
Wheatbelt (Robinson and Emmott, 2005). The relevance of this approach to the yield
from plantations in the Avon Valley and the northern regions is not known. As the
natural stands are likely to have been older and grown more slowly, there might be a
significant difference from material grown in plantations in a higher rainfall zone.
In a local cutting trial in the Avon Valley, 16 plants were cut and weighed at four
times of the year in a harvest at age 6 years. The plants were about 1.5 m in height
and of good form and are believed to have been acceptable for brushwood production.
The green weight for the whole 16 plants at each sampling date varied from 52 to 69
kg. Using the mean weight of 60.5 kg per 16 plants, the average green weight of each
plant was 60.5/16=3.8 kg, say 4 kg. In the absence of additional data, this figure has
been used in resource calculations. Clearly, this figure requires further validation. It is
possible that the weight per plant of the regrowth Brushwood after harvesting will be
higher than that for the initial growth. Ideally, growers should have a graph of green
weight against age of plant for each broad soil type used for plantations.
Data have recently become available from a machine harvesting trial at Southern
Brook. The mean plant weight varied from 5.6 to 7.3 kg, although the latter
plantations were a little beyond the ideal height for harvesting. It would seem,
therefore, that mean plant weight in a commercial harvesting operation could be 6 kg.
A rigorous approach to resource determination would also require an inspection of
each individual plantation. This was not possible during the development of this plan,
so assumptions have to be made. The development of some sort of acceptability index
for brushwood manufacture would be a useful aid for growers. Such an index might
be based on assessments of stem height, stem straightness and freedom from forking,
along the lines suggested in Table 2 above.
To develop resource estimates, several assumptions have had to be made. For the
purposes of the estimates set out below, two scenarios have been produced, assuming
a mean plant weight of either 4 kg or 6 kg. Depending on the planting data supplied, a
survival rate of 80% or a stocking figure of 2500 plants per ha has been assumed and
all usable plantations have been assumed to be available for harvesting at age 6. The
harvest age of 6 is based on growth rates on a limited number of plantations, and will
no doubt, in practice, vary from site to site.
For private growers we have data on the number of seedlings or the area planted each
year, but we have very little data on the MIS plantations. There has been some
harvesting of the latter in 2008 and it is believed that some varieties planted are
unlikely to be commercially acceptable. It is assumed that the MIS effective
plantations still available are :
50 ha planted in 2003
50 ha planted in 2004
50 ha planted in 2005
It is further assumed that the effective stocking is 2500 plants/ha, allowing for some
The 2006 planting was the first in this region and comprised 3 unknown subspecies
of M.uncinata. It is likely that much of this year’s planting comprised unsuitable
genotypes. Some is also likely to have been established on sites that were not optimal
for growth. For the purposes of this estimate it is assumed that only one third of the
area planted is suitable for brushwood manufacture.
The total estimated number of plants surviving at year 1 is estimated to be 235,000,
based on data received from Georgie Troup. One third of this figure is 78,000. At the
nominal harvest age of 6 years, this resource would be available in 2012 assuming all
grows at about the same rate.
In 2007, the planting used only M.atroviridis and M.hamata, both of which are
believed to be suitable for brushwood. The net surviving number of plants is
estimated to be 489,000 and all plantings are assumed to be commercially acceptable,
and would become available in 2013.
The expected planting for 2008, using the same average 80% survival rate as in
previous years, the net number of surviving plants is 346,000, implying a resource for
These data have been combined to produce the table below.
Somewhat different assumptions were made for the Avon Valley resource, due to the
different way the data were available. All the seedlings were assumed to be either
M.atroviridis or M.hamata.
Table 5. Estimated Avon Valley Resource Availability (tonnes)
Some loss of resource will result from undesirable genetic material being used in the
first few years of planting, but this is not significant overall.
Planting of Brushwood also took place around Katanning in 2007 and 2008. It is also
assumed that all plantings are either M.atroviridis or M.hamata.
Table 6. Estimated Southern Region Resource Availability (tonnes)
We can combine the data for all regions to produce an overall estimate of Brushwood
resources, shown below.
Figure 5. Estimated Brushwood resource availability 20092014 (tonnes/year)
from private grower sources (plant weight 4 kg)
The graphs indicate that there will be only a small resource from private growers until
2012, then a moderate level of resource until 2013, when it suddenly escalates to over
3000 t/year at 4 kg/plant and over 5000 t/year at 6 kg/plant.
In practice, however, it will be possible to smooth out the supply to some extent by
bringing forward the harvest of some fastergrowing areas and postponing harvest of
slowergrowing areas. Nevertheless, the potential for overproduction is clear, unless a
large market can be found. Planning a harvest program will therefore require quite
detailed knowledge about each plantation. A process for gathering these data and
carrying out the planning process needs to be developed.
Using assumptions for available area set out above, namely, 50 ha available in each
year, the MIS plantings are estimated to yield either 500 t/year (4 kg/plant) or 750
t/year (6 kg/plant) in each of 2009, 2010 and 2011. Parts of the MIS plantation that
were harvested in 2008 may yield a second crop in 2014, estimated again to be 500 or
750 t. If this resource is added to the private grower resource, the yield is considerably
increased in the next 3 years, as shown in Figures 7 and 8 below.
Figure 7. Combined private and MIS Brushwood resource (tonnes, plant weight
There is very limited value in retaining older crops of Brushwood beyond the
optimum harvest age, as continued growth means that they no longer meet the
brushwood standard specification, especially if machine harvesting is to be used.
Hand harvesting could cope with older crops much better, as the cutter can vary the
height of the cut for each plant as required.
These estimates are only a general indication of the resource situation, given the
caveats expressed above. We cannot produce a reliable estimate of resource
availability until there is an assessment of each individual crop against the standard
and more data is obtained from the MIS sector. Nevertheless, the possibility of an
oversupply of Brushwood is there.
The question to be considered now is how this resource flow equates to possible