ArticlePDF Available

Preliminary carbon sequestration modelling for the Australian macadamia industry


Abstract and Figures

There is a need to accurately estimate the carbon sequestration potential of many of our agricultural and horticultural industries now that the Australian Government has introduced the Carbon Farming Initiative and is planning to introduce an emissions trading scheme in 2015. This study estimates that the carbon sequestration of macadamia plantations is around 3t CO2e/ha/yr, and provides a methodology to assess the carbon footprint of the Australian Macadamia Industry. This study attempts to estimate the growth rate, and subsequently the sequestration rate of plantation grown Macadamia spp. through regression analysis of stem characteristics of destructively sampled Macadamia integrifolia var. 344. A volume increment curve was also derived using three common genetic varieties (A4, A16 & A42). This curve is used to extrapolate a carbon sequestration rate for the national macadamia plantation estate. Once volume estimates and sequestration rates are determined, an economic benefit of the carbon sequestration can be estimated by auditing the amount of carbon produced by activities such as “on farm” fuel use, fuel used in transport, and energy used in producing the product. In this way, a life cycle carbon budget can be developed that will aid the sustainable development of the macadamia and horticultural industries in Australia through the production of carbon credits from the carbon stored in the trees.
Content may be subject to copyright.
Preliminary carbon sequestration modelling
for the Australian macadamia industry
Tim Murphy Graham Jones Jerry Vanclay
Kevin Glencross
Received: 20 September 2011 / Accepted: 5 December 2012
ÓSpringer Science+Business Media Dordrecht 2012
Abstract There is a need to accurately estimate the
carbon sequestration potential of many of our agricul-
tural and horticultural industries now that the Austra-
lian Government has introduced the Carbon Farming
Initiative and is planning to introduce an emissions
trading scheme in 2015. This study estimates that the
carbon sequestration of macadamia plantations is
around 3t CO
e/ha/yr, and provides a methodology
to assess the carbon footprint of the Australian
Macadamia Industry. This study attempts to estimate
the growth rate, and subsequently the sequestration
rate of plantation grown Macadamia spp. through
regression analysis of stem characteristics of destruc-
tively sampled Macadamia integrifolia var. 344. A
volume increment curve was also derived using three
common genetic varieties (A4, A16 & A42). This
curve is used to extrapolate a carbon sequestration rate
for the national macadamia plantation estate. Once
volume estimates and sequestration rates are deter-
mined, an economic benefit of the carbon sequestra-
tion can be estimated by auditing the amount of carbon
produced by activities such as ‘‘on farm’’ fuel use, fuel
used in transport, and energy used in producing the
product. In this way, a life cycle carbon budget can be
developed that will aid the sustainable development of
the macadamia and horticultural industries in Austra-
lia through the production of carbon credits from the
carbon stored in the trees.
Keywords Carbon sequestration Macadamia
industry Modelling Carbon markets Informing
policy development
This paper is the first part of an investigation into the
carbon sequestration potential of the Australian mac-
adamia industry, and has arisen from a need to develop
information to enable sequestration modelling for
plantation grown Macadamia spp. This research is
needed to establish and enhance the sustainability of
the industry from both economic and environmental
perspectives, and assist the development of an appro-
priate industry policy in relation to climate change
regulation. This in turn may improve the viability of an
important Australian export market, and assist the
ongoing development of Australia’s international
reputation for climate change action. The research
for this paper is being conducted in parallel with
T. Murphy J. Vanclay K. Glencross
Centre for Forestry, School of Environment, Science and
Engineering, Southern Cross University, Lismore, NSW
2480, Australia
G. Jones (&)
School of Environment, Science and Engineering,
Southern Cross University, Lismore, NSW 2480,
Agroforest Syst
DOI 10.1007/s10457-012-9589-2
research designed to provide an initial estimate of the
farm gate carbon emissions of macadamia production.
When combined with an industry emissions audit, this
is expected to demonstrate that the macadamia
industry does not follow the general trend of Austra-
lian agricultural emissions, and instead sequesters
more carbon than it emits. This information is also
likely to be relevant to Australia’s other horticultural
industries, and will ideally result in the development
of an appropriate carbon audit framework and protocol
for the wider Australian horticulture industry.
Macadamia integrifolia and M. tetraphylla are well
known as horticultural species, despite being (in the
case of M. integrifolia) considered rare in the wild
(Harden et al. 2000). A range of natural and controlled
hybrids with M. tetraphylla exist, with a wide variety
of intermediate forms being apparent. They have been
largely ignored to date by forestry circles (Bootle
1998; NCCP 1998; Ilic et al. 2000; Dimitriadis 2005)
due probably to their low significance as a timber
species. As a result, the information required to
reliably estimate sequestration rates in macadamia
plantations has not previously been available.
M. integrifolia (Maiden & Betche) is a member of
the Proteaceae family. Tree form is a tree or tall shrub
to 6–18 m tall, with a spread of 13 m. Adult leaves are
whorled in 3’s; margins are undulate, entire to spinose
(an important differentiator from M. tetraphylla), and
up to 14 cm long, as well as being sessile or on very
short petioles. Flowers are a creamy pink to mauve,
occurring in large racemes. The fruit consists of a
fleshy green husk enclosing one seed; nuts are usually
elliptical or spindle-shaped. (NCCP 1998; Harden
et al. 2000). Both natural and controlled hybridisation
with M. integrifolia generate numerous intermediate
forms varying in spininess of leaves, colour of flower,
size of nut and thickness of shell (NCCP 1998; Harden
et al. 2000).
The Australian Macadamia industry
Macadamia nuts are Australia’s only commercially
significant native food product with the industry
located principally around the northern New South
Wales towns of Alstonville, Lismore and environs,
and in Queensland around Gympie, Bundaberg,
Rockhampton and the Atherton Tablelands. The
industry is the largest of Australia’s tree nut industries,
and represents around 13 % (by tree numbers) of
Australia’s fruit and nut production (Australian
Bureau of Statistics (ABS) 2008). It is also well run
and it is thought that it has a small carbon footprint.
According to 2005 data, the Australian Macadamia
industry consists of over 900 growers who collectively
farm 3.3 million nut-bearing trees on 17,000 hectares
in Queensland and NE New South Wales (FAR 2005;
ABS 2008; Australian Macadamia Society (AMS)
2008). This was worth a total of $146 million in 2004,
but has declined to less than half this value according
to 2007 estimates (Australian Macadamia Society). It
is an ideal industry to assess the carbon stored in the
trees with a view to monetizing this stored carbon
through agricultural offsets through emerging markets
within Australia such as the Carbon Farming Initiative
Agroforestry systems and horticulture
There are numerous examples around the world of
how agroforestry systems can be used to sequester
carbon in annual and perennial crop-tree combina-
tions, and how this carbon sequestration can provide
payment for environmental services (Montagnini and
Nair 2004). However, as far as we are aware this is the
first time such an approach (carbon sequestration) has
been described for macadamia plantations with a view
to attaching a monetary value to this sequestration.
Stem volumes, wood density, root-to-shoot ratios are
all well described; as are the forestry allometric
equations that can be applied to macadamia trees
(West 2004).
Carbon farming initiative
Carbon Farming is a carbon offsets scheme or carbon
credit scheme being established by the Australian
Government to provide new economic opportunities
for farmers, forest growers and landholders and help
the environment by reducing carbon pollution (Aus-
tralian Government Department of Climate Change &
Energy Efficiency). The Carbon Farming Initiative
(CFI) in Australia includes: (1) Legislation to establish
a carbon crediting mechanism, (2) Fast-tracked devel-
opment of methodologies for offset projects, and (3)
Information and tools to help farmers and landholders
Agroforest Syst
benefit from carbon markets. Carbon Offsets or
Carbon Credits represent an abatement of GHGs
which can be achieved by: (1) Reducing or avoiding
emissions, for example, through capture and destruc-
tion of methane emissions from landfill or livestock
manure; or (2) Removing carbon from the atmosphere
and storing it in soil or trees, for example, by growing a
forest or reducing tillage on a farm in a way that
increases soil carbon. The carbon credits or offsets are
usually purchased and used by individuals or compa-
nies to cancel out or offset the emissions they generate
during their day-to-day life or normal course of
business, for example, by consuming electricity or
catching a plane. Carbon credits can be used to offset
emissions voluntarily or to meet regulatory require-
ments. Offset projects established under the CFI will
need to apply methodologies approved by the Austra-
lian Government. They will contain the detailed rules
for implementing and monitoring specific abatement
activities and generating carbon credits under the
scheme. Methodologies can be developed and pro-
posed by private project proponents, as well as
government agencies. The Australian Government is
working with industry and other stakeholders, state
government officials and technical experts to develop
offset methodologies that have broad application.
These methodologies are expected to be approved and
rolled out progressively from November 2011. An
independent expert committee, the Domestic Offsets
Integrity Committee, has been established to assess
offset methodologies proposed under the scheme and
provide recommendations to the Minister for Climate
Change and Energy Efficiency on their approval. The
Committee will ensure that methodologies are rigor-
ous and lead to real abatement. Once approved CFI
methodologies will be published on the CFI website.
Australia announced the framework for a national
emissions trading scheme in December 2008 and has
introduced a carbon tax in 2011 as a precursor to an
emissions trading scheme in 2015. The framework
establishes that forests or plantations that were estab-
lished after 1990 on land that was cleared before 1990
may be eligible to sell sequestration that is generated
after 2008. Macadamia industry census (2008) data
establishing the post 1990 macadamia plantings is
presented in Fig. 1. It is assumed for this estimate that
all plantings occurred on previously cleared land and
are therefore eligible for inclusion in the trading
scheme. The aim of this study is therefore to establish
a robust methodology for estimating carbon storage in
the macadamia industry by using forestry allometric
equations that estimate tree biomass and growth,
together with density and carbon conversions to
estimate carbon sequestration in the macadamia trees.
This sequestration can then be used by the macadamia
industry to apply for carbon offsets if the industry
wishes to do so. This is part of a broader aim to assess
and decrease industrial emissions of GHGs through
agricultural and forestry offsets.
Materials and methods
The study was conducted at Deenford Macadamia
Plantations near Knockrow, north-eastern NSW. One
25 year old stand of M. integrifolia var. 344 was
sampled using destructive measurement techniques,
with material being obtained from a thinning opera-
tion. A second stand of 17 year old var. 344 was also
surveyed for height and diameter growth, as was a
mixed 10 year old stand of vars. A4, A16 & A42. The
Var. 344 stands were planted with a stocking rate of
357 stems per hectare, and the A series hybrids were
planted in contiguous rows with a stocking rate of 250
stems per hectare (Greg James, pers. comm.). The A
series hybrids had not yet achieved a sufficient degree
of growth to close their canopies at either their actual
stocking rates, or at the stocking rate of their
neighbouring stands. Their diameter data has therefore
been directly compared with the older and more
densely stocked stands without modification, on the
assumption that basal area increments remain compa-
rable due to lack of competition effects. Height data is
Tree Numbers by Year Planted
Number of Trees (x 1000)
Fig. 1 Macadamia tree census according to year planted (AMS
Agroforest Syst
also compared directly, ignoring any varietal variabil-
ity, on the assumption that early height increment is
likely to be more similar than not.
The soil of the study site forms part of the
Wollongbar landscape group, consisting of deep, well
drained kraznozems, probably of Gn3.11 or Gn4.11
Northcote soil group (Morand 1994). The terrain
consists of an East facing slope with a fall of
approximately 1, and experiences an autumn domi-
nated rainfall of 1,730 mm per annum (Bureau of
Meteorology, BOM). The average temperature is
19.3 °C with a daily range of 10.2 °C, an average
annual maximum of 28.2 °C, and an annual average
minimum of 8.4 °C (BOM). Temperature records
from 1963 to 2009 from the Centre for Tropical
Horticulture at Alstonville, northern New South
Wales, near the study site indicate a strong seasonality
in warming, with winters increasing by 1.5 °C over
this period (*0.31 °C/decade)(Olsen 2011). Flush
development for macadamia’s trees hedged 25 days
prior to the winter solstice decreased by about 17 days
from 1963 to 2008, whilst for hedging at the summer
solstice there was no trend.
Diameter and height measurements were recorded
in the non-destructively sampled stands by systemat-
ically measuring every 2nd tree, excluding edge trees,
using a diameter tape and an inclinometer. Thinning in
the 25 year old stand was conducted in a systematic
manner, removing every second tree from every
second row, resulting in a 25 % thin. The 30 cm
height was marked with paint prior to felling, and
subsequently used to permit accurate direct height
measurements. Heights and diameters (over bark) of
36 thinned trees were measured post felling using a
tape measure and diameter tape. Heights of standing
trees were estimated using a Sunto inclinometer, with
measurement difficulties meaning that only top
heights were measured for multi stemmed trees.
Sample disks were removed from thinned trees at
heights of 1.3 m and 5.3 m, and re-sawn to provide 2
outer and 2 inner-wood blocks of approximately
30 mm
. The blocks were oriented across the disk to
capture the greatest degree of eccentricity within the
log as was possible, resulting in an ability to charac-
terise variation due to uneven growth characteristics.
Blocks from the centre of the disk were cut directly
adjacent to the pith, whilst perimeter blocks were
obtained from 3/4 of the distance from pith to bark in
order to obtain a weighted density sample suitable for
averaging with the centre samples. No sapwood
banding was visible, but the medullary ray tissue
was noticeably spongy when directly adjacent to the
bark. This spongy tissue extended to a depth of around
5 mm, and was not included in the basic density
Bark samples were obtained with a chisel from
heights of 1.3 and 5.3 m above ground, and then
trimmed to retain only material with a complete
profile. Bark thickness was measured using a Vernier
calliper on a cut surface prior to bark removal.
Samples were weighed on an electronic balance with
a resolution of 0.001 g to obtain green mass and
volume (using the water displacement method), and
subsequently dried at 103 °C in a wood drying oven
until no further reduction in mass was evident
(*3 days). Samples were then re-weighed to obtain
oven-dry mass.
Stem volume was calculated using a combination of
Newtons and Smalians formulae to the 5.3 m mark,
and assuming the remainder of the stem followed a
conical form (after West 2004). The 0–0.3 sectional
volume was calculated as a cylinder, with the presence
of micro-buttressing/soil erosion making direct mea-
surement difficult. This will likely cause an under-
estimation of volume, however the short length of the
butt section is expected to minimise this error (Philip
1994). Bark volume was obtained by calculating the
basal area of the bark as the difference between the
over and under bark diameter measurements and
calculating volume in the same way as stem volume.
Values in between measurement points were extrap-
olated using separate linear functions for the lower and
upper stem, and assuming that bark thickness was zero
at the tip, in a manner similar to that proposed by
Husch et al. (2002) and Specht et al. (2000).
Macadamia timber has been reported as being hard
and dense, with a Basic Density (BD) of 800–1,000 kg/
, of fine texture, and with pronounced medullary
rays (see Dimitriadis 2005), indicating that macadamia
trees can store significant amounts of carbon. Mea-
surements of Basic Density (BD) was conducted
according to the water displacement method (Ilic
et al. 2000), using the formulae tabled in Table 1
(Bootle 1998). The basal area of multiple leader trees
was calculated using the quadratic summed diameter at
breast height (QSDBH). Samples for bark density (BD)
were obtained at breast height in accordance with Ilic
et al. (2000). Samples were also obtained from a height
Agroforest Syst
of 5.3 m (approximately half way up the stem) in order
to identify the degree of variability in BD due to height.
A formula for estimating stem volume (over bark
and under bark) as well as bark volume of the 25 year
old stand was developed using multivariate regression
of the form: y¼b1x1þb2x2...þe. Factors were
identified with a backwards stepwise elimination
approach (Vanclay 1994), with elimination criteria
being p[0.1. Eliminated factors were reintroduced
and re-eliminated where co-linearity was suspected.
Potential parameters were deliberately restricted to
those which are readily obtainable using basic mea-
surement tools, with parameters being considered in
both unmodified form and in various transformations.
Regression statistics were calculated in MS Excel and
the null hypothesis ðH0¼DxDyÞwas tested at both
regression and parameter level using the Fand P
statistics from the regression analysis table, subject to
acceptable model evaluation as recommended in
Vanclay (1994). The over-bark volume model was
then recalibrated against a sub-dataset, and bench-
marked against the remaining data. The under-bark
volume was not benchmarked, as it followed the same
form as the over-bark model, and was based on
modified over-bark measurements. The derived for-
mulae were then applied to survey results from two
neighbouring stands to further benchmark the models.
The mixed 10 year old M. tetraphylla hybrid stand
was not adequately represented by the models, and
stem volumes for this stand were estimated using the
formulae of a cone and cylinder. The default expan-
sion factors for estimating carbon volume from stem
volume are tabled in Table 2. Above and below
ground expansion factors were not assessed for this
study, and the default factors are therefore applied. A
factor of 0.5 is the generally accepted biomass to
carbon conversion ratio in Australia (Gifford 2000;
Australian Greenhouse Office (AGO) 2002), however
this estimate is based on a limited suite of predom-
inantly Eucalyptus species (Gifford 2000). This author
noted in particular that the carbon contents of
sapwood, heartwood and bark warranted further
exploration in a wider range of species. Fourteen
random samples of wood and bark were therefore
submitted to the Southern Cross University Environ-
mental Analysis Laboratory (EAL) for analysis using a
LECO CNS2000 instrument. The wood was not
differentiated into sap and heartwood due to the
absence of a visually distinct sapwood band.
It is recognized that the carbon sequestration value
is a preliminary estimation and is based upon a number
of assumptions. A key assumption is the non-species
specific nature of the default expansion factors used in
this study. We have assumed they are the same for all
species of tree.
Wood/bark carbon content and basic density
LECO carbon analysis results were 46.3 % carbon
(±0.28 %)
95 % confidence
for wood, and 47.4 %
(±0.46 %)
95 % confidence
for bark. These results show
that the carbon fraction of the wood is 3.7 % less than
the generally accepted value (50 %), with the bark
fraction being 2.6 % less.
Basic density of the wood was 650 kg/m
(±4.78 kg/m
95 % conf
at 1.3 m, and 668 kg/m
(±5.56 kg/m
95 % conf
at 5.3 m. No significant dif-
ference was identified between the outer and inner
samples of 1.3 m and 5.3 m disks using a 2-tailed,
paired ttest. A significant difference was identified
Table 1 Formulae used in analysis of Macadamia stem form
and properties
Green density Green Density ¼Green mass
Green volume
Basic density BD ¼Oven dry mass
Green volume
Quadratic summed diameter
Table 2 Summary of default formulae and factors used to
estimate carbon sequestration
Stem volumeðm3Þ¼
pDiameter cm under barkðÞðÞ=200ðÞ
Above ground
Above ground volume (m
)=1.25 9stem volume
Root to Shoot
Total volume (m
)=1.25 9above ground volume
Basic Density Biomass (t) =total volume 90.5 t/m
Biomass to
C(t) =biomass 90.5
(t) =C93.67
Agroforest Syst
between the pooled 1.3 m and 5.3 m samples how-
ever, with basic density increasing by 2.8 % at the
5.3 m measurement point (approximately half way up
the stem). The bark Basic Density was 533 kg/m
531 kg/m
for 1.3 m and 5.3 m samples, ±0.40 and
0.33 kg/m
, respectively at the 95 % confidence level.
A high degree of variation existed between the sample
pairs, and a significant difference also existed between
the 1.3 m and 5.3 m Green Density readings. A two
tailed, paired ttest found no difference comparing the
1.3 m and 5.3 m bark samples however. One outlier
lay above 700 kg/m
, and 2 outliers lay below 400 kg/
, and were removed due to identification as
measurement errors.
Stem volume
Stem diameters (over bark) were measured every
meter, from 0.3 m through to 5.3 m. Multiple leaders
were converted to an equivalent single stem diameter
using the quadratic summed diameter transformation
tabled in Table 1. The diameter data ranges are plotted
in Fig. 2.
The outliers shown in Fig. 2were assessed to
ascertain their cause. Both data points are assessed as
valid according to Chauvenet’s criterion. A further
assessment of stem form indicated that the 2.3 m
outlier (Tree 22) appeared to be a legitimate data point,
whilst the 1.3 m outlier generated by Tree 27 appears
to be caused by a measurement anomaly (data not
shown). It is likely that this outlier was caused by a
measurement or transcription error, and as this data
point has a high leverage in the regression, it was
excluded from further analysis.
The diameter at breast height over bark (DBHOB)
and height distributions for the 3 surveyed stands are
shown in Fig. 3. The increase in diameter of the
17 year old stand over the 25 year old stand was found
to be significant at the 95 % confidence level, but its
cause was not ascertained. The mean stem height of
the 25 year old M. integrifolia var. 344 stand was
10.2 m, ±0.25 m
. Twenty-eight percent of sam-
pled trees contained double leaders at 1.3 m, and the
average height difference of these leaders was 0.38 m,
or 3.56 % less than the top height.
The nature of the crowns was such that the accurate
measurement of multiple leaders on standing trees is
likely to be highly difficult for taller trees. Stem
volume analysis was therefore restricted to the height
from the tallest stem in order to maximise the
useability of the results.
Over bark volume: 25 yo M. integrifolia
Over bark volume was found to be related to
diameter and height through the regression (Fig. 4):
y¼0:000386842 QSDBHðobÞ2þ0:008927399
The null hypothesis was rejected, with combined
regressor’s of QSDBH and top height returning an r
value of 0.937. Individual coefficient of significance
levels are tabled in Table 3. The plot of the over bark
volume regression is shown in Fig. 4, with the trend
line of the fit following a 1:1 slope, and passing
approximately through the origin. The data distribution
does not suggest the presence of heteroskedasticity,
Diameter & Taper: M. integrifolia var. 344
0.3 1.3 2.3 3.3 4.3 5.3
Measurement height (m)
Diameter (cm)
Quartiles Mean values Outliers
Fig. 2 Quadratic Mean Diameter (over bark) versus height for
25 year old Macadamia integrifolia var. 344. [Outliers repre-
sent data lying more than 1.5 9the inter quartile range from the
1st and 3rd quartiles]
Fig. 3 Diameter and height distribution of sampled 10, 17 and
25 year old Macadamia stands. [Error bars denote the
maximum and minimum range, or 1.5 9the inter quartile
range, whichever is the lesser value. The 10 year old stand
consisted of 3 different tetraphylla/integrifolia hybrids, whilst
the 17 & 25 yo stands consisted of M. integrifolia var 344]
Agroforest Syst
with stem volume vs height regressions and statistical
residual analyses providing no further suggestion of a
serious violation of regression assumptions.
The average double sided bark thickness of the
25 year old stand was 7.9 mm ±0.39 mm
95 % conf.
at 1.3 m, and 5.8 mm ±0.38 mm
95 % conf.
at 5.3. The
relative thickness of the bark increased from 4 % of
stem diameter at 1.3 m to 5.7 % of stem diameter at
5.3 m.
Under bark volume: 25 yo M. integrifolia
Under Bark Volume was also found to be significantly
related to diameter and height, with the regression
following the form.
y¼0:0093005 QSDBHðubÞ2þ0:0003578 h
Regression analysis of these coefficients returned
an r
value of 0.939 and a Psignificance value of
(data not shown). Stem volume versus height
and QSDBH regressions with an analysis of residuals
(data not shown) indicate that the data may not be
normally distributed, but the normal probability plot
shows that the deviation from normality is not great,
and resultant errors are not likely to be significant.
Regression model evaluation
The regression model was evaluated by recalibrating
the over bark model to a sub-dataset and fitting the
revised model to the remaining data points. The data
set was split at n=20, with 15 data points being
retained for validation (data not shown). No sorting or
selection criteria were applied. The residuals of the
resulting model were assessed for gross modelling
assumption violations with the mean of the 15
residuals being -0.0012, with r=0.0174. The
normal probability plot and histogram show that the
smaller data set resulted in a poorer approximation of a
normal distribution, however this error was not found
to invalidate the model. The model was found to fail
completely in trees significantly smaller than those
measured however, with y
ˆ=0 when d =10 cm and
h=4.8 m. The model was not tested against com-
pletely independent data due to the lack of a suitable
An estimate of the carbon sequestration rate
of Macadamia plantations
The model was used to estimate the average stem
volume in the 10, 17, and 25 year old stands. Stem
volume in the 10 year old stand was estimated using
frusta of cylinder below breast height, and cone above
breast height, whilst its bark volume was calculated as
a constant percentage (8.58 %) of total stem volume.
The results are shown in Table 4and default values for
expansion factors, root to shoot ratio, basic density,
carbon fraction and conversion factor (Table 5).
Table 6summarises the conversion factors used to
Measured vs. predicted over bark volume
0.050 0.150 0.250 0.350 0.450
Predicted Volume (m3)
Measured volume (m
Fig. 4 Plot of calculated over bark volume versus predicted
Table 3 Over bark volume regression coefficients and values
Multiple RR
Adj R
SE Obs FSig F
0.969784319 0.940481626 0.936761728 0.017236866 35 252.8245497 2.4798E-20
Coefficients SE tstat Pvalue Lower 95 % Upper 95 %
Intercept -0.113647657 0.041543189 -2.735650747 0.01006968 -0.198268365 -0.0290269
QSDBH^2 0.000386842 2.02104E-05 19.14070415 4.37442E-19 0.000345675 0.00042801
Top height 0.008927399 0.004229135 2.11092776 0.042683373 0.000312932 0.01754187
Agroforest Syst
derive total tonnes CO
/tree from bark volume and
under bark stem volume. This data was used to
produce an MMF model equation (Table 6) with the
resulting data fitted using Curve Expert (Hyams 2005)
as shown in Fig. 5.
Applying the sequestration vs age estimate in Fig. 5
to the post 1990 planting statistics suggested that
Australian Macadamia plantations will sequester
around 51 thousand tonnes of CO
e in 2008/09 and
56 thousand tonnes in 2010/11, assuming no new
Over 50 thousand tonnes of CO
e is being sequestered
by post 1990 Macadamia plantations each year, with
an average sequestration rate of about 3 tonnes of
e/ha/yr. This is equivalent to about 5 % of the total
GHG emissions from the horticulture industry (Aus-
tralia’s National Greenhouse Accounts 2009). Whilst
the potential for converting this sequestration into
carbon credits may be limited with fewer than 10 % of
growers managing plantations of more than 10,000
trees, we have estimated that 10 t CO
e/ha/yr is also
sequestered by the shell and husk of the macadamia
nut (Murphy et al. 2008). This improves the seques-
tration to 13 t CO
e/ha/yr. At $23 a tonne the potential
value of this sequestration is over $5 million dollars
per annum. The prevalence of smaller operators within
Table 4 Stem volume and total CO2 estimates in tonnes for plantation Macadamia trees
Volume (m
/tree) CO
Over bark Under bark Bark
10yo Mean 0.0198 0.0181 0.0017 0.033705
Confidence (95 %) n/a n/a n/a
17yo Mean 0.1625 0.1481 0.0143 0.276387
Confidence (95 %) 17.00 % 17.37 % 13.21 %
25yo Mean 0.2020 0.1852 0.0167 0.343904
Confidence (95 %) 11.31 % 11.48 % 9.47 %
The 17 and 25 year old trees were M. integrifolia, whilst the 10 year old trees consisted of several M. tetraphylla hybrids. [The
factors in 5 were used to derive total CO
n/a = not available
Table 5 Summary of conversion factors used to convert stem
volume to total CO2
Wood Bark
Above ground expansion factor 1.25
Root to shoot ratio 1.25
Basic density 0.65 0.533
Carbon fraction 46.30 % 47.40 %
conversion factor 3.67
Combined factor 1.72576 1.44874
Table 6 MMF model coefficients for estimating stem volume
versus age
MMF model: y=(a*b?c*x^d)/(b?x^d)
Coefficient Over bark Under bark
a= 0 0
b= 47591923 31688386
c= 73.55737 67.67044
d= 6.70408 6.54456
Volume (over bark) and CO2increment rates
0 5 10 15 20 25
Volume increment (m3/stem)
CO2increment (t CO2
Fig. 5 Stem volume (over bark) and CO
increment rates. The
Current Annual Increment (CAI) is the estimated growth
occurring in a given year, and the Mean Annual Increment
(MAI) is the average growth rate to a given year
Agroforest Syst
the industry is likely to add to compliance and
management costs of running sequestration projects,
and would require careful consideration. However, the
carbon sequestration could be managed by a company
which audits the sequestration for the industry as a
whole, through the CFI after subtracting emissions.
The key sources of direct and indirect emissions
from horticulture are; (1) fuel and electricity =
*70 % of total emissions; (2) nitrogenous fertilizers
and animal manures =*20 %; and (3) Waste and
refrigerant loss to the atmosphere =*10 % of total
emissions (unpublished). To accurately monetise the
real economic value of the carbon stored in the
macadamia plantations these emissions must be sub-
tracted from the tree carbon and carbon stored in the
husk and shell. However, we have roughly estimated
that fuel and electricity usage from two farms in the
northern rivers region is equivalent to about 1 t CO
ha/yr and if carbon stored in the shell and husk can be
valued and stored in soils for example the total
sequestration should be around 12 t CO
e/ha/yr giv-
ing an economic return of about $4.7 million dollars
per annum if this methodology is accepted by the
Government Regulator.
The real economic value of the sequestration may
however lie in the ability to assist market positioning
of macadamia produce. The market price of macad-
amia nuts has dropped due to a current global
oversupply, and Australian nuts are therefore compet-
ing on quality and desirability. The ability to confi-
dently demonstrate good environmental credentials
may assist this end. A search of the literature indicates
that further research is also required to inform accurate
carbon modelling in macadamia plantations. The
results of basic density testing show that this variety
of Macadamia has a significantly lower BD than was
suggested by the available literature. The lack of
referenced sources negated the ability to determine the
provenance from which the reports came however,
leaving the possibility that 0.65 t/m
represents an
appropriate value for plantation grown macadamia
spp, and that the previous data came from wild
provenances. The estimates of carbon content for both
wood and bark were both close to 3 % of the
recommended default value. Further research into this
value would improve statistical confidence for the
species; however the proximity of the results to the
default value suggests that greater gains in accuracy
may result from other research avenues. One such area
likely to benefit from further research is in the
estimation of expansion factors. Macadamia spp. are
known to have a large network of fine surface roots,
accompanied by a proteoid root structure, and a
shallow taproot on grafted species (Firth et al. 2003).
Keith et al. (2000) note that massive variation in root
to shoot ratios may occur, and failed to identify any
previous research into appropriate values for either the
Proteaceae family as a whole, or any of the macad-
amia species in particular. This appears to be the first
instance of a published growth model for any of the
Macadamia species. It is not expected that this model
is sufficiently accurate for detailed modelling. It is
however believed that it provides a starting point for
further validation and adaptation.
This estimate of sequestration appears to be the first
instance of a horticultural industry in Australia acting
to capture the potential value of the sequestration it
generates. This is worth around $1.3 million annually
for the carbon stored in the trees, and higher if carbon
in the shell and husk is included. The Australian
horticultural industry has a strong international repu-
tation for quality and high standards. This sequestra-
tion estimate represents the first step of an industry
carbon audit process that may help secure the repu-
tation and industry value in the face of increasing
competition and recognition of agriculture’s effects on
the climate (IPCC 2007). However, further research on
default expansion factors, variations in growth of
different species and hybrids used, and research into
the management and regional variability of soils and
the trees is warranted. This study also does not
consider the emissions profile of the industry in any
detail. This emissions profile will require consider-
ation before making any specific environmental
claims for macadamia orchards, beyond the stated
sequestration estimate. Also this carbon sequestration
estimate will vary over the coming decades as
temperatures increase further and variations in photo-
synthesis affect carbohydrate reserves in macadamia
orchards (Olsen et al. 2008).
Acknowledgments This research has been greatly aided by
help from Andrew Heap and Kim Jones (Australian Macadamia
Society), Greg and Cliff James of Deenford Macadamia farm in
Agroforest Syst
the Northern Rivers region of NSW, and Peter Leihn from the
NSW Greenhouse Gas Abatement Scheme/Green Power.
Australia’s National Greenhouse Accounts (2009) National
GreenhouseGas Inventory accounting for the KYOTO target
May 2009. Pub. Department of Climate Change. ISBN:
Australian Bureau of Statistics (ABS) (2008) Agricultural
commodities Australia. No. 7121. Australian Bureau of
Statistics, Canberra
Australian Greenhouse Office (AGO) (2002) Field measure-
ment procedures for carbon accounting. Australian
Greenhouse Office, Canberra
Australian Macadamia Society (AMS) (2008) Australian Mac-
adamia Society Macadamia Farm Census. Australian
Macadamia Society, Lismore
Bootle KR (1998) Wood in Australia. McGraw Hill, Roseville
Commonwealth of Australia (2008) Carbon pollution reduction
scheme: green paper. Australian Government, Canberra
Dimitriadis E (2005) Proteaceae—the forgotten Aussie oaks.
Agroforestry news. University of Melbourne, Tatura
FAR (2005) The Australian Macadamia industry. Industry note,
food & agribusiness research—horticulture Australia
Firth DJ, Whalley RDB et al (2003) Distribution and density of
the root system of macadamia on krasnozem soil and some
effects of legume groundcovers on fibrous root density.
Aust J Exp Agric 43(5):503–514
Gifford RM (2000) Carbon contents of above-ground tissues of
forest and woodland trees. Australian Greenhouse Office,
Harden GJ, Hardin DW et al (2000) Proteaceae of NSW. UNSW
Press, Sydney
Husch BT, Beers W et al (2002). Forest mensuration. Wiley
Hyams D (2005) Curve expert. Available at http://userpages.*dhyams/cmain.htm
Ilic J, Boland D et al (2000) Wood density phase one—state of
knowledge. Australian Greenhouse Office, Canberra
IPCC (2007) Climate change 2007. Synthesis report. Contribu-
tion of working groups I, II and III to the fourth assessment
Keith H, Barett D et al (2000) Review of allometric relationships
for estimating woody biomass for New South Wales, the
Australian Capital Territory, Victoria, Tasmania and South
Australia. National Carbon Accounting System Technical
Report No. 5B. Dept. Climate Change, Canberra
Montagnini F, Nair PKR (2004) Carbon sequestration: an un-
derexploited environmental benefit of agroforestry sys-
tems. Agrofor Syst 61:281–295
Morand DT (1994) Soil landscapes of the Lismore-Ballina 1:100
000 sheet. Soil Conservation Service of NSW, Sydney
Murphy T, Jones GB, Leihn P, Vanclay J, James G, Glencross K
(2008) Climate change and carbon trading: implications for
the Macadamia industry. Conference proceedings of the
Australian Macadamia Society, annual meeting, Ballina,
October 2008, pp 27–30
NCCP (1998) ‘Macadamia integrifolia Maiden & Betche, &
Macadamia tetraphylla L. Johnson.’’ Retrieved 2nd
March, 2008, from
Olesen T, Roberton D, Muldoon S, Meyer R (2008) The role of
carbohydrate reserves in evergreen tree development, with
particular reference to macadamia. Sci Hortic 117:73–77
Olsen T (2011) Late 20th century warming in a coastal horti-
cultural region and its effects on tree phenology. N Z J Crop
Hortic Sci 39(2):119–129
Philip MS (1994) Measuring trees and forests. CABI,
Specht A, McBeth B, Macgregor-Skinner J (2000) Measure-
ment of carbon sequestration in small non-industrial forest
plantations. Australian Greenhouse Challenge Project,
Southern Cross University, 26 pp
Vanclay J (1994) Modelling forest growth and yield. CABI,
West PW (2004) Tree and forest measurement. Springer, Berlin
Agroforest Syst
... Environmental performance was determined using a multi-criteria approach based on six key indicators: three positive indicators (carbon sequestration potential, soil stabilisation, soil enrichment) and three negative indicators (impact on biodiversity, fertiliser application, water use). Carbon sequestration potential of each species was determined by converting total biomass to carbon and applying an expansion factor (Gifford, 2000;Snowdon et al., 2000;Murphy et al., 2013). The impact on biodiversity rating was higher for scenarios with more than two species because of the association benefit provided by different species characteristics that provide a variety of habitat or resources (Lindenmayer et al., 2003;Kelty, 2006). ...
Full-text available
Agricultural land uses can contribute to land degradation, water quality decline, and loss of ecosystem function and biodiversity in the surrounding catchment. Trees can assist in catchment management, and re-afforestation strategies have been implemented in an effort to mitigate agricultural impacts and improve degraded land and waterways worldwide. Re-afforestation strategies often target private land, and their success relies on landholder participation. Landholders’ decisions about land-use allocation are driven primarily by the private financial costs and benefits associated with different farming strategies. This research assesses the private on-farm financial impact and the public environmental benefit of land use transition from beef grazing to a mixed beef grazing-forestry system in the Richmond River catchment on the east coast of Australia. GIS analysis identified more than 30% of the catchment as beef grazing land potentially available for re-afforestation, across a variety of soil types and geomorphic characteristics. We used a farm-scale financial model to assess the costs and benefits associated with transition from grazing to a variety of cattle-forestry mixtures that were determined on the basis of their suitability to soil type and slope in different parts of a catchment. We also used a multi-criteria approach to assess the environmental outcomes associated with each transition. The results demonstrate that diversification to a mixed beef grazing-forestry system consistently provides environmental benefit, but the financial impact on landholders varies depending on soil type. Landholders on ferrosol and vertosol soils in this catchment have favourable options that can simultaneously deliver private and public benefits, whereas landholders on kurosol and dermosol soils are more restricted, with environmental improvements possible only as a trade-off with farm financial performance. Based on these results, we suggest that different policy mechanisms are required to encourage graziers in different parts of the catchment to shift towards mixed cattle-forestry systems.
... Often this process involves the combustion of materials such as liquefied petroleum gas in order to provide a continuous stream of hot air to massive silos or batch column driers over a period of several days. 25 Therefore, although in theory macadamia nut plantations are a form of carbon sequestration, 26 the actual cracking process is significantly polluting. The shells and other waste (ca. ...
The effect of ionic liquids upon the mechanical and (bio)chemical integrity of macadamia nut shells (from Macadamia integrifolia) has been investigated. Whole macad-amia nuts-in-shell are notoriously difficult to crack, and the Australian macadamia nut shells used in this study required 2240 ± 430 N of force to crack. Ionic liquids were screened for their solubility values, with 1-ethyl-3-methylimidazolium acetate ([Emim][OAc]) able to dissolve 5.5 ± 0.5 wt % macadamia nut shell. Treatment with small quantities of [Emim][OAc] resulted in weakened whole nut-in-shells that could be cracked with only ca. 46% of the displacement (0.67 ± 0.16 mm), ca. 34% of the force (760 ± 240 N) and ca. 15% of the energy (0.25 ± 0.10 J per shell) relative to no treatment. Further treatment by dissolution and precipitation of macadamia nut shell, followed by enzymatic hydrolysis with cellulase, resulted in the release of 80 ± 15% of the expected glucose content, relative to 1.3 ± 1.0% before any pretreatment. Ionic liquids (ILs) are increasingly difficult to define, due to their increasing diversity in chemistry, structure and function. A general definition of ILs is that they are liquid below 100 °C, and at this point, it should be an ionic compound. A staggering number of IL structures are possible. 1 Of these, a relatively small number of ILs possess the ability to act as nonderivatizing (to a degree 2) near-universal solvents for a wide range of lignocellulosic biomasses, spanning from soft and hard woods to rice husks, bagasse, straws and grasses, as examples. 3 With weight percentage solubility values exceeding 50% for lignin 4
Full-text available
Agroforestry systems demonstrate carbon sequestration potentials far higher than widely promoted practices like conservation agriculture and managed grazing. Their lifetime carbon stocks are also significantly higher. In addition, perennial staple crops offer an opportunity to greatly improve carbon sequestration on cropland. The potential of these staple crops has been little explored. In tropical regions, yields of many perennial staple crops are competitive with those of annual staples. This is not yet the case for temperate and boreal climates. Both agroforestry and perennial staple crops offer additional co-benefits in the form of ecosystem services and benefits to farmers. Many agroforestry systems already incorporate perennial staple crops. The addition of trees to agricultural mitigation practices like conservation agriculture or managed grazing can increase sequestration rates by 5–10 times, and increase soil carbon stocks by a factor of 3–10. Tropical multistrata systems feature the highest carbon sequestration rates and carbon stocks of any food production system. Many tropical multistrata systems incorporate perennial staple crops with yields similar to those of annual staple crops. Given their potential impact, these practices and crops should be prioritized in agricultural climate mitigation efforts.
Full-text available
Forests must be measured, if they are to be managed and conserved properly. This book describes the principles of modern forest measurement, whether using simple, hand-held equipment or sophisticated satellite imagery. Written in a straightforward style, it will be understood by everyone who works with forests, from the professional forester to the layperson. It describes how and why forests are measured and the basis of the science behind the measurements taken.
Full-text available
Agroforestry has importance as a carbon sequestration strategy because of carbon storage potential in its multiple plant species and soil as well as its applicability in agricultural lands and in reforestation. The potential seems to be substantial; but it has not been even adequately recognized, let alone exploited. Proper design and management of agroforestry practices can make them effective carbon sinks. As in other land-use systems, the extent of C sequestered will depend on the amounts of C in standing biomass, recalcitrant C remaining in the soil, and C sequestered in wood products. Average carbon storage by agroforestry practices has been estimated as 9, 21, 50, and 63 Mg C ha−1 in semiarid, subhumid, humid, and temperate regions. For smallholder agroforestry systems in the tropics, potential C sequestration rates range from 1.5 to 3.5 Mg C ha−1 yr−1. Agroforestry can also have an indirect effect on C sequestration when it helps decrease pressure on natural forests, which are the largest sink of terrestrial C. Another indirect avenue of C sequestration is through the use of agroforestry technologies for soil conservation, which could enhance C storage in trees and soils. Agroforestry systems with perennial crops may be important carbon sinks, while intensively managed agroforestry systems with annual crops are more similar to conventional agriculture. In order to exploit this vastly unrealized potential of C sequestration through agroforestry in both subsistence and commercial enterprises in the tropics and the temperate region, innovative policies, based on rigorous research results, have to be put in place.
Full-text available
Forest growth models attempt to quantify the growth of a forest, and are commonly used for two principal purposes: to predict the future status of a forest and the nature of any harvests from that forest, and to help consider alternative cultivation practices. Models may also find other uses, such as in education, communicating information, etc. Depending on the purpose of the model, modelers may choose to emphasize physiological detail or statistical efficiency, but generally seek both biological and statistical accuracy.
We take three approaches to more clearly define the role of carbohydrate (CHO) reserves in the development of evergreen trees. First, we examine the lychee and macadamia literature to develop a whole tree carbon budget to show that current photosynthate makes a greater contribution to the carbon for new growth than CHO reserves. Second, we show that the presence of leaves is sufficient for the production of a functional new shoots on small, girdled branches with few CHO reserves. Third, we use a shade experiment to show that short-term suppression of whole-tree photosynthesis can severely affect new shoot development. Subsequent decapitation of all branches on both the shaded and control trees resulted in faster bud release in the shaded trees, notwithstanding lower CHO reserves.Overall we argue that the main role of CHO reserves is to buffer the pool of current photosynthate. Although such buffering can be quite strong during periods of high carbon demand, even then CHO reserves are a secondary source of carbon.
Whole-tree excavations, root-core and minirhizotron studies indicate that the grafted macadamia tree root system is relatively shallow and spreading, with a short taproot and most of the fibrous root system near the soil surface, while ungrafted trees have a longer taproot. The length of fibrous roots diminished with depth and distance from the trunk. This pattern is consistent with other fruit trees, in that the highest density is generally within 1 m of the trunk. Values obtained in core samples in this study were 4.97 (± 0.43) cm/cm3 and 1.67 (± 0.45) cm/cm3 for 0–10 cm and 10–20 cm at 0.5 m from the trunk, and 2.34 and 1.08 cm/cm3, respectively, at 1 m from the trunk at Clunes. These values were similar to those obtained in separate studies in 1991–93, involving assessments at 5 cm depth increments down to 15 cm, where mean root length densities were 2.0–3.5 cm/cm3 and 1.3–1.9 cm/cm3 at 0–5 cm and 5–15 cm depth, respectively, 1.4 m from the trunk. Root length under old trees in bare soil at Dorroughby and Clunes, using minirhizotrons (0.25–0.40 cm/cm2) and soil cores (1.14 and 3.50 cm/cm3, respectively), was similar to that found at other sites in the study area (minirhizotrons 0.28–0.33 cm/cm2; soil cores 1.25–2.80 cm/cm3). There is an apparent lower rate of decrease in root length density with increasing distance from the trunk at 10–20 cm compared with 0–10 cm. New root growth occurred predominantly in autumn, but some new fibrous roots were produced in early winter and spring. Proteoid roots were found in abundance in soil cores and adjacent to minirhizotron tubes and there were more of them in the root systems of younger trees at Clunes than with older trees at Dorroughby. Proteoid roots were found at a greater depth than previously recorded for other Proteaceae species, and appeared to retain their function in relatively dry conditions for more than a year. Non-proteoid fibrous roots at the minirhizotron surface appeared to be functional for about 1.5 years in relatively dry conditions, before decay after the onset of wet soil conditions.The effects of 2 newly established perennial legume groundcovers on the root systems of younger and older macadamia trees were studied over 2.5 years. In general, the presence of groundcover either had no effect on the growth of the macadamia roots or increased the root length density at some sampling dates and some depths. At Clunes, where the proteoid root length density was higher than at Dorroughby, the presence of groundcover was associated with higher proteoid root length density than that with bare ground. Arachis pintoi cv. Amarillo generally had a lower root length density than Lotus pedunculatus.
Temperature records (1963–2009) from Alstonville, northern New South Wales, Australia, were examined for long-term trends using two methods: annual smoothing using a sine curve; and temperature-based phenological models applied to different parts of the year. Both methods indicated strong seasonality in warming, with winters increasing by 1.5°C over the period, but summers largely unchanged. Estimates of spring custard apple flower development time decreased by c. 13 days (20%) from 1963/64–2008/09; estimates of autumn litchi flush development decreased by c. 8 days (7%); estimates of winter macadamia flush development decreased by c. 17 days (12%) but summer flush development was not affected; and estimates of mango fruit development decreased by c. 12–16 days (7–8%) depending on variety.
Measurement of carbon sequestration in small non-industrial forest plantations
  • A Specht
  • Mcbeth
  • Macgregor
  • J Skinner
Carbon contents of above-ground tissues of forest and woodland trees
  • Gifford