Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and the Grains Research & Development Corporation 1
http://archive.agric.wa.gov.au/OBJTWR/imported_assets/content/fcp/cu09_plenary.pdf
Building soil carbon for productivity and
implications for carbon accounting
Jeff Baldock, CSIRO Land and Water, Adelaide, SA
KEY MESSAGES
Composition of soil organic carbon
• Soil organic carbon is composed of a wide range of different materials with different chemical
and physical properties and different extents of decomposition.
Functions of organic carbon/organic matter in soil
• Soil organic matter contributes to a variety of biological, chemical and physical properties of
soils.
• Chemical—cation exchange, pH buffering, reduces effects of sodicity.
• Physical—water retention, soil structural stability, soil wettability, soil temperature.
• Biological—energy for microbes, provision of nutrients and resiliency.
• Increasing soil organic carbon content can result in an increase in the ability of a soil to hold
water and thereby lead to enhanced productivity.
Calculating changes in soil organic carbon content
• Soil carbon content represents the balance between inputs and outputs.
• Values are required for the depth, bulk density and carbon content of the soil layer you are
interested in to determine how much carbon is present.
• Changes in soil carbon content are slow and typically require at least 5 years to be detectable.
• Simulation models can be used to predict the likely outcomes of management practices on soil
carbon content.
$$ from sequestration—fact or fiction?
• Maximising crop productivity will maximise carbon inputs and soil organic carbon content.
• At current prices, it is hard to justify modifying management practices for the sole purpose of
selling carbon credits.
SOIL ORGANIC CARBON: WHAT IS IT?
Soil organic carbon is a complex and heterogeneous mixture of materials. These materials vary in
their physical size, chemical composition, degree of interaction with soil minerals and extent of
decomposition. Although determining the impact of management practices on soil organic carbon
contents is important, it does not tell us anything about the type of organic carbon present. For
example, is the organic carbon dominated by pieces of plant residue or more recalcitrant charcoal? It
is therefore important to determine the composition of soil organic carbon to gain an appreciation for
the implications of management practices and changes in organic carbon content on soil productivity.
We now recognise four different types of soil organic carbon:
• Crop residues—shoot and root residues > 2 mm residing on and in soil.
• Particulate organic carbon—individual pieces of plant debris that are smaller than 2 mm but
larger than 0.053 mm.
• Humus—decomposed materials less than 0.053 mm that are dominated by molecules stuck to
soil minerals.
• Recalcitrant organic carbon—dominated by pieces of charcoal.
FUNCTIONS OF ORGANIC CARBON/ORGANIC MATTER IN SOIL
Organic carbon/organic matter contributes to a variety of functions in soils. These functions can be
broadly classified into three types: biological, chemical and physical (Figure 1). Strong interactions
(represented by the grey arrows) often exist between these different functions. For example, the Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
2
biological function of providing energy that drives microbial activity also results in improved structural
stability and creates organic materials that can contribute to cation exchange and pH buffering.
Figure 1 Functions performed by organic matter present in soils.
Future predictions of climate change suggest that most regions of Australia will be become warmer
and drier based on increases in the concentrations of greenhouse gases (carbon dioxide, methane
and nitrous oxide) in the atmosphere. Under such conditions it will become even more important than
it currently is to be able to maximise the storage of plant available water in soils. The amount of plant
available water a soil can hold is determined by two parameters:
1. the lower limit—the amount of water in the soil that plants cannot extract; and
2. the upper drained limit—the amount of water that can be held against drainage.
The difference between the upper and lower limit defines the potential available water holding capacity
(PAWC) of a soil. If this value can be increased even marginally it will help to maintain or enhance
potential productivity by allowing the soil to retain more water each time it rains.
In the absence of subsoil constraints such as salinity, the lower limit is defined by soil clay content and
increases with increasing clay content (Figure 2). Evidence from WA sands suggests that increases in
organic carbon content can also increase the amount of water present at the lower limit. The upper
limit is also affected by the content of clay and organic carbon clay (Figure 2). For any given clay
content, as organic carbon increases the upper limit, and therefore PAWC, of the soil increases. An
analysis of the influence of increasing soil organic carbon content by 1% of total soil mass (e.g. from
0.7% to 1.7%) on the soil PAWC was completed using data collected from red brown earths of the
mid-north region of SA. This analysis indicated that such an increase in carbon content in the surface
0−10 cm soil layer would increase PAWC from its original value by 2 to 4 mm with the effect
diminishing as soil clay content increased (Figure 3). Although the change in PAWC for sandy-low clay
content soils is predicted to be larger than for clay soils, it would be much more difficult to increase soil
carbon content on sands relative to clays.
Figure 2 Changes in upper and lower limits of soil water content with changes in clay content. The light grey area
defines the plant available water holding capacity of the soil. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
3
Figure 3 Change in soil water holding capacity (WHC) induced by increasing soil organic carbon content by 1% of
total soil mass for red-brown earths of the mid-north of SA.
WHAT DETERMINES SOIL ORGANIC CARBON CONTENT?
The amount of carbon in a soil results from the balance between inputs (plant residues) and losses
(microbial decomposition and associated mineralisation). Figure 4 the bucket represents the amount of
carbon a soil could potentially hold. This amount will vary with factors such as soil clay content, soil
depth, and bulk density and is not influenced by management. The bucket will be smaller for a sand
than a clay soil.
Figure 4 Inputs and losses define soil organic carbon content.
To increase the content of organic carbon in soil, the input of residues must be increased and/or the
rates of loss of carbon must be decreased.
Inputs are controlled by the type and amount of plant residue added to the soil. Any practice that
enhances productivity and the return of plant residues (shoots and roots) to the soil opens the input
tap and will result in an increase in soil carbon. For example, appropriate use of fertilisers to maximise
productivity also maximise returns of organic residues to the soil. However, an upper limit to the input
of residues exists in Australian dryland agriculture because of the limitation that the availability of
water places on potential plant productivity. Maximum soil carbon contents will be obtained for any
management system if productivity and capture of carbon are maximised by attaining 100% water use
and nutrient use efficiencies.
Losses of carbon from soil result from decomposition and conversion of carbon in plant residues and
soil organic materials into carbon dioxide. Processes that accelerate decomposition open the losses
tap further; while those that slow down the rate of decomposition will close the losses tap and help to
maintain or increase soil carbon. Reducing the extent of cultivation has been suggested to result in
enhanced soil carbon levels. Results from Australian studies suggest that while this may be true for
some soils, for other soils such an effect has not been observed (Figure 5). It should also be noted
that even shifting to a zero tillage system is unlikely to result in soil carbon values equivalent to those
that would be obtained under a pasture system. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
4
CALCULATING CHANGES IN SOIL ORGANIC CARBON CONTENT
The amount of organic carbon found in a soil can be calculated using values for the depth (cm) of the
soil layer of interest, the soil bulk density (g/cm3
) and the soil carbon content (%) (Equation [1]). Using
Equation [1], indicates that a 30 cm layer of soil having a bulk density of 1.2 g/cm3
and a carbon
content of 1.2% contains approximately 43 Tonnes of C/ha.
3 Organic C (T C/ha) Depth (cm) Bulk density (g/cm ) Carbon content (%) =× × [1]
Figure 5 Change in soil carbon induced by different levels of tillage for several different Australian soil types. The
number of samples of each type of soil are given by n.
Suggestions have been put forward that altering management practices can increase soil organic
carbon content from 2% to 4% in 5 years. Is this really possible?
If we use the same bulk density as above (1.2 g/cm3
) and restrict our calculations to the top 10 cm of
soil where organic carbon is most easily increased, at 2% carbon the soil would contain 24 tonnes
C/ha. At 4% carbon the same soil layer would contain 48 tonnes C/ha. This indicates that 24 tonnes of
C/ha would have to be added to the soil. Since plant residues contain approximately 45%C this would
equate roughly to 50 tonnes/ha of dry matter (DM). If this increase was to occur over 5 years, then an
additional 10 tonnes DM/ha/year above that currently being added would be required if no
decomposition occurs. If half of the residues added were in the form of roots below ground, then we
would have to add an additional 5 tonnes shoot DM/ha/year. Since we know that at least 50% of the
added plant residues will decompose, annual additions of approximately 10 tonnes shoot DM/ha/year
above that currently being added would be required to achieve an increase in soil organic carbon
content from 2% to 4% in five years.
Under dryland conditions typical of the Australian cereal belt, increases in returns of shoot dry matter
of this magnitude are unlikely and thus it is hard to substantiate such changes in C content. However,
in specific locations where rainfall may not be used efficiently to produce agricultural crops/pastures
(particularly regions with significant amounts of summer rainfall and where annual crops are being
produced) significant increases in crop production and residue returns are possible by modifying
existing management practices. Conversion of annual to perennial pastures and altering grazing
practices from set stocking to rotational grazing will enhance plant dry matter production and increase
soil carbon content.
PREDICTING THE INFLUENCE OF MANAGEMENT ON SOIL CARBON
CONTENT
Soil organic carbon content changes very slowly. When this fact is considered along with the annual
variability in rainfall normally experienced at any given location, measurements of soil organic carbon
over several decades may be required to accurately define the effects of particular management
treatments on soil organic carbon contents. We have used a soil carbon model (RothC) to predict the
likely soil organic carbon content that would be obtained under wheat production using average Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
5
climatic conditions and retaining all crop stubble. At Roseworthy the water limited grain yield was
calculated using the French-Schultz approach (slope = 20 kg grain/mm water and slope = 110 mm
water). To define the potential long term soil carbon content (equilibrium soil carbon content), wheat
production was set to 75% of the water limited potential was used along with a harvest index of 0.37
and a root:shoot ratio of 0.43 to calculate the crop residue addition rate including roots. The
equilibrium soil C content predicted for the 0−30 cm layer was 86 tonnes C/ha. It should be noted that
in these modelling analyses a clay content of 15% was used.
In Figure 6 the estimated changes in the amount of soil organic carbon content stored in the 0−30 cm
soil layer is presented for different levels of wheat production defined in terms of water use efficiency
(WUE). The predicted wheat grain yields (for Roseworthy) are given in parentheses after each WUE
value in the legend. The changes in soil carbon associated with a 25 year time frame are given in
Table 1. The model predictions suggest that if we were to move wheat yields to an average of 4.5 t
grain/ha (100% WUE), soil carbon would increase by about 11 tonnes C/ha over the 25 year period,
and by about 32 tonnes C/ha if we could obtain wheat yields equivalent to 150% of current water use
efficiencies. These data show that carbon changes will be slow but enhancing productivity, if it can be
maintained, will result in increased soil carbon levels. Other management scenarios (e.g. conversion
to pasture production) may provide larger increases; however, if such management changes are
made, attempting to claim carbon credits in a carbon trading scheme will limit future options for land
use because the new levels of carbon attained would have to be maintained.
Figure 6 Changes in the amount of carbon stored in the 0−30 cm soil layer at Roseworthy, SA predicted using the
RothC soil carbon cycling model for different water use efficiencies (WUE).
Table 1 Change in soil carbon after 25 years for different levels of wheat productivity
Total amount of carbon stored in the
Wheat grain 0−30 soil layer (t C/ha) yield (t/ha)
Water use efficiency
(% water limited potential) 0 years 25 years Change
1.1 0.25 86 65 -21
2.3 0.50 86 75 -11
3.4 0.75 86 86 0
4.6 1.00 86 97 11
5.7 1.25 86 107 21
6.8 1.50 86 118 32
$$ FROM SEQUESTRATION—FACT OR FICTION?
There is no doubt that soils could potentially hold more carbon. The challenge is to be able to do this
while still maintaining an economically viable farm enterprise. Some potential options include:
• enhancing the proportion of perennial vegetation in pastures or conversion of portions of
cropped paddocks that continually give negative returns to perennial vegetation; Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
6
• increased retention of crop residues, reduced stocking rates and increased use of green
manure crops to return more plant material to the soil;
• optimise farm management inputs to maximise water use efficiency and thus maximise the
return of crop residues to soil but be careful not to generate other greenhouse gases in the
process which may offset any benefits.
At a price of < $20 per tonne of sequestered carbon and the slow potential rates of soil carbon
change, it will be hard to economically justify modifying management practices for the purpose of
selling carbon credits alone. Under such pricing, carbon credits should be considered as a secondary
benefit that may be realised whilst attempting to enhance soil productivity by building soil carbon
content.
Reviewed by: Bill Bowden Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
7
Fact or Fiction: Who is telling the truth and how to
tell the difference?
Doug Edmeades, agKnowledge Ltd, PO Box 9147, Hamilton
AIMS
In most farm budgets fertiliser is normally the largest item of discretionary expenditure. How a farmer
spends the fertiliser dollar can and does have a major impact on the financial bottom line. The fertiliser
industry is also large in terms of total revenues and profits and hence there is strong motivation to ‘get
a piece of the action’. There are now many players in the fertiliser market.
We also live at a time when the dominant political philosophy is ‘laissez faire’. People in western
democracies want their governments at arms length and preferably not interfering with their desire to
make a buck. Thus, were possible government rules and regulations are abandoned in favour of
‘caveate emptor’—let the buyer beware! There are no rules to control the behaviour of the various
players.
Making matters worse, at least for the farmer, science itself is under threat. There was a time when
science was the authority and that authority was based on evidence. Truth was defined by the balance
of the evidence. This has been eroded by post-modern philosophy: now the truth is defined by
opinion—what you feel is your truth. And, importantly Political Correctness demands that all opinions
must be given equal weight, irrespective of the balance of the evidence. It is this environment which
nurtures and encourages belief in things like organic farming and homeopathy which are not evidence
based but belief based. They both depend on dogma.
The consequence of all these modern forces is that farmers today are inundated with information,
much of it unsolicited, contradictory and of dubious quality. Not surprising farmers are very confused.
Who do they believe and who can they turn to?
In this talk I want to give you some tools that I hope will enable you to tell fact from fiction, and to
lessen the risks of legal action by those who may feel threatened by what I have to say I make my
motivation clear.
“Those who are fortunate enough to have chosen science as a career have an obligation to
inform the public about voodoo science.” Robert Park
“The special responsibility of scientists is to inform the world of its choices.” Robert Park
WHAT IS SCIENCE?
There are two common questions I get from the public about science and they indicate to me that we
(i.e. scientists) must do more to enhance science literacy in society.
1) Scientists are always arguing—who am I to believe? This is normal, healthy and essential for
science to progress. Science is about testing ideas against the evidence and scientist must
debate and argue and test again. As more and more evidence come to light we can have more
and more confidence that we a getting nearer the truth. This is best seen in hindsight. For
example we all now agree that the sun is the centre of the solar system, that the earth is not flat
and that atoms are not solid. But these were matters of public debate in their time resolved only
by getting enough evidence. A problem for the public arises when new areas of science emerge
and the subsequent scientific arguments spill into the public arena (e.g. climate change, stem
cell research).
2) “If science is so good how come scientists do not know everything?” Science will never know
everything for the simple reason that everything question has not been asked and every
conceivable experiment has not been done. Science evolves raising new questions and new
techniques develop so that new types of measurements can be made the important point is that
the more mature the science the more confident we can be. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
8
3) “You have not tested product ‘A’ so how can you say it does not work?” As science develops
theories are formulated and then tested and if they stand the test of time (i.e. more and more
evidence) we say we have a law. Some common ‘laws’ which are useful in the science of
fertilisers are: a) Liebigs Law of the minimum; b) the principle of cause and effect; and c) there
are 16 nutrients required for plant growth. By applying such laws to a given product we can
deduce whether a product will be effective of otherwise.
By making use of our knowledge of science—what it is, how it should be conducted and what scientific
laws to apply—we can construct a list of tests that can be applied to information about fertiliser
products to help us decide what weight to place on the available evidence.
TESTS FOR SCIENCE?
Test 1 (Plausibility Test)
In this test we apply the Principle of Cause and Effect. Things do not happen by chance. If there is an
effect there must be a cause. This universal principle applies also to nature and hence to soils and
plants. Related to this we must ask the question: what is the mechanism by which this product works,
or is claimed to work, and is it plausible? Be very cautious if the mechanism claimed for the product
defies a well established principle of science.
Test 2 (Credibility Test)
Examine the advertising and promotional information you are given about a product or service. If you
detect one or a combination of the following, the product or service is not likely to be credible.
a) Is the product/service promoted on the basis of a doomsday message? “We are ruining our
soils, polluting our water, poisoning our stock, endangering human health.”
b) Does the company literature suggest a conspiracy? “You cannot trust the Universities or the
Department of Agriculture—they are in the pocket of the big fertiliser companies.”
c) Is the product/service promoted solely on testimonials?
d) Is the product/service promoted because it is natural or a very old practice only recently
rediscovered?
e) Is the product/service so new and exciting that it is ahead of science or beyond science or
requires a new paradigm of understanding?
f) Is the product/service developed by a lone genius, overlooked by science?
Test 3 (Evidence test)
This test is in essence the ‘acid test’—where is the evidence?
a) What are the specific claims made for the product/service?
b) Beware of products/services for which very general non-specific claims are made.
c) Beware of products/service which make multiple claims.
d) Where is the evidence for the claim(s)? Is it published in a reputable peer reviewed science
publication?
e) If it is not published in the scientific literature ask who conducted the research. Is there a conflict
of interest? Were the trials properly designed and conducted? Get it checked by an independent
scientist.
f) Is there supporting evidence for the product/service such as other trials by other independent
agencies in other countries?
g) Ignore anecdotal evidence (testimonials).
h) Soils vary—a product may work in some situations and not in others!
“The only antidote to pseudo science is science itself.” Carl Sagan Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
9
Test 4 (Common Sense test)
Use you common sense when the salesman calls. Ask the obvious question: If what you are told is
true and it is indeed a good product/service and the claimed benefits are true then every farmer would
be using the product and service? Apply Test 2 as you listen to the answer.
“If it sounds too good to be true it probably is.” Dr J Roche
Test 5 (Reality Test)
Many products and services are sold on basis that we are destroying our planet, our soils and our
health. Many today believe that science is the cause of these dilemmas. So let us remind ourselves
how successful science and its close cousin technology have been. We live longer now than at any
time in our history, we grow more food than at any time on our history and our food is abundant and
healthy. This is clear evidence not of destruction but of science and its success.
Paper reviewed by: Bill Bowden Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
10
Four decades of crop sequence trials in Western
Australia
Mark Seymour, Department of Agriculture and Food, Western Australia
KEY MESSAGES
A database that collates crop sequence trial data from over 160 trials is now available.
On average, wheat after lupin out yielded wheat after wheat by 0.5 t/ha. Greater increases in wheat
yields following lupin occurred after the availability and widespread use of good, selective herbicides
coinciding with the uptake of no-till seeding in the 1990s.
AIMS
Collate all of the available trial results from crop sequence experiments conducted in WA and use the
database to quantify the rotational benefits of break crops. In addition determine the need for crop
sequence research in the future.
METHOD
Over 150 crop sequence experiments have been conducted throughout WA since the 1960s to
determine the rotation effects of leguminous or oilseed crops in cereal based rotations. The vast
majority of these experiments have been conducted by the DAFWA. The results of these experiments
and the limited number conducted by other organisations are available in various formats, but have
never been collated in the one place in a uniform way. This paper briefly describes the production of a
Microsoft Access database that collates the available information and provides a summary of break
crop effects in WA.
The database currently holds 10 191 records consisting of trial x year x this year’s crop x previous
crop(s) x nitrogen combinations. The results of 167 trials appear in the database, around 165 are
DAFWA experiments, one trial had CSIRO as the lead agency with DAFWA input and the remaining
experiment is the rotation trial run by the Facey Group at Wickepin.
The database is available for DAFWA staff at \\Agessrdc01\users\Seymour\Break crop rotation
database. People outside of DAFWA should contact the Mark Seymour (mseymour@agric.wa.gov.au)
for copies.
Notation for crop sequences and rotations used in this paper and on occasions in the database are as
follows:
• Abbreviations for major crops are—wheat (W), barley (B), canola (N), lupin (L), field pea (Fp),
linseed (Li), oats (O), fallow (Fa), vetch (V), chickpea (K), faba bean (H), and mustard (Mu).
• Crop sequences are listed in order, e.g. LWW refers to lupin followed by wheat followed by
wheat.
• Reference to the particular part or year of the crop sequence uses the notation/n. For example,
for a LWW sequence LWW/1 refers to the first crop, lupin. LWW/2 refers to the first wheat after
lupin and LWW/3 refers to the third crop, which in this case is the second wheat after lupin.
The base data
Fields that appear in the main database include trial information such as trial number, major personnel
involved, site (farmer’s name), location (nearest town), agzone, soil type, year(s) of experiment, the
current year’s crop and sometimes which variety was used, nitrogen application rate (kg N/ha), details
of the previous 6 crops if available, some coding for rotation types and phase (incomplete), general
comments, and some brief information on other treatment applied such as: ripping, fertiliser, time of
sowing. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
11
Crop traits in the database include grain yield, grain yield of previous crop if available, dry matter—
usually peak or harvest biomass (noted if otherwise) and grain protein. Plants include: barley, canola,
cereal rye, chickpea, faba bean, fallow, field pea, lentil, lathyrus, linseed, narrow-leafed lupin, albus
lupin, yellow lupin, oats, serradella, sub. clover, medic, volunteer pasture, summer crops, triticale,
vetch and wheat. Distinctions are made between harvest, green or brown manured, ploughed in, not
harvested or stubble removed treatments, mixes of species and other variations.
Additional information linked to the database include rainfall records for the nearest meteorological
station to the experiment from which annual rainfall, growing season rainfall (May to October) and
stored water have been calculated. Stored water is estimated by using the formula: 10% of the
previous November and December rainfall plus 20% of January and February rainfall, plus 55% of
March rainfall plus 75% of April rainfall. Total water available to the plant (mm) was then estimated as
rain falling during the growing season plus stored water.
RESULTS
As part of the GRDC project “Increasing the Profitability of Cropping Systems in Western Australia
using Lupins, Oats, Oilseeds and Pulses” a detailed report is being prepared which summarises some
of the database results. An extract of the section of this report that deals with narrow-leafed lupin is
given below.
Wheat after narrow-leafed lupin
Narrow-leafed lupin has been the most widely examined break crop species with over 150 trials x year
combinations available in the database. If we look at the raw data from all of the trials (Figure 1) we
can see the range of yields obtained in the trials. The majority of wheat on wheat (WW/2) yields are
less than 2.5 t/ha indicating that in the trials conducted to date it has been difficult to achieve yields
higher than 2.5 t/ha with wheat sown after wheat.
In general it is also noticeable that the majority of wheat after lupin responses above the 1:2 ratio line
occur when wheat on wheat yields are below 1.5 t/ha, indicating an agronomic issue with wheat-wheat
which the inclusion of lupin helps to remediate. Invariably these issues have been identified in
individual trials to be the presence of Take-all or high levels of annual ryegrass or brome grass. The
outlier on the y-axis of yields of lupin-wheat at or above 4 t/ha when wheat-wheat yields less than
1.0 t/ha are from the trial 91KA111 at West Katanning in which Take-all was a factor that severely
limited the yield of wheat on wheat and a wide range of break crops such as lupin, field pea and
canola provided a good break from the disease. Similarly the outlier where WW/2 yields close to zero
and LW/2 yields 2.5 t/ha is from a trial at South Carrabin in 1995 where brome grass became very
difficult to control in the wheat on wheat plots.
Similarly there are occasions where the lupin sequence fails. For example, the outlier on the x-axis
where WW/2 yields 1.8 t/ha and LW/2 yields close to zero are from a trial in 1983 at Nabawa (78C1)
where wild radish was not able to be controlled in the lupin phase and the weeds swamped the
following cereal crop. In later years the availability of diflufenican solved this issue, although in recent
times wild radish has again become harder to control in the lupin year with selective herbicides.
Overall though wheat sown after lupin out yields wheat sown after wheat. A linear relationship can be
fitted to the response of wheat after lupin compared to wheat after wheat over a wide range of wheat
on wheat yields. If the outliers discussed earlier are removed this relationship is: GY of LW/2 = 0.9
(GY of WW/2) + 0.6, r² = 0.58, P < 0.001, GY = grain yield. If we were to constrain the regression
through the origin the regression would become: GY of LW/2 = 1.34(GY of WW/2), r² = 0.45,
P < 0.001. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
12
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Yield of wheat after wheat (WW /2, t/ha)
Yield of wheat after lupin (LW /2, t/ha)
1:2
1:1
Figure 1 Relationship between the yield of wheat sown after wheat (WW/2) and the yield of wheat sown after
lupin (LW/2) in 88 trials (167 trial x year combinations) in experiments conducted throughout WA since 1974.
Linear curves indicate 1:1 and 1:2 ratios.
Another way to look at the data set is to consider the magnitude of the difference in yield (Ydiff)
between WW/2 and LW/2 and the frequency in which various levels of Ydiff occur. In the first instance
we will look at Ydiff averaged across all rates of nitrogen applied to the second year of wheat. Figure 2
shows that whilst there are relatively few instances where Ydiff is more than 1.5 t/ha, in 10% of
instances Ydiff is less than or equal to 0 t/ha, and the distribution is centred around 0−500 kg/ha range
with the mean increase in yield being 540 kg/ha.
4%
5%
24%
20%
12%
15%
9%
4%
8%
0%
5%
10%
15%
20%
25%
30%
< -0.25 -0.25 to 0 0 to 0.25 0.25 to 0.5 0.5 to 0.75 0.75 to 1.0 1.0 to 1.5 1.5 to 2.0 > 2.0
LW /2 - WW /2 (t/ha)
Relative frequency (%)
Figure 2 Relative frequency (%) in which the difference in yield (Y diff, t/ha) between wheat following lupin (LW/2)
and wheat following wheat (WW/2) falls into 9 yield categories. Data are from 167 trials x year combinations in 86
trials conducted in WA since 1974.
As seasons influence the magnitude of any break effect it can be useful to compare the upper limit of
the water use efficiency of the different rotations. To do this we calculated modified French and Shultz
figures for the two rotations. We then fitted by eye a boundary line encompassing most of the data
points (data not shown). Using this method the potential water use efficiency for wheat after lupin was
19 kg/ha/mm and was 15 kg/ha/mm for wheat after wheat. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
13
Response to nitrogen
It is widely known that one of the major effects of lupin will be the residual nitrogen they supply to the
following crop. Results discussed so far have been averaged across all the nitrogen fertiliser rates
applied to the following wheat crop. In order to evaluate the effect of residual nitrogen on Ydiff we first
grouped rates of applied nitrogen fertiliser into five groups labelled 0N, 25N, 50N, 100N, 150N, where
0N = all treatments where no nitrogen fertiliser was applied, 25N = where up to 25 kg N/ha was
applied, 50N = 25 to 50 kg N/ha, 100N = 50 to 100 kg N/ha, and 150N = more than 100 kg N/ha. We
then restricted the dataset to the 31 trials that included at least four of these five groups so that n = 67
for all N groups except 150N which had 44 observations. Residual maximum likelihood (REML)
models were then fitted using Genstat 10 with N group as the fixed effect and Trial.Year as the
random effect.
Overall nitrogen applied as fertiliser had a significant (P < 0.001) but small effect on Ydiff (data not
shown). The largest Ydiff was 556 kg/ha and occurred when no nitrogen fertiliser was applied. Ydiff
decreased as the rate of nitrogen fertiliser increased so, the highest nitrogen fertiliser group (150N)
produced a Ydiff of 396 kg/ha.
Do changes over time affect the response to nitrogen?
The difference in yield between wheat after wheat and wheat after lupin appears to change over time
with a gradual rising trend from 1974 up to 1990 when the difference in yield between LW/2 and WW/2
increases dramatically and then drops off again after 1993 (Figure 3).
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Year
Ydiff (LW /2 - WW /2, t/ha)
Figure 3 The difference in yield between wheat after wheat (WW/2) and wheat after lupin (LW/2) over time.
To investigate this observation further we restricted the data set to the years when the most number of
lupin Agzones had trials, which was the period 1983 to 1995. This showed a relatively flat period from
1983 to 1991 and then an increase in the period following. We then considered the changes in the
1990s that led to this unprecedented increase in the difference in yield between LW/2 and WW/2. Was
it environmental, such that we had a run of years that suited wheat after lupin more so than wheat
after wheat? Or were there changes in agronomic practices that were of benefit to wheat after lupin or
made lupin a better break crop?
To separate the effect of rainfall from management we compared the water use efficiency of the two
sequences (data not shown). This showed that the difference between the WUE of LW/2 and WW/2
was, for the first time, consistently above 3 kg/ha/mm from 1990. Around that period of time there was
a shift to no-till machinery both on farms and for experimental purposes. There was also a wider use of
more effective herbicides for in-crop control of grass and broadleaf weeds in lupin crops, and rotations
shifted to more continuous cropping as sheep numbers declined throughout WA. In general,
comments from trials in the period 1990−95 indicated that the lupin plots were generally free from
weeds and there were few reports of poor lupin growth in the trials. Thus these changes seemed to be
of benefit both for the lupin crop and the following cereal. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
14
This is further demonstrated if we group the data into two periods: ‘1983−89’ and ‘1990−95’. The Ydiff
was 0.4 t/ha in the 1983−89 period and 0.9 t/ha 1990−95 period (P < 0.001) and this difference
remained even when the peak year of 1993 was removed from the analysis. Of interest then was to
see if the agronomic changes also changed the response to fertiliser nitrogen. To do this we had to
further reduce the dataset because there were few trials that included rates of fertiliser nitrogen above
100 kg/ha (n = 6).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 25 50 100
Nitrogen group (kg N/ha)
Ydiff (LW /2 - WW /2, t/ha)
1983-89 1990-95
Figure 4 Response of Ydiff (LW/2−WW/2 in t/ha) to nitrogen fertiliser application in the wheat year for the periods
prior to and after 1990.
Prior to 1990, as the rate of nitrogen increased the difference in yield between LW/2 and WW/2
decreased (Figure ). However, since 1990 nitrogen has no effect (P > 0.05) on the difference in yield.
It appears that since 1990 wheat after lupin continues to respond to increasing rates of nitrogen
whereas in the previous period wheat after lupin did not respond to increasing rates of nitrogen whilst
wheat on wheat did.
CONCLUSION
A database has been collated from all of the available crop sequence experiments conducted in WA.
Over 10 000 records representing the results of over 160 experiments conducted since 1966 appear in
the database, allowing for rigorous interrogation of rotation effects over a long period of time. In the
experiments conducted to date continuous wheat was rarely as productive or economically viable as
rotations that included either a pasture or break crop, regardless of amount of nitrogen fertiliser
applied.
In general terms, since 1990, both the yield of wheat on wheat and the likelihood of a response to
lupin in the following year have increased at all levels of applied nitrogen. This corresponds to a period
where more effective herbicides were used, rotations shifted to more continuous cropping and trials
were more likely to be sown with no-till machinery
If changes to crop management in the past have influenced the size of the break crop effect then we
need to consider the implications of even more recent changes to crop management. In particular the
benefits of break crops are likely to be influenced by the modern use of more effective fungicides,
inter-row seeding to avoid last years crowns and roots, metering out of nutrition throughout the
growing season, and the reduced effectiveness of weed control.
KEY WORDS
crop sequence, break crop, lupin Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
15
ACKNOWLEDGMENTS
Thanks to: Pam Burgess for helping to collate the data, Andrew Van Burgel for statistical advice, and
all the previous researchers who provided the data. Funds for the work are provided by DAFWA and
GRDC.
Project No.: DAW161
Paper reviewed by: Peter White
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and the Grains Research & Development Corporation 1
http://archive.agric.wa.gov.au/OBJTWR/imported_assets/content/fcp/cu09_plenary.pdf
Building soil carbon for productivity and
implications for carbon accounting
Jeff Baldock, CSIRO Land and Water, Adelaide, SA
KEY MESSAGES
Composition of soil organic carbon
• Soil organic carbon is composed of a wide range of different materials with different chemical
and physical properties and different extents of decomposition.
Functions of organic carbon/organic matter in soil
• Soil organic matter contributes to a variety of biological, chemical and physical properties of
soils.
• Chemical—cation exchange, pH buffering, reduces effects of sodicity.
• Physical—water retention, soil structural stability, soil wettability, soil temperature.
• Biological—energy for microbes, provision of nutrients and resiliency.
• Increasing soil organic carbon content can result in an increase in the ability of a soil to hold
water and thereby lead to enhanced productivity.
Calculating changes in soil organic carbon content
• Soil carbon content represents the balance between inputs and outputs.
• Values are required for the depth, bulk density and carbon content of the soil layer you are
interested in to determine how much carbon is present.
• Changes in soil carbon content are slow and typically require at least 5 years to be detectable.
• Simulation models can be used to predict the likely outcomes of management practices on soil
carbon content.
$$ from sequestration—fact or fiction?
• Maximising crop productivity will maximise carbon inputs and soil organic carbon content.
• At current prices, it is hard to justify modifying management practices for the sole purpose of
selling carbon credits.
SOIL ORGANIC CARBON: WHAT IS IT?
Soil organic carbon is a complex and heterogeneous mixture of materials. These materials vary in
their physical size, chemical composition, degree of interaction with soil minerals and extent of
decomposition. Although determining the impact of management practices on soil organic carbon
contents is important, it does not tell us anything about the type of organic carbon present. For
example, is the organic carbon dominated by pieces of plant residue or more recalcitrant charcoal? It
is therefore important to determine the composition of soil organic carbon to gain an appreciation for
the implications of management practices and changes in organic carbon content on soil productivity.
We now recognise four different types of soil organic carbon:
• Crop residues—shoot and root residues > 2 mm residing on and in soil.
• Particulate organic carbon—individual pieces of plant debris that are smaller than 2 mm but
larger than 0.053 mm.
• Humus—decomposed materials less than 0.053 mm that are dominated by molecules stuck to
soil minerals.
• Recalcitrant organic carbon—dominated by pieces of charcoal.
FUNCTIONS OF ORGANIC CARBON/ORGANIC MATTER IN SOIL
Organic carbon/organic matter contributes to a variety of functions in soils. These functions can be
broadly classified into three types: biological, chemical and physical (Figure 1). Strong interactions
(represented by the grey arrows) often exist between these different functions. For example, the Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
2
biological function of providing energy that drives microbial activity also results in improved structural
stability and creates organic materials that can contribute to cation exchange and pH buffering.
Figure 1 Functions performed by organic matter present in soils.
Future predictions of climate change suggest that most regions of Australia will be become warmer
and drier based on increases in the concentrations of greenhouse gases (carbon dioxide, methane
and nitrous oxide) in the atmosphere. Under such conditions it will become even more important than
it currently is to be able to maximise the storage of plant available water in soils. The amount of plant
available water a soil can hold is determined by two parameters:
1. the lower limit—the amount of water in the soil that plants cannot extract; and
2. the upper drained limit—the amount of water that can be held against drainage.
The difference between the upper and lower limit defines the potential available water holding capacity
(PAWC) of a soil. If this value can be increased even marginally it will help to maintain or enhance
potential productivity by allowing the soil to retain more water each time it rains.
In the absence of subsoil constraints such as salinity, the lower limit is defined by soil clay content and
increases with increasing clay content (Figure 2). Evidence from WA sands suggests that increases in
organic carbon content can also increase the amount of water present at the lower limit. The upper
limit is also affected by the content of clay and organic carbon clay (Figure 2). For any given clay
content, as organic carbon increases the upper limit, and therefore PAWC, of the soil increases. An
analysis of the influence of increasing soil organic carbon content by 1% of total soil mass (e.g. from
0.7% to 1.7%) on the soil PAWC was completed using data collected from red brown earths of the
mid-north region of SA. This analysis indicated that such an increase in carbon content in the surface
0−10 cm soil layer would increase PAWC from its original value by 2 to 4 mm with the effect
diminishing as soil clay content increased (Figure 3). Although the change in PAWC for sandy-low clay
content soils is predicted to be larger than for clay soils, it would be much more difficult to increase soil
carbon content on sands relative to clays.
Figure 2 Changes in upper and lower limits of soil water content with changes in clay content. The light grey area
defines the plant available water holding capacity of the soil. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
3
Figure 3 Change in soil water holding capacity (WHC) induced by increasing soil organic carbon content by 1% of
total soil mass for red-brown earths of the mid-north of SA.
WHAT DETERMINES SOIL ORGANIC CARBON CONTENT?
The amount of carbon in a soil results from the balance between inputs (plant residues) and losses
(microbial decomposition and associated mineralisation). Figure 4 the bucket represents the amount of
carbon a soil could potentially hold. This amount will vary with factors such as soil clay content, soil
depth, and bulk density and is not influenced by management. The bucket will be smaller for a sand
than a clay soil.
Figure 4 Inputs and losses define soil organic carbon content.
To increase the content of organic carbon in soil, the input of residues must be increased and/or the
rates of loss of carbon must be decreased.
Inputs are controlled by the type and amount of plant residue added to the soil. Any practice that
enhances productivity and the return of plant residues (shoots and roots) to the soil opens the input
tap and will result in an increase in soil carbon. For example, appropriate use of fertilisers to maximise
productivity also maximise returns of organic residues to the soil. However, an upper limit to the input
of residues exists in Australian dryland agriculture because of the limitation that the availability of
water places on potential plant productivity. Maximum soil carbon contents will be obtained for any
management system if productivity and capture of carbon are maximised by attaining 100% water use
and nutrient use efficiencies.
Losses of carbon from soil result from decomposition and conversion of carbon in plant residues and
soil organic materials into carbon dioxide. Processes that accelerate decomposition open the losses
tap further; while those that slow down the rate of decomposition will close the losses tap and help to
maintain or increase soil carbon. Reducing the extent of cultivation has been suggested to result in
enhanced soil carbon levels. Results from Australian studies suggest that while this may be true for
some soils, for other soils such an effect has not been observed (Figure 5). It should also be noted
that even shifting to a zero tillage system is unlikely to result in soil carbon values equivalent to those
that would be obtained under a pasture system. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
4
CALCULATING CHANGES IN SOIL ORGANIC CARBON CONTENT
The amount of organic carbon found in a soil can be calculated using values for the depth (cm) of the
soil layer of interest, the soil bulk density (g/cm3
) and the soil carbon content (%) (Equation [1]). Using
Equation [1], indicates that a 30 cm layer of soil having a bulk density of 1.2 g/cm3
and a carbon
content of 1.2% contains approximately 43 Tonnes of C/ha.
3 Organic C (T C/ha) Depth (cm) Bulk density (g/cm ) Carbon content (%) =× × [1]
Figure 5 Change in soil carbon induced by different levels of tillage for several different Australian soil types. The
number of samples of each type of soil are given by n.
Suggestions have been put forward that altering management practices can increase soil organic
carbon content from 2% to 4% in 5 years. Is this really possible?
If we use the same bulk density as above (1.2 g/cm3
) and restrict our calculations to the top 10 cm of
soil where organic carbon is most easily increased, at 2% carbon the soil would contain 24 tonnes
C/ha. At 4% carbon the same soil layer would contain 48 tonnes C/ha. This indicates that 24 tonnes of
C/ha would have to be added to the soil. Since plant residues contain approximately 45%C this would
equate roughly to 50 tonnes/ha of dry matter (DM). If this increase was to occur over 5 years, then an
additional 10 tonnes DM/ha/year above that currently being added would be required if no
decomposition occurs. If half of the residues added were in the form of roots below ground, then we
would have to add an additional 5 tonnes shoot DM/ha/year. Since we know that at least 50% of the
added plant residues will decompose, annual additions of approximately 10 tonnes shoot DM/ha/year
above that currently being added would be required to achieve an increase in soil organic carbon
content from 2% to 4% in five years.
Under dryland conditions typical of the Australian cereal belt, increases in returns of shoot dry matter
of this magnitude are unlikely and thus it is hard to substantiate such changes in C content. However,
in specific locations where rainfall may not be used efficiently to produce agricultural crops/pastures
(particularly regions with significant amounts of summer rainfall and where annual crops are being
produced) significant increases in crop production and residue returns are possible by modifying
existing management practices. Conversion of annual to perennial pastures and altering grazing
practices from set stocking to rotational grazing will enhance plant dry matter production and increase
soil carbon content.
PREDICTING THE INFLUENCE OF MANAGEMENT ON SOIL CARBON
CONTENT
Soil organic carbon content changes very slowly. When this fact is considered along with the annual
variability in rainfall normally experienced at any given location, measurements of soil organic carbon
over several decades may be required to accurately define the effects of particular management
treatments on soil organic carbon contents. We have used a soil carbon model (RothC) to predict the
likely soil organic carbon content that would be obtained under wheat production using average Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
5
climatic conditions and retaining all crop stubble. At Roseworthy the water limited grain yield was
calculated using the French-Schultz approach (slope = 20 kg grain/mm water and slope = 110 mm
water). To define the potential long term soil carbon content (equilibrium soil carbon content), wheat
production was set to 75% of the water limited potential was used along with a harvest index of 0.37
and a root:shoot ratio of 0.43 to calculate the crop residue addition rate including roots. The
equilibrium soil C content predicted for the 0−30 cm layer was 86 tonnes C/ha. It should be noted that
in these modelling analyses a clay content of 15% was used.
In Figure 6 the estimated changes in the amount of soil organic carbon content stored in the 0−30 cm
soil layer is presented for different levels of wheat production defined in terms of water use efficiency
(WUE). The predicted wheat grain yields (for Roseworthy) are given in parentheses after each WUE
value in the legend. The changes in soil carbon associated with a 25 year time frame are given in
Table 1. The model predictions suggest that if we were to move wheat yields to an average of 4.5 t
grain/ha (100% WUE), soil carbon would increase by about 11 tonnes C/ha over the 25 year period,
and by about 32 tonnes C/ha if we could obtain wheat yields equivalent to 150% of current water use
efficiencies. These data show that carbon changes will be slow but enhancing productivity, if it can be
maintained, will result in increased soil carbon levels. Other management scenarios (e.g. conversion
to pasture production) may provide larger increases; however, if such management changes are
made, attempting to claim carbon credits in a carbon trading scheme will limit future options for land
use because the new levels of carbon attained would have to be maintained.
Figure 6 Changes in the amount of carbon stored in the 0−30 cm soil layer at Roseworthy, SA predicted using the
RothC soil carbon cycling model for different water use efficiencies (WUE).
Table 1 Change in soil carbon after 25 years for different levels of wheat productivity
Total amount of carbon stored in the
Wheat grain 0−30 soil layer (t C/ha) yield (t/ha)
Water use efficiency
(% water limited potential) 0 years 25 years Change
1.1 0.25 86 65 -21
2.3 0.50 86 75 -11
3.4 0.75 86 86 0
4.6 1.00 86 97 11
5.7 1.25 86 107 21
6.8 1.50 86 118 32
$$ FROM SEQUESTRATION—FACT OR FICTION?
There is no doubt that soils could potentially hold more carbon. The challenge is to be able to do this
while still maintaining an economically viable farm enterprise. Some potential options include:
• enhancing the proportion of perennial vegetation in pastures or conversion of portions of
cropped paddocks that continually give negative returns to perennial vegetation; Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
6
• increased retention of crop residues, reduced stocking rates and increased use of green
manure crops to return more plant material to the soil;
• optimise farm management inputs to maximise water use efficiency and thus maximise the
return of crop residues to soil but be careful not to generate other greenhouse gases in the
process which may offset any benefits.
At a price of < $20 per tonne of sequestered carbon and the slow potential rates of soil carbon
change, it will be hard to economically justify modifying management practices for the purpose of
selling carbon credits alone. Under such pricing, carbon credits should be considered as a secondary
benefit that may be realised whilst attempting to enhance soil productivity by building soil carbon
content.
Reviewed by: Bill Bowden Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
7
Fact or Fiction: Who is telling the truth and how to
tell the difference?
Doug Edmeades, agKnowledge Ltd, PO Box 9147, Hamilton
AIMS
In most farm budgets fertiliser is normally the largest item of discretionary expenditure. How a farmer
spends the fertiliser dollar can and does have a major impact on the financial bottom line. The fertiliser
industry is also large in terms of total revenues and profits and hence there is strong motivation to ‘get
a piece of the action’. There are now many players in the fertiliser market.
We also live at a time when the dominant political philosophy is ‘laissez faire’. People in western
democracies want their governments at arms length and preferably not interfering with their desire to
make a buck. Thus, were possible government rules and regulations are abandoned in favour of
‘caveate emptor’—let the buyer beware! There are no rules to control the behaviour of the various
players.
Making matters worse, at least for the farmer, science itself is under threat. There was a time when
science was the authority and that authority was based on evidence. Truth was defined by the balance
of the evidence. This has been eroded by post-modern philosophy: now the truth is defined by
opinion—what you feel is your truth. And, importantly Political Correctness demands that all opinions
must be given equal weight, irrespective of the balance of the evidence. It is this environment which
nurtures and encourages belief in things like organic farming and homeopathy which are not evidence
based but belief based. They both depend on dogma.
The consequence of all these modern forces is that farmers today are inundated with information,
much of it unsolicited, contradictory and of dubious quality. Not surprising farmers are very confused.
Who do they believe and who can they turn to?
In this talk I want to give you some tools that I hope will enable you to tell fact from fiction, and to
lessen the risks of legal action by those who may feel threatened by what I have to say I make my
motivation clear.
“Those who are fortunate enough to have chosen science as a career have an obligation to
inform the public about voodoo science.” Robert Park
“The special responsibility of scientists is to inform the world of its choices.” Robert Park
WHAT IS SCIENCE?
There are two common questions I get from the public about science and they indicate to me that we
(i.e. scientists) must do more to enhance science literacy in society.
1) Scientists are always arguing—who am I to believe? This is normal, healthy and essential for
science to progress. Science is about testing ideas against the evidence and scientist must
debate and argue and test again. As more and more evidence come to light we can have more
and more confidence that we a getting nearer the truth. This is best seen in hindsight. For
example we all now agree that the sun is the centre of the solar system, that the earth is not flat
and that atoms are not solid. But these were matters of public debate in their time resolved only
by getting enough evidence. A problem for the public arises when new areas of science emerge
and the subsequent scientific arguments spill into the public arena (e.g. climate change, stem
cell research).
2) “If science is so good how come scientists do not know everything?” Science will never know
everything for the simple reason that everything question has not been asked and every
conceivable experiment has not been done. Science evolves raising new questions and new
techniques develop so that new types of measurements can be made the important point is that
the more mature the science the more confident we can be. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
8
3) “You have not tested product ‘A’ so how can you say it does not work?” As science develops
theories are formulated and then tested and if they stand the test of time (i.e. more and more
evidence) we say we have a law. Some common ‘laws’ which are useful in the science of
fertilisers are: a) Liebigs Law of the minimum; b) the principle of cause and effect; and c) there
are 16 nutrients required for plant growth. By applying such laws to a given product we can
deduce whether a product will be effective of otherwise.
By making use of our knowledge of science—what it is, how it should be conducted and what scientific
laws to apply—we can construct a list of tests that can be applied to information about fertiliser
products to help us decide what weight to place on the available evidence.
TESTS FOR SCIENCE?
Test 1 (Plausibility Test)
In this test we apply the Principle of Cause and Effect. Things do not happen by chance. If there is an
effect there must be a cause. This universal principle applies also to nature and hence to soils and
plants. Related to this we must ask the question: what is the mechanism by which this product works,
or is claimed to work, and is it plausible? Be very cautious if the mechanism claimed for the product
defies a well established principle of science.
Test 2 (Credibility Test)
Examine the advertising and promotional information you are given about a product or service. If you
detect one or a combination of the following, the product or service is not likely to be credible.
a) Is the product/service promoted on the basis of a doomsday message? “We are ruining our
soils, polluting our water, poisoning our stock, endangering human health.”
b) Does the company literature suggest a conspiracy? “You cannot trust the Universities or the
Department of Agriculture—they are in the pocket of the big fertiliser companies.”
c) Is the product/service promoted solely on testimonials?
d) Is the product/service promoted because it is natural or a very old practice only recently
rediscovered?
e) Is the product/service so new and exciting that it is ahead of science or beyond science or
requires a new paradigm of understanding?
f) Is the product/service developed by a lone genius, overlooked by science?
Test 3 (Evidence test)
This test is in essence the ‘acid test’—where is the evidence?
a) What are the specific claims made for the product/service?
b) Beware of products/services for which very general non-specific claims are made.
c) Beware of products/service which make multiple claims.
d) Where is the evidence for the claim(s)? Is it published in a reputable peer reviewed science
publication?
e) If it is not published in the scientific literature ask who conducted the research. Is there a conflict
of interest? Were the trials properly designed and conducted? Get it checked by an independent
scientist.
f) Is there supporting evidence for the product/service such as other trials by other independent
agencies in other countries?
g) Ignore anecdotal evidence (testimonials).
h) Soils vary—a product may work in some situations and not in others!
“The only antidote to pseudo science is science itself.” Carl Sagan Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
9
Test 4 (Common Sense test)
Use you common sense when the salesman calls. Ask the obvious question: If what you are told is
true and it is indeed a good product/service and the claimed benefits are true then every farmer would
be using the product and service? Apply Test 2 as you listen to the answer.
“If it sounds too good to be true it probably is.” Dr J Roche
Test 5 (Reality Test)
Many products and services are sold on basis that we are destroying our planet, our soils and our
health. Many today believe that science is the cause of these dilemmas. So let us remind ourselves
how successful science and its close cousin technology have been. We live longer now than at any
time in our history, we grow more food than at any time on our history and our food is abundant and
healthy. This is clear evidence not of destruction but of science and its success.
Paper reviewed by: Bill Bowden Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
10
Four decades of crop sequence trials in Western
Australia
Mark Seymour, Department of Agriculture and Food, Western Australia
KEY MESSAGES
A database that collates crop sequence trial data from over 160 trials is now available.
On average, wheat after lupin out yielded wheat after wheat by 0.5 t/ha. Greater increases in wheat
yields following lupin occurred after the availability and widespread use of good, selective herbicides
coinciding with the uptake of no-till seeding in the 1990s.
AIMS
Collate all of the available trial results from crop sequence experiments conducted in WA and use the
database to quantify the rotational benefits of break crops. In addition determine the need for crop
sequence research in the future.
METHOD
Over 150 crop sequence experiments have been conducted throughout WA since the 1960s to
determine the rotation effects of leguminous or oilseed crops in cereal based rotations. The vast
majority of these experiments have been conducted by the DAFWA. The results of these experiments
and the limited number conducted by other organisations are available in various formats, but have
never been collated in the one place in a uniform way. This paper briefly describes the production of a
Microsoft Access database that collates the available information and provides a summary of break
crop effects in WA.
The database currently holds 10 191 records consisting of trial x year x this year’s crop x previous
crop(s) x nitrogen combinations. The results of 167 trials appear in the database, around 165 are
DAFWA experiments, one trial had CSIRO as the lead agency with DAFWA input and the remaining
experiment is the rotation trial run by the Facey Group at Wickepin.
The database is available for DAFWA staff at \\Agessrdc01\users\Seymour\Break crop rotation
database. People outside of DAFWA should contact the Mark Seymour (mseymour@agric.wa.gov.au)
for copies.
Notation for crop sequences and rotations used in this paper and on occasions in the database are as
follows:
• Abbreviations for major crops are—wheat (W), barley (B), canola (N), lupin (L), field pea (Fp),
linseed (Li), oats (O), fallow (Fa), vetch (V), chickpea (K), faba bean (H), and mustard (Mu).
• Crop sequences are listed in order, e.g. LWW refers to lupin followed by wheat followed by
wheat.
• Reference to the particular part or year of the crop sequence uses the notation/n. For example,
for a LWW sequence LWW/1 refers to the first crop, lupin. LWW/2 refers to the first wheat after
lupin and LWW/3 refers to the third crop, which in this case is the second wheat after lupin.
The base data
Fields that appear in the main database include trial information such as trial number, major personnel
involved, site (farmer’s name), location (nearest town), agzone, soil type, year(s) of experiment, the
current year’s crop and sometimes which variety was used, nitrogen application rate (kg N/ha), details
of the previous 6 crops if available, some coding for rotation types and phase (incomplete), general
comments, and some brief information on other treatment applied such as: ripping, fertiliser, time of
sowing. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
11
Crop traits in the database include grain yield, grain yield of previous crop if available, dry matter—
usually peak or harvest biomass (noted if otherwise) and grain protein. Plants include: barley, canola,
cereal rye, chickpea, faba bean, fallow, field pea, lentil, lathyrus, linseed, narrow-leafed lupin, albus
lupin, yellow lupin, oats, serradella, sub. clover, medic, volunteer pasture, summer crops, triticale,
vetch and wheat. Distinctions are made between harvest, green or brown manured, ploughed in, not
harvested or stubble removed treatments, mixes of species and other variations.
Additional information linked to the database include rainfall records for the nearest meteorological
station to the experiment from which annual rainfall, growing season rainfall (May to October) and
stored water have been calculated. Stored water is estimated by using the formula: 10% of the
previous November and December rainfall plus 20% of January and February rainfall, plus 55% of
March rainfall plus 75% of April rainfall. Total water available to the plant (mm) was then estimated as
rain falling during the growing season plus stored water.
RESULTS
As part of the GRDC project “Increasing the Profitability of Cropping Systems in Western Australia
using Lupins, Oats, Oilseeds and Pulses” a detailed report is being prepared which summarises some
of the database results. An extract of the section of this report that deals with narrow-leafed lupin is
given below.
Wheat after narrow-leafed lupin
Narrow-leafed lupin has been the most widely examined break crop species with over 150 trials x year
combinations available in the database. If we look at the raw data from all of the trials (Figure 1) we
can see the range of yields obtained in the trials. The majority of wheat on wheat (WW/2) yields are
less than 2.5 t/ha indicating that in the trials conducted to date it has been difficult to achieve yields
higher than 2.5 t/ha with wheat sown after wheat.
In general it is also noticeable that the majority of wheat after lupin responses above the 1:2 ratio line
occur when wheat on wheat yields are below 1.5 t/ha, indicating an agronomic issue with wheat-wheat
which the inclusion of lupin helps to remediate. Invariably these issues have been identified in
individual trials to be the presence of Take-all or high levels of annual ryegrass or brome grass. The
outlier on the y-axis of yields of lupin-wheat at or above 4 t/ha when wheat-wheat yields less than
1.0 t/ha are from the trial 91KA111 at West Katanning in which Take-all was a factor that severely
limited the yield of wheat on wheat and a wide range of break crops such as lupin, field pea and
canola provided a good break from the disease. Similarly the outlier where WW/2 yields close to zero
and LW/2 yields 2.5 t/ha is from a trial at South Carrabin in 1995 where brome grass became very
difficult to control in the wheat on wheat plots.
Similarly there are occasions where the lupin sequence fails. For example, the outlier on the x-axis
where WW/2 yields 1.8 t/ha and LW/2 yields close to zero are from a trial in 1983 at Nabawa (78C1)
where wild radish was not able to be controlled in the lupin phase and the weeds swamped the
following cereal crop. In later years the availability of diflufenican solved this issue, although in recent
times wild radish has again become harder to control in the lupin year with selective herbicides.
Overall though wheat sown after lupin out yields wheat sown after wheat. A linear relationship can be
fitted to the response of wheat after lupin compared to wheat after wheat over a wide range of wheat
on wheat yields. If the outliers discussed earlier are removed this relationship is: GY of LW/2 = 0.9
(GY of WW/2) + 0.6, r² = 0.58, P < 0.001, GY = grain yield. If we were to constrain the regression
through the origin the regression would become: GY of LW/2 = 1.34(GY of WW/2), r² = 0.45,
P < 0.001. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
12
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Yield of wheat after wheat (WW /2, t/ha)
Yield of wheat after lupin (LW /2, t/ha)
1:2
1:1
Figure 1 Relationship between the yield of wheat sown after wheat (WW/2) and the yield of wheat sown after
lupin (LW/2) in 88 trials (167 trial x year combinations) in experiments conducted throughout WA since 1974.
Linear curves indicate 1:1 and 1:2 ratios.
Another way to look at the data set is to consider the magnitude of the difference in yield (Ydiff)
between WW/2 and LW/2 and the frequency in which various levels of Ydiff occur. In the first instance
we will look at Ydiff averaged across all rates of nitrogen applied to the second year of wheat. Figure 2
shows that whilst there are relatively few instances where Ydiff is more than 1.5 t/ha, in 10% of
instances Ydiff is less than or equal to 0 t/ha, and the distribution is centred around 0−500 kg/ha range
with the mean increase in yield being 540 kg/ha.
4%
5%
24%
20%
12%
15%
9%
4%
8%
0%
5%
10%
15%
20%
25%
30%
< -0.25 -0.25 to 0 0 to 0.25 0.25 to 0.5 0.5 to 0.75 0.75 to 1.0 1.0 to 1.5 1.5 to 2.0 > 2.0
LW /2 - WW /2 (t/ha)
Relative frequency (%)
Figure 2 Relative frequency (%) in which the difference in yield (Y diff, t/ha) between wheat following lupin (LW/2)
and wheat following wheat (WW/2) falls into 9 yield categories. Data are from 167 trials x year combinations in 86
trials conducted in WA since 1974.
As seasons influence the magnitude of any break effect it can be useful to compare the upper limit of
the water use efficiency of the different rotations. To do this we calculated modified French and Shultz
figures for the two rotations. We then fitted by eye a boundary line encompassing most of the data
points (data not shown). Using this method the potential water use efficiency for wheat after lupin was
19 kg/ha/mm and was 15 kg/ha/mm for wheat after wheat. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
13
Response to nitrogen
It is widely known that one of the major effects of lupin will be the residual nitrogen they supply to the
following crop. Results discussed so far have been averaged across all the nitrogen fertiliser rates
applied to the following wheat crop. In order to evaluate the effect of residual nitrogen on Ydiff we first
grouped rates of applied nitrogen fertiliser into five groups labelled 0N, 25N, 50N, 100N, 150N, where
0N = all treatments where no nitrogen fertiliser was applied, 25N = where up to 25 kg N/ha was
applied, 50N = 25 to 50 kg N/ha, 100N = 50 to 100 kg N/ha, and 150N = more than 100 kg N/ha. We
then restricted the dataset to the 31 trials that included at least four of these five groups so that n = 67
for all N groups except 150N which had 44 observations. Residual maximum likelihood (REML)
models were then fitted using Genstat 10 with N group as the fixed effect and Trial.Year as the
random effect.
Overall nitrogen applied as fertiliser had a significant (P < 0.001) but small effect on Ydiff (data not
shown). The largest Ydiff was 556 kg/ha and occurred when no nitrogen fertiliser was applied. Ydiff
decreased as the rate of nitrogen fertiliser increased so, the highest nitrogen fertiliser group (150N)
produced a Ydiff of 396 kg/ha.
Do changes over time affect the response to nitrogen?
The difference in yield between wheat after wheat and wheat after lupin appears to change over time
with a gradual rising trend from 1974 up to 1990 when the difference in yield between LW/2 and WW/2
increases dramatically and then drops off again after 1993 (Figure 3).
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Year
Ydiff (LW /2 - WW /2, t/ha)
Figure 3 The difference in yield between wheat after wheat (WW/2) and wheat after lupin (LW/2) over time.
To investigate this observation further we restricted the data set to the years when the most number of
lupin Agzones had trials, which was the period 1983 to 1995. This showed a relatively flat period from
1983 to 1991 and then an increase in the period following. We then considered the changes in the
1990s that led to this unprecedented increase in the difference in yield between LW/2 and WW/2. Was
it environmental, such that we had a run of years that suited wheat after lupin more so than wheat
after wheat? Or were there changes in agronomic practices that were of benefit to wheat after lupin or
made lupin a better break crop?
To separate the effect of rainfall from management we compared the water use efficiency of the two
sequences (data not shown). This showed that the difference between the WUE of LW/2 and WW/2
was, for the first time, consistently above 3 kg/ha/mm from 1990. Around that period of time there was
a shift to no-till machinery both on farms and for experimental purposes. There was also a wider use of
more effective herbicides for in-crop control of grass and broadleaf weeds in lupin crops, and rotations
shifted to more continuous cropping as sheep numbers declined throughout WA. In general,
comments from trials in the period 1990−95 indicated that the lupin plots were generally free from
weeds and there were few reports of poor lupin growth in the trials. Thus these changes seemed to be
of benefit both for the lupin crop and the following cereal. Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
14
This is further demonstrated if we group the data into two periods: ‘1983−89’ and ‘1990−95’. The Ydiff
was 0.4 t/ha in the 1983−89 period and 0.9 t/ha 1990−95 period (P < 0.001) and this difference
remained even when the peak year of 1993 was removed from the analysis. Of interest then was to
see if the agronomic changes also changed the response to fertiliser nitrogen. To do this we had to
further reduce the dataset because there were few trials that included rates of fertiliser nitrogen above
100 kg/ha (n = 6).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 25 50 100
Nitrogen group (kg N/ha)
Ydiff (LW /2 - WW /2, t/ha)
1983-89 1990-95
Figure 4 Response of Ydiff (LW/2−WW/2 in t/ha) to nitrogen fertiliser application in the wheat year for the periods
prior to and after 1990.
Prior to 1990, as the rate of nitrogen increased the difference in yield between LW/2 and WW/2
decreased (Figure ). However, since 1990 nitrogen has no effect (P > 0.05) on the difference in yield.
It appears that since 1990 wheat after lupin continues to respond to increasing rates of nitrogen
whereas in the previous period wheat after lupin did not respond to increasing rates of nitrogen whilst
wheat on wheat did.
CONCLUSION
A database has been collated from all of the available crop sequence experiments conducted in WA.
Over 10 000 records representing the results of over 160 experiments conducted since 1966 appear in
the database, allowing for rigorous interrogation of rotation effects over a long period of time. In the
experiments conducted to date continuous wheat was rarely as productive or economically viable as
rotations that included either a pasture or break crop, regardless of amount of nitrogen fertiliser
applied.
In general terms, since 1990, both the yield of wheat on wheat and the likelihood of a response to
lupin in the following year have increased at all levels of applied nitrogen. This corresponds to a period
where more effective herbicides were used, rotations shifted to more continuous cropping and trials
were more likely to be sown with no-till machinery
If changes to crop management in the past have influenced the size of the break crop effect then we
need to consider the implications of even more recent changes to crop management. In particular the
benefits of break crops are likely to be influenced by the modern use of more effective fungicides,
inter-row seeding to avoid last years crowns and roots, metering out of nutrition throughout the
growing season, and the reduced effectiveness of weed control.
KEY WORDS
crop sequence, break crop, lupin Agribusiness Crop Updates 2009
Crop Updates is a partnership between the Department of Agriculture and Food, Western Australia and
the Grains Research & Development Corporation
15
ACKNOWLEDGMENTS
Thanks to: Pam Burgess for helping to collate the data, Andrew Van Burgel for statistical advice, and
all the previous researchers who provided the data. Funds for the work are provided by DAFWA and
GRDC.
Project No.: DAW161
Paper reviewed by: Peter White
