Hit The Order Button To Order A **Custom Paper**

>> CLICK HERE TO ORDER 100% ORIGINAL PAPERS FROM AustralianExpertWriters.com <<

26 Aug
2019

SITE-SPECIFIC CROP MANAGEMENT | Good Grade Guarantee!

A Process for Implementing
SITE-SPECIFIC
CROP MANAGEMENT
A General Introduction to Precision Agriculture
1
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
A Process for Implementing
Site-Specific Crop Management
from the
Australian Centre for Precision Agriculture
Obviously we don’t farm to intentionally loose money and in general this is not
the case. But if we consider farming over a short time frame (say a growing
season) then financial losses do occur. Incorporating Site-Specific Crop
Management into farm management will be no gaurantee against future losses,
but the risk of short-term financial losses may be minimised by using the
information gained and optimising the product input/output ratio. All the while,
we also profit from progress in long-term improvements in operability, landscape
and environmental management, product marketing, storage of knowledge
relevant to enterprise management and our contribution to society.
STEPPING THROUGH THE PROCESS
Site-Specific Crop Management (SSCM), should be considered as part of the
continuing evolution in arable land management. Recent developments in technology
(satellite navigation systems, geographic information systems, real-time crop and
soil sensors) have essentially improved the scale at which we can observe variability
in production.
Obviously, the variability found on individual farms and paddocks will be related to
the location and previous management, but we can provide a generalised outline of
how SSCM may be introduced to a farming system (Table 1).
In Table 1, the steps are to be considered in numerical order so that the most benefit
is gained with the least additional cost. This does not mean they cannot be applied in
conjunction, but each additional step in this process does require some new tools or
techniques to be aquired and applied.
Steps 2 and 3 are where most work is concentrating now in an effort to identify
practical ways to quantify and respond to observed variability.
A General Introduction to Precision Agriculture
2
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Steps
Tools & Techniques that PA can offer
1. Optimise average crop
management
Crop scouting and soil sampling tools, vehicle
guidance and auto-steering, simple paddock
experimentation tools.
2. Determine the magnitude,
extent and responsiveness
of spatial and temporal
variability
Crop scouting and soil sampling tools, yield
monitors, soil sensors and remote sensing,
more advanced experimentation and
analytical tools.
3. Optimise the production
input/output ratio for quantity
and quality
(to maximize gross margin and
minimize environmental footprint)
Crop scouting and soil sampling tools, crop
yield and quality monitors, soil sensors and
remote sensing, vehicle guidance and auto
steering, advanced experimentation,
analytical and decision support tools, variable
rate controllers
4. Output quality control and
product marketing
Crop quality monitors and segregation tools,
variable-rate controllers, application map
recording, electronic information tagging
5. Maintaining resource-base
and operation information
Crop scouting and soil sampling tools,
mapping capabilities and specialized storage
software
Table 1. Generalised steps to making progress with SSCM.
For SSCM to be tested/accepted/adopted across the agroclimatic zones in Australia,
it is important that cost-effective, practical systems be offered to assess the withinfield variability in crop production. Such systems should aim at investigating causal
relationships between soil/crop factors and yield at the within-field scale along with
the extent to which these relationships vary across the field. This information should
be used to determine whether the observed variability warrants differential treatment
and if so, direct a route through a SSCM decision methodology.
DIFFERENTIAL TREATMENT OPTIONS
In implementing differential treatment, rate-based operations that influence crop yield
can be targeted to achieve desired yield goals with the minimum input of resources.
Such governing operations occur at nearly all phases of the crop growth cycle. The
array of variable-rate control designs available or proposed range from simple control
of flow rate to more complex management of rate, chemical mix and application
pattern. The control segment of any variable-rate application should optimise both
the economic and environmental product of the field and should ensure that estimates
of operational accuracy and dynamics are included in the application process. This
is an important point, as incorrect spatial application may be economically and
environmentally detrimental.
In all the operations that are under consideration presently, the control commands
may be instigated by accessing a map of application rates and locations (e.g. VRA
A General Introduction to Precision Agriculture
3
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
map for Lynx controller), combining real-time data with the real-time use of a response
algorithm (e.g. Yarra N-Sensor), or a combination of both techniques. For the
majority of cropping industries the important areas of managerial intervention would
include:
~ Fertiliser application (quantity and mix)
~ Gypsum/lime application
~ Sowing rates and depth
~ Crop variety
~ Pesticide application
~ Irrigation water
~ Soil tillage implements and depth of operation
~ Crop growth regulator
DECISION METHODOLOGY
The decision methodology may follow a tree-structure of questions which require a
positive or negative answer to decide on a progress path. The information gathered
using SSCM technologies would provide the basis for the answers. An example of
the logic pathway required for a decision support methodology is presented in Figure
3. This model begins with the premise that variability in crop yield is the initial signal
that variable-rate treatment might be warranted. Another model might begin with the
observation of soil variability or crop reflectance.
In this model, differential treatment is then examined as an option based on:
~ the degree of variation
~ the cause/s of variation
~ suitability for management intervention
Uniform treatment, continuously variable treatment or the division of a paddock into
potential management sub-units (management classes) are the considered options.
Ultimately the assessment and treatment of variability would be undertaken in realtime and the scale of treatment effectively restricted only by the functional specifications
of the application equipment (i.e. continuously variable treatment). For the present,
the state of agronomic and technological developments probably dictates that the
most practical approach for Australian conditions is the identification and assessment
of ‘broad’ management classes within a paddock using relevant layers of information.
If significant production differences can be identified between classess and if the
class differences in requirements and responses to the input/s under consideration
for VRA can be understood, then PA will be qualified to enter the practical management
of cropping systems.
A General Introduction to Precision Agriculture
4
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Figure 3. Management decision tree for SSCM – a simple model based on
uniform, management class or continuous crop treatment.
Determine class yield models
for variables of interest
Does crop yield variability
warrant treatment?
YES
Can the cause/s of
variability be determined
and modelled?
YES NO
Can the cause/s of variability
be managed?
NO Uniform field
treatment
Can the cause/s be used to
determine management classes?
YES NO
Uniform field
treatment
Will the cause/s be used to
determine management classes?
NO YES
Determine
management classes
Set yield goals
Is variability suited to
continuous management?
NO
Uniform field
treatment
YES
YES
NO YES
Are management classes to be
treated uniformly for other
variables
Apply continuous
yield models
NO Uniform field
treatment
Instigate differential action
based on class mean value of
the variable of interest
Instigate continuous
differential action
A General Introduction to Precision Agriculture
5
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
MANAGEMENT CLASS DETERMINATION
Essentially, the management classes should partition the variability within the field
so that:
~ within-class variability is reduced below whole field variability.
~ mean within-class variability is significantly different between
management classes
~ the reduction in variability will also be expressed in important attributes
that have not been used to make the management classes.
A Brief History
A In the United States, VRA began prior to the advent of yield mapping, using the
analytical testing (chemical analysis of nutrients) of topsoil samples collected on a
100-yard grid. This approach is expensive (in Australian terms) and may be logically
flawed. The idea presupposes that all areas in a paddock have the same yield potential
and in order to reach that potential the optimum amount of fertiliser has to be applied
at each point. Research in Europe and Australia (and only recently in the US) has
suggested that it would be better to recognise areas within paddocks which have
different yield potentials (and therefore management requirements), but which may
be managed uniformly within the defined boundaries. These areas, called
management classes are in essence, small fenceless paddocks within much bigger
paddocks. This approach may be regarded as a risk-averse compromise between
uniform management with little or no spatial information and continuous management
of cropping variability.
There have been a number of techniques used in the delineation of potential
management classes. They include:
~ Polygons hand-drawn on yield maps or imagery.
~ Classification of remote sensed imagery from an aerial or satellite
platforms using both supervised and unsupervised procedures.
~ Identification of yield stability patterns across seasons at fixed map
nodes using correlation co-efficients, weighted taxonomic distance,
temporal variance, normalised yield classification.
~ Fuzzy multivariate cluster analysis using seasonal yield maps.
~ Morphological filters or buffering.
~ Spectral filters using Fast Fourier Transform.
~ Multivariate analysis by hard k-zones.
Other options that have been raised are the classification of a soil fertility index
calculated by factor analysis and the simple use of standard deviation and the
frequency distribution to partition yield/soil maps or imagery.
A General Introduction to Precision Agriculture
6
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
A few studies have been undertaken to compare strategies for management unit
delineation. Grid sampling at a fine scale (approx 50m) often proves more successful
than using existing soil unit maps in delineating units with differing yield potentials
but the cost of grid sampling always means that this option was is not the most
profitable.
In some instances, aerial imagery of crop reflectance has produced more accurate
and precise estimation of soil unit delineations than final yield maps. Importantly, the
aerial photographs must be taken at the correct time of season to truly represent the
yield variability induced by soil variability. The period just prior to flowering (anthesis)
is suggested as the optimum window for cereals.
But most studies suggest that intensive grid sampling of soil attributes is the most
accurate method of determining management classes (at least for single nutrient
fertiliser application). The expense and labouriousness of the sampling regime has
fostered the examination of alternative methods. Intuitively, management classes
which are developed with the inclusion of data layers that represent an integrative
attribute such as crop yield or vegetative index should be more robust for the
application of a range of differential treatments.
Relevant Data Layers For Australia
Layers of accurate, spatially-dense, georeferenced information are required to begin
the process. Maximising practicality and minimising cost are the major constraints.
Crop yield maps obviously contain information on seasonal production that is essential
to this process. Beginning this process without information on the spatial variability in
the saleable product would appear to be financially imprudent.
It is, however agronomically sensible to include some information on soil and landscape
variability in the decision process. Many studies have shown that the most dominant
influences on yield variability (other than climate) are the more static soil physical
factors such as soil texture, soil structure, and organic matter levels. These are known
to indirectly contribute to cation exchange capacity, nutrient availability and moisture
storage capacity of the soil.
Gathering direct data on these attributes at a fine spatial scale is problematic, but a
number of correlated attributes can be gathered relatively swiftly. Apparent electrical
conductivity of the soil (ECa) has been shown to provide corroboration to the spatial
yield pattern in many fields, and correlation with a number of deterministic physical
soil parameters. Paddock topography has also been shown to provide an indirect
indication of variability in soil physical and chemical attributes – again usually due to
a high correlation with a deterministic attribute such as soil texture. Topography also
provides indirect information on microclimate attributes that influence crop production
potential.
These soil attributes are, however, extremely difficult or impractical to amend in the
A General Introduction to Precision Agriculture
7
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
short-term. However if the more rigid factors are going to limit yield then it would
seem prudent to allow these to influence the application rates of any inputs/ameliorants
in the field. Intuitively, factors contributing to variability in the soil moisture regime
and physical properties controlling soil water movement and nutrient supply may be
the most significant causal factor in the spatial variability of crop yield in the majority
of cereal growing regions in Australia. Many of the more easily adjusted soil factors
such as available nutrient levels and pH could be expected to vary based on the
consequences of variation in the physical properties of the soil. Using the variation in
the indicator factors – crop yield, soil ECa and elevation – as a basic data set to
delineate areas of homogeneous yield potential may prove useful. The response of
inputs/ameliorants to these factors will of course be site-specific, but the significance
of their influence may not. Of course other data layers that may be gathered at the
same spatial scale may be included if warranted.
At the ACPA, research suggests that a number of years yield data in combination
with soil EC
a and elevation provides a very sound basis for management unit
determination when subject to a multivariate clustering process.
How Are We Doing It?
The general approach we have been using is:
~ Measuring spatial variability in the paddock (at present best simply
described by soil ECa maps, crop yield maps, and digital elevation
models)
~ Determine number and location of potential management classes if the
variation is deemed suitable.
~ Direct soil/crop sampling and analysis within the management classes
to investigate practical causes of variation.
~ Interpret test results and instigate remedial action if indicated, or design
within-paddock experimentation for input response measurement which
can be used in the future with basic seasonal prediction information.
Growers now routinely gather yield data using their own or contract harvesters and
those with autosteer systems can collect data for the DEM during all navigation
operations (tillage, sowing, spraying etc). The soil ECa maps are generally gathered
using a local contractor who uses an Electromagnetic Induction (EMI) instrument
such as the EM38 or an Electrical Resistivity (ER) instrument such as the Veris 3100.
It is from this stage that the process takes on a bit more complication, and while a
number of growers are taking on the tasks themselves, the techniques and software
being used at present take time to master.
A Method For Delineating Potential Management Classes With Some Certainty
All attributes to be used in the ‘classification’ process for each paddock are predicted
onto a single, 5-metre grid through local block kriging with local variograms using
VESPER. With all attributes on a common grid, multivariate k-means clustering is
A General Introduction to Precision Agriculture
8
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
used to delineate the potential management classes. This is an iterative method that
creates disjoint classes by estimating cluster means which maximise the difference
between the means of the classes and minimises the variation within the cluster
groupings.
Of the available data layers, crop yield (or the income derived there from) has the
greatest bearing on farm management and practices at present. Potential
management classes, however they are derived, should therefore display significant
differences in yield for VRA to be worthwhile. However, ensuring that the differences
displayed in crop yield maps are genuine, let alone significant is difficult. Fortunately,
the block kriging process provides an estimate of the prediction standard deviation at
each point in a yield map, and we use the median value (σkrig (median)) to calculate
the confidence interval (95% C.I.) surrounding the mean yield estimate within a
paddock (μ) (Equation 1).
95%C.I. = μ +/- (σkrig (median) x 1.96) Equation 1
And the absolute difference between mean class yields (|Yclass1 (mean) – Yclass2 (mean)|)
should then follow Equation 2 for the classes to be considered representative of
regions of significantly different yield (p<0.05).
|Yclass1 (mean) – Yclass2 (mean)| > (σkrig (median) x 1.96) x 2 Equation 2
This gets us to the point where we can decide the number of potential management
classes and set out sampling points within each class. The sampling is a vital point
as it allows us to explore what may be causing the variability we have been seeing in
our data layers.
Directed Soil Sampling
The basic layers used in determining the potential management classes provide an
integrated assessment of changes in production potential using soil, landscape and
yield attributes. The next step requires that the classes be interrogated for the cause
of the observed yield variability. For SSCM, there are 4 propositions to consider:
~ Whether one (or a correlated combination of) static factor/s can be identified
that dominates the changes in yield potential in a field.
~ Whether there is a transient, manipulable factor that is restricting zones of the
field reaching seasonal yield potential.
~ Whether complex interrelationships between observable factors need to be
analysed and modeled.
~ Whether the yield variability is caused by a change in the production process
that was not measured (e.g. unobserved, localised pest damage or disease).
A General Introduction to Precision Agriculture
9
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
The first two proposals simplify management responses. The third may be optimal in
terms of optimising yield and environmental benefits, but economically unviable (at
present). The fourth would probably show up in a correlation with a static factor
unless there was a breakdown in normal standard of agronomy management.
At present, soil sampling is undertaken using a form of stratified random sampling
with the potential management classes as the strata. Constraints on the random
allocation of sample points are imposed to avoid strata boundaries and to target
class means. A minimum of 3 separate spatial locations, with segregated samples
from the top soil (0-0.3m) and subsoil (0.3 – 0.9m (max)) are initially targeted for
each potential zone. The depth of sampling can be adjusted to suit local agronomic
testing regimes if need be.
Analysis of the soil test data should provide us with some explanation or highlight
were we need to look further. If an amelioration issue arises (e.g. pH or sodicity
problem) then VRA can take place based on the soil test results or further experiments
can be laid out within the classes.
The whole process can be described in the flow diagram below:
Relevant Data Layers : Yield, soil conductivity, elevation
k-means clustering using all relevant layers to delineate
production classes
Spatial prediction onto a single grid using block kriging
Utilise the mean kriging variance for yield to determine
Confidence Interval (C.I.) for class partitioning
Direct soil sampling into management classes to interrogate
observed production variation
A General Introduction to Precision Agriculture
10
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
PRACTICAL EXAMPLES OF MANAGEMENT CLASS DETERMINATION
Paddock 1
Data Layers
In this 75ha field, the data layers used are sorghum and chickpea yield in successive
growing seasons (Figures 4a-4b), soil electrical conductivity (Figure 4c) and elevation data (Figure 4d) all collected on a similar spatial scale. The data was collected
using (respectively) an Agleader yield monitoring system, the Veris® 3100 conductivity array and an AshtechTM single frequency plus C/A-code RTK GPS with postprocessing.
(a) (b)
(c) (d)
Figure 4. Data layers from a 75 ha paddock in northern NSW – (a) Sorghum
yield (b) Chickpea yield (c) soil ECa (d) elevation.
Sorghum Yield (t/ha) ChickPea Yield (t/ha)
A General Introduction to Precision Agriculture
11
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Potential Management Classes
Two and three potential management classes were delineated (Figure 5) for the purposes of testing the validity of the multivariate clustering and significance procedures
through subsequent soil analysis. The delineation of classes using this procedure has
provided a C.I. for the two crops in question (Table 2).
(a) (b)
Table 2. Class means for the data layers used in the delineation process.
Values for 2 and 3 class scenarios are shown along with C.I. values.
Sorghum
Yield (t/ha)
Chickpea Yield
(t/ha)
ECa
(mS/m)
Elevation
(m)
2 Classes
Class 1
Class 2
3 Classes
Class 1
Class 2
Class 3
5.8
4.8
1.4
1.1
185
156
371
375
5.9
4.7
5.5
1.4
1.1
1.2
189
155
173
374
375
363
C.I. (+/- t/ha) 0.2 0.1 13.6
Management Class Management Class
A General Introduction to Precision Agriculture
12
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Concentrating on sorghum, a C.I. of +/- 0.2t/ha means that a difference of at least 0.4
t/ha between the mean sorghum yields in the potential classes should be seen to
negate the possibility that the variability carried through the mapping and classification
procedures is incorrectly depicting the spatial patterns. From Table 2, the 2-class
difference is 1.0 t/ha and the smallest three-class difference is 0.4 t/ha. This suggests
that a split into 3 classes is on the border of being justified based on the mean
sorghum yield differences. For chickpea, a difference of 0.2 t/ha between the mean
yields in the potential classes should be seen to warrant further investigation. This is
clearly the case for 2 classes but if we increase the number of classes to 3 the
differences are not large enough.
Directed Soil Sampling
The classes have been delineated using production information gathered in great
detail. Soil sampling sites have been directed within each of the 3 classes in Figure
5b in an attempt to explore causes for the yield differences (Tables 5 and 6). In
Tables 3 and 4, the sample sites have been reallocated to one of 2 classes described
in Figure 5a.
In the case of 2 potential classess, analysis of the top soil (Table 3) shows that class
2 has produced lower crop yields despite a higher CEC and a lower sand fraction
than class 1. Soil nitrate is also double in class 2. An examination of the soil below
0.3m (Table 4) shows that the CEC and clay content of class 2 are significantly lower
than in class 1, and the soil nitrate remains double. The difference in the physical
properties of the subsoil, combined with the fact that the soil is on average 40%
shallower in class 2 conspires to restrict the quantity of available moisture in the
profile compared to class 1. This relative limitation in soil moisture in class 2 would
limit crop yield and therefore reduce the nitrogen requirement. Under uniform fertiliser management, accumulation of soil nitrogen reserves (as evident in nitrate and
total N levels in Tables 3 and 4) would be expected.
If the field is broken into 3 potential classes, the process essentially divides the previous
class 1 into 2 classes. The soil analysis (Tables 5 and 6) shows that the partitioning
is reflected in a more refined separation of texture, CEC, depth, soil profile moisture
content and nitrogen reserves between all 3 classes. Combining this information
with the uncertainty analysis would suggest that in this instance, 3 classes are probably
warranted for cereal crops where nitrogen is applied.
A General Introduction to Precision Agriculture
13
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Table 4. 2 classes – soil test results for the 0.3-0.9m soil layer.
Table 3. 2 classes – soil test results for the 0-0.3m soil layer.
Soil Attribute
Class
(Red)
1
Class 2
Paddock
(Green)
Mean
pH (CaCl) 7.5 7.6 7.6
O.C. (%C) 0.7 0.9 0.8
N03 (mg/kg) 15.0 30.4 22.7
P (mg/kg) 4.5 5.3 4.9
K (meq/100g) 0.7 0.6 0.7
Ca (meq/100g) 45.9 62.3 54
Mg (meq/100g) 20.2 13.2 16.7
Na (meq/100g) 0.8 0.2 0.5
Total N (mg/kg) 868 1026 947
CEC (meq/100g) 67 76 72
Ca/Mg 2.3 4.8 3.6
ESP % 1.13 0.25 0.69
Sand % 14 10 12
Silt % 13 15 14
Clay % 73 75 74
E.C. 137 163 150
Soil Attribute
Class
(Red)
1
Class 2
Paddock
(Green)
Mean
pH (CaCl) 7.9 7.7 7.8
O.C. (%C) 0.7 0.8 0.8
N03 (mg/kg) 8.7 14.7 11.7
P (mg/kg) 2.8 3.7 3.3
K (meq/100g) 0.6 0.42 0.51
Ca (meq/100g) 42.9 42.1 42.5
Mg (meq/100g) 23.3 9.5 16.4
Na (meq/100g). 2.4 0.3 1.3
Total N (mg/kg) 610 887 749
CEC (meq/100g) 69 53 61
Ca/Mg 1.9 5.2 3.6
ESP % 3.5 0.7 2.1
Sand % 13 17 15
Silt % 11 17 14
Clay % 76 66 71
E.C. 159 126 143
Soil Depth (m) 1.22 0.71 0.97
Profile avail. H20
at sampling (mm) 118 68 93
A General Introduction to Precision Agriculture
14
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Table 5. 3 classes – soil test results for the 0-0.3m soil layer.
Table 6. 3 classes – soil test results for the 0-0.3m soil layer.
Soil Attribute
Class 1
(Red)
Class 2
Class 3
Paddock
(Green)
(Purple)
Mean
pH (CaCl) 7.8 7.6 7.2 7.5
O.C. (%C) 0.6 0.9 0.8 0.8
N03 (mg/kg) 10.6 30.4 19.3 20.1
P (mg/kg) 2.7 5.3 6.3 4.8
K (meq/100g) 0.5 0.6 0.9 0.7
Ca (meq/100g) 51.3 62.6 40.5 51.5
Mg (meq/100g) 22.1 13.2 18.3 17.9
Na (meq/100g) 1.0 0.2 0.5 0.6
Total N (mg/kg) 658 1026 1079 921
CEC (meq/100g) 75 77 60 70
Ca/Mg 2.3 4.8 2.2 3.0
ESP % 1.35 0.25 0.92 0.84
Sand % 12 10 16 13
Silt % 13 15 13 14
Clay % 75 75 71 74
E.C. 136 163 138 145
Soil Attribute
Class 1
(Red)
Class 2
(Green)
Class 3
Paddock
(Purple)
Mean
pH (CaCl) 8.0 7.7 7.8 7.8
O.C. (%C) 0.6 0.8 0.7 0.7
N03 (mg/kg) 5.6 14.7 11.9 10.7
P (mg/kg) 2.5 3.7 3.0 3.1
K (meq/100g) 0.48 0.42 0.65 0.5
Ca (meq/100g) 47.0 42.1 38.9 42.7
Mg (meq/100g) 24.9 9.5 21.5 18.6
Na (meq/100g) 2.7 0.3 2.1 1.7
Total N (mg/kg) 532 887 687 702
CEC (meq/100g) 74.8 52.3 63.4 63.5
Ca/Mg 1.9 5.2 1.8 3.0
ESP % 3.6 0.7 3.2 2.5
Sand % 11 18 15 15
Silt % 11 17 11 13
Clay % 78 65 74 72
E.C. 155 126 162 148
Soil Depth (m) 1.24 0.68 1.17 1.03
Profile avail. H20
at sampling (mm) 128 68 108 101
A General Introduction to Precision Agriculture
15
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Paddock 2
Data Layers and Potential Management Classes
For this 325 ha paddock , wheat yield , soil ECa and elevation were collected as
described earlier. The mean results from delineating 2 classes (C1 yield = 3.7 t/ha,
ECa = 114 mS/m ; C2 yield = 4.9 t/ha, ECa = 140 mS/m) and 3 classes (C1 yield =
3.4 t/ha, ECa = 112 mS/m ; C2 yield = 4.9 t/ha, ECa = 132 mS/m ; C3 yield = 5.0 t/ha,
ECa = 144 mS/m) suggest that there is little increase in management opportunity
revealed by the 3 classes. The C.I. calculation (+/- 0.35 t/ha) adds weight to this
assessment. Figure 6 shows the delineation patterns for 2 classes (a) and 3 classes
(b) respectively.
(a) (b)
Directed Soil Sampling
The results for soil sampling into the 3 classes are shown in Tables 7 and 8. The most
striking class deviations from the estimated paddock mean show up in the ESP%,
clay content and profile available moisture. If an ESP% >6 is taken as indicating
problematic soil structure, sampling for an average would suggest the paddock was
not yet in need of treatment. Class sampling, however, identifies class 1 as having a
much higher ESP% than the other clases, and importantly, above critical limits in the
topsoil (where treatment is more practical). The high ESP% can be hypothesised to
be contributing to surface-sealing and reduced infiltration in class 1. A lower clay
content helps magnify the difference in the ability of this class to store moisture, as
seen in Table 7.
The C.I. calculation suggested that 2 classes were likely warranted in this paddock
6657400
6657600
6657800
6658000
6658200
6658400
6658600
6658800
6659000
6659200
188300 188800 189300 189800 190300
easting
6657400
6657600
6657800
6658000
6658200
6658400
6658600
6658800
6659000
6659200
188300 188800 189300 189800 190300
easting
Figure 6. (a) 2 and (b) 3 potential management classes as defined by
multivariate k-means clustering. Class 1 = red, Class 2 = green,
Class 3 = blue.
A General Introduction to Precision Agriculture
16
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
and this has been born out by subsequent, directed soil sampling. The similarity of
soil conditions in clases 2 and 3 reflect the closeness in mean yield observed in the
wheat yield map. VRA of gypsum, or directed deep-ripping offer potential remedies.
Soil Attribute
Zone 1
(Red)
Zone 2
(Green)
Zone 3
(Blue)
Field
Mean
pH (CaCl) 7.8 7.8 7.9 7.8
N03 (mg/kg)
P (mg/kg)
K (meq/100g)
Ca (meq/100g)
Mg (meq/100g)
Na (meq/100g)
Total N (mg/kg)
CEC (meq/100g)
Ca/Mg
ESP %
Sand %
Silt %
Clay %
E.C.
9.2
9.7
0.71
17.7
11.3
2.4
501
32.1
1.5
8.1
31
22
47
0.143
12.2
10.3
1.03
21.4
14.0
1.8
600
38.2
1.5
4.7
16
19
64
0.113
15.1
8.7
0.97
26.8
12.8
2.0
496
42.7
2.1
4.7
16
23
60
0.137
12.2
9.6
0.9
22.0
12.7
2.1
532
37.7
1.7
5.8
21
21
57
0.131
Soil Attribute
Zone 1
(Red)
Zone 2
(Green)
Zone 3
(Blue)
Field
Mean
pH (CaCl) 8.3 8.3 8.3 8.3
N03 (mg/kg)
P (mg/kg)
K (meq/100g)
Ca (meq/100g)
Mg (meq/100g)
Na (meq/100g)
Total N (mg/kg)
CEC (meq/100g)
Ca/Mg
ESP %
Sand %
Silt %
Clay %
E.C.
6.0
21.2
0.64
17.2
14.1
6.5
275
38.5
1.2
17.3
27
20
53
0.373
6.4
12.0
0.81
18.6
17.6
5.1
339
42.1
1.1
12.1
13
23
64
0.233
9.7
9.7
0.81
22.5
15.2
5.4
419
43.9
1.5
12.2
15
22
63
0.256
7.4
14.3
0.75
19.4
15.6
5.7
344
41.5
1.3
14.1
18
22
60
0.287
Soil Depth (m) 0.8 0.85 0.8 0.82
Profile avail. H20
at sampling (mm) 24 58 56 46
Table 7. 3 classes – soil test results for the 0-0.3m soil layer.
Table 8. 3 class – soil test results for the 0.3-0.9m soil layer.
A General Introduction to Precision Agriculture
17
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
SUMMARY
In the Australian dryland environment, it is not unexpected that factors controlling the
interaction between crops and the climatic environment should be prominently
influential in the variability displayed in crop yield maps. For management, this
suggests that it will be necessary to use this class information in conjunction with
early season environmental indicators and crop response models (or simpler, empirical
budget models) to guide differential action decisions.
These decisions should not focus on treating a paddock to produce a uniform yield
unless the potential is uniform. The benefits from this type of analysis will only be
realised by acknowledging diversity in yield potential and environmental conditions
when formulating paddock management operations. For example, well-documented
areas of low yield potential may be removed from production, have the land-use
changed or have their inputs reduced to minimise potential financial losses.
The process of potential management class delineation described here offers a
relatively simple, practical approach to using production data gathered at a fine spatial
scale. The directed soil sampling should identify whether there is a/are manipulatable
limitation/s on production or definable variability in crop yield potential. The process
described here is not designed to correct poor traditional (managing to the average)
agronomy. Farmers will get greater financial gains by ensuring uniform management
is reasonable before venturing down the SSCM path. For those ready to explore
improvement on uniform management.
When contemplating the number of agronomically significant classes, care must also
be taken to consider and test for the major limiting factors in each zone. Much research
will be required to understand the agronomy of response at the within-field scale,
under site-specific conditions.
WITHIN-CLASS EXPERIMENTATION
Where there are no amelioration issues, field scale experiments can be established
to estimate the response in each identified potential management class to a single
input. The choice of input for experimentation in each field will be made on the basis
of results obtained from the strategic sampling missions within the potential
management classes. A marked build-up or depletion in a soil parameter between
classes could be used as a criterion along with the magnitude of contribution the
associated input makes to the variable costs of production. A zero rate treatment
should be included in all trials while the alternative treatments can be multiples of the
farmer’s uniform application rate. The design of the experiments should consider
application equipment capability and size, spatial constraints due to management
class pattern and a desire to minimize the area/financial impact of the experiment.
The classes must also be interrogated for the cause of the observed yield variability
and the results carefully considered before contemplating any within-field
experimentation. What a farmer would be looking for is a managerial significant
difference in indigenous soil nutrients, soil restrictions or crop growth/disease
I
A General Introduction to Precision Agriculture
18
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
parameters. When the data suggests that response experimentation within the classes
is an option, a ‘strip’ or ‘fleck’ design is proposed here, whereby randomised block
experimentation is performed with spatial constraints and economic considerations.
Strip or Fleck Design for Experimental Fertiliser Application
The treatment and plot-layout designs have dimensional and orientation constraints
imposed by the harvesting operations/equipment. Specifically:
~ Treatments must be laid out in the direction of sowing and harvesting.
~ Physical dimensions of each treatment plot should be at least three harvest
widths wide to ensure that at least one full harvest width can be achieved
from each treatment without the possibility of contamination from adjoining
treatments. Therefore the minimum plot width will be controlled by the
minimum multiple of the application machinery width that will meet this
target.
~ The minimum length of each treatment plot shall be constrained by the
operational mechanics of the harvesters. With grain mixing within the
harvester occuring along the direction of operation, yield data gathered at
the beginning and end of each treatment plot should be regarded as
contaminated by surrounding treatments (usually standard paddock
treatment). The plots should be a minimum of 80 m long, and a generic
rule of thumb suggests 100 m would ensure most mechanical set-ups are
covered. It is suggested that data from the first and last 20m of each
treatment plot be discarded from response analysis.
An economic constraint is also included, based on the desire to minimise any penalty
to the farmer’s expected profit by using potentially sub-optimal application rates over
much of the field. Most of the field can have an initial uniform treatment which the
manager considers his best practice. Data from the whole field treatment can be
used in the analysis.
EXAMPLES OF EXPERIMENTAL LAYOUTS
Paddock 44
A 130ha paddock located near Yarrawonga in Victoria (Figure 7a). The experiment
was established without variable-rate controlling equipment which reduced the
treatment level options. A zero:single:double rate design was implemented by marking
the plot locations with a DGPS and shutting off the spreader for the zero rate and
making two passes for the 200 kg urea/ha rate. Mean deep soil nitrogen levels (DSN)
prior to sowing in 2003 and 2004 are listed in Tables 1 and 2 respectively. The paddock
was sown to canola (Brassica napus) in 2003 and wheat (Triticum aestivum) in 2004.
A General Introduction to Precision Agriculture
19
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Bill’s Paddock
A 50ha paddock located near Crystal Brook in the north-west region of the Yorke
Peninsula , South Australia (Figure 7b). The paddock was delineated into 3 potential
management classes. A three treatment rate (0, 30, 45 kg N ha-1) with two replicate
experimental design was established with a Zynx variable-rate controller. The rest of
the padock received 15 kg N ha-1. Mean deep soil nitrogen levels (DSN) prior to
sowing in 2003 and 2004 are listed in Table 3. The field was sown to wheat (Triticum
aestivum) in 2003 and barley (Hordeum vulgare) in 2004.
(a) (b)
Analysis
Yield estimates were obtained using on-harvester yield monitors. The estimates were
spatially predicted onto a whole-of-field 5 metre grid using local block kriging with
localvariograms. The yield data was then extracted for each treatment plot and spatially
trimmed to a central kernel by removing 20 metres from the leading/trailing edges
and 10 metres from the remaining two sides. This left a 60 metre long, by at least one
harvest comb width (depending on the original plot widths), strip of data for analysis.
The average yield from each treatment plot was calculated.
Figure 7. Experimental design for 2 fields, the potential management classes
are designated 1,2 and 3; (a) Field 44 (130 ha): 0 kg urea/ha = black,
200 kg urea/ha = cross hatch, rest of field = 100 kg urea/ha (b) Bill’s
Field (50 ha): 0 kg N/ha = black, single hatch = 30 kg N/ha, cross
hatch = 45 kg N/ha, rest of field = 15 kg N/ha.
A General Introduction to Precision Agriculture
20
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
Paddock 44
The nitrogen response functions for the two years are shown in Figure 8. The urea
rate for maximum yield and also economic optimum in each class using a marginal
rate analysis is shown in Table 9. In 2003, the response data shows that the input/
output ratio from the different classes would have been economically optimised by
applying different average rates in each. In 2004, the results suggest that the whole
paddock may have been economically optimized with 0 kg urea/ha.
The 2003 season was considered excellent for the region with an annual rainfall of
523mm (mean annual = 516mm) and 303mm distributed fairly evenly during the
growing season (June – Nov.). Annual rainfall for 2004 was restricted to 365mm with
243mm falling during the growing season and only 5mm falling during the crucial
October grain filling period.
In 2003, the presowing DSN figures suggested that for a target yield of 2.5 t/ha
canola, class 1 was adequately supplied with indigenous nitrogen and the other two
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
2.3
2.4
2.5
Yield (t/ha)
0 100 200
Applied Urea (kg/ha)
Class 1
Class 2
Class 3
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
Yield (t/ha)
0 100 200
Applied Urea (kg/ha)
Class 1
Class 2
Class 3
Figure 8. Nitrogen response functions for Paddock 44 (a) canola season 2003
(b) wheat season 2004. Dashed line shows the uniform application
rate for the field (100 kg urea/ha).
Table 9. Urea rates to achieve maximum yield and economic optimum per
potential management class in 2003 and 2004.
ClassPresowing
DSN
to maximise
to maximise
DSN
to maximise
to maximise
2003
returns (kg/ha)
yield (kg/ha)
2004
returns (kg/ha)
yield (kg/ha)
2003 urea rate
2003 urea rate
Presowing
2004 urea rate
2004 urea rate
1 209 0 0 186 0 0
2 99 169 237 89 0 0
3 151 72 151 150 0 200
A General Introduction to Precision Agriculture
21
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
classes would probably benefit from the addition of nitrogen fertilizer. The experimental
results bear testament to these expectations. At the time of urea application in 2004,
the target yield was 3.5 t/ha wheat and again the DSN suggested that class 1 was
adequately supplied compared with the other two classes. The results from 2004
show that the classes in the paddock maintained their potential production relationships
(1>3>2) from 2003. However, the final outcome was the result of a crop with good
initial nitrogen status, sustaining good vegetative growth, only to be restricted in access
to moisture in the final growth stages. The differences in the class response to this
moisture restriction is noted as function of soil ECa (mean class ECa: 1 = 61, 2 = 20,
3 = 34) and position in the landscape (data not included).
Using the response functions from the typical 2003 season, it is possible to make a
simple estimate of what gains or losses in gross margin would have been made if
this information had been used to formulate fertiliser decisions at the beginning of
the season. Table 10 documents a comparison with the paddock average treatment
of 100 kg Urea/ha. As can be seen in Table 10, in 77ha of the field there was more
fertiliser than required, and in 53ha of the field an extra application of 69 kg/ha would
have brought in over 5 tonne more canola. The total waste in this scenario is A$3028
or A$23.29 per hectare.
Table 10. Paddock 44: analysis of gross margin differences between
variable-rate and uniform (100 kg urea/ha field average) fertilizer
application.
Bill’s Paddock
The nitrogen response functions for the two years are shown in Figure 9. The urea
rate for maximum yield and economic optimum for each class using a marginal rate
analysis is shown in Table 11. In 2003, the response data shows that the whole field
would have been economically optimized with an application of 0 kg N/ha. In 2004,
the response data shows that the input/output ratio from the different clases would
Fertiliser waste ha x kg = t x $400/t =$A
Class 1 18 x 100 = 1.8 720
Class 3 59 x 28 =1.65 660
Yield loss x $400/t =$A
Class 2 53 x 100 =5.3 2120
Yield gain
Class 3 59 x 20 = 1.18 472
Total Wastage 3028 (23.29/ha)
A General Introduction to Precision Agriculture
22
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
have been economically optimised by applying different average rates in each. The
2003 season was not a good season for the region (mean annual rainfall = 400mm)
even with the annual rainfall reaching 383mm. 199mm of this fell in the growing
season (June – Nov.) however 78% of this fell in the first 3 months leaving a dry finish
to the crop. 2004 saw average yearly rainfall (401mm) achieved, 232mm of which fell
during the growing season, but again 77% arrived in the first 3 months and only 4mm
during the crucial October grain filling period.Economic analysis of the 2003 wheat
season shows that the uniform treatment at 15 kg N/ha produced a gross wastage of
A$ 2417 (A$48.34/ha). Yield loss and fertilizer wastage account for 73% and 27% of
this figure respectively. This particularly negative response was induced by both the
seasonal weather conditions and the fact that the experimental design was laid down
as a side-dress following the uniform application of 30 kg N/ha at sowing.
In 2004 the experimental design was established after crop establishment and no N
fertilizer was applied at sowing. As can be seen in Table 12, 18.5ha of the paddock
2
2.5
3
3.5
4
Yield (t/ha)
0 10 20 30 40 50
Applied Urea (kg N/ha)
Class 1
Class 2
Class 3
2
2.5
3
3.5
Yield (t/ha)
0 10 20 30 40 50
Applied Urea (kg N/ha)
Class 1
Class 2
Class 3
Figure 9. Nitrogen response functions for Bill’s paddock (a) wheat season 2003
(b) barley season 2004. Dashed line shows the uniform application
rate for the field (15 kg N/ha).
ClassPresowing
DSN
2003
maximise
returns
(kgN/ha)
maximise yield
(kgN/ha)
DSN
2004
maximise
returns
(kgN/ha)
maximise yield
(kgN/ha)
2003 N rate to
2003 N rate to
Presowing
2004 N rate to
2004 N rate to
1 76 0 0 42 17 32
2 60 0 0 39 6 15
3 54 0 0 39 27 31
Table 11. Urea rates to achieve maximum yield and economic optimum per
potential management class in 2003 and 2004.
A General Introduction to Precision Agriculture
23
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
was given more fertiliser than required, and in the remaining 31.5 ha of the paddock
an extra application of 17 kg/ha would have brought in 3.9 tonne more wheat. The
total waste in this scenario is A$574 or A$11.48 per hectare.
Table 12. Bill’s Field: analysis of gross margin differences between variablerate and uniform (15 kg N/ha field average) fertilizer application
SUMMARY
The confirmation of potential site-specific yield response functions is not new.
However, the condition of minimal soil moisture limitation that accompanies these
assessments is rarely met in Australia. The response function information presented
here shows that variability in N response can be expected in Australia. A very basic
partition of the gross margin analysis helps to highlight the potential for environmental
as well as financial gains in the Australian environment.
All paddocks on all farms can provide the information relevant for individual
management. Input response data from individual padocks may then be used directly
or as a replacement for generic models in crop simulation programs. More
sophisticated spatial analysis of the N response data, along with intensive grain protein
data, will improve its usefulness.
Fertiliser waste ha x kgUrea = t x $400/t =$A
Class 2 18.5 x 20 = 0.37 -148
Yield loss x $130/t =$A
Class 1 12 x 14 = 0.17 -22
Class 3 19.5 x 193 = 3.76 -489
Yield gain
Class 2 18.5 x 35 = 0.65 +85
Total Wastage -574 (11.48/ha)
BRETT WHELAN & JAMES TAYLOR
Australian Centre for Precision Agriculture
www.usyd.edu.au/su/agric/acpa
A General Introduction to Precision Agriculture
24
www.usyd.edu.au/su/agric/acpa www.usyd.edu.au/su/agric/acpa
Site-Specific Crop Management
USEFUL SOFTWARE
Vesper
Spatial prediction software for mapping irregularly sampled data onto a regular grid.
(http://www.usyd.edu.au/su/agric/acpa)
JMP
Statistical analysis and visualisation software that performs k-means clustering.
(http://www.jmp.com)
Geod
Coordinate converter for Australia. Transforms data between geographic and cartesian
coordinates and between reference datums.
(http://www.lands.nsw.gov.au/Records/Surveying/GDA/GEODSoftware.htm)
Yield Editor
A filtering program that allows a number of indicators to be used to clean up raw yield
data files. It also converts between geographic and cartesian coordinates using the
UTM projection.
( h t t p : / / w w w . a r s . u s d a . g o v / s e r v i c e s / s o f t w a r e /
download.htm?softwareid=208modecode=36-20-15-00)
Splus
Programmable analytical software for basic and advanced statistics and clustering.
(http://www.insightful.com/products/splus/default.asp)

QUALITY: 100% ORIGINAL – NO PLAGIARISM.

  • **REMEMBER TO PRECISE PAGE NUMBER**
  • Hit The Order Button To Order A **Custom Paper**

>> 100% ORIGINAL PAPERS FROM AustralianExpertWriters.com <<