View
2
Download
0
Category
Preview:
Citation preview
Increasing the Adoption of Precision Agriculture in Australia
Precision Ag Conference
Friday 9th September 2011
Technology Park, Mawson Lakes SA
Contents A review of the history of Precision Agriculture in Australia and some future opportunities – Brett Whelan GRDC project Victorian Update Andrew Whitlock Innovative Practices for Efficient and Profitable Use of N Inputs – Craige MacKenzie Our Journey with PA – farmer case studies
- Todd Matthews - Mark Bender - Robin Schaefer
Innovative Irrigation – PA related irrigation tools – John Hornbuckle Precision Agriculture in Grazing Systems – Mark Trotter
Surveying with Sensors for Soil Mapping Michael Wells The challenge of reducing information and learning costs in Precision Agriculture – Frank D’Emden More confidence in making decision using N and P – What PA tools to consider – Peter Treloar Acknowledgements This event has been made possible by the generous support of industry. SPAA wishes to thank the following organisations and business for their financial assistance in putting this event together, and assisting with the travel arrangements of our key note speakers. GRDC, Government of South Australia, Landmark, Incite Pivot, John Deere, Case IH, New Holland, Omnistar, Topcon, MEA, Trimble, PA Source, precisionagriculture.com.au, Spatial Scientific, Outline imagery and the Stock Journal. SPAA also thank its corporate supporters Rabobank, Viterra and MGA insurance for support.
1
Conference Program
8.30am Registration 9.15am Welcome address Randall Wilksch, SPAA
President 9.20am Word from our GOLD sponsors
Landmark Operations Ltd and Incitec Pivot Ltd 9.30am A review of the history of Precision Agriculture in
Australia and some future opportunities Brett Whelan, ACPA
9.55am GRDC SPAA Project update
Sam Trengove, Leighton Wilksch and Andrew Whitlock
10.25am Questions to the panel 10.40am Sponsor presentation Case IH & Incitec Pivot Ltd
Morning tea at 11.00am with trade exhibitors
Farmer panel session 11.30pm Sponsor presentation New Holland & Omnistar 11.50pm Innovative Practices for Efficient and Profitable
Use of N Inputs Craige Mackenzie (New Zealand) 12.15pm Farmer panel featuring; - Todd Matthews (Eyre Peninsula, SA) - Mark Bender (Riverine Plains, NSW) - Robin Schaefer (Mallee, SA)
LUNCH 1.00-1.30pm with trade exhibitors
2
1.30pm SPAA AGM Randall Wilksch
Learning from others 2.20pm Innovative Irrigation – PA related irrigation tools
Dr John Hornbuckle, CSIRO 2.40pm Precision Agriculture in Grazing Systems Dr Mark Trotter, UNE - PARG 3.00pm Sponsor presentation
Landmark Operations Ltd & John Deere
AFTERNOON TEA 3.20pm with trade exhibitors
Consultant’s session 3.45pm Surveying with sensors for soil mapping – EM,
gamma & NIR on the go Michael Wells, Precision Cropping Technologies
4.10pm The challenge of reducing information and
learning costs in Precision Agriculture Frank D’Emden, precision agronomics Australia
4.30pm More confidence in making decision using N & P – what PA tools to consider Peter Treloar, Vision Ag
4.50pm CLOSING address Randall Wilksch, SPAA President
PA Connections – networking from 5pm
Address by CBH Grain and IPL
3
A review of the history of Precision Agriculture in Australia
and some future opportunities
Brett Whelan
Australian Centre for Precision Agriculture, University of Sydney
Senior Research Fellow
Australian Technology Park 1 Central Avenue Eveleigh NSW 2015
+61 2 8627 1132
brett.whelan@sydney.edu.au
sydney.edu.au/agriculture/acpa
Key Findings/Take Home Messages:
Precision Agriculture (PA), in its current form, has a long history of
innovators and pioneers with a single aim of improving agricultural
management. Australia remains at the forefront of the development of
PA tools and practical applications, not the least because of our unique
range of production conditions.
Undoubtedly the application of PA continues to be a general success in
Australia despite the fact that some products or techniques have not
been adopted or remain ahead of their time.
Development opportunities are naturally opening in areas which better
quantify small-scale variation and allow such information to be usefully
integrated into management decisions. Sensing systems, analytical
procedures, software, agronomic understanding, robotics, human
resources will all provide areas for PA to continue the transformation of
agricultural management into an increasingly resource efficient, less
risky, societal endeavor.
Precision Agriculture (PA)
A philosophy aimed at increasing long term, site-specific and whole farm
production efficiency, productivity and profitability while minimising unintended
impacts on the environment.
4
PA in Australia
In essence, PA has been evolving in Australia since 1788 when the first wheat
crop was grown by Henry Dodd at the site of the now Royal Botanical
Gardens in Sydney. It mostly failed and by early 1789 he had identified the
site was unsuitable for economic production and moved to a more suitable
site in Parramatta.
Since then, farmers, scientists and agribusiness have been learning about,
modifying and redesigning management systems to suit the spatial and
temporal variability in agricultural production conditions presented across
Australia. Australian examples from the period up to the early 1990’s that
show people have been considering the economic, environmental and social
benefits that understanding variability can bring are easily found . For
example, 1934 wheat yield maps produced by Fairfield Smith in Canberra
from hand harvesting small sections of crop; the Concorde Detectspray,
broadacre weed spot spraying system commercialised from work by Warwick
Felton in the mid 1980’s.
But the leap to the current PA situation, here and around the world, came
with the introduction in 1992/3 of civilian access to the US Defense force
Global Satellite Navigation System (GNSS). This now ubiquitous tool is known
as the Global Positioning System (GPS). Australian’s have since developed
equipment to directly utilise the GPS and other GNSS information for vehicle
navigation and implement operation, and integrated GNSS positioning
information into diverse systems and techniques to spatially describe and
manage production variation.
Some Significant Developments in Australia
(Particularly grain-centric, to some extent personally blinkered and, no doubt,
incomplete)
GPS
GPS receivers operate with selective availability from 1993 until
May 2000, meaning a correction signal was vital for all operations.
5
Reliable correction signals available from a variety of sources in
1995, including coast guard beacon, FM sideband, geostationary
satellite, local base station.
High accuracy and autosteer vehicle navigation systems using
local base stations developed in Australia in late 1990’s and sold
internationally from around 2000.
Wide area corrections increase in accuracy from sub metre to sub
decimeter by 2005
High accuracy CORS networks available to agriculture from 2009.
Other hardware
Grain yield monitors available in 1992/3 used with dead reckoning
for location.
Yield monitors for horticulture and viticulture available from 1995.
Variable-rate controllers available before GPS, but initial uptake in
Australia not until the late 1990’s.
Soil ECa field measurement instruments mobilized and applied to
PA in the late 1990’s. Gamma radiometrics application at the within-
field scale begins around 2000.
First on-harvester protein sensing system trialed in Australia at the
end of the 1990’s.
Broad acre spot spraying systems using plant reflectance
commercially available in the early 1990’s, but not considered widely in
Australia until 2005/6.
Boom and planter section control arrives in 2005.
Crop reflectance sensors available for nutrition management in
2003/4, but not widely available in Australia until 2006.
Implement steering available in 2009.
Software
Basic software available with manufacturer’s yield monitors from
1993/4 allowed yield map construction.
Australian and international companies begin to produce
independent spatial farm management software from 1995/6, but PA
6
software with intuitive GUI and GIS capabilities begin to appear in
2000.
Hand held mobile mapping software introduced from 1996 and
developed specifically for PA by 2000.
Management options
In 1995/6, yield data gathered at 1 second intervals within-fields
sparks wider interest in dealing with production variability across what
are becoming increasingly large paddocks.
In the same time period, the grid soil sampling concept to
investigate causes of yield variation is brought to Australia by
commercial entities.
First all-in-one hardware and software solution for yield mapping,
field navigation for sampling and variable-rate control released in 1996.
Late 1990’s saw the introduction of the management class/zone
concept to direct sampling and manage inputs.
Early 2000’s shows that management class construction using
high accuracy elevation, soil ECa and yield maps proves useful across
much of Australia.
By mid 2000’s variable-rate nutrient and ameliorant application
within management classes is used in numerous agricultural industries
From 2001, vehicle navigation accuracy continues to improve and
controlled traffic/swath farming takes off by 2006. Increasing
positioning accuracy brings reduced input application from section
control by 2007, interrow sowing by 2009 and automated implement
control by 2010.
In 2003, the first general conference for PA in livestock
management (held in Europe) signals the beginning of fine-scale
spatial management to join precision feeding and animal handling
operations.
Plant reflectance sensors applied to manage in-fallow weeds from
2007 offer significant reductions in herbicide applications.
7
In 2008, plant reflectance sensors used to monitor crop
vigour/health/nutrition begin to be used by innovators for fine-scale N
nutrient management.
Future Opportunities in Cropping
PA in cropping is generally moving towards ever smaller spatial units to
increase production efficiencies. As sensing/analytical systems develop and
become more cost-effective, then finer scale management should be less
wasteful and less financially risky than uniform applications or even
management class applications. General areas for development include:
fine scale, real-time, cost-effective estimation of crop/soil nutrient
levels;
fine scale, real-time, cost-effective estimation of profile soil
moisture content;
localised weather predictions;
crop yield monitors for more crops;
efficient, integrated crop quality monitors;
spatial yield prediction/simulation models;
combining crop reflectance sensors with an independent biomass
sensor;
understanding agronomic impact of fine-scale resource variability
and interactions;
autonomous weeding;
public-funded research targeting PA for increased water-use
efficiency and improved farm C and N emission management;
secondary and tertiary education;
improving PA GIS capabilities;
integrating multiple data layers for real-time decision making for
nutrient/irrigation applications;
product tracking and production information traceability; and
more plug and prosper.
8
GRDC Training and Demonstration PA in Practice
Update of Victorian PA Groups
Andrew Whitlock
PrecisionAgriculture.com.au
Precision Agriculture Consultant
16 Queen Victoria St, Ballarat, VIC 3350
0458 312 589
andrew@precisionagriculture.com.au
www.precisionagriculture.com.au
Key Findings/Take Home Messages:
Opportunities exist for farmers to benefit from a range of precision agriculture
technologies and the GRDC funded project has enable the fast-tracking of this
awareness among the Birchip Cropping Group, Southern Farming Systems
and Riverine Plains Group.
Introduction/Background:
Andrew Whitlock is the Victorian facilitator of the four Victorian PA discussion
groups, supporting each group with designing workshops, identifying guest
speakers and coordinating on-farm demonstration sites. The four groups
spread from the Victorian Mallee down to the south-west and across the
north-east. Each group has chosen to investigate their topics of interest and it
has been a pleasure working with all of them.
Presentation Content/ Results:
The groups have tested and demonstrated a range of PA technologies
including:
Crop senor mapping (Green Seeker & Crop Circle)
Satellite imagery (80cm, 5m and 30m pixel resolutions)
EM-38 mapping
Weed seeker
Elevation maps generated from farmers autosteer data
9
Yield monitors and computer programs (predominantly Farmworks and
SMS)
Image 1: Satellite derived NDVI map with 30cm contour overlay clearly
identifies the key driver of paddock variability. This was a consistent response
for most paddocks last year which suffered from excess water.
Demonstration sites and trials have focused on:
Nitrogen response
Potash trials
Variable rate lime
10
Image 2: Satellite derived NDVI with Murate of Potash trial layouts overlay. As
seen here there was no crop response to the potash despite soil test levels
indicating very low soil potassium levels. This demonstration highlights the
power of on-farm trials in order to understand a likely return of investment.
Conclusions:
Thank you to all the farmers who attend the PA groups and the local industry
providers and partnering research organisations who generously support the
associated trials and activities.
Precision Agriculture is obviously site specific in nature and thus it is difficult
to make conclusions other than to say that farmers who are regularly
measuring their farms through PA technologies place themselves in a better
position to make sound economic management decisions.
11
Innovative Practices for Efficient and Profitable Use of N Inputs
Craige Mackenzie
Arable and Dairy farmer from Methven, S.I., New Zealand.
craige.mackenzie@agrioptics.co.nz
Background Roz and I farm near Methven at the foothills of the Canterbury Plains. We run
a 200 ha intensive irrigated arable farm growing a wide range of crops
including wheat, ryegrass, faba beans, carrots, radish, pak choi and hemp,
with almost all of our production for seed.
In 2006 while looking for options for diversity we had an opportunity to
become dairy farmers in a 220 ha, 850 cow high-input and high-output dairy
farm. In 2010 we expanded the operation by leasing a further 110ha and this
season we have increased milking numbers to 1220. The farm is run by a
variable order share-milker at a 24% level and has a staff of 5. The dairy farm
is a pasture-based system and is supplemented with grain, canola meal and
silage.
In 2010 along with our daughter Jemma we started Agri Optics New Zealand
Ltd, a precision agricultural company providing services for farmers, fertiliser
companies, and machinery companies. We provide EM soil survey and data
management services to clients as well as being NZ distributors for Trimble’s
GreenSeeker®, WeedSeeker® and FarmWorks software.
In 2010 we also started a research company to help with new initiatives in the
development and the patenting of our Smart-N fertiliser application system.
Nutrient Management Deep nitrogen tests (mineral N plus an estimate of mineralisable N) are
undertaken on the cropping farm in most crops in the early spring. With these
results we can accurately assess the amount of applied N we need to have
the total amount of N for each crop to reach its potential.
12
Nutrient budgets are undertaken in conjunction with our fertiliser company.
They include all imported nutrients, enabling us to accurately assess our
nutrient requirements. We also use them when looking at our profitability and
where we may be able to reduce fertiliser costs.
The theoretical nutrient removals that scientists have been promoting in the
past are now being tested as we can closely monitor the removal.
Variability On farm we are now grid soil testing to locate the variability that exists in our
paddocks, some of this is natural but a significant amount man-made. We
have had lime spreading that has been less than accurate as often trucks may
not have returned to the same areas when returning with additional truck
loads. Or they may have run loads out when there has been excess product
on the trucks, resulting in some areas with very high pH levels.
The removal of fences and the amalgamation of multiple paddocks for the
development of irrigation and large centre pivots also created variability within
paddocks. Grid soil sampling has identified the issues in these and allows us
to fix all the base variability with the use of variable rate fertiliser applications.
This gives us the ability to deal with man made issues. In the situation in the
following page we used 10 tonne of lime instead of the 75 tonnes we would
have used had we walked the normal trancests. Lime was applied only to
where it was needed. Targetted application for a targetted result.
13
Figure 1 pH map from fields joined together with old fence lines shown in black Mapping EM soil surveys are showing us soil variability, identifying areas that have
different production potential. Because of this we have the ability to reduce N
application in areas that cannot perform as well or are able to avoid areas
such as gate ways, stock camp areas and streams, with confidence, using
prescription maps. We also plan to install variable rate irrigation this year on
our dairy farm based on our EM maps. The variable rate irrigation will
minimise wet lanes reducing lameness while matching our water application
to soil type, using our irrigation resources more efficiently.
GreenSeeker® equipment is now being used to map the variation in N
throughout season to be able to reduce the amount of N that we require in
each individual field. We are seeing vast changes in the field, some because
of soil variability but often because of stock behaviour and movements.
We are also using GreenSeeker® for bare soil imagery, variable rate growth
regulators, variable rate fungicides and variable rate insecticides.
14
GreenSeeker® optical sensor technology enables us to measure, in real time,
a crop's nitrogen levels and variably rate apply the "prescribed" nitrogen
requirements. GreenSeeker® also predicts yield potential for the crop using
the agronomic vegetative index (NDVI). The N recommendation is based on
in-season yield potential and the responsiveness of the crop to N.
GreenSeeker® is similar to satellite and aerial imagery zone management
imagery programmes, however, it is in real time.
Figure 2: GreenSeeker® map identifying areas of high N loading due to uneven effluent spreading patterns under centre pivot. Water/ Irrigation Water will be the biggest issue that will face the world in the future so its
efficient use is very important on a range of levels, politically, environmentally
and financially.
On farm, the efficient use water allows us to reduce our impact on the
environment. With careful irrigation management we can control the nutrient
levels in the soil, matching demand to the plants requirements, keeping
fertiliser applications to a minimum. If farmers are not allowed to use water for
irrigation there will be a resulting increase in the soil’s nutrient bank in the
extended dry periods, through the deposits of urine, dung and applied N. In
heavy rain events this high nutrient bank creates a huge potential to pollute
the environment through its increased potential for nitrate leaching and nitrous
oxide fluxes.
15
Smart-N Using crop-sensing technology (Weedseeker®) we have developed a process
where we can apply liquid N (urea solution with 26% N) between the urine
and dung patches and not on them. This system has up to a 30% saving in
fertiliser N per application. Along with the 30% cost saving through reduced N
application there is also the related reduction in nitrate leaching and reduction
in nitrous oxide (N2O), providing the potential for reduced greenhouse gas
(GHG) emissions without reduced production. This same system can also be
used to apply nitrification inhibitor directly to urine patches if required.
WeedSeeker® can also be used to selectively apply fungicides, insecticides,
fertilisers and other inputs to targeted plants instead of ‘weeds’ in a range of
applications.
Research Initiatives We work closely with regional and central government and other research
providers with research trials looking at improving nutrient efficiency and
monitoring of environmental conditions. With the National Institute of Water &
Atmospheric Research Ltd (NIWA) we have installed an eddy co-variance
tower on the dairy farm to measure CO2 and N2O emissions on a working
farm. Another multi-party research programme has installed a lysimeter to
research the drainage of nutrients to the ground water under an irrigated dairy
farm.
These research programmes will allow us to have real measurements for a
working farm in NZ and provides more accurate data for an Emissions
Trading Scheme (ETS) if agriculture is included in the future.
We need to measure our agricultural practises so we can model them. This
will then allow us to mitigate our impact or perceived impacts on the
environment.
As we increase production output while reducing our inputs through the use
of things like precision agriculture, we will reduce our emissions intensity per
16
kg of product produced. This should help farmers meet our requirements
under a pending ETS and improve profitability on farm.
Profitability and good environmental practises go hand in hand so we need to
be proactive in this area. If farmers are profitable then it will be possible to
invest in technology to help in all areas of sustainability.
17
My experience with Precision Ag Todd Matthews
c/- PO Kyancutta SA 5651
Mobile 0428 563 352
Email: todd.waterbottle@gmail.com
2003
Purchased a KEE X15 on marine beacon and section control for our Hardi
goosekneck boomspray.
2006
Purchased a Autofarm 2cm unit for the seeding tractor, the main driver for this
purchase was to inter-row sow with our 62ft flexicoil on 9 foot spacing. A New
Holland Flexicoil SC380 box with Topcon variable rate was also purchased.
No variable rate maps were created the first year and the urea tank was
turned on and off manually on the sand hills. This was not always done and
as a results we had sand hills that went from dark green to yellow to dark
green. A trimble ez-guide plus and ez-steer were also purchased for the
boom spray.
2007
During the summer we drove around all of our paddocks and mapped the
different soil types. This was a very large job and the maps are not perfect as
some small areas of sand were missed. The maps were then processed using
Topcon’s Quickmap software and a three layer prescription was created.
Small software glitches were noticed and Topcon sent out updates for the
variable rate software.
Problems experienced during seeding were when the target rate was zero the
zynx would flick the tank on and off continuously. Only seed rate and urea
was varied. The first year of inter-row sowing worked well and we were
impressed with the performance of the steering system.
18
2008
New seed rate controller software was installed on the X15 that fixed the rate
zero issue.
Purchased a mobile processor for our JD 9650 sts header to enable us to
yield map. The first years yield maps are split into hundreds of small files and
were burnt to a cd to be processed at a later date.
2010
Made the decision to switch to liquid fertiliser, purchased a variable rate
capable liquid cart from Western Australian manufacturer Techfarm
Engineering. Great care was taken to purchase components (such as flow
meters and flow control valves) which were fully polypropylene or
SS316; which is phosphoric acid resistant. Topcon supplied wiring diagrams
and we were able to control the cart with the fourth channel on our variable
rate MDECU. Having one master on/off switch which is capable of timing each
product so it hits the ground at the same time is very useful.
2011 and into the future...
- Purchase more powerful data management software.
- Dual zone EM38 the entire farm. This data will be using in making
decisions in the future as to whether we should delve, clay spread or
spade non wetting sand. This data may also lead to variable rate maps.
- Purchase a fertiliser spreader and convert it to variable rate, in the
future we are going to use more sulphate of ammonia before seeding
especially on the sand which can take more nitrogen then our flats.
- Variable rate will allow us to do this in one pass. N sensors may also
be an option in-crop although little work has been done in our area.
- I would like to seed 2 cm steering used on all of our equipment.
- Especially the sprayer so all in-crop spraying is done on the same
tracks. The problem at the moment is we have multiple brands of PA
equipment and compatibility with signal type.
19
My PA journey
Mark Bender
WK & LC Bender
The Wattles Lockhart NSW 2656
Introduction/Background:
Our property is situated in Lockhart in southern NSW where I run a cropping
operation with my mum, dad & brother Lindy, Keith & Brent. We grow wheat
barley & canola on our 2000Ha property,
Presentation Content/ Results:
Our PA journey started back in 2002 when we purchased a 46ft FlexiCoil
ST820 bar with 7.2 inch spacings & 2640 air cart that has a variable drive
system, although we didn’t have the capacity or the knowhow to use this we
wanted to keep our future options open. The same year we also brought a
new 2388 & had it setup for yield mapping & we started drawing pretty
pictures.
In 2004 with the purchase of an autonomous EZ Guide Plus lightbar to spray
& sow with, we changed our cropping operations from round & round to up &
back.
In 2005 we invested in an EZ Steer unit to be moved between our tractors &
header.
In 2006 with 4 years of map data & a higher yielding year we noticed our yield
maps had flip flopped, with this we decided to upgrade the air cart & work on
things that were in our control like creating zones around paddock trees
gullies & tree lines.
In 2007 we purchased a second EZ Steer unit & also started using an EZ
Boom auto section boom control system, we also upgraded our seeding GPS
accuracy to 10cm.
20
In 2008 after going through yet another drought year we made the move to
press wheels & decided to head towards 3 meter tramline farming. We moved
our tyne spacings out to 10 inch. Although not a good year we upgraded our
header to a 2588 with a pro 600 screen, with this we setup our STX 450
Steiger with a Case IH Accuguide RTK autosteer system.
In 2010 we changed our seeder to 13.5m wide & changed the tynes to 13 inch
for trash clearance to inter row sow.
In 2011 the decision was made to trade our seeder bar & go to a 12m unit,
our reason was to be able to buy a smaller more affordable header & to be
able to spread the chaff better, we are now setup in a 12m CTF system.
Conclusions:
- Look ahead in any machinery purchases for what you think you may
want to do in the future.
- Ask lots of questions (the best mistake is someone else’s)
- Don’t over complicate things.
21
‘BULLA BURRA’ OUR PA JOURNEY
Robin Schaefer
Bulla Burra Operations
PO Box 182 Loxton SA 5333
Phone: 0417877578
robinschaefer@bigpond.com
Key Findings/Take Home Messages:
Aim for a single PA system across all machines
If you need to harvest with two headers, each with a different PA
system the best way to produce good maps is to harvest pass for pass
Does you PA provider give you good backup, can you ring at 2.00am in
the morning to get help with a problem
You don’t need all the bells and whistles to start PA
Introduction/Background:
Bulla Burra is a collaborative farming venture between my family farming
business ‘Schaefer Enterprises’ and John & Bronny Gladigau. It was
established in 2009 and is situated in South Australia’s Northern Mallee. We
crop 4000ha of our own land and 4000ha of leased and share farmed land.
Our properties are spread from the township of Loxton to 65km SW at the
furtherest point. The land class is typically dune swale, red sandy loam soils.
The sandy rises have rooting depths of up to 2m, the swales tend to have
their rooting depth constrained to 60cm by transient salinity or sodicity. We
also have areas constrained by stone. Annual rainfall is 275 to 290mm and
growing season 180 to 200mm.
Our business specializes in dryland cropping, growing wheat, barley, canola,
rye, and for the first time this year lentils & lupins.
22
We use conservation farming practices, with 97% of our crop sown using No-
Till and the remainder minimum till. We do not run any livestock, though some
of the businesses we share farm for do have a livestock component.
Current Main Machinery
TRACTOR: JD 8430 (PA Platform - JD 2600 Screen RTK)
TRACTOR: JD9100 (PA Platform - JD 2600 Screen RTK & Farmscan canlink
3000)
SEEDER: 12.6 m JD 1870 Conservapac and 1910 Commodity Cart (JD PA
platform)
SEEDER: 12.6 m Horwood Bagshaw Scaribar and 7000 l Aircart (Farmscan
PA platform)
HARVESTER: JD 9760 12.6m Honeybee front (JD PA platform)
HARVESTER: JD 9860 11m 936D front (JD PA platform)
HARVESTER: Case 2166 9m front (Farmscan PA platform) (For Sale)
SPRAYER: JD 4930 SP 36m boom (JD PA platform) (For Sale – changing
over)
23
Presentation Content/ Results:
My PA Journey began in 2002 under our Family business ‘Schaefer
Enterprises’, when we purchased a triple bin variable rate airseeder with a
farmscan controller. Without GPS we initially manually adjusted our nitrogen
fertilizer to target higher rates on our sandy rises. It was a start, but relied on
the operators’ memory and knowledge of the field. In 2004 we upgraded our
harvester, which gave us the ability to yield map. This was done using John
Deere Mapping software. We also fitted a GPS to our seeder enabling us to
automatically variable rate. As we upgraded machinery, our long term aim
was to move towards a controlled traffic system. In 2005 we had roughly
achieved this, most of our wheels were tracking on the same area but we
were still manually steering. Due to erosion risk in our very sandy soils all of
our tramlines were sown. We had two sowing rows at a closer spacing in the
centre of the seeder to mark each run for the header then used the header
tracks in the stubble as our tramlines for spraying and sowing. This allowed
me to summer spray at night without GPS.
With the ability now to yield map and variable rate seed we began, in
conjunction with Mallee Sustainable Farming Inc (MSF), to look at how we
could use variable rate seed and fertilizer application to strategically manage
our inputs. Research in the Mallee through MSF showed strong correlations
24
between EM maps and yield variation across our paddocks, enabling us to
use the EM maps to develop management zones. We had a number of
paddocks EM mapped and used these maps to better target our Nitrogen and
seed inputs. On farm trials and extensive soil testing also showed that we
could move to Phosphorus (P) replacement across our cropping programme.
Initially I had a consultant clean my yield maps and prepare my P replacement
maps, but as my knowledge and confidence grew I began to do this myself. It
was always a challenge though manipulating yield maps in the John Deere
software “APEX” and generating application maps in Farmscans
Datamanager. A number of paddocks also had old potato pivot sites in them,
with very high P levels. This added an even greater level of complexity to
generating the maps especially when the surrounding paddock was low in P.
With a run of dry seasons it was difficult to find the cash flow to increase the
area that had been EM mapped, however during that period we had a number
of years with dry springs which along with some ground truthing, enabled us
to develop pretty good zone maps without the cost of EM mapping. In 2007
we had our seeder, header and sprayer equipped with autosteer via a JD
universal steering kit, which improved the accuracy of our controlled traffic
system.
The formation of Bulla Burra in 2009 generated a new range of PA
challenges. Right from the start, Bulla Burras’ aim has been to maximize
efficiencies and economies of scale and use the latest technology to increase
profitability. However this had to be done in a staged plan as cash flow
allowed. Part of this involved putting aside the controlled traffic system on my
home property for a few years.
The challenges that I had experienced using two different PA platforms (John
Deere & Farmscan) significantly influenced how we approached PA in Bulla
Burra. We wanted to work toward having a common PA platform across all
machines. I could see that it would not only make my PA management easier
but it would also make staff management of PA easier and decrease the
clutter in machine cabins.
25
We were very happy with the backup and support of the local John Deere
dealer and were impressed with the low cost 24hr phone support provided by
John Deere ‘Stellar Support’ so we decided to adopt the John Deere PA
system on all our machines.
Though this was our aim, it is not what we had to work with at our initial setup,
in fact we moved to an even more complex situation. Each of our two
headers had different PA platforms (Farmscan and John Deere) and likewise
both seeders. This made my job very difficult.
For efficiencies we needed to run our two harvesters in the same paddock but
I discovered that trying to merge the farmscan and John Deere yield maps
was a nightmare and did not produce an accurate map. Through this
journey, I discovered that if we harvested run for run with the different
harvesters I could use the data from only one harvester to produce a pretty
good map. We also had a policy that in any paddock if one harvester was
faster or broke down we would aim to only have a chunk of data on one
machine. This significantly helped in managing the data. Most of the data
would usually come from the John Deere harvester so I would only have to
work in Apex to clean the yield maps and build the application maps. There
were still plenty of challenges, moving the maps between Apex and
Datamanager.
Another area we had to be particularly careful with when operating two
harvesters in the same field is their initial calibration and post calibration
checking.
We have found that if you harvest each machine on level ground to the point
when each box is just about to overflow, then weigh the grain and use this for
your calibration you will have a lot better idea how well each machine is
calibrated to the other. If you are both harvesting in the same paddock, in the
same crop, and each machines bin is at the point of overflow with a
proportionately similar amount of tonnes in the box to the initial calibration
load, you know the calibration between the machines is correct. This is most
26
critical as you cannot adjust it later. We also try to check this is correct near
the start of most paddocks so if we do have to adjust the calibration most of
the paddock will be pretty accurate.
As you could imagine with a large operation logistics at harvest time make it
difficult to post verify the calibration for each paddock. Often we site our
loading point where four or more paddocks may be harvested into one place.
It is always difficult to estimate the tonnes remaining in the Mother bin and
field bins when we change paddocks. I have discovered a more accurate way
is to check your calibration per loading point. So if we are harvesting from
four paddocks to one point the harvester tonnes from those four paddocks are
compiled to compare with the total tonnes delivered from the four paddocks.
This needs to be done to ensure that our P replacement maps are as
accurate as possible. Checking the Hectolitre weight of the loads as they are
delivered is also a useful way to determine if you may need to recalibrate
Last year we added a spreader boom to our JD 1910 commodity cart. This
has enabled us to variable rate Sulphate of Ammonia post emergence on our
deep sands and also Urea.
Our 4930 sprayer is equipped to spray at a variable rate as well, though this is
a feature I do not use regularly, I have used it from time to time when we have
wanted to apply a higher or lower rate to a different weed spectrum or size in
a particular zone.
With the introduction of RTK 1cm accurate autosteer on our seeder tractors in
2010 we are working toward inter row sowing. The stubble load and in
particular crops that had gone down during harvest last year made it difficult
to achieve this year. However, where these factors were not an issue it
worked well.
Conclusions:
Looking back over the last 9 years we have had an interesting PA journey.
There has been a lot of speed bumps along the way and at times it has been
very frustrating but it has also been very rewarding. The savings and yield
27
increases we have achieved through improved management of our inputs
have been substantial. The reduced stress levels and increased productivity
that auto steer has bought is great. The opportunity to put unskilled labour
onto seeders which can still apply inputs at the right rate to the right area of
the paddock makes finding labour that much easier. In short we have come a
long way in that time, so what does the future hold for Bulla Bura?
We are aiming to integrate weed seeker into our weed management system.
We plan to change over the Horwood Bagshaw seeder bar and cart to have a
common PA platform across all our machines.
We will continue to work toward a controlled traffic system as we upgrade
machinery.
One of the properties that we lease has a requirement restricting the
chemicals we are able to use, with this in mind me we are researching inter
row weed spraying with knockdowns in cereals.
There is still plenty to do and I’m sure there’ll be plenty of headaches along
the way but gee it’s great fun when you have great toys, which actually save
you money.
28
Innovative Irrigation – PA related irrigation tools
John Hornbuckle
CSIRO Sustainable Agriculture Flagship
Irrigation Research Scientist
Research Station Road, Griffith, NSW, Australia, 2680
Phone 0429862920
Email Address john.hornbuckle@csiro.au
Website www.irrigateway.net and
http://www.csiro.au/people/John.Hornbuckle.html
Key Findings/Take Home Messages:
Increased opportunity for management intervention in irrigation
systems and water scarcity is likely to see an increase in the use of irrigated
PA tools and techniques into the future
Innovative tools which have been developed in the irrigation industry
now provide low cost, site specific spatial information on crop water demands.
Previously, this detailed spatial crop water demand information has been
missing, and is a key component for implementing irrigated PA
Introduction/Background:
In 2011 irrigation still remains essential to the production of food and fibre on
a global scale. Water scarcity across the globe has seen increased needs for
improved management of the limited water resources available. It is clear that
gains can be made to water use efficiency by precisely matching the spatially
distributed crop water needs. These demands are a function of the impacts
imparted on the crop from a number of sources such as soils, climate, rainfall
and indeed irrigation variations across fields. These effects and associated
variability may be due to natural (soils) or manmade causes (irrigation
systems). Understanding how these factors combine and ultimately effect
crop production potentially allows irrigators to match crop water demands to
maximize the potential of each individual plant within the system to maximize
yield and/or maximize water use efficiency. This is provided that adequate
29
infrastructure in terms of tools and irrigation systems are capable of turning
this knowledge into practice. This paper presents two innovative PA based
tools which can be used to better understand the variability induced by
irrigation systems and secondly determine crop water needs spatially across
fields.
Understanding Irrigation System Variability:
A reduction in seasonal water application through improvements in irrigation
system uniformity is one of the key opportunities for improving water use
efficiency in irrigated agricultural systems. Compared to irrigation systems
with high distribution uniformity (DU), systems with low uniformity require
increased irrigation supply to ensure all plants receive adequate water and
achieve maximum yields. This increased irrigation leads to unnecessary water
losses via deep drainage, soil evaporation and cover crop transpiration.
Similarly, if growers schedule irrigation to avoid excessive vegetative growth
or employ deficit irrigation strategies the risks of yield penalties and other
adverse effects of water stress are increased when DU is low.
‘DU Calculator’ (www.irrigateway.net/tools/du/) is a practical and versatile tool
developed by CSIRO to provide users of micro-irrigation systems with spatial
information regarding irrigation system performance based on point
measurements of emitter rates (Hornbuckle et al. 2009). Growers are
provided with instructions for measuring emitter rates and recording GPS
coordinates of emitters and block boundaries. This data, and basic block
information, is entered in the DU website. A report in .pdf format is
automatically generated and emailed back to the user which contains maps of
the emitter application rates, seasonal applied water, seasonal applied
fertigation and calculates the DU for their field. Spatial patterns of emitter
rates revealed in these maps are of particular value in determining possible
causes of poor system performance (Hornbuckle et al. 2008). For example,
decreasing emitter rates at increasing distance from the supply lateral indicate
a pressure problem or design flaw; low rates associated with a particular line
could indicate a hole or kink in that line and a random occurrence of low rates
throughout a block may indicate blocked or damaged emitters.
30
Tools such the DU Calculator allow irrigators to begin to investigate potential
site specific impairments to crop production caused through irrigation system
issues which might be maintenance or design issues and take active steps to
correct the problems.
Figure 1 presents drip irrigation emitter variability across five vineyards
spread throughout Australia over a three year period measured during a
national GWRDC field study. What can be clearly seen is that with all systems
there is considerable variation across single irrigation management units. As
an example, at Griffith, the non-uniformity of emitter rates has a significant
effect on the total volume of irrigation water applied to the vineyard over the
season even with a system DU of 85%. At the extreme points minimum
irrigation application was calculated as being 3 ML/ha/season and at the
maximum 5.1 ML/ha/season. The measured DU of 85% indicates that the
lowest quarter of the block received on average 25% less water than the
highest quarter of the paddock. This corresponds to a 25% difference in the
volume of water applied over a season. This non uniformity also applies to
fertilizer. The fertigation strategy aimed to apply 25 kg/ha of Nitrogen (N26)
through the season, consisting of 12.5 kg/ha during November and 12.5 kg/ha
during April. Dosages were based on the design emitter rate of 5.25 L/hr/vine.
Due to the non-uniformity of water application there is an equivalent non-
uniformity associated with the applied nitrogen. Seasonal application of
nitrogen ranged from a minimum of 16 kg/ha to a maximum of 26 kg/ha. The
combined effects of water and fertiliser have a large impact on vine growth.
31
Figure 1 Map of emitter rates in five vineyard blocks across Australia over a three year period, collected as part of the GWRDC five sites project.
Figure 2 shows the spatial relationship between a satellite derived vegetation
index and the drip application rates across two of these vineyards. It can be
seen that there is a strong relationship between these two measurements.
This is expected due to the influence of water application on the growth of the
vines. It can be seen in Figure 2 that high NDVI areas generally correspond
to areas that have high drip application rates. There appears to be a useful
relationship between drip emitter application rates and the amount of
32
vegetation. This demonstrates that the variation in emitter rates has a large
effect on the performance of the crop. Indeed, this information also provides
an insight into the potential benefits of PA practices in irrigation systems if this
variability can be managed for in terms of meeting crop water demands.
Figure 2 Relationship between drip application rate and satellite Normalised Difference Vegetation Index (NDVI) at the Griffith and Tatura sites
33
Providing spatial crop water demand information:
In order to provide irrigators tools for implementing PA practices into everyday
management a critical need is the ability to understand spatial water
requirements. Few existing systems for irrigation scheduling have this
capability. The most widely adopted irrigation scheduling tool, the soil
moisture sensor is generally a point source measurement and does not meet
the requirements for providing extensive spatial information on crop water
demand. More recently new integrated sensor and data management systems
which make use of satellite data and on-ground weather stations have been
developed which begin to provide a viable foundation for implementing
Irrigated PA practices. Figure 3 presents one such system known as Irrigate
which has been developed in Australia to provide low cost, site specific
irrigation water management information in both a spatial and temporal
context. At the centre of the system is the Irrigate server which acts as a data
collection portal for the various data feeds and a calculation engine to convert
these data into useable irrigation scheduling information in a spatial context.
Figure 3 Overview of the IrriSAT system showing information data flow
The use of weather station information in irrigation scheduling has long been
used in the scientific domain for predicting crop water requirements and
34
scheduling irrigations (Allen et al. 1998). The general approach that is widely
accepted is given in Allen et al. (1998) and is based on the application of
reference station evapotranspiration figures which are collected over a grass
reference surface (ETo). This reference evapotranspiration is used to
represent the climatic conditions under which evapotranspiration takes place,
which is then used to calculate actual evapotranspiration (Etc.) for specific
crops by multiplying the Eton by a specific crop coefficient (Kc). Generally four
Kc values are used over the growing season, - initial, mid, full and late. The
crop coefficient takes into account differences in canopy cover, stomata
characteristics, aerodynamic properties and albedo, which affect the rate at
which crops evapotranspire compared to the reference crop ETo.
A major limitation of this approach has been that the crop coefficient (Kc) is
specific to a particular crop, irrigation system, soil and management. There
have been a number of approaches used to derive crop coefficients which
measure actual crop evapotranspiration and compare this to reference
evapotranspiration allowing a Kc to be developed for that crop. However,
these methods (e.g. eddy covariance, bowen ratio, water balance) are
expensive and require a high level of expertise to implement and again only
produce a single site specific kc. As slight changes in agronomic
management, soils and irrigation regimes affect the crop coefficient this also
makes it difficult to derive a specific crop coefficient for an individual crop.
A number of authors have reported on strong correlations between vegetation
indexes and canopy cover in irrigation contexts (Hornbuckle et al. 2008).
Canopy cover is a direct driver of crop water use and hence allows a direct
relationship to be developed between NDVI satellite derived values and crop
coefficients which take into consideration specific agronomic and
management conditions for individual crops. This allows a specific crop
coefficient to be derived on an area as small as 30x30m when data from a
suitable satellite platform such as the Landsat Thematic Mapper satellite is
used (Hornbuckle et al. 2010). This information, when combined with ETo
weather station data commonly available throughout Australia provides
35
relevant spatial irrigation crop water demand information which can be used
for irrigation scheduling.
The IrriSAT server directly sources weather station information and NDVI
satellite information which are data feeds provided to the server. It then uses
this information for determining actual crop water use for the irrigators specific
crop and management situation and runs a daily spatial water balance model.
This data is then provided directly to an irrigator on a daily basis. Tools such
as IrriSAT are now providing information which can be used to begin to
implement irrigated PA techniques. The challenge is now how to integrate
these tools into an infrastructure (irrigation system) and management package
that can be used to maximise the benefits of PA approaches in irrigation
systems.
Conclusions:
Application of PA practices to irrigation systems offer many potential
advantages. Critical to understanding the benefits of PA in irrigation systems
are tools and techniques which allow the integration of PA thinking into the
everyday management of irrigation systems through controlling irrigation
water which is generally the single biggest driver of crop yield. This paper has
presented two innovative tools which contribute to this understanding.
Considering the increased opportunity for management intervention in
irrigated systems and the increased pressures being placed on water
resources the use of Irrigated PA (IPA) techniques is likely to see increase
interest in the coming decade.
References:
Allen, R.G., Pereira, L.S., Raes, L.S., Smith, M. (1998) Crop evapotranspiration - Guidelines for computing crop water requirements - FAO Irrigation and drainage paper 56, http://www.fao.org/docrep/X0490E/X0490E00.htm Hornbuckle, J.W., E. Christen, D. J., Smith, L. McClymont, I. Goodwin, and D.M. Lanyon (2007) Measuring drip irrigation distribution uniformity in irrigated vineyards and understanding its effects on vine vigour, ASVO PROCEEDINGS • WATER, FRIEND OR FOE? https://www.asvo.com.au/proceedings/?action=view&id=29 Hornbuckle, J.W., Car, N., Christen, E.W. & Smith, D.J. (2008) Large scale, low cost irrigation scheduling – making use of satellite and ET0 weather station information, IAL Conference, Melbourne, May 2008
36
http://www.irrigation.org.au/assets/pages/75D4C8BD-1708-51EB-A6816CD6992AC045/45%20-%20Hornbuckle%20Paper.pdf Hornbuckle, JW, Car, J. Destombes, D. Smith, E.W. Christen , I. Goodwin & L.McClymont (2009) Measuring, Mapping and Communicating the Effects of Poor Drip Irrigation Distribution Uniformity with Satellite Remote Sensing and Web Based Reporting Tools, Swan Hill, October, 2009 http://www.irrigation.org.au/IAL_IDC_Conf_2009/Hornbuckle,%20John%20Abstract%2057.pdf Hornbuckle, J.W., Car, N.J., Christen, E.W., Stein, T.M. and Williamson, B. (2009). IrriSatSMS - Irrigation water management by satellite and SMS - A utilisation framework. CRC for Irrigation Futures Technical Report No. 01/09 and CSIRO Land and Water Science Report No. 04/09. http://www.irrigateway.net/publications/irrisatsms_v_60_finalwAppendix.pdf Hornbuckle, J., Christen, E., Car, N. & Smith, D. (2010) Convenient and low cost irrigation scheduling – an opportunity for irrigators, Australian irrigation Conference 2010, Darling Harbour Sydney, 8-11th June 2010 http://www.irrigation.org.au/assets/pages/44FD5F05-95F8-8584-4F8E654E66264702/Convenient%20and%20low%20cost%20Irrigation%20scheduling%20-%20Hornbuckle%20.pdf
37
Precision Agriculture in Grazing Systems
Mark Trotter
University of New England Precision Agriculture Research Group
Research Lecturer – Precision Agriculture
Address: PARG Stokes Building (C24) UNE Armidale NSW 2351
Phone: 0447 441 841
Email Address: mtrotter@une.edu.au
Website: www.une.edu.au/parg
Facebook: Precision Agriculture Research Group University of New England
Australia
Key Findings
1) Precision Agriculture in grazing systems is a rapidly evolving field of
research and development with significant investments being made from both
research agencies and commercial entities.
2) One of the key technologies, the ability to monitor the location and
behaviour of livestock is becoming a commercial reality with several real-time
monitoring systems currently being trialled. This technology is essentially the
“yield monitor” of the grazing industry.
3) The integration of information from soil, plant and animal sensors could
lead to a number of potential benefits including the monitoring and
management of spatial variability in grazing pressure, site specific nutrient
management strategies, more precise timing of grazing rotations and better
understanding of the impacts of livestock in mixed farming systems.
4) The benefits of PA in grazing systems will be limited by the capacity to deal
with the large volumes of data being generated by the sensors. Research is
required into data management and analysis techniques that will enable
producers to glean meaningful information from these systems.
Introduction
The cropping and horticultural industries have been the beneficiaries of
Precision Agriculture (PA) technologies for many years; however there has
been comparatively little work done examining the potential for PA in grazing
systems. Despite this, there is growing recognition amongst researchers and
38
producers that there is potential for PA to increase productivity and efficiency
in animal and pastoral systems (Hacker et al., 2008; Schellberg et al., 2008;
Virgona and Hackney, 2008; Trotter, 2010).
This paper briefly reviews the sensors and technologies available for
monitoring the various components of the grazing system and relates these to
some of the PA management applications that are currently being
investigated in this context. The challenges and limitations that face the
grazing industry in terms of development and adoption of PA sensor and
management strategies are also discussed.
Sensing technologies for precision grazing
There are several technologies available to the grazing industry to help
understand spatial and temporal variability in the soil, pasture and animal
subsystems.
Soil monitoring technologies for pastures
Electromagnetic induction (EMI) instruments have been extensively used in
cropping and the derived apparent electrical conductivity (eCa) is known to
have a relationship with a number of soil properties including soil moisture,
soil texture, soil depth and ion content. This sensor has clear potential for
application in understanding variability in pastoral soils. The limited reports of
their application in pastures demonstrated some relationships between eCa
and plant species, soil characteristics and pasture productivity (Guretzky et
al., 2004; Serrano et al., 2010).
Remote sensing technologies for pastures
Vegetation monitoring tools are probably the most common and commercially
mature PA tools available to pasture and rangeland managers with
commercial remote sensing products currently on the market, for example the
Pastures From Space (PFS). A large amount of research has been
undertaken concerning the use of satellite based remote sensing, primarily
using low resolution multispectral (Boschetti et al., 2007) and hyperspectral
(Numata et al., 2008) systems. More recently, high resolution systems
(Dutkiewicz, 2006) have been investigated.
39
Proximal sensing technologies for pastures
Active optical sensors like GreenseekerTM and Crop CircleTM offer alternative
ground-based platforms for deriving similar measures as the PFS system
(Flynn et al., 2008; Trotter et al., 2010a). Other proximal plant sensors
investigated for dairy pasture systems include ultrasonic and optical plant
height sensors (Yule et al., 2006; Awty, 2009). Digital image analysis of plant
morphology is also gaining interest as a means of weed identification
(Schellberg et al. 2008). Recent work has investigated the potential for optical
sensors to predict pasture quality parameters (Pullanagari et al., 2011).
Animal monitoring technologies
In recent years there has been a rapid growth in research and development
activity in monitoring the spatial behaviour of livestock. This is largely a result
of low-cost global navigation satellite systems (GNSS) tracking technology
(Trotter et al., 2010c). Whilst simple store-on-board collar tracking units are
currently used in research, a number of real-time tracking systems are known
to be in development for commercial application; particularly based around an
ear tag form factor (Stassen, 2009; Andrews, 2010). The application of spatial
livestock monitoring ranges from simply reporting the current location of stock,
recording movement and grazing pressure to health and welfare monitoring
(Trotter, 2011).
Potential PA management applications in grazing systems
Understanding and managing spatial variability in grazing systems
Researchers have found large variations in spatial landscape utilization by
livestock (Trotter and Lamb, 2008). Whilst this comes as no surprise to
producers, there is a great deal of interest in how this information might best
be exploited to increase production and efficiency of grazing systems. Spatial
livestock monitoring systems are essentially the “yield monitor” of the grazing
industry (Trotter et al., 2009) and enable producers to quantify this variability
enabling informed management decisions. Research is currently underway
which is examining how producers might best use this information and
manipulate livestock using strategic paddock design, supplementary feed and
water placement.
40
Site specific management of nutrients
Of particular interest to the grazing industry is the potential to use the data
generated by PA sensing technologies to manage the spatial variability in soil
nutrients. Whilst soil and vegetation sensors may be used to provide an
indication of underlying nutrient status of an area, animal monitoring
technologies can indentify grazing events (nutrient uptake) and urination and
defecation events (nutrient redistribution). By understanding how nutrients are
moved around a farm by animals producers can start to explore site specific
fertiliser management strategies (Trotter et al., 2010b) or even targeted
application of nitrification inhibitors (Betteridge and Costall, 2010).
Initial results from trials linking ALSM with urine sensors suggest that there
can be significant spatial variability in the deposition of urine onto dairy
pastures during grazing events, the extent of which may warrant consideration
of site specific nutrient management strategies (Draganova et al., 2010).
Optimizing pasture utilization in rotational systems
The integration of plant vegetation sensors and livestock monitoring systems
may also assist in better managing livestock in rotational grazing systems.
Research is currently underway investigating the potential for spatial livestock
monitoring tools to predict the grazing stage of cattle. Preliminary results
indicate that the activity of livestock changes over the period of time they are
grazing a paddock and this can be measured using spatial monitoring
systems (Roberts et al., 2010). This could ultimately lead to systems which
predict when animals should be rotated to new paddocks to optimize animal
production or pasture persistence.
Monitoring the effects of livestock in mixed farming systems
Integrating soil, plant and livestock data may also help producers better
understand the spatial variability apparent in mixed farming systems. Trials
are currently underway which examine the spatial variability in livestock and
how this affects soil compaction (Guppy et al., 2011).
Limitations and challenges
One of the key challenges that has become apparent in the development of
precision grazing systems is the capacity for producers, researchers and
industry to manage the data. PA in cropping systems is notorious for
generating large volumes of data; however livestock systems are likely to
41
dwarf this by many orders of magnitude. As well as base data layers such as
soil EM38, remote and proximal vegetation sensors the livestock data which
can be generated is enormous. For example one herd of 100 cows having
their position logged every 5 minutes over 1 year results in over 10 million
data points being logged. All this data needs to be stored, processed and
analysed to provide producers with meaningful information that enable the
management strategies previously discussed.
Another key issue for the industry is the availability of experts to assist
graziers. There is already some suggestions from technology developers in
this field that obtaining people with the necessary skills to deal with the
hardware, software and data issues is difficult and likely to get harder as
demand outstrips supply. There is a need for education and training at all
levels of the industry to facilitate adoption of PA technologies however a
particular focus should be made on training on information technology in
agriculture at the tertiary level to provide the experts required to work as
professionals in this field.
References
Andrews, C. (2010) Translating industry research into farm profit: a commercially viable approach to precision livestock and remote monitoring in Australia. In: Trotter, M., Lamb, D.W., Trotter, T.F. (Eds.), 1st Australian and New Zealand Symposium on Spatially Enabled Livestock Management Precision Agriculture Research Group, University of New England, Armidale, NSW, Australia. Awty, I. (2009) Taking the guess work out of feeding cows. In: Trotter, M.G., Garraway, E.B., Lamb, D.W. (Eds.), 13th Annual Symposium on Precision Agriculture in Australasia. Precision Agriculture Research Group, The University of New England Armidale, Australia, p. 73. Betteridge, K., Costall, D., (2010) Why does it matter where animals urinate? In: Trotter, M., Lamb, D.W., Trotter, T.F. (Eds.), 1st Australian and New Zealand Spatially Enabled Livestock Management Symposium. Precision Agriculture Research Group University of New Engand Armidale, p. 1. Boschetti, M., Bocchi, S., Brivio, P.A., (2007) Assessment of pasture production in the Italian Alps using spectrometric and remote sensing information. Agriculture Ecosystems and Environment 118, 267-272. Draganova, I., Yule, I.J., Hedley, M., Betteridge, K., Stafford, K., (2010) Monitoring dairy cow activity with GPS-tracking and supporting technologies. In: Kholsa, R. (Ed.), 10th International Conference on Precision Agriculture. Colorado State University, Denver USA.
42
Dutkiewicz, A.L., Megan M, Ostendorf, Bertram Franz, (2006). Per-paddock mapping of perennial lucerne with spot imagery. 13th Australasian Remote Sensing and Photogrammetry Conference. ARSPC, Canberra. Flynn, E.S., Dougherty, C.T., Wendroth, O., (2008). Assessment of pasture biomass with normalised difference vegetation index from active ground-based sensors. Agronomy Journal 100, 114-121. Guppy, C., Trotter, M., Flavel, R., Schneider, D., Roberts, J., Jasper, S., Young, I., (2011) Preliminary soil compaction results from McMaster Research Station GRDC Research Update. Guretzky, J.A., Moore, K.J., Burras, C.L., Brummer, E.C., (2004) Distribution of Legumes along Gradients of Slope and Soil Electrical Conductivity in Pastures. Agron J 96, 547-555. Hacker, R., Thompson, T., Murray, W., Alemseged, Y., Timmers, P., (2008) Precision pastoralism - advanced systems for management and integration of livestock and forage resources in the semi-arid rangelands in south eastern Australia. 8th International Rangelands Congress (a joint meeting with the 21st International Grassland Congress). Guangdong Peoples Publishing House, Hohhot, Inner Mongolia, China. Numata, I., Roberts, D.A., Chadwick, O.A., Schimel, J.P., Galvão, L.S., Soares, J.V., (2008) Evaluation of hyperspectral data for pasture estimate in the Brazilian Amazon using field and imaging spectrometers. Remote Sensing of Environment 112, 1569-1583. Pullanagari, R., Yule, I.J., King, W., Dalley, D., Dynes, R., (2011) The use of optical sensors to estimate pasture quality. International Journal on Smart Sensing and Intelligent Systems 4, 125-137. Roberts, J., Trotter, M.G., Lamb, D.W., Hinch, G.N., Schneider, D.A., (2010) Spatiotemporal movement of livestock in relation to available pasture biomass. Food Security from Sustainable Agriculture 15th Australian Society of Agronomy Conference. Australian Society of Agronomy, Lincoln, New Zealand. Schellberg, J., Hill, M.J., Gerhards, R., Rothmund, M., Braun, M., (2008) Precision agriculture on grassland: Applications, perspectives and constraints. European Journal of Agronomy 29, 59-71. Serrano, J., Peça, J., Marques da Silva, J., Shaidian, S., (2010) Mapping soil and pasture variability with an electromagnetic induction sensor. Computers and Electronics in Agriculture. Stassen, G., (2009) Sirion, the new generation in global satellite communications: livestock GPS tracking and traceback. In: Trotter, M.G., Garraway, E.B., Lamb, D.W. (Eds.), 13th Symposium on Precision Agriculture in Australasia: GPS Livestock Tracking Workshop. Precision Agriculture Research Group The University of New England, Armidale, Australia, pp. 68-70. Trotter, M., (2011) Applications of autonomous spatial livestock monitoring in commercial grazing systems. In: Gonzalez, L., Trotter, M. (Eds.), 2011 Spatially Enabled Livestock Management Symposium. Society for Engineering in Agriculture, Gold Coast, Australia.
43
Trotter, M.G., (2010) Precision agriculture for pasture, rangeland and livestock systems. In: Dove, H., Culvenor, R. (Eds.), Food Security from Sustainable Agriculture15th Australian Agronomy Conference. Australian Society of Agronomy, Lincoln, New Zealand. Trotter, M.G., Lamb, D.W., (2008) GPS tracking for monitoring animal, plant and soil interactions in livestock systems. In: Kholsa, R. (Ed.), 9th International Conference on Precision Agriculture, Denver, Colorado. Trotter, M.G., Lamb, D.W., Donald, G.E., Schneider, D.A., (2010a) Evaluating an active optical sensor for quantifying and mapping green herbage mass and growth in a perennial grass pasture. Crop and Pasture Science 61, 389-398. Trotter, M.G., Lamb, D.W., Hinch, G.N., (2009) GPS livestock tracking: A pasture utlisation monitor for the grazing industry. In: Brouwer, D., Griffiths, N., Blackwood, I. (Eds.), News South Wales Grasslands Society Conference. New South Wales Department of Primary Industries, Taree, Australia, pp. 124-125. Trotter, M.G., Lamb, D.W., Hinch, G.N., Guppy, C., (2010b) GNSS tracking of livestock: towards variable rate fertilizer strategies for the grazing industry. In: Kholsa, R. (Ed.), 10th International Conference on Precision Agriculture. Colorado State University, Denver, Colorado USA. Trotter, M.G., Lamb, D.W., Hinch, G.N., Guppy, C.N., (2010c) Global Navigation Satellite Systems (GNSS) livestock tracking: system development and data interpretation. Animal Production Science 50, 616–623. Virgona, J.M., Hackney, B., (2008) Within-paddock variation in pasture growth: landscape and soil factors. In: Unkovich, M. (Ed.), 14th Australian Agronomy Conference. Australian Society of Agronomy, Adelaide, South Australia. Yule, I.J., Fulkerson, W.J., Lawrence, H.G., Murray, R., (2006) Pasture Measurement: The First Step Towards Precision Dairying. 10th Annual Symposium on Precision Agriculture Research and Application in Australasia. Australian Centre for Precisision Agriculture, University of Sydney, The Australian Technology Park, Eveleigh, Sydney, p. 6.
44
Surveying with Sensors for Soil Mapping Michael Wells
Precision Cropping Technologies.
Crystal Brook, SA
michael@pct-ag.com
Background
Collecting and mapping good quality yield data is an important step in
gathering information and building knowledge about variability within
paddocks. They show where variability is occurring, how significant it is and
importantly the impact it has on profit. With this can arise many questions
about what caused the variability, can I fix it and how should I change my
management. The variation in crop performance can be due to many reasons
both agronomic and management driven. However, factors contributing to the
variability in the soil moisture environment and importantly the physical and
chemical soil properties that impact on water storage and use efficiency are
often the most significant influences of variations in yield.
Our understanding of where soil conditions change over a field and their
potential relationship with yield can be greatly enhanced if soil properties can
be mapped.
Typically in Australia Electro Magnetic Induction (EM Survey) has been used
for mapping of spatial soil variability. There are several types of instruments
being used currently with many sensing multiple depths simultaneously. Due
to the EM sensor response being strongly influenced by agronomically
important soil properties (soil texture, water content and salt levels) the
outcomes from an EM survey, in many instances, delivers good results in that
it corroborates with patterns in yield maps and has explained a majority of the
variability in production.
In some regions in Australia where the Gamma Radiometrics technology is
commercially available it has been found to provide valuable information
about soil variability as it can detect and discern between specific soil
45
properties that are different to the EM technology. In these situations the two
technologies can complement each other and combine to enhance the overall
ability to map important soil variability on a local scale.
Assessing the applications for both soil sensing technologies on South
Australian soils for managing variability is one of the components of a 3 year,
SAGIT funded project ‘Increasing economic returns of agronomic
management using Precision Agriculture’.
Gathering the Information.
How is the data collected?
In most instances the EM instrument is towed on a sled or suspended above
the ground behind a 4WD or ATV at an approximate speed of 15-20km/hr.
The survey area is covered in parallel transects which vary according to the
agricultural enterprise and intended application of the survey. As such the EM
instrument does not need to come into contact with the soil.
The Gamma Radiometrics (GR) data is
generally collected simultaneously with the
EM data on a multi sensor platform. In
figure 1 the GR instrument is mounted on
the front of the vehicle. The information
from the respective instruments is
combined with GPS coordinates and logged
into an on board computer.
Figure 1 Multiple Sensing for EM and Gamma Radiometrics combined with RTK GPS for additional elevation
data logging (photo courtesy of Precision Agronomics Aust)
What do the soil sensors measure?
EM: The Electro Magnetic sensors are electrically powered and when turned
on this power runs through induction coils and creates an energy field in the
space all around the sensor. When in contact with the ground the transmitter
coil emits an electromagnetic signal which passes down through the soil
profile. This generates a second magnetic field in the soil that is detected by a
46
receiver on EM instrument. The strength of the second magnetic field varies
with the degree conductivity of the soil and is referred to as the apparent
electrical conductivity (ECa).
The EM instrument in responding to the conductivity of the soil is most
strongly influenced by changes in clay content, soil water content and soil
salinity. The electrical conductivity of the profile is never governed by a single
soil property rather by a combination, in constantly varying proportions.
Gamma Radiometrics: Soil profiles contain natural radioactive isotopes which
in turn provide gamma ray emissions from the soil. These gamma ray
emissions are detected the Gamma Radiometrics instrument. As the
instrument is moved over the survey area it detects changes in these gamma
ray emissions in approx the top 40cm of the soil profile and provides
information about soil forming parental geology of the soil, changes in clay
and levels of gravel. The main elements measured for use in agriculture are
Potassium (K), Thorium (Th) and Uranium (U) and Gamma Radiometrics
Total Count.
The technologies working together.
Whilst the use of the Gamma Radiometrics technology is relatively new
compared to the EM there are already some identified situations where they
complement each other. These are few examples.
Where soil conductivity levels are very low the Gamma Radiometrics
distinguishes between deep sand and gravel profiles.
Where the soil profiles are clay with areas of gravel the EM will tend to
distinguish between these better.
The Gamma Radiometrics can help separate clay profiles from those
that are saline which are both high conductivity for the EM
instrument.
On the Northern sand-plain areas of WA the Gamma Total Count has
been useful in identifying soil profiles with better water holding
capacity.
47
Figure 2. Shows the four major elements from the Gamma Radiometrics survey used for PA applications.
Ground Truthing and Interpreting the information.
When the soil sensor moves over the ground it detects a change in soil
conditions by recording a different number which is then used to create a
map. When viewing the EM/Gamma maps it is indicating apparent differences
in soil profile conditions over the surveyed area and that one area is different
to another but it doesn’t provide information on exactly what it is about the soil
that is changing. It is not reliable to use findings from other districts or farms to
interpret your own soil sensor maps. This can only be achieved by ground
truthing with soil testing and is a critical and essential step in understanding
the nature of the soil changes on your farm and therefore gaining the best
long term use of your soil sensor survey.
An advantage of the soil sensor maps is that in detecting changes in soil
profile conditions they provide a guide to where to collect the soil cores in
different locations.
When collecting soil cores, the depth to which the respective soil sensors
detect apparent soil change should be considered. Soil cores for interpreting
48
the EM survey are generally collected to a depth of 0-60 or 0-90cm and those
for Gamma Radiometrics the top 30-40cm. These cores are segmented and
analysed for a range of soil properties.
Farmers and agronomists have used this combined information to manage a
specific issue they were interested in from the outset like targeting gypsum
applications to address sodicity, liming or perhaps in deciding where to clay
spread or delve. The ground truthing process can sometimes discover new
opportunities to manage variability on the paddock or farm but importantly
when there is a relationship between the soil sensor survey map and the yield
maps for the same paddock they are equipped with the knowledge of what is
driving the variation in production and therefore can make an informed
decision on how to manage it.
The following are interesting early findings from the initial EM and Gamma
Radiometrics survey on paddock B3 at Mark Modra’s property, ‘Shadow
Brook’, Wanilla. This is one of locations for the SAGIT supported project
investigating the use of soil sensor surveys for managing local agronomic
issues with PA technologies.
Figure 3. DualEM Shallow and the Gamma Radiometrics Thorium layer with Canola yield from 2006.
GIS software was used to analyse these two soil layers with the canola yield
to investigate if the soil variation they have mapped is influencing the
variations in productivity.
49
Figure 4 shows a trend of
the 2006 canola yield
increasing as the
Thorium response
increases over the
paddock.
Figure 4.
4
1
Zone 1
Zone 4
Figure 5. Gamma Thorium zones
The Gamma Thorium layer was then further segregated into zones (figure 5)
to enable analysis with other seasons of yield to assess if similar trends
existed. Table 1 provides a summary showing that in each of the seasons that
yield data was available the Gamma Thorium zones ranked the same for yield
outcome. Initial investigative soil coring revealed the low Thorium/lowest yield
zone as having sand over heavy sodic clay and coupled with low landscape
position and slope, poor drainage was a major limiting factor. In Zone 4 the
highest Thorium/highest yield the profile was light sandy/gravel loam, deeper
profile and greater slope and subsequent better drainage.
50
Table 1. Yield summary of zones
AreaBarley 2005
Canola 2006
Canola 2009
Wheat 2010
1 19.9 1.84 0.25 0.62 2.492 27.3 2.27 0.40 1.16 3.503 28.6 2.63 0.64 1.57 4.124 16.6 2.89 0.80 1.96 4.27
Zone
Take home messages:
1. Take the time to collect good quality yield data - you only have one
opportunity to capture it and it is such a valuable layer of information.
2. Map your soil variability to understand your soil type’s and their
production capacity starting with a soil sensor survey.
3. Ground truth the soil sensor maps using soil coring. The soil sensor
maps are site specific and require local interpretation to build
knowledge that is relative to your farm, agronomic and management
planning.
4. Focus on local issues on your farm and use the best PA tools for
mapping and managing it better. Start with just one issue like targeted
gypsum or lime or perhaps phosphorus management.
5. Use the soil sensor mapping to create zones for conducting simple on-
farm trials using the yield monitor/mapping to evaluate different
treatments in zones and the economics of targeted inputs using VRT.
Acknowledgements: South Australian Grains Industry Trust Fund (SAGIT), Precision Agronomics Australia, Esperance, WA, Craig Topham, Agrarian Management, Geraldton.
51
The challenge of reducing information
and learning costs in Precision Agriculture
Frank D’Emden
Precision Agronomics Australia
Technology Development Manager
PO Box 2418 Esperance WA 6450
0488 917 871
frank.demden@precisionag.com.au
www.precisionag.com.au
Key Findings/Take Home Messages:
The economic benefits of Precision Agriculture (PA) are generally a result of
more efficient production through reduced input risks and increased labour
efficiency. Recent projects undertaken by Precision Agronomics Australia
reveal benefits from variable rate applications of lime, gypsum, potash and
phosphorous ranging from $10/ha to $25/ha.
However, the information and analysis (learning) costs required to achieve
these benefits are a barrier to further adoption by growers and agronomists. It
could be argued that while opportunities exist to reduce these information
costs through the development of specialized precision agricultural consulting
services, existing service providers are currently under-prepared to meet the
growing demand for such services.
Further economic gains from PA are likely to be achieved through improved
analysis and understanding of existing spatial data (e.g. EM, radiometrics,
landscape, soil properties, biomass and yield) on a region-by-region basis,
and how this data fits together to inform agronomists and growers about
optimal variable rate decisions. The development of simplified yet powerful
spatial analysis software and variable rate hardware technologies are required
to handle large datasets and reduce the costs of operator downtime.
52
Introduction/Background:
Adoption of PA, particularly automated variable rate (VR) application of
fertiliser and soil ameliorants is emerging from the innovator to early adopter
phase. The early majority of potential adopters is becoming more familiar with
the technology required for automated VR applications, however actual
uptake is slow. The area of PA adoption research is gaining increasing
attention from R&D agencies1 and is likely to reveal key barriers to adoption.
Previous studies have conclusively found significant economic benefits from a
range of PA technologies currently being applied in Australian cropping
systems2,3,4,5,6,7. These studies cover a range of approaches to investment
appraisal (discounted cashflow analysis, gross margins, investment analysis,
Real Options etc) and PA technologies (GPS guidance, autosteer, tramlining,
variable rate fertiliser and soil ameliorants).
The case studies by Robertson et al. (2009) and McCallum (2008)8 reveal a
wide range of benefits ($1/ha/yr to $37/ha/yr) attributable to variable rate (VR)
and navigation aids. Recent projects conducted by Precision Agronomics
Australia have resulted in benefits from variable rate lime, gypsum, potassium
and phosphorous ranging from $10 to $30/ha depending on site variability and
baseline comparisons.
1 Robertson et al. (2011). Adoption of variable rate fertiliser application in the Australian grains industry: status, issues and prospects. Precision Agriculture (in press). 2 D’Emden, F.H. (2008). Optimising resource-use efficiency through precision agriculture. Soils2008 Conference, Australian and New Zealand Soil Science Societies, Palmerston North, New Zealand. 3 D’Emden, F and Knight, Q. (2009). Optimising gypsum applications through remote sensing and Variable Rate Technology. Agribusiness Crop Updates Proceedings: Soils. Perth, WA. 4 D’Emden et al. (2010) Variable rate prescription mapping for lime (and potassium) inputs based on electromagnetic surveying and deep soil testing 5 Tozer, P (2009) Uncertainty and investment in precision agriculture – Is it worth the money? Agricultural Systems, 100. 6 Robertson et al. (2009) Economic benefits of variable rate technology: case studies from Australian grain farms. Crop and Pasture Science, 60. 7 Oliver and Robertson (2009) Quantifying the benefits of accounting for yield potential in spatially and seasonally responsive nutrient management in a Mediterranean climate. Australian Journal of Soil Research, 47. 8 McCallum, M (2008). Farmer Case Studies on the Economics of PA Technologies. GRDC Updates, Ballarat.
53
However, few (if any) Australian studies provide a comprehensive
assessment of the costs of data acquisition and analysis, information
management, learning and consulting costs (herein referred to as information
costs) that are attributable to good PA decisions9.
Presentation Content/Results:
Various Australian studies into the economics of PA have concluded that
ongoing benefits are attributable to more efficient allocation of inputs through
reduced overlap and zone management (e.g. Robertson et al., 200710;
McCallum, 2008). There are also the somewhat intangible benefits of reduced
operator fatigue and ability to use unskilled labour to implement variable rate
tasks. Other intangible benefits linked to growers’ and consultants’
understanding of agronomic systems include the ability to conduct trials,
knowledge of variability and increased confidence in decision making
(Robertson et al., 2009). Other benefits that are likely to become more
tangible in future are reduced environmental impacts (e.g. greenhouse gas
emissions, fertiliser run-off and leaching etc) through targeted fertiliser use.
The combined factors of reduced PA hardware costs and increased input
costs since these studies were conducted means that the benefits cited in
these studies are now generally higher. Indeed, cost and lack of profitability
were minor reasons for not adopting variable management in a recent survey
of Victorian growers (Robertson et. al., 2011).
While it appears there is little debate about the potential benefits attributable
to variable rate and guidance technology, the cost side of the equation
appears to be less well understood. Two regional surveys and recent
unpublished research conducted by SPAA confirms that software and
hardware technical issues and data complexity were the most commonly cited
9 Robertson et al. (2009) gathered cost data for consulting fees and time to set up equipment; however these cost components are not detailed in their results. 10 Robertson, M. J., Isbister, B., Maling, I., Oliver, Y., Wong, M., Adams, M., et al. (2007). Opportunities and constraints for managing within-field spatial variability in Western Australian grain production. Field Crops Research, 104, 60–67.
54
constraints to further adoption, even among groups with higher existing levels
of VR adoption (Robertson et al., 2011).
These constraints can be generally categorized as costs involved with
information management and analysis. This is not a new problem, indeed it
has been 10 years since Cook and Bramley11 noted that “If the major benefit
of [site]-specific interpretation (of agronomic complexity) is accuracy and
relevance, the major cost is the need for direct delivery to growers”. The
phrase ‘direct delivery to growers’ is interpreted as meaning the delivery of an
agronomic solution to growers in the form of a prescription that is spatially
relevant within existing paddock boundaries in a format that is conversant with
existing variable rate software.
In the case of PA, the particular transaction costs that appear to be
constraining further adoption are those regarding learning specific skills
related to geographical information systems (GIS) software and the analysis
and interpretation of geophysical data that can be used to more accurately
define soil management zones. It is unreasonable to expect these learning
costs to be borne by the majority of growers when it forms a minor part of their
overall operational requirements. The question remains as to whether
agronomy consultants are prepared to bear these costs and in order to
provide a premium PA service to their clients.
The significance of these costs in constraining adoption of PA was confirmed
by a recent Grains Research and Development Corporation (GRDC) PA Think
Tank report (Blumenthal, pers. comm.). The GRDC responded to some of the
fundamental issues identified as constraining PA adoption by providing
training opportunities for growers and agribusiness. However there are still
large gaps in the knowledge about the best approaches to developing site-
specific agronomic solutions on a region-by-region basis.
11Cook, S. and Bramley, R. (2001). Is agronomy being left behind by precision agriculture? Proceedings of the 10th Australian Agronomy Conference.
55
This implies that the major economic gains to agriculture from wider adoption
and application of PA will not necessarily be achieved through ground-
breaking new sensor technologies or layers of data, but more likely in better
understanding of the existing technologies and layers of data, and how they fit
together to inform agronomists and growers about optimal variable rate
decisions.
Conclusions
The economic benefits of PA are apparent to those who have invested in the
data acquisition, analysis and hardware required to enable informed PA
decisions. Wider adoption and further economic gains from PA are likely to be
achieved through improved analysis and understanding of existing spatial
data on a region-by-region basis, and the development of simplified yet
powerful software and hardware technologies for variable rate decision
making and application.
56
More confidence in making decision using N and P
What PA tools to consider
Peter Treloar
Precision Ag Consultant
Minlaton, SA
Mob: 0427 427 238
Email: pete.pas@internode.on.net
Take Home Messages
Use PA technology to its fullest – on farm trials are essential
Don’t get information overload – keep it simple s….. (KISS)
Base N decisions on yield potential and risk exposure
Introduction
Yield mapping is the first step most farmers take in managing spatial
variability. It forms one of the most important layers in a successful Variable
Rate program by providing both a layer to base VR on and the layer used to
measure any successes.
Variable Rate gives farmers the opportunity to make significant returns from
PA, by matching inputs to areas of greatest return farmers not only improve
returns, but they can also reduce risk and improve efficiency.
To make better decisions about N and P farmers and consultants need to
treat VR as a tool of agronomy. Very rarely does a map or data layer provide
a straight solution that farmers can implement without further investigation.
Most layers just indicate there is a difference between point A and point B, it is
up to the farmer or consultant to work out what that difference is, if it affects
yield and whether it can it be responded to economically.
57
Decision Making
It is easy to get overwhelmed with data in PA, so it is best to start with a single
layer and work your way through. Some layers can confuse the issue by
giving similar values for completely different reasons, for example EM can blur
the line between sand and limestone.
This is why ground truthing is a crucial part of any successful VR program,
whether it is with soil testing, tissue testing or simply even comparing to
previous Yield maps can help understand a layer.
Phosphorus Decision Making
P Replacement, where fertiliser is added according to how much yield is taken
off, is the most common method used for VR phosphorus. This has the benefit
of been simple and easy to understand, it also helps balance a farm budget
by reducing fertiliser costs after poor years and while higher costs occur after
better years.
P Replacement relies on good soil P levels, as you don’t want to the reduce P
in a low yielding area if that is the cause of the poor yield. Also a triple bin is
recommended so a base level of N can be maintained across the different
treatments. This is particularly important in cereal on cereal crops as seeding
nitrogen can have a big impact on yield, particularly if it is a good year
following a poor year.
Nitrogen Decision Making
Decisions about N in VR are exactly the same as you make without VR, we
just have a few more options when it comes to implementing those decisions.
For example if you ask what is my yield potential and the answer for the whole
paddock may be 3t/ha, but if there are sections in that paddock where it is
6t/ha, VR can be used to reach those potentials without the risk of over
fertilising other parts of the paddock.
In most of my work N decisions come down to soil water and risk exposure.
With the huge range of tools available to farmers looking at soil water,
58
permanent zones based on soil water characteristics should be able to be
established.
By having zones based on soil water farmers can take decisions with risk as a
key factor, for example zones with high subsoil constraints are high risk areas
and seasonal conditions need to be above average to reduce that risk. But
areas with large ‘buckets’ of available water provide opportunities for farmers
to maximize returns from N investments.
Trial Strips
On Farm trials are the most powerful feature of VR and Yield Mapping, they
give the farmer the ability to trial nearly anything on their own farm using their
farming system. Trials are an essential part of any successful VR program as
both a measure of success and as a source of information for improvement.
On Farm trials are easy to establish and should be setup to provide maximum
information, i.e. with replications, large enough to guarantee full passes of the
header and kept simple so results are easy to access.
Other factors affecting Yield
It is important to ask what the drivers of Yield in your district are. Is it subsoil
constraints or does elevation and water logging have the greatest impact, are
we better applying VR at seeding or are the greatest gains available in late
season applications of N. Is NDVI showing us nitrogen effects across the
paddock or are parts of the paddock related to Trace Elements?
This is why ground truthing is an essential part of VR; it will help avoid
spending good money after bad.
Conclusions
Successful decisions on VR Nitrogen and Phosphorus follow the same
decision tree as traditional blanket approaches. It’s just PA can provide more
information (sometimes too much) in making the decision and provide more
options to implement that decision.
59
The best advice is to start simple and use the technology to its fullest through
on farm trials, because every season is different and what worked in a dry
year may not work in a wet year.
60
2011 SPAA Precision Agriculture Conference 61
Notes
Disclaimer The information presented in this publication is provided in good faith and is intended as a guide only. Neither SPAA nor its editors or contributors to this publication represent that the contents are accurate or complete. Readers who may act on any information within this publication do so at their own risk.
61
PO Box 3490 Mildura | Victoria 3502 P 0437 422 000 | F 1300 422 279
www.spaa.com.au
62
Recommended