Articles, Blog

Future Farm


Almost a decade ago South Australian grain
grower Mark Branson was interviewed for GRDC’s Groundcover TV. Back in the 90’s we were doing 3 to 3 and
a half tonne a hectare; early two thousands we were doing 4 tonne a hectare and last year
we hit a jackpot and got a 6.6 tonne average wheat yield. What contributed to that increasing yield
trend was Mark’s use of Precision agriculture and crop sensor technology, together with
the adoption of Nitrogen Rich strips. A crop fertiliser management practice Mark
brought home from the USA after a Nuffield study-tour in 2005. And this idea of the Nitrogen Rich strip with
the GreenSeeker is able to confirm whether I need to apply Nitrogen or not, that’s been
a big saving on this farm. N-Rich strips are a spreader-width early application
of Nitrogen applied at a luxury rate, which is typically twice the rate a grain grower
would apply. It’s a reference strip, the rest of the paddock
can be compared too. The recommendations I’m getting from the paddock
in behind me is that I’m going to be needing 30 kilograms of Urea in one part of the paddock
and 140 kilogram of Urea in another part of the paddock. It’s one progressive farmer’s approach to
what the grains industry sees as a much broader opportunity. Identifying ways time poor farmers can use
automated systems and tools such as precision agriculture and sensor technology, to improve
profitability. So the Grains Research and Development Corporation
invested in the Future Farms project. Now in its second phase, where 4 farm-scale
core sites and 9 satellite trials are providing data sets for computer modelling, CSIRO and
its research partners are working to enhance the algorithms used in sensor technology to
make them more applicable to Australian farming systems. But the problem with those sensors is that
the algorithms that accompany those that allow you to generate a decision first and foremost
they haven’t been developed under Australian conditions but the other thing is that they
assume that the only thing you need in order to make the nitrogen decision is the set of
numbers that get spat out of the sensor. This research is advancing the thinking that
a Nitrogen fertiliser decision is a univariate decision based on the sensor values alone. The idea being pursed by the Future Farm team
is a better prediction of the likelihood of a crop responding to additional nitrogen will
result from combining a number of resources with the crop sensor’s data. Resources such as soil moisture sensors, farm
scale strip trials and publicly available weather data. And also collecting plant and soil samples
to calibrate the sensors. This combined data will better reflect the
multivariate nature of a Nitrogen fertiliser decision. And allow the team to be more certain that
a fertiliser application will deliver the desired response. To help assess that likelihood of responsiveness
Mark Branson’s paddock was given N-Rich strips and zero-N plots for reference and an array
of sensors are being used to scan the entire paddock. Now on the Kubota that we’ll probably see
going up and down in the background we’ve got a number of sensors which sense the so
called Normalised Difference Vegetation Index the NDVI and variations on that theme, there
are a number of different indices that we can sense. Precision Ag researcher Andre Coloaco now
works for CSIRO, managing data collection and sensor calibration at both the South Australian
and Western Australian trial sites. These sensors are measuring reflectance so
basically reflected light from the crop. So how do we turn what these sensors are measuring
into something that is useful for the farmer to recommend Nitrogen? By comparing the readings that we are doing
with the sensors in both strips, the zero and the N-Rich can potentially predict the
final response of the crop in terms of yield and protein concentration and so on. Other valuable sources of data include Mark
Branson’s paddock histories and the part played by different soils types. The role of this other variable, the soil
information that we’re collecting soil moisture and soil physical characteristics it will
adjust the sensor algorithm to that environment. So that’s why we are hoping it will work across
different environments and in different seasons and in different climates because we are collecting
all this extra information. And with that information the farmer can then
hopefully make as close as possible to an optimum decision in terms of his or her decision
to invest in Nitrogen and get a reasonable expectation of the expected return from that
investment in Nitrogen. Before these outcomes can be applied on-farm
the science needs to deliver the analytical tools and extend its scope beyond Nitrogen
application. And while there are a couple of years for
the project to run Rob Bramley believes the innovators could be applying the research
sooner. A farmer like Mark Branson who’s been playing
around with this technology himself for sometime he could potentially adopt what we’re doing
quite quickly particularly as he’s got his own soil moisture sensor along the fence here
assuming the project is successful there’ll certainly need to be a little bit of development
and extension beyond the research that we’re doing at the moment before things can get
adopted. What the Future Farm project is trying to
do is exactly what I’ve been trying to do for the last 10 years. Its also introduced some new ideas to me like
the zero strip and I can see a lot of future in that because that gives an indication of
whether the crop needs Nitrogen earlier than what a N-Rich strip does. For grain growers Nitrogen fertiliser is one
of the greatest input costs they face and as Mark Branson told GCTV back in 2011 his
use of precision ag and sensor technology allowed him to make a big savings on Nitrogen
expenditure. Future farm could potentially deliver a similar
result for a lot more grain growers.

Leave a Comment

Your email address will not be published. Required fields are marked *