Key points

• The Future Farm project is developing the methods and processes necessary for automated analysis of on and off-farm data, including remote sensing layers
• Initial research has confirmed the need to account for multiple variables when trying to accurately predict the response to applied nitrogen, not just a single vegetation index.

Efficient use of nitrogen fertiliser is an important driver of the profitability of cereal enterprises across Australia, yet selecting the most efficient and effective rate is no easy task.

On an industry-wide basis, the Australian grains sector applies approximately one million tonnes of nitrogen fertiliser annually.

While precision agriculture advocates the '4 Rs' - putting the right amount of the right product in the right place at the right time - it takes a lot of time and effort to acquire and interpret the data needed to be right and the process is not well integrated.

Now GRDC's Future Farm project aims to change the way growers make on-farm decisions by linking multiple high-tech sensors with sophisticated data analysis - and this approach is supported by a growing list of collaborating growers.

The project brings together skills from CSIRO, the universities of Sydney and Southern Queensland, Queensland University of Technology and Agriculture Victoria.

Automating data processes

The team aims to develop site-specific analytics through better use of in-season field-monitored data, historic on-farm data and external public and private data, such as remote sensing, soil mapping and weather data.

Their challenge is to automate the process from data acquisition and analysis to help growers evaluate their decision options using the benefits of on and off-farm data sources.

With expenditure on nitrogen fertiliser being one of grain growers' biggest annual variable costs, the team has initially focused its efforts on applied nitrogen decisions.

Our ultimate goal is to develop an automated, sensor-based approach to the delivery of site-specific decision support. - CSIRO senior principal research scientist Dr Rob Bramley

A four-year field program was established in 2018 to evaluate the value of automated crop and soil sensors and other on and off-farm data sources.

On the ground

There are 'core' sites in each region and these are located at:

• Tarlee in South Australia;
• Kalannie in Western Australia; and
• Narrabri in New South Wales.

The field program includes opportunistic 'satellite' sites, which are grower-initiated nitrogen strips - with simpler monitoring - to assist with mid-season fertiliser decisions.

While the 'core' sites enable detailed measurements and analysis, the 'satellite' sites offer the chance to calibrate the sensors over a broader range of soils and climates.

At the Tarlee 'core' site, three nitrogen regimes were established as fertiliser strips in a 64-hectare paddock - with final total-applied nitrogen rates of:

• 195 kilograms per hectare for the 'N-rich' strip;
• 110kg/ha for the 'paddock'; and
• 45kg/ha for the 'N-minus' strip
• (See Figure 1, below).

Figure 1: Trial layout at the 64-hectare South Australian core site near Tarlee. Source Dr André Colao

The strips crossed different management zones in the paddock, based on historical yield and soil electrical conductivity maps.

Soil and plant samples were taken from 21 target locations spread across zones, and strips for crop sensor calibration and soil moisture probes (sensing to one-metre depth) were installed in each management zone - in addition to one already installed in the paddock.

Results from the 2018 analysis showed that the crop response to nitrogen varied along the strip and was more pronounced for grain protein than for yield.

No single vegetation index was able to provide valid predictions of grain yield, grain protein or mid-season nitrogen requirements.

Decision support

Soil and plant samples were taken to calibrate the crop sensors. Photo: Dr Rob Bramley

The next step is to investigate the benefit of combining more prediction variables, such as soil moisture, into multivariate models to support improved, site-specific nitrogen fertiliser decision-making.

The ultimate aim is to develop a framework to accurately predict the response to applied nitrogen using various layers of on and off-farm data.

The project is also working to develop an automated, sensor-based approach for the delivery of site-specific information to inform application decisions in real-time.

More information: Dr Rob Bramley, 08 8303 8594, rob.bramley@csiro.au