Mapping N supply using mid-infrared

Figure 1: Standard laboratory detection of biological soil N supply

By Dr Daniel Murphy, Dr Nui Milton and Gina Pemberton, Centre for Land Rehabilitation, The University of Western Australia

Understanding the capacity of soil microbes to produce nitrogen naturally is becoming a key to better nitrogen (N) management in cropping soils.

Anywhere from 40 to 80 percent of the crop N requirements can be met through microorganisms breaking down residues and organic matter to release plant-available N (biological soil N supply). The remaining N requirement is met through fertilisers.

Improving N fertiliser management is therefore a matter of knowing the timing and location of biological soil N supply, so that fertiliser N is only applied when and where necessary.

Splitting applications of fertiliser N at strategic plant growth stages is already becoming common. This maximises crop uptake at the right time and minimises the risk of nutrient leaching.

Growers can also spatially adjust fertiliser rates in the field using information from yield mapping (knowing the best/worst performing areas of a paddock) and by soil type.

The next extension of this approach is to utilise spatial soil maps that tell us about the soil’s capacity for biological N supply, together with information on the soil’s chemical and physical fertility.

For example where biological soil N supply is high, less fertiliser N may be needed to achieve optimum yields. Alternatively where biological soil N supply is low, there is a greater reliance on fertiliser N.

To develop a practical decision-support tool for fertiliser application rates, we are exploring the use of midinfrared technology to develop calibration curves for a range of soil properties – biological, chemical and physical.

The advantage of the midinfrared technology is that once calibrated, soil samples can be collected from the field and scanned rapidly – two minutes per sample.

Mid-infrared is not as accurate as measuring each soil property by standard techniques, but it does have a place in the development of soil spatial maps to identify sources of crop constraint.

In soil that was collected under an oat crop in 2003 at Dangin, Western Australia, using a 25m x 25m sampling grid (180 separate sampling points over 10ha), biological soil N was measured using standard laboratory methods (figure 1) and also predicted using mid-infrared technology (figure 2).

Figure 1: Standard laboratory detection of biological soil N supply – data from soil samples (0-10 cm) collected on a 25m x 25m sampling grid. Colours represent data categorised into 4 ranges, where red = very low biological soil N supply, yellow = low, light blue = moderate and blue = high.

Figure 2: Mid-infrared predicted biological soil N supply – data for the same 10ha area as that shown in Figure 1 (ie, the pattern of colour on Figure 2 would be identical to that on Figure 1 if the mid-infrared prediction was 100% accurate). The same colour groupings apply.

Figure 2: Mid-infrared predicted biological soil N supply

There was strong agreement between measured (figure 1) and mid-infrared predicted (figure 2) values of biological soil N supply.

The data suggests optimum crop yields would require extra fertiliser in the red and yellow areas. In a good rainfall year, low fertiliser N application would also benefit the light blue area. In the dark blue areas, soil N is already sufficient.

This work, however, is still at the development stage, and more research is needed to test the transferability of calibration curves among different regions and soil types.

For more information:
Dr Daniel Murphy, 08 9380 7083, dmurphy@agric.uwa.edu.au

GRDC RESEARCH CODE UWA395, program 4