Grains Research and Development

Date: 01.01.2000

On farm test is best

Bar Graphs comparing the yield per hectare for a variety of wheat strains, for stubble and no-stubble farming

A study from the Maranoa, in the Roma district of Queensland, has shown how important it can be to test trial results from research sites on the farm before wider adoption. The interaction of soil and weather can affect yield results in a fairly localised way.

The study showed healthy increases in Janz and Hartog wheat yields on six common soils in response to adding nitrogen and phosphorus fertilisers and retaining stubble. But there were some surprises that did not fit with what was generally known about the soils.

"In some cases, we actually saw a decline in wheat yields as a result of adding fertiliser, or retaining stubble instead of burning it," says Brett Robinson, Systems Agronomist at the Queensland Department of Natural Resources.

"We suspect that where we retained stubble on the grey clay soil, that has been cropped continuously, disease problems may have caused the lower yield," says Mr Robinson

Brown clay example

The lower yield on the brown clay after fertilising was also unexpected, especially since the soil had only about half the original soil nitrogen of the other soils (35 kg/ha of available nitrogen compared to 50-75 kg/ha for the other soils). (See graphs this page for comparison.)

Local site factors affect yield. "The brown clay can surface-seal during rains, and this affected emergence in one of the trial years, changing the crop's ability to take up the additional nitrogen. The improvement in yield when stubble was retained suggests that this could be the case."

Another possible explanation offered by Mr Robinson for the lower-than-expected yields is the dry finish to the growing season.

Also, "you can lose yield under fertiliser if there is a dry finish, as it can produce more leaf at the expense of grain development".

A negative response to fertiliser in a dry year doesn't mean that yields will not increase after fertilising in a wet year, as shown by longer-term research trials at Nindigully. It could mean that, for some soils, the best response from fertiliser will be obtained only if the soil moisture is full on sowing.

"What these results do show," says Mr Robinson, "is the value in on-farm testing, or at least in a very similar paddock nearby, using the same machinery and crop varieties. In some cases, it might be necessary to test for a few years to understand what will happen over a range of seasons."

The study involved trials on six soils common in the Maranoa:

  • grey clay, originally with Mitchell grass and scattered trees, known as the open downs
  • red earth of good quality, originally with a mixed woodland of ironbark, cypress pine and wilga
  • solodic soil, originally with poplar box woodland
  • brown cracking clay, shallow in parts with ironstone nodules, originally with belah woodland
  • deep-surface solodic soil originally with belah and poplar box woodland
  • red cracking clay, deep and of very high quality, originally with belah open woodland.

This work was supported by growers through the GRDC and is based on the data collected by Evan Thomas as part of a Master's thesis aimed at predicting land suitability from soil, terrain and climate factors. Other initial work was conducted by Lindsay Boaler.

Program 3.5.1

Contact: Mr Brett Robinson 07 4688 1343

Rainfall (mm) at Roma during 1989 and 1990, and the long-term average

Rainfall during fallow or crop stage (mm)
Fallow (Nov. - April)Early (May - July) Mid (Aug. - Sept.)Late (Sept. - Oct.)
1989411136167
1990519972720
Long-term average3771203735

Effects of stubble cover on wheat yield at six sites with different soil types. Only two sites recorded a significant yield increase, The numbers below the columns indicate the plant-available water content of the soil.

Region North