Securing yields on the edge

GroundCover Live and online, stay up to date with daily grains industry news online, click here to read more

Even on the eastern edge of the Western Australian wheatbelt, in an area known affectionately as the ‘western goldfields’, there is potential to lift grain yields

Yield gains of 0.5 to 1.5 tonnes per hectare are available for the taking on some soils in the eastern region of the Western Australian wheatbelt, according to CSIRO benchmarking research carried out as part of the GRDC-funded Water Use Efficiency Initiative.

Project leader Dr Yvette Oliver says the secret to lifting grain yield in the region is to determine which soils have the highest and most reliable potential to yield more across seasons and focus effort and inputs on these areas, rather than those with little prospect of yield improvement.

Dr Oliver and her CSIRO colleagues have developed a new approach to yield benchmarking. It combines commonly available information such as soil testing records, farm rainfall and yield data with grower knowledge about specific paddock histories to determine which soils are consistently underperforming in terms of potential yield and the reasons for this.

The value of the approach is that it provides a whole-farm analysis rather than individual paddock analysis and enables a grower to develop a strategic approach to lifting yields across the farm over time. The approach not only targets underperforming areas but also attempts to determine why they are underperforming and the likelihood of lifting yields via soil amelioration or increased inputs.

John and Tim Butcher examine seedlings on their WA property

Growers John and Tim Butcher have lowered inputs and improved production efficiencies by better matching yield potential to soil types across their 4000-hectare property on the edge of the Western Australian wheatbelt.

 PHOTO: Nicole Baxter

Case-study farm

Bodallin growers John and Tim Butcher worked with the CSIRO research team to characterise the yield potential of their low-rainfall 4000-hectare grain enterprise 330 kilometres east of Perth. 

The research has highlighted the importance of focusing inputs and time on areas most likely to deliver a yield return rather than wasting effort and money on poor-performing areas destined never to improve.

For the Butchers this has meant redirecting inputs away from poor-performing areas and using variable rate technology to target more reliable but nevertheless underperforming areas. In doing so they still spend the same amount per hectare on inputs but with the redirected funds have been able to afford to implement a liming program in areas destined to produce a much higher yield benefit.

John and Tim say the project enabled them to put a number on the yield and variation across their property – which they knew existed but had not had the time to quantify themselves.

Throughout the project the Butchers became more and more intrigued about the yield variation and what was causing it. They are now much more mindful of their soil types and their input requirements and through yield mapping and soil tests have gained a more thorough and detailed understanding of their farm’s productive potential. Between 2004 and 2009 average farm yield from the Butchers’ wheat program ranged from 0.47 to 1.97t/ha. According to potential yield calculations based on farm rainfall records the yield gap ranged from 0.6t/ha in the dry year of 2007 to 1.5 t/ha in the high-yielding year of 2005. 

Dr Oliver analysed yield and positional data collected from 32 paddocks during harvest – an area totalling 4466ha under cereal production for the 2004–09 seasons.

The yield and positional data were then converted into a whole-farm yield map using a grid size of 25 square metres. Records from the farm’s network of five rainfall stations were used to calculate the potential yield possible for each of the grid locations across the six seasons. From this the relative yield potential of each grid and season was determined by dividing the actual yield (as measured via the John Deere harvester yield software) by the potential yield.

Cereal yields from the 32 paddocks over the six seasons were then characterised as either below average (less than 50 per cent of relative yield potential), average (50 to 75 per cent of relative yield potential) or above average (greater than 75 per cent of relative yield potential).

Using the new benchmarking approach Dr Oliver and colleagues determined that just under a third of the farm was yielding below average in the ‘normal’ seasons of 2004, 2005, 2008 and 2009 and 70 per cent was performing below average in the weedy and dry seasons of 2006 and 2007.

Across the normal seasons, 10 to 30 per cent of the farm was performing near yield potential while 40 to 50 per cent was operating at less than 50 per cent of the relative yield potential (average performance).

Linking yield to soil type

Soil maps drawn by the Butchers combined with additional soil test information were overlaid on the whole-farm yield map to determine the soil constraints holding back yield.

The best-performing areas were the sandy earth and deep loamy duplex soils, which make up 40 per cent of the farm. While the majority of these soils were performing at average-to-above-average relative yield potential, about 500ha was achieving only 50 per cent of its potential yield – with no obvious soil constraint evident.

The poorest-performing areas were associated with the gravel, loamy earth and shallow sandy duplex loamy soils, which make up about 20 per cent, or 900ha, of the farm. These soils had a combined relative yield potential of just 48 per cent (below average) with about 200ha yielding only 38 per cent of the relative yield potential and the best 200ha yielding just above average at 60 per cent of the relative yield potential. Many of the soils in this category had intractable soil constraints.

The average-performing areas make up 40 per cent of the farm and were associated mostly with deep yellow sands and acid sands and the shallow loamy duplex soils. About a quarter of this area achieved a relative yield potential of only 48 per cent, with the top quarter achieving about 70 per cent of its relative yield potential.

While a deep-yellow sand can have a plant-available water capacity of 100 millimetres, the water capacity of the same sand with an acidity problem can be reduced to just 30mm. Overcoming low soil pH with lime can deliver a significant increase in the yield potential of such a soil.

By focusing on these average-yielding soil types with a potential for amelioration (such as the deep acid sands), Dr Oliver determined that there was potential for grain yield to be realistically and reliably improved on slightly more than a quarter of the farm. This amounted to about 1000ha.

While the Butchers had known intuitively about their soil acidity and structural issues, the research has given them concrete information to target soil testing, prioritise inputs and implement a long-term soil amelioration program.

Tim Butcher says the past few dry seasons have confirmed the need for growers in drier regions to adjust inputs and management to the season that is delivered. For the Butchers their “soil isn’t just the dirt that flows between the tines on an air-seeder”, it is the basis of the potential yield possible in any particular growing season – and by adjusting inputs according to soil type and rainfall they hope to continue achieving greater production efficiencies and better returns.

More information:

Dr Yvette Oliver, CSIRO,
08 9333 6469

Next: Pre-breeding for water use efficiency

Previous: Gypsum improves yield on responsive soils

GRDC Project Code WWL00002

Region West