Grains Research and Development

Date: 01.10.2004

Progress in identifying and managing subsoil issues in the Mallee

Figure 1. Soil water content (0.4 to 1.6 m) determined from calibrated EM38 (vertical mode) measured at sowing and harvest at Loxton during 2003. Note the signifi cantly drier profile at harvest (lighter shade).

By Garry O"Leary, David Roget and Victor Sadras

Yield losses due to subsoil constraints of 20 to 25 percent have been estimated for the low rainfall (less than 350 millimetres) Mallee district, equating to losses of about $50 per hectare. These estimates are based on an assessment of Mallee Sustainable Farming Inc. focus farms across South Australia, Victoria and NSW.

Most farmers know the better-performing paddocks from the poorer performers. The reasons are many, but can often be traced to soil differences. But how can knowledge of the soil lead us to better manage our subsoil constraints?

Broadly, subsoil constraints can be classified into either physical or chemical constraints. Such constraints limit root growth, water use, nutrient uptake and ultimately crop yield. It all boils down to water and nutrient management, with the key factor being water.

The management of scarce water and expensive nitrogen fertiliser is critical to farm profitability. With conventional technologies, both resources are managed on a whole-paddock basis. But with the access to accurate position and time data with GPS, old technology such as EM (electromagnetic) mapping becomes a powerful new tool to better manage our water and nitrogen resources.

So, coupled with improved agronomy (for example, better cultivars and disease control) the spatial management of our water resource offers new benefits to croppers.

How can this be achieved? Three steps are needed: to identify the severity and extent of the problem areas; to determine the water-use pattern and available water and nitrogen; and to plan differential input to better match the available water and nitrogen.

CSIRO researchers have found serious compaction (a physical constraint) on some Mallee soils. Such compaction has been ameliorated by tactical ripping with consequent increases in water use, yield and grain protein content of wheat.

There is uncertainty about how widespread this problem is and farmers are using ripped test strips to assess the problem. It is likely that an assessment of subsoil compaction can be made with a simple penetrometer (welding rod!) but other tests may be needed.

At experimental sites at Loxton and Caliph, ripping dramatically reduced soil penetration resistance between 0.10 and 0.3 to 0.4 metres. Control crops yielded between 1.2 and 2.9 tonnes per hectare and yield improvement attributable to easing of soil compaction ranged from nil to 43 percent; yield response to ripping remained for two seasons and has largely been attributed to increased water use.

All soil and crop responses to ripping were more marked in sand hills than in sandy loam fl ats.

By far the more widespread subsoil constraints in the Mallee are the chemical ones such as high salt and boron. The researchers have found that surveys using EM38, an apparatus for measuring soil salinity through electromagnetic induction, can help detect such problem areas of fields.

This is because EM38 measures the soil electrical conductivity and this is strongly related to the soil water, salt and clay content. Its ability to sense water allows the scientists to determine areas within paddocks where subsoil constraints limit crop water use.

Unlike physical soil compaction that can be alleviated with ripping, there is currently no practical option to remove chemical subsoil constraints. Improved management of these soils will rely on three steps:

Two EM maps! Researchers are experimenting with using two EM maps, one around sowing time and the other after harvest. Two maps allow more precise calculation of crop water-use and therefore more accurate estimates of yield potential.

Figure 1 shows average soil water content from the 0.4 to 1.6m soil layer determined at sowing and harvest at the Loxton site.

Figure 1. Soil water content (0.4 to 1.6 m) determined from calibrated EM38 (vertical mode) measured at sowing and harvest at Loxton during 2003. Note the significantly drier profile at harvest (lighter shade).

CSIRO"s soil water-content maps are determined by on-site calibration of EM38 readings with measured soil water contents. The calibration relationship is not perfect because of the interference of salt and clay content, with a mean error of around 15 to 20mm water for a 1-metre soil profile.

The research team views this error as acceptable for defining different management zones within the paddock, provided the management history is similar.

In addition to soil EM38 mapping, yield or biomass (derived from satellite imagery) maps can also provide a measurement of paddock variability. The more consistency there is between the variability identified on an EM map and a yield map, the more confidence there is that it is subsoil issues driving yield variability.

Conversely, where there is little consistency between the two maps, then other significant yield limiting factors are involved. A collection of yield or biomass maps over the years from the same paddock will reveal consistently poor areas as well as the areas that behave differently in wet and dry years.

While a set of several yield or biomass maps for the paddock helps in interpreting the causes of yield variation, its management requires consideration of the economics of production. One way this can be considered is to map the gross margin. The question, however, remains - can higher profits be achieved at such sites from, say, the reduction of inputs in the salt-affected areas or can some agronomic treatment boost profits in these areas?

While a collection of yield or biomass maps provide useful information on the likely subsoil constraints where no - or a limited collection of - yield maps exist, two EM maps offer alternative information, particularly on the high and low water-use areas.

At the Loxton field site, the water-use map was compared to the yield map (Figure 2).

Figure 2. Water use and yield map of wheat grown at Loxton during 2003. Note that the water-use map is very different to a single water-content map (Figure 1) and the greater resemblance of the water-use map to the yield map.

Figure 2. Water use and yield map of wheat grown at Loxton during 2003. Note that the water-use map is very different to a single water-content map (Figure 1) and the greater resemblance of the water-use map to the yield map.

Subtracting the soil water content measured by EM38 at harvest from the soil water content measured at sowing, and adding seasonal rainfall, gives us a good estimate of crop water use. A strong resemblance between crop water use and yield is seen (Figure 2).

Note the mid-slope areas showing high water use (dark blue) and yield (blue), conversely, the low water use and yield areas are coloured red. Therefore, management plans based on the potential water supply can be developed and tested.

Basing management on water supply should be more reliable than using other single factors since water supply is the major determinant of grain yield in Australia.

CSIRO is testing the value of knowing where the subsoil constraints are by developing site-specific fertiliser plans for a number of the focus paddocks in the Mallee Sustainable Farming network across the Murray Mallee.

Fertiliser is one of the high-cost items for Mallee farmers and if cost savings and more efficient fertiliser use can be achieved, then some of the losses attributed to subsoil constraints may be clawed back.

EM mapping offers new ways to detect the extent and severity of subsoil constraints in Mallee soils that should help develop more precise and profitable farming strategies in areas with often unseen subsoil constraints.

EM mapping should also be useful in the higher rainfall areas.

For more information:
Garry O"Leary, 03 5091 7200, fax 03 5091 7210, garry.o"leary@csiro.au

GRDC Research Code: CSO 216