Subsoil manuring

About the project

Soil and site

An ARC linkage project began in 2005 at Yallock Estate near Ballan and worked with sodosols, which are duplex soil with clay sodic subsoil, which are sodic. The physical soil constraints that restricted root growth was the low macro-porosity(< 10%) and water movement with very low hydraulic conductivity. There was also little root growth below 30 cm. These factors restricted water storage and use in the subsoil, which can have severe effects on crop performance in a dry spring during the grain filling stage of the crop.


Ballan is located in between Melbourne and Ballarat (longitude 144.23; latitude 37.86 S) at a height of 508.7 m above sea level. The long term average rainfall at the site is 576 mm falling primarily in winter and spring (from April to October). The mean temperature is 13.5°C in winter, 16.5°C in spring 24.5°C in November and December. The cool spring and cool summer climate enables long season winter crop varieties to be grown and to produce high yield.

Experimental approach

Ameliorating this dense sodic subsoil was a major challenge. We considered ameliorating the subsoil in near 10 cm wider stripes with stripes can be near 80 cm apart from each other. Initial lab incubation experiments and a study with clay and sand mixtures showed that 20% organic manure should be essential to instantaneously improve the macroporosity and hydraulic conductivity of the soil. Considering this rate to be mixed in a soil layer 10 cm wide and approximately 20 cm deep, we decided on an application rate of 20t/ha of organic material.

To place organic materials 30 cm below the soil surface, we applied the material via a pipe behind a deep ripper, which was pulled by a tractor. Our first approach was to use an air seeder to feed the organic material from the tank to the pipe. We used organic materials in pellet form (Dynamic Lifter and lucerne pellets), which can easily used with an air seeder attachment or gravity fed manually through the pipe. Unfortunately, the air seeder was only able to deliver a maximum of 4t/ha of organic material. So we manually dropped the required rate of organic materials through the pipe as shown.


Subsoil manuring theory

The major physical constraint of these dense sodic subsoils is low macroporosity (>10%). So even after soil reaches field capacity after draining, air filled pores (macro pores that can be drained under gravity after saturation of soil) remain less than 10%. Ten percent aeration is the critical level below which roots cannot breathe, so roots cannot grow in this soil.

When we apply organic manures 30 cm to 40 cm below the soil surface behind the deep ripper, this increases the macroporosity and hydraulic conductivity of the soil in nearly 15 cm wide and 10 cm deep soil volumes. Roots start growing in the improved subsoil and so start extracting water, thereby drying the soil. This creates higher soil water metric potential in these areas.  It also leads water movement from the surrounding areas, increasing aeration so that roots can then expand. Ultimately, roots can grow in between the rip lines.

The growth of roots will initiate different physical and biological process such as wet and dry cycles, secretion of root mucilages and other exudates, and the movement of organic compounds from organic manures with water movement. These support the growth of bacteria in the rhizosphere, which in turn produce extra-cellular polysaccharides. The mucilages and polysaccharides are the cementing agents that stabilize the aggregates. The aggregation was enhanced by wetting and drying cycles caused by the roots by extracting water from soil.

Experiment design

Two paddocks were selected to apply the treatments. One paddock had five years of lucerne history (Medicago sativa cv Cimaron.) during 1999-2003, followed by a canola crop in 2004. Lucerne was sprayed with roundup (glyphosate) in August 2003. The other paddock adjacent to the lucerne paddock was under continuous cropping (canola-wheat-barley-wheat-canola). Both paddocks have been under permanent raised beds (1.7 m wide centre to centre) for the last eight years.

The field trials were established in 2005. The trial on each paddock was a randomized block design with nine treatments in four blocked replicates. The size of each plot was a 5 m long raised bed (1.7 m wide), with a buffer bed in between two treated side-by-side beds, and a 2 m buffer between the length sides of each treated bed. Two beds were left as buffers between blocks with a 3 m buffer along the beds in each block.

Treatment Description Amount of amendment added and tillage?
1 Control Direct sowing
2 Deep ripping only To 40 cm depth
3 Gypsum 10 t ha-1 incorporated in 30-40 cm depth
4 MAP 100 kg ha-1 of mono-ammonium phosphate (MAP) incorporated in 30-40 cm depth
5 Lucerne pellets 20 t ha-1 incorporated in 30-40 cm depth
6 Dynamic Lifter® 20 t ha-1 incorporated in 30-40 cm depth
7 Sand 20 t ha-1 incorporated in 30-40 cm depth
8 Gypsum + MAP 10 t ha-1 gypsum and 100 kg ha-1 MAP incorporated in 30-40 cm depth
9 Lucerne pellets + gypsum +MAP 20 t ha-1 lucerne pellets +10 t ha-1 gypsum + 100 kg t ha-1 MAP incorporated in 30-40 cm depth

The amendments were poured manually down a 15 cm diametre feeder pipe that was attached behind a large ripper fitted to a tractor. The base of the feeder pipe bent backwards with its basal opening in the vertical plane, which meant that the amendments were accurately placed in a continuous stream behind the ripper foot. Organic amendments were in pellet form as they were easy to apply manually. Dynamic Lifter® had 4% N, 2.2% P and 1.9% K and lucerne pellets had 2.8% N, 1.4% K, 0.9% P. There were two strip lines on each 1.7 m bed (centre to centre). Treatments were applied one week before the crop was sown.

The 2005 crop of wheat

Crop biomass growth

Plants started growing faster in organic amendment treated plots once the roots grew below 30 cm and the temperature was optimal for fast growth (near stem elongation stage). Biomass accumulation at anthesis was nearly 70% higher in organic manure ameliorated plots than in control or deep ripped only plots. Interestingly, coarse sand application at the non-lucerne site also produced nearly 70% higher biomass at anthesis.

Grain yield

Marked increases in grain yield occurred in both experiments when high rates of organic amendment were incorporated at 30-40 cm depth in the subsoil. The high yields of 11-13 t/ha have not previously been reported for wheat crops in Australia.

What contribute for high yield?

The basis for the high yields with the organic amendments were the high numbers of wheat kernels produced per unit area. The high yielding plots were able to produce around 25,000 to 30,000 kernels/m2. High kernel numbers per unit area are prerequisites for high grain yield (Fischer, 1985) and were achieved in these experiments by combinations of increased ear numbers and grains ear-1, compared to the control treatments. Clearly, the high yielding wheat plants were well supplied with water and nitrogen, enabling this wheat genotype to produce sufficient assimilate at critical stages of growth.  This enabled many floret primordia in the developing spikelets to survive and produced kernel during the grain filling phase.

Table 1 Grain yield and yield parameters under different treatments at the non-lucerne history site

Grainyield (t/ha)
Harvest Index
Ear number (per m2)
Deep ripping
Inorganic amendment
Gypsum + MAP
Course sand
Organic amendment
Dynamic lifter
Lucerne + MAP + gypsum
LSD (P=0.05)

Increases in the order of 55-60% occurred between the average kernel numbers for the Dynamic Lifter® and lucerne pellet treatments, and the average numbers for the control and deep ripped treatment. A major part of the increase was achieved by increased ear density (ears/m2). The high kernel numbers per square metre with the organic amendments can partly attributed to the high N status of these plants and the extra water supply.

Extra soil water accessed below 40 cm by crop particularly at grain filling time

Wheat plants with the deep incorporation of organic amendments were able to extract greater amounts of water below 40 cm during crop growth, compared to those from the control treatment. This was a very advantageous outcome as this subsoil water can be used very efficiently by crop plants.  Passioura (1976) points out that this water is accessed late in the growing season, when the products of photosynthesis are being translocated straight to the developing grain, with minimum respiratory losses.

Table 2 Loss of soil water (mm) from soil profiles between sowing and crop maturity under wheat (var. Amarok) grown in selected treatments at the two experimental sites

20 cm
40 cm
60 cm
80 cm
Lucerne + MAP + gypsum
LSD (P=0.05)

Further evidence that highlights the value of subsoil water comes from the results of a recent field experiment in southern NSW.  In this study, Kirkegaard et al. (2007) were able to show that subsoil water, used by wheat plants after anthesis, resulted in an extra 60 kg of grain yield per ha for each mm of subsoil water used by the crop, which is three times that for total seasonal water use. If the subsoil water in this study – which we designate as soil water below 40 cm – was used to produce grain with similar efficiency, then the extra 50 mm of subsoil water used by the organic amendment plants at the non-lucerne site would account for three of the four tonne/ha yield difference between the organic amendment and the control plants (Table 2).

A further striking result from the organic amendment treatment was the change in water extraction patterns of wheat plants. The change was from a pattern where 60% of the profile water at sowing was extracted from the top 40 cm, as occurred with control plants at the non-lucerne site and gypsum plants at the lucerne site (Table 2), to that where 60% of the soil water at sowing was extracted from below 40 cm, as occurred with the organic amendment treatments at both site. This ability to increase the extraction of subsoil water suggests that this approach to subsoil amelioration has the potential to deliver real increases in water use efficiency.

Yield cause

Remarkably high Harvest Index values were achieved with deep organic manure placements

Harvest index (HI) values for the winter wheat cultivar Amarok were remarkably high at the two sites in 2005, with many values exceeding 0.60 (Tables 1).  While values of 0.60 were considered by Austin et al. (1980) to be the upper biological limit for HI, there have been reports of high values of around 0.60 occurring for winter wheat cultivars in the UK.  In southern Australia however, most field studies report lower HI values for modern wheat cultivars. Yet there are a number of studies indicating that as the proportion of the crop’s water use that is used after anthesis increases, the HI values also increase.

There are several reasons why higher HI values result from greater post-anthesis water use. The first is that any increase in post-anthesis water use would result in more assimilation during the grain filling period, with more photosynthate moving to the developing grain, increasing the mass of grain produced by the canopy. The second reason is that more post-anthesis water use would tend to prolong the grain filling period, providing more time for pre-anthesis assimilates to be translocated to the developing grain.

Plants stayed green for longer periods

The deep incorporation of high rates of N-rich organic amendment in the subsoil resulted in keeping the flag leaf remaining greener for longer. This is perhaps the key to the high grain yields from these treatments. The green flag leaf scores at the hard dough stage show how all organic amendment treatments were able to delay flag leaf senescence and extend the duration of green leaf area during the final stages of the grain filling period.  Observations made at this time indicate that the flag leaves in these treatment plots remained greener than those in the control plots for a period of at least 10 days. The data from Ruske et al. (2003) show that for every extra day that the flag leaf remained green, the grain yield increased by between 84 and 210 kg ha-1.  Such yield improvements, associated with delayed senescence and extended green leaf duration during the grain filling period, are observed with the ‘stay green’ genetic variants that occur with most crop species. Spano et al. (2003) were able to produce ‘stay green’ mutants in durum wheat using a chemical mutagen; these plants produced higher grain yields than their parental lines.


Results from the 2005 season indicate that higher N and water supply produced more fertile tillers (higher number of heads/m2) and more available water and N at grain filling stage kept flag leaf greener for longer with continuously filling the extra grains efficiently. This led to a 70% higher grain yield in organic manure ameliorated plots. 

2006 - Soil properties changes

In February 2006 soil physical properties and root growth (from 0 to 40 cm deep) were measured between the rip lines.  Macroporosity of organic manure treated plots at 40 cm depth increased 2.5 times from control untreated plots from less than 10%, which is the critical limit for root growth to more than 20%. Addition of organic manure in the subsoil also increased the saturated hydraulic conductivity of the soil nearly 50 times. Improvements in soil physical conditions led to better root growth at 40 cm depth. Lower limits of water at permanent wilting point in the organic-manure-treated upper subsoil layer showed that roots were able to extract more water from the soil.

Table 1 Change in soil physical properties and root growth at 30 to 40 cm soil depth after 9 mounth of subsoil manuring intervention.

Macroporosity (%)
Bulk Density (g/cm3)
Saturated hydraulic conductivity (cm/hr)
Root density
Deep rip
Gypsum + MAP
Lucerne pellets
Dynamic lifter
LSD (P=0.05)

These measurements were taken between the rip lines, indicating that soil physical properties across the plot at a depth of 30 to 40 cm were improved. In 2009 we recovered soil from the organic-manure-treated plots and untreated control plots. We found that subsoil at a depth 0f 30 to 40 cm in the organic manure treated plots was now well aggregated with appearance of surface soil.

Summer fallow recharge in 2006

An important outcome from sub-soil manuring is to store extra water in soil profile during the summer fallow (summer fallow efficiency).  There was 200 mm of rainfall between the harvest of the 2005 crop to the sowing of 2006 crop and most of this rainfall fell in amounts greater than 10 mm. The improvement in hydraulic conductivity of the soil and the drying of the soil profile up to a depth of one meter by wheat roots in 2005 in organic-manure-treated plots led to nearly double the summer fallow efficiency at the non lucerne site. Subsoil manuring also increased the depth at which the majority of the soil water was stored in the soil profile during the fallow period in 2006.  At the non-lucerne site for example, a total of 96mm or almost 2/3 of the soil water captured in the 0-80 cm soil profile during the summer, was stored below the depth of 40 cm, when averaged over the two subsoil-manured treatments. In contrast, only 36 mm or less than ½ of the rainfall captured in the profile of the control treatment was stored below 40 cm.

Table 2 Treatment effects on soil water (mm) in the 0- 80 cm soil profile at different times during the cropping cycle, and on soil water recharge in the soil profile during the summer fallow period

At 2005 harvest
At 2006 sowing
Lucerne + MAP + gypsum
LSD (P=0.05)

Wheat performance in 2006 draught year

A significant increase in grain yield occurred in this second wheat crop in 2006 following the subsoil manuring 18 months earlier.   The yield increase for the average of the 3 organic amendment treatments over that of the control and deep-ripped control treatments, was 49% or 1.9 t/ha, despite the very dry year.  The basis for this grain yield increase was the 43% increase in the number of fertile tillers, with an extra 126 ears per m2 producing a 49% increase in the number of kernals per m2, compared to the controls.


Print version Close