Subsoil manuring

Background and experimental approach

The results of the ARC linkage project at Ballan were very convincing. However, we needed to test this subsoil manuring technique in other areas of Victorian High Rainfall Zone. The Ballan site was 450 m above sea level with cool temperatures in spring and summer, which makes it very suitable for long season winter wheat varieties. The major benefits of subsoil manuring were to significantly increase summer fallow efficiency and capacity for storing extra water deep in soil profile, which can be accessed at grain filling time.

We thought it would be better to test our technique over a range of climate and soil types to assess the benefits of this technology at different sites. We established different sites in 2009, 2010 and 2011 with support from the Grains Research and Development Corporation (GRDC).

Our sites

Subsoil manuring is being tested on five sites in Victoria:


This site was established in 2009. Poultry manure was applied manually in rip lines with the help of a pipe behind the deep ripper, as at Ballan sites. In 2009 there was a severe heat wave in November at grain filling time. Plants remained greener in subsoil manuring treatments for more than 10 days compared to untreated control plots. The highest yield was from the subsoil-manured plots that received 20 t/ha of manure, which was 58% higher than control plots and 36% higher than nutrient-only plots (Table 1). Overall the yield was less in this site compared to Derrinallum.It may be due to the high rainfall (nearly 100 mm) in each month from July to September causing waterlogging in the plots.

Table 1 Grain yield at Penshurst for different treatments in different years

Treatment2009 Wheat2010 Canola2011 Wheat2012 Canola
Deep rip4.
10 T/ha of manure5.61.410.02.9
20 T/ha of manure7.
Deep nutrients6.
10 T/ha of manure + deep nutrients5.
LSD (P=0.05)1.6ns2.10.7

The second year in 2010 was a wet year which caused water-logging particularly in the August to September period when 230 mm of rain fell. This water-logging caused lots of variability in canola yield in the plots. The 20 t/ha subsoil-manured plots withstood the water-logging better and yielded twice that of the control plots. Because of high variability it was not statistical significant (Table 1).

The third year in 2011 showed very interesting results. The wheat crop was sown late (in the second week of June). Conditions were wet and cold in winter and so there was not much growth during winter. The crop started growing at the start of spring but the spring was cool and anthesis started in December and there was little rain during grain filling time (59 mm). So the crop relied on water uptake from the subsoil. After anthesis,the subsoil-manured plots extracted 51 mm of extra water than the control plots and most of this water (32 mm) was extracted from below 50 cm in the soil profile. Subsoil manuring plots produced 66% higher yield compared to untreated plots. Nutrient-only plots did not produce better than control plots (Table 1). This showed that soil water availability in the subsoil was the main factor driving the yield increase.

In 2012 canola was sown at the site. There was good growing season rainfall (523 mm) but less rain than average fell during pod filling (90 mm). Grain yield was almost double in 20 t/hac subsoilmanured plots compared to control (Table 1). The 10 t/ha subsoil-manured plots also yielded nearly 50% higher than control. Subsoil manuring plots extracted 51 mm extra water from soil profile compared to the control plots.

The four years of results from this site show that basis for the grain yield increases with the full rate of subsoil manuring can be explained by the increased extraction of deep subsoil water by the crop during the grain filling period. This occurred for each crop at Penshurst from 2010 to 2012. The differences in soil water in the 50-100 mm subsoil layer were significant in each of these years.  It is interesting to note that the differences in the deep subsoil water extraction between the subsoil-manured and control plots were greater with the 4th successive crop of canola, with the average finish with 95 mm of rain between October and November in 2012, than with the 3rd crop of wheat in 2011 with a drier finish with 59 mm of rainfall. Clearly the impact from subsoil manuring on deep water extraction did not diminish in this 4th year.


This site was established in 2009. Poultry manure was applied manually in rip-lines with help of the pipe behind the deep ripper as at the Ballan sites. In 2009 there was a severe heat wave in November at grain filling time. The plants remained greener with subsoil manuring treatments for more than 10 days compared to the control, untreated plots. There were 40% extra heads produced in the subsoil-manured plots (582/m2) compared to the control plots (340/m2). This led to a 96% higher yield in subsoil-manured plots (9.8 t/ha) compared to the control plots (5t/ha). The plots where we applied deep N and P fertiliser to match the nutrients supplied in 20 t of poultry manure, yielded 7.1t/ha. This shows that the 2.8 t/ha of extra yield compared to the nutrients-only plots, could be explained by the extra soil water supplied by the subsoil manuring treatment.

The second year in 2010 was a very wet year with an annual rain of more than 900 mm, with a growing season rainfall of 632 mm. The site was severely waterlogged throughout winter so we did not find any response between treatments.

Table 1 Grain yield (t/ha) at Derrinallum with different treatments in different years

Treatment2009 Wheat2010 Canola2011 Wheat2012 Wheat
Deep rip6.
10 T/ha of manure7.
20 T/ha of manure9.80.87.410.4
Deep nutrients7.
10 T/ha of manure + deep nutrients8.
LSD (P=0.05)1.3ns1.81.1

The 2011 season was an average rainfall year, with an annual rainfall of 650 mm and a growing season (April to November) rainfall was 384 mm. The soil profile was also full of water at sowing time and 130 mm of rain fell during grain filling time (between September and November). So water supply was not a limiting factor for crop production. There were higher yields produced in the subsoilmanured and nutrients-only plots (7.4 and 7.3 t/ha) compared to the control plots (5 t/ha). This also showed that the yield response was due to increased nutrient supply and not because of any extra supply of water.

However in the drier finish in 2012, there was a 4.1 t wheat/ha yield increase with the full subsoil manuring treatment. There was also a very large, significant increase in the deep water extraction between the full manuring treatment and all other treatments.

The magnitude of the yield response was again related to the rainfall during the grain filling period from October to November. The large yield increments in 2009 and 2012, or 4.8 and 4.1 t wheat/ha filled during the period when 104 and 109 mm of rain fell. In both years the subsoil manuring treatment yielded significantly more than the full nutrient treatment.  However in 2011, the more modest yield increment of 2.4 t wheat/ha was lower, and this grain filled with 120 mm of rainfall in the October to November period. In this year there was no grain yield difference between the subsoil-manured and full nutrient treatment.


The Wickliffe site was established in 2010. Treatments were applied with the custom-built subsoil manuring machine. The site was under 2m raised beds.

The 2010 year was a very wet and we received a 27% higher yield (11.6 t/ha) in all subsoil manuring plots (10 t or 20 t/ha of poultry manure). It was a wet year so water was not a limiting factor. There were no differences in soil water uptake from soil profile. The yield increases may be due to relief from waterlogging stress, as there was no yield increase with the deep nutrients-only treatment.

Table 1 Grain yield at Wickliffe with different treatments in different years

Treatment2010 Wheat2011 Wheat2012 Faba bean
Deep rip8.64.04.0
10 T/ha of manure12.04.56.2
20 T/ha of manure11.64.96.3
Deep nutrients9.35.04.5
10 T/ha of manure plus deep nutrients11.34.65.0
LSD (P=0.05)1.7ns1.5

In 2011 the soil profile was full of water at the time of sowing. Crop establishment was poor, and the paddock was over-run with herbicide-resistant annual ryegrass, so we did not observe any significant differences between treatments.

The large yield increase in faba bean yield of 2.7 t/ha with the full and the half- subsoil manuring rates, above the control yield (Table 1) at the Wickliffe site in 2012, was striking. This was the first pulse crop that we had witnessed, that had been grown on subsoil-manured land.  The yield increase was associated with an average finish when 85 mm fell early in the finish period.


The Stewarton site and the nearby Dookie sites in north east Victoria were the driest sites in this project, with an average annual rainfall of 553 mm. They also differed from the three sites in south west Victoria by the absence of raised beds, and their location north of the great dividing range, meant that they had a generally shorter, hotter and drier finish to the growing season in October and November.

So the issue would be whether the subsoil manuring intervention would be effective on the two duplex soils at these sites.

There were modest increases in wheat yield of 2.4 and 2.8 t wheat/ha, for the two manuring treatments over the control treatment, at the Stewarton site in 2011. The two manuring treatments involved 20 t/ha of poultry litter, with the litter being either incorporated in the subsoil at a depth of 30-40 cm, or spread evenly over the soil surface. The surface manuring treatment was imposed in response to a criticism of the subsoil manuring research, where there was never any surface application control, to test whether it was really necessary to incorporate the litter in the subsoil.

The decile-8 year at Stewarton in 2011 resulted in a full profile of soil water at sowing, and at anthesis. This combined with a somewhat below-average grain-fill rainfall of 74 mm of rain, compared to the average of 90 mm, meant that there was not a lot of soil water deficit stress during the finish, with generally above average control yields of 5.7 t/ha.

Table 1 Grain yield at Stewarton with different treatments in different years

Treatment2011 Wheat2012 Wheat
Deep rip5.95.7
Deep nutrients7.25.6
20 T/ha of manure in subsoil8.19.4
10 T/ha of manure on the soil surface8.56.7
LSD (P=0.05)2.21.3

In 2012 there was less rainfall during the grain-fill period, and this was associated with a large increase in wheat yield of 4.5 t wheat/ha with the full subsoil manuring rate (20t/ha) over the control. However the there was no longer any equivalent yield response with the surface manuring treatment in this drier second year at Stewarton. An increase in grain protein in the wheat grain occurred with the manuring treatments in 2011, but there were no differences in the grain protein concentrations between treatments in 2012, perhaps due to the large increase in grain yield with the full subsoil manuring treatment in 2012.

The absence of any increases in subsoil water extraction differences between treatments in 2011 (Table 1) is consistent with the likely lack of soil water deficit stress during the grain fill period.  However the opposite occurred in 2012, when there was an extra 40 mm of subsoil water extracted from the subsoil manuring plots and this extraction was associated again with the large increase in wheat yield of 4.5 t/ha. The lower subsoil water extraction rates with the surface manuring and with the deep fertilizer treatments in 2012, were associated with the smaller non-significant differences in grain yield, compared to the control treatment.


Although the Dookie site was established in 2011 we have not presented the yield data in this first year.  The issue was with the reliability of the yield results, which were harvested by collaborators using a small-plot harvester. Despite the visual growth responses, and the prolonged greening in the canopy of the subsoil-manured plots, compared with the control plots – both of which have been associated with large yield increases with the subsoil-manured plots, there were no differences in grain yields measured by the harvester.

One issue here is that there were no raised beds at Dookie, and the front of the harvester has the same width as the twin-row sub-soiling machine.  There needs to be perfect alignment of the harvester, exactly over the 10 m long treated plots, otherwise large yield errors will result.  This alignment can be difficult without the guidance from furrows between raised beds.

There was a large increase in grain yield of 4.1 t wheat/ha at the Dookie site, which was very similar to the increase at the nearby Stewarton site.  The deep fertilizer nutrient treatment, although significantly greater than the control yield, was significantly less than the yield for the manured treatment. Again there were no differences in grain protein despite the large grain yield increase.

The loss in deep subsoil water below 50 cm was generally greater at the Dookie site in 2012, compared to the nearby site at Stewarton.  Nevertheless there was a similar increase in the volume of deep subsoil water that was extracted with the subsoil-manured treatment in 2012 compared to the control treatment: the volume was 40 mm at Stewarton and 34.8 mm at Dookie. Again there were no differences in the water extracted by the deep fertilizer treatment, compared to the control, as was the case at the Stewarton site.

Four of these sites are spread in western districts from Mt Pollock to Penshurst and two sites are in north east (Dookie). Sites selected have rainfall ranging from 550 mm to 750 mm and has different soil types with different level of subsoil constraints.