The strategic use of tillage within conservation farming

How much damage is done to soil by occasional tillage, strategically applied, in an otherwise direct drilled system?

Methods

We selected three contrasting sites where some form of cultivation made sense — following limestone application within a cropping phase, following a five year pasture phase and before canola, and following a green manure crop. The sites were at Thuddungra between Young and Grenfell, Berthong near Cootamundra, and Daysdale near Corowa. A fourth site was the CSIRO long term trial at Harden which had 20 years of contrasting tillage and stubble management. 

We measured a range of chemical, physical, and at two sites, biological properties of the soils so as to assess what effects the implementation of tillage had on the soil. We also measured a range of agronomic variables so as to assess impacts on plant production. We followed these soil properties and plant yield variables over 3-5 years to measure the changes in time following tillage. For example, if there were detrimental effects from tillage, how long does it take for the soil to ‘recover’?

In this paper, we document the soil physical property known as Wet Aggregate Stability (WAS). This is a measure of the soil’s structural stability, in particular how it behaves as it wets up following rain or irrigation. Macro-aggregates are defined as being greater than 250 micron diameter (µm)and a high proportion of these is considered desirable. Micro-aggregates are defined as being less than 50 micron diameter and are not considered desirable (effectively dust).

Here, we show the results from one on-farm site and from the CSIRO long term trial at Harden.

Results and discussion

Daysdale

The design at Daysdale involved three tillage treatments — on-going direct drilling (DD), one-pass with a scarifier (Scar) or one-pass with offset discs (Offs). Each tillage was applied in 2012 and again in another set of plots in 2013, with the tillage treatments for both years randomly distributed within the trial area so that we could assess any effect of the year of tillage. There were, therefore, six treatments — three different tillage operations by the two years of tillage, each with four replicates. We measured the wet aggregate stability (%WAS) of the soil in each plot at two depths — 0-5 and 5-10cm. 

The statistical analysis showed very few significant effects at this site (Figure 1). The macro-aggregates (greater than 250µm) at 0-5cm depth showed lots of variability and no significant differences amongst the six treatments over the years. The micro-aggregates (less than 50µm) at 0-5cm depth showed one anomalous effect in April 2013, where offset discing produced a significantly (P=0.03) lower % of micro-aggregates (less dust) than ongoing direct drilling or scarifying. It was accompanied by a higher % of macro-aggregates but which failed to reach statistical significance. From 2014 to 2015, the use of offset discs in 2012 tended to produce a lower % of macro-aggregates and a higher % of micro-aggregates. Although not significant because of the variability in the data, the trend serves as a caution that small detrimental effects from the use of offset discs might have occurred.

At 5-10cm depth, there were three statistically significant differences amongst the six treatments over the period 2012-2015. In March 2013, the ongoing direct drilling produced better macro-aggregation than the scarifier or offset discs. In June 2012, ongoing direct drilling produced more micro-aggregates at 5-10cm (this was post sowing). In May 2013, the use of offset discs produced more micro-aggregates (and less macro-aggregates though difference is not significant) than the other two treatments. There were no differences between the six treatments at 5-10cm depth from July 2013 to April 2015, indicating that any of these minor effects of tillage on WAS had been overcome.

Overall, the effects of tillage on WAS at this site were minor and short lived, indicating that there were no detriments of tillage to WAS that are likely to be of practical relevance.

Figure 1. Change in WAS over three years at two depths at the Daysdale site.
The upper lines and symbols in each graph describe the macro-aggregate data, while the lower lines and symbols describe the micro-aggregate data. The treatment key is given on each graph. Filled symbols describe the 2012 site and hollow symbols the 2013 site. Error bars are for tillage main effects.

Harden long term site

The CSIRO long term experiment at Harden contained treatments that had been under direct drilling for 20 years, with stubble that had been either burnt or mulched. We split the plots and tilled half of each with a rotary hoe. This gave us four treatments, a 2x2 factorial: direct drilling with stubble retained (DD SR) or with stubble burnt (DD SB), and a single cultivation by rotary hoe with either stubble retained (RH SR) or stubble burnt (RH SB). In addition, we undertook measurements on two test strips — one strip under a double rotary hoeing, which we then sowed to pasture (ryegrass — subclover) and the other under a five year old pasture (phalaris) across the trial fence in the grower’s paddock.

Where the soil had been under DD SR for over 20 years, about 70% of the surface soil (0-5cm depth) was in stable macro-aggregates (Figure 2). The five year old pasture in the grower’s paddock had a similar proportion of macro-aggregates. However, when the soil received two passes of a rotary hoe, the soil’s macro-aggregation decreased to approximately 50%WAS, indicating a large increase in susceptibility to erosion. However, after two years of pasture, the soil recovered from this loss of aggregate stability. Conversely, the increased micro-aggregation (dust) due to rotary hoeing was rectified after 12 months. Hence the damage caused by a severe cultivation was rectified in 12 to 24 months where a pasture was grown after the cultivation.

Under continuous cropping, the recovery of the soil from cultivation was not so clear (Figure 3). The treatments were all previously under DD but were split for a single pass of a rotary hoe, not two passes as in Figure 2. The damage from the single pass was not so large, averaging about an 8% decrease in macro-aggregate stability. The DD SR treatment maintained the best macro-aggregate stability for the full four years of measurement. Both the burning of stubble and cultivation of soil lowered the stability of the soil’s macro-aggregates. The RH SB treatment produced the lowest macro-aggregate stability as expected, at about 60%. After two years (2011 to 2013), only the DD SR treatment was significantly better than the other treatments, though tillage and stubble effects re-emerged in 2014 only to disappear again in 2015. From a practical perspective, the DD SR treatment maintained better soil structure in the surface soil over the four years of the trial, while the other three treatments converged to a lower value of %WAS over 2-4 years.

Figure 3B shows the macro-aggregate stability at 5-10cm depth for the same four treatments. Here the macro-aggregate stability of the soil under DD was less than for the cultivated treatments. This indicates that some of the ’damage’ caused by tillage was simply due to mixing of the 0-5 and 5-10cm layers of soil by tillage. Again, by two years there was no significant difference between treatments. 

Given that macro-aggregate stability is most important for the surface soil, as protection from erosion, Figure 3A indicates the need for caution when contemplating tillage. The impact of rotary hoeing on the surface soil causes an increased susceptibility to erosion, albeit only a 5-10% increase in risk on this soil.

Figure 2. The wet aggregate stability (%WAS) of the pasture soils at Harden over four years (2011-2015). The upper two lines represent macro-aggregates (greater than 250µm) whilst the lower two lines represent micro-aggregates (less than50µm). The error bars are for the stubble by tillage interaction.

Figure 3. The wet aggregate stability (%WAS) of macro-aggregates (greater than 250µm) for the cropped soils at Harden over four years (2011-2015).
The error bars are for the stubble by tillage interaction at each depth.

Brief comment on the other sites

The Berthong site, like the Harden site, showed some longer term detriment from tillage to macro-aggregation in the surface 0-5cm. The Thuddungra site showed some initial damage to soil structure at 0-5cm depth from offset discs but recovery occurred within two years.

Conclusions from the data

Tillage can do damage to soil structural stability. The damage is generally small, 5-10% in terms of WAS, but can exceed 15% loss of WAS in the instance of rotary hoeing.

Recovery time for WAS ranged from one to more than four years depending on the severity of the tillage and the rotation and stubble management that followed. Including a pasture phase in the farming system is likely to repair any damage to soil structure resulting from strategic tillage during the cropping phase.

Practical message

• It is not frequent, i.e. less than once every four years based on the above.
• The well-known risks for erosion are considered such as slope, groundcover, and current weather, and the soil moisture at the time of the proposed tillage. Hence we suggest leaving the tillage as late as possible so as to avoid the chance of erosive rainfall between tillage and crop establishment.