Evaluation of deep tillage in cohesive soils of Queensland, Australia

Evaluation of deep tillage in cohesive soils of Queensland, Australia

With the rapid global trend towards mechanized, continuous and dense cropping systems that provide agricultural efficiency to meet consumer demand, soil compaction has become a recognized problem. Soil compaction under modern machines has had immense impact on productive land’s physical, chemical and biological properties, including soil-water storage capacity, fertiliser use efficiency, and plant root architecture. As a result, farms are experiencing substantially reduced crop yields and economic returns. The percentage of soil compaction increases with increased soil clay fraction. Numerous investigations have been conducted to evaluate the technical, economic and soil-crop efficiency of compaction mitigation strategies, but deep tillage has not received sufficient consideration, particularly in relation to high clay content soils.

This study was conducted to technically and economically evaluate a range of deep ripping systems, and study the effect of tillage on soil and crop grown on cohesive soils. A series of field experiments were conducted to parametrise a soil tillage force prediction model, previously developed by Godwin and O’Dogherty (2007) and the Agricultural Productions Systems sIMulator (APSIM) developed by the Agricultural Production Systems Research Unit in Australia (Holzworth et al., 2014; Keating et al., 2003). The behaviour of soil physical properties, power requirements of ripping operations and cost, and agronomic and economic performance of sorghum and wheat were assessed at the University of Southern Queensland’s research ground in Toowoomba, Queensland (Australia) over two consecutive seasons (2015-16 and 2016-17). The work was conducted by replicating the soil conditions commonly found in non-controlled or ‘random’ traffic farming systems, referred to as RTF. Sorghum was also grown at a commercial farm located in Evanslea near Toowoomba, under controlled traffic (CTF) conditions (a farm system based on a permanent lanes for machinery traffic) during the 2018 summer crop season.

The soil types at the two sites are Red Ferrosol (69.1% clay, 10.0% silt, and 20.9% sand) and Black Vertosol (64.8% clay, 23.4% silt, and 11.8% sand). Three levels of deep ripping depth, namely, Deep Ripping 1 (D1= 0-0.3 m), Deep Ripping 2 (D2= 0- 0.6 m), and Control (C= no ripping) were applied using a Barrow single tine ripper at the Ag plot site - USQ, and a Tilco eight-tine ripper was used at the Evanslea site. The tillage operations were performed at 2.7 km/h. A predetermined optimum N fertiliser rate was applied after sorghum and wheat sowing at the Ag plot site. The field experiments were conducted according to the randomized complete block design (RCBD). The Statistical Package for Social Scientists (SPSS) software was utilized to analyse the significance of the differences between the variables at the probability level of 5% as the least significant difference (LSD).

The statistical analysis results showed that the D2 treatment significantly reduced soil bulk density and soil strength by up to 5% and 24% for Red Ferrosol soil, and by up to 6% and 40% for Black Vertosol soil respectively, and increased water content compared with the D1 and C treatments. Overall results showed that D2 was superior in ameliorating the properties of both soils. In both soils, energy requirement results showed that tillage draft force and tractor power requirements were dependent on tillage depth, but for both tillage treatments, energy consumption was slightly lower for the CTF system (Evanslea site) than the RTF system at Ag plot site.

Crop performance results showed that at the Ag plot site, the grain and biomass yields were highest by up to 19% for sorghum and by up to 30% for wheat when the D2 treatment was applied, compared to the D1 and C treated crop yield components. Also, the grain and biomass yields were highest for fertilised soil by up to 10% for sorghum and by up to 16% and 25% for wheat respectively, in comparison with the nonfertilised treatments soils yield. Fertilising of D2 treated soil produced the highest significant yield of sorghum grain (5360 kg/ha), biomass (13269 kg/ha), wheat grain (2419 kg/ha), and biomass (5960 kg/ha) compared to the yield of the other treatment interactions. However, at Evanslea site, the D1 treatment showed significantly higher yield and yield components for sorghum compared with C practice (by up to 17% higher yield), and no differences were observed for treatment D2.

Economically, the D1 treatment required the lowest total operational cost at both sites, which was estimated at AUD125/ha and AUD25.8/ha at the Ag plot and Evanslea sites, respectively. These results compare to AUD139.3/ha (Ag plot) and AUD30.8/ha (Evanslea) for the D2 ripping system. With regard to economic returns, at the Ag plot site, D2 yielded the highest sorghum gross benefit (AUD1422/ha) and net benefit (AUD1122/ha), wheat gross benefit (AUD590/ha) and net benefit (AUD482.3/ha), 2017 season gross benefit (AUD 2011.7/ha) and 2017 season net benefit (AUD 1604.7/ha), compared to D1 and C soil benefits. The economic fertiliser application at this site achieved the highest gross benefit for sorghum (AUD1384.2/ha), wheat (AUD555.6/ha), and 2017 season (AUD1939.8/ha) respectively, in comparison with the non-fertilised soils’ total return. Also, fertilised D2 treated soil resulted in the highest sorghum gross benefit (AUD1512.9/ha) and net benefit (AUD1170.3/ha), wheat gross benefit (AUD633.7/ha) and net benefit (AUD492.4/ha), 2017 season gross benefit (AUD2146.6/ha), and net benefit (AUD1662.7/ha) compared to other interactions’ benefits. At the Evanslea site, D1 significantly increased sorghum gross benefit and net benefit by up to 17% (AUD2277.9/ha) and by up to 20% (AUD1825.5/ha), respectively compared to C benefits, and no differences were observed with treatment D2.

The average of APSIM derived results for the long-term (1980-2017) at the Ag plot site showed that the D2 treatment reported consistently higher grain sorghum (4192 kg/ha), biomass (11454 kg/ha), wheat grain (3783 kg/ha), and biomass (10623kg/ha), compared to the D1 and C treatments’ yields under the same long-term conditions. However, at the Evanslea site, for long-term (1980-2018), APSIM simulation showed that D1 treatment increased the yield of sorghum grain and biomass significantly by up to 10% (5823 kg/ha) and 11% (12171 kg/ha), respectively compared to C treatment’s production, but these increases were found not significant with the D2 yields’ components. APSIM model simulation of field experiment conditions during 2017 season at the Ag plot site showed that the D2 treatment also had the highest significant yield of sorghum grain (5284 kg/ha), biomass (12488 kg/ha), wheat grain (2341 kg/ha) and biomass (6081 kg/ha) compared to the C and D1 crop yields. Similarly, APSIM model simulation of field experiment circumstances during the 2018 season at the Evanslea site showed that the D1 treatment produced the highest yield of sorghum grain (7129 kg/ha), biomass (13364 kg/ha) yields, compared to the C and D1 crop yields.

Overall, both the long and short-term model outputs were in good agreement with experimental data, suggesting beneficial effects of deep tillage in improving cereal crops’ productivity in this region. Moreover, in comparison with the study findings, the model prediction error rate was ±7, which indicates that the developed model approach is valid and calibrated during this study.

Results derived from the G&O soil tillage mechanics model under the Ag plot and Evanslea soil conditions showed that the required tractive force increases with the increasing operation working depth. Furthermore, the D1 was superior, requiring the lowest draft force at Ag plot (7.48 kN) and Evanslea (19.65 kN) soils, compared to the D2 required forces which were 43.28 kN and 41.41kN at both sites, respectively. In general, the model values were in line with the experiments' draft forces and when compared with the study readings, the model prediction error rate was ±8, which indicates that it is also valid and calibrated during this study.

Finally, the study provides conclusions and recommendations that contribute to crop production improvement in the face of recurrent and increasing challenges, as well as emphasizing the necessity of correct management and cultivation of economically important crops after the application of deep ripping to produce accurate results that serve decision-making in the agricultural sector.